Project

PDF-Introduction-to-Java-Programming-and-Data-Structures-Comprehensive-11th.pdf

Home >Computer Science homework help >Project

ISSLJC-SETUP-ALIEN-PAREU-BEGUN-LIKES

Digital Resources for Students

Your new textbook provides 12-month access to digital resources that may include VideoNotes (step-by-step video tutorials on programming concepts), source code, web chapters, quizzes, and more. Refer to the preface in the textbook for a detailed list of resources.

Follow the instructions below to register for the Companion Website for Daniel Liang’s Introduction to Java™ Programming and Data Structures, Comprehensive Version, Eleventh Edition, Global Edition.

1. Go to www.pearsonglobaleditions.com/liang 2. Enter the title of your textbook or browse by author name. 3. Click Companion Website. 4. Click Register and follow the on-screen instructions to create a login name and password.

Use a coin to scratch off the coating and reveal your access code. Do not use a sharp knife or other sharp object as it may damage the code.

Use the login name and password you created during registration to start using the digital resources that accompany your textbook.

IMPORTANT: This prepaid subscription does not include access to MyProgrammingLab, which is available at www.myprogramminglab.com for purchase.

This access code can only be used once. This subscription is valid for 12 months upon activation and is not transferable. If the access code has already been revealed it may no longer be valid.

For technical support go to https://support.pearson.com/getsupport

Liang_11_1292221879_ifc_Final.indd 1 17/11/17 9:19 PM

Digital Resources for Students

Use a coin to scratch off the coating and reveal your access code. Do not use a sharp knife or other sharp object as it may damage the code.

Use the login name and password you created during registration to start using the digital resources that accompany your textbook.

IMPORTANT: This prepaid subscription does not include access to MyProgrammingLab, which is available at www.myprogramminglab.com for purchase.

This access code can only be used once. This subscription is valid for 12 months upon activation and is not transferable. If the access code has already been revealed it may no longer be valid.

For technical support go to https://support.pearson.com/getsupport

Liang_11_1292221879_ifc_Final.indd 1 17/11/17 9:19 PM Digital_Resources_for_Students.indd 1 1/17/18 8:14 PM

IntroductIon to

Java ProgrammIng and

data StructureS comPrehenSIve verSIon

Eleventh Edition

Global Edition

Y. daniel Liang Armstrong State University

™

330 Hudson Street, NY NY 10013

A01_LIAN1878_11_GE_FM.indd 1 1/2/18 11:57 PM

To Samantha, Michael, and Michelle

Java™ and Netbeans™ screenshots ©2017 by Oracle Corporation, all rights reserved. Reprinted with permission. Credits and acknowledgments borrowed from other sources and reproduced, with permission, in this textbook appear on the appropriate page within text. Microsoft and/or its respective suppliers make no representations about the suit- ability of the information contained in the documents and related graphics published as part of the services for any purpose. All such documents and related graphics are provided “as is” without warranty of any kind. Microsoft and/ or its respective suppliers hereby disclaim all warranties and conditions with regard to this information, including all warranties and conditions of merchantability, whether express, implied or statutory, fitness for a particular purpose, title and non-infringement. In no event shall Microsoft and/or its respective suppliers be liable for any special, indi- rect or consequential damages or any damages whatsoever resulting from loss of use, data or profits, whether in an action of contract, negligence or other tortious action, arising out of or in connection with the use or performance of information available from the services. The documents and related graphics contained herein could include techni- cal inaccuracies or typographical errors. Changes are periodically added to the information herein. Microsoft and/or its respective suppliers may make improvements and/or changes in the product(s) and/or the program(s) described herein at any time. Partial screen shots may be viewed in full within the software version specified.

Pearson Education Limited KAO Two KAO Park Harlow CM17 9NA United Kingdom

and Associated Companies throughout the world

Visit us on the World Wide Web at: www.pearsonglobaleditions.com

The rights of Y. Daniel Liang to be identified as the author of this work have been asserted by him in accordance with the Copyright, Designs and Patents Act 1988.

Authorized adaptation from the United States edition, entitled Introduction to Java Programming and Data Structures, Comprehensive Version, 11th Edition, ISBN 978-0-13-467094-2 by Y. Daniel Liang, published by Pearson Education © 2018.

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without either the prior written permission of the publisher or a license permitting restricted copying in the United Kingdom issued by the Copyright Licensing Agency Ltd, Saffron House, 6–10 Kirby Street, London EC1N 8TS.

All trademarks used herein are the property of their respective owners. The use of any trademark in this text does not vest in the author or publisher any trademark ownership rights in such trademarks, nor does the use of such trademarks imply any affiliation with or endorsement of this book by such owners.

British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library

10 9 8 7 6 5 4 3 2 1

Typeset by SPi Global ISBN-10: 1-292-22187-9

Printed and bound by Vivar in Malaysia ISBN-13: 978-1-292-22187-8

Senior Vice President Courseware Portfolio Management: Marcia J. Horton

Director, Portfolio Management: Engineering, Computer Science & Global Editions: Julian Partridge

Higher Ed Portfolio Management: Tracy Johnson (Dunkelberger)

Portfolio Management Assistant: Kristy Alaura Managing Content Producer: Scott Disanno Content Producer: Robert Engelhardt Web Developer: Steve Wright Assistant Acquisitions Editor, Global Edition:

Aditee Agarwal Assistant Project Editor, Global Edition: Shaoni

Mukherjee

Manager, Media Production, Global Edition: Vikram Kumar

Senior Manufacturing Controller, Production, Global Edition: Jerry Kataria

Rights and Permissions Manager: Ben Ferrini Manufacturing Buyer, Higher Ed, Lake Side

Communications Inc (LSC): Maura Zaldivar-Garcia Inventory Manager: Ann Lam Marketing Manager: Demetrius Hall Product Marketing Manager: Bram Van Kempen Marketing Assistant: Jon Bryant Cover Designer: Lumina Datamatics Cover Image: Eduardo Rocha/ shutterstock.com Full-Service Project Management: Shylaja Gattupalli,

SPi Global

A01_LIAN1878_11_GE_FM.indd 2 1/2/18 11:57 PM

Dear Reader,

Many of you have provided feedback on earlier editions of this book, and your comments and suggestions have greatly improved the book. This edition has been substantially enhanced in presentation, organization, examples, exercises, and supplements.

The book is fundamentals first by introducing basic programming concepts and techniques before designing custom classes. The fundamental concepts and techniques of selection statements, loops, methods, and arrays are the foundation for programming. Building this strong foundation prepares students to learn object-oriented programming and advanced Java programming.

This book teaches programming in a problem-driven way that focuses on problem solv- ing rather than syntax. We make introductory programming interesting by using thought- provoking problems in a broad context. The central thread of early chapters is on problem solving. Appropriate syntax and library are introduced to enable readers to write programs for solving the problems. To support the teaching of programming in a problem-driven way, the book provides a wide variety of problems at various levels of difficulty to motivate students. To appeal to students in all majors, the problems cover many application areas, including math, science, business, financial, gaming, animation, and multimedia.

The book seamlessly integrates programming, data structures, and algorithms into one text. It employs a practical approach to teach data structures. We first introduce how to use various data structures to develop efficient algorithms, and then show how to implement these data structures. Through implementation, students gain a deep understanding on the efficiency of data structures and on how and when to use certain data structures. Finally, we design and implement custom data structures for trees and graphs.

The book is widely used in the introductory programming, data structures, and algorithms courses in the universities around the world. This comprehensive version covers fundamen- tals of programming, object-oriented programming, GUI programming, data structures, algo- rithms, concurrency, networking, database, and Web programming. It is designed to prepare students to become proficient Java programmers. A brief version (Introduction to Java Pro- gramming, Brief Version, Eleventh Edition, Global Edition) is available for a first course on programming, commonly known as CS1. The brief version contains the first 18 chapters of the comprehensive version.

The best way to teach programming is by example, and the only way to learn programming is by doing. Basic concepts are explained by example and a large number of exercises with various levels of difficulty are provided for students to practice. For our programming courses, we assign programming exercises after each lecture.

Our goal is to produce a text that teaches problem solving and programming in a broad context using a wide variety of interesting examples. If you have any comments on and suggestions for improving the book, please email me.

Sincerely,

Y. Daniel Liang [email protected] www.pearsonglobaleditions.com/Liang

fundamentals-first

problem-driven

data structures

comprehensive version

brief version

Preface

A01_LIAN1878_11_GE_FM.indd 3 1/2/18 11:57 PM

4 Preface

ACM/IEEE Curricular 2013 and ABET Course Assessment The new ACM/IEEE Computer Science Curricular 2013 defines the Body of Knowledge organized into 18 Knowledge Areas. To help instructors design the courses based on this book, we provide sample syllabi to identify the Knowledge Areas and Knowledge Units. The sample syllabi are for a three semester course sequence and serve as an example for institutional customization. The sample syllabi are accessible from the Instructor Resource Center.

Many of our users are from the ABET-accredited programs. A key component of the ABET accreditation is to identify the weakness through continuous course assessment against the course outcomes. We provide sample course outcomes for the courses and sam- ple exams for measuring course outcomes on the Instructor Resource Center.

What’s New in This Edition? This edition is completely revised in every detail to enhance clarity, presentation, content, examples, and exercises. The major improvements are as follows:

■■ The book’s title is changed to Introduction to Java Programming and Data Structures with new enhancements on data structures. The book uses a practical approach to introduce design, implement, and use data structures and covers all topics in a typical data structures course. Additionally, it provides bonus chapters that cover advanced data structures such as 2-4 trees, B-trees, and red-black trees.

■■ Updated to the latest Java technology. Examples and exercises are improved and simplified by using the new features in Java 8.

■■ The default and static methods are introduced for interfaces in Chapter 13.

■■ The GUI chapters are updated to JavaFX 8. The examples are revised. The user interfaces in the examples and exercises are now resizable and displayed in the center of the window.

■■ Inner classes, anonymous inner classes, and lambda expressions are covered using practi- cal examples in Chapter 15.

■■ More examples and exercises in the data structures chapters use lambda expressions to simplify coding. Method references are introduced along with the Comparator interface in Section 20.6.

■■ The forEach method is introduced in Chapter 20 as a simple alternative to the foreach loop for applying an action to each element in a collection.

■■ Use the default methods for interfaces in Java 8 to redesign and simplify MyList, MyArrayList, MyLinkedList, Tree, BST, AVLTree, MyMap, MyHashMap, MySet, MyHashSet, Graph, UnweightedGraph, and WeightedGraph in Chapters 24–29.

■■ Chapter 30 is brand new to introduce aggregate operations for collection streams.

■■ FXML and the Scene Builder visual tool are introduced in Chapter 31.

■■ The Companion Website has been redesigned with new interactive quiz, CheckPoint ques- tions, animations, and live coding.

■■ More than 200 additional programming exercises with solutions are provided to the instructor on the Instructor Resource Center. These exercises are not printed in the text.

A01_LIAN1878_11_GE_FM.indd 4 1/2/18 11:57 PM

Preface 5

Pedagogical Features The book uses the following elements to help students get the most from the material:

■■ The Objectives at the beginning of each chapter list what students should learn from the chapter. This will help them determine whether they have met the objectives after completing the chapter.

■■ The Introduction opens the discussion with a thought-provoking question to motivate the reader to delve into the chapter.

■■ Key Points highlight the important concepts covered in each section.

■■ Check Points provide review questions to help students track their progress as they read through the chapter and evaluate their learning.

■■ Problems and Case Studies, carefully chosen and presented in an easy-to-follow style, teach problem solving and programming concepts. The book uses many small, simple, and stimulating examples to demonstrate important ideas.

■■ The Chapter Summary reviews the important subjects that students should understand and remember. It helps them reinforce the key concepts they have learned in the chapter.

■■ Quizzes are accessible online, grouped by sections, for students to do self-test on programming concepts and techniques.

■■ Programming Exercises are grouped by sections to provide students with opportunities to apply the new skills they have learned on their own. The level of difficulty is rated as easy (no asterisk), moderate (*), hard (**), or challenging (***). The trick of learning program- ming is practice, practice, and practice. To that end, the book provides a great many exer- cises. Additionally, more than 200 programming exercises with solutions are provided to the instructors on the Instructor Resource Center. These exercises are not printed in the text.

■■ Notes, Tips, Cautions, and Design Guides are inserted throughout the text to offer valuable advice and insight on important aspects of program development.

Note Provides additional information on the subject and reinforces important concepts.

Tip Teaches good programming style and practice.

Caution Helps students steer away from the pitfalls of programming errors.

Design Guide Provides guidelines for designing programs.

Flexible Chapter Orderings The book is designed to provide flexible chapter orderings to enable GUI, exception handling, recursion, generics, and the Java Collections Framework to be covered earlier or later. The diagram on the next page shows the chapter dependencies.

A01_LIAN1878_11_GE_FM.indd 5 1/2/18 11:57 PM

C ha

pt er

3 9

Ja va

Se rv

er F

ac es

C ha

pt er

4 0

R M

C ha

pt er

4 1

W eb

S er

vi ce

C ha

pt er

4 4

T es

ti ng

U si

ng J

U ni

C ha

pt er

3 8

Ja va

Se rv

er P

ag es

C ha

pt er

2 5

B in

ar y

Se ar

ch T

re es

C ha

pt er

2 6

A V

L T

re es

C ha

pt er

2 9

W ei

gh te

d G

ra ph

s an

d A

pp lic

at io

C ha

pt er

2 8

G ra

ph s

an d

pp lic

at io

C ha

pt er

2 1

Se ts

a nd

M ap

C ha

pt er

2 2

D ev

el op

pi ng

f� ci

en t A

lg or

it hm

C ha

pt er

1 I

nt ro

du ct

io n

C om

pu te

rs , P

ro gr

am s,

a nd

av a

C ha

pt er

2 E

le m

en ta

P ro

gr am

m in

C ha

pt er

5 L

oo ps

C ha

pt er

7 S

in gl

e- D

im en

si on

A rr

ay s

C ha

pt er

8 M

ul ti

di m

en si

on al

rr ay

C ha

pt er

4 M

at he

m at

ic al

un ct

io ns

, C ha

ra ct

er s,

nd S

tr in

P ar

t I : F

un da

m en

ta ls

o f

ro gr

am m

in g

C ha

pt er

3 S

el ec

ti on

C ha

pt er

9 O

bj ec

ts a

nd C

la ss

C ha

pt er

1 7

B in

ar y

I/ O

N ot

e: C

ha pt

er s

1– 18

a re

in th

e br

ie f v

er si

on o

f t hi

s bo

ok .

N ot

e: C

ha pt

er s

1– 30

a re

in th

e co

m pr

eh en

si ve

v er

si on

N ot

e: C

ha pt

er s

31 –4

4 ar

e bo

nu s

ch ap

te rs

a va

ila bl

e fr

om th

e C

om pa

ni on

W eb

si te

C ha

pt er

1 0

T hi

nk in

g in

O bj

ec ts

C ha

pt er

1 1

In he

ri ta

nc e

an d

ol ym

or ph

is m

C ha

pt er

1 2

E xc

ep ti

H an

dl in

g a

nd T

ex t I

C ha

pt er

1 3

A bs

tr ac

t C la

ss es

nd I

nt er

fa ce

C ha

pt er

6 M

et ho

P ar

t I I:

O bj

ec t-

O ri

en te

P ro

gr am

m in

C ha

pt er

3 2

M ul

ti th

re ad

in g

an d

P ar

al le

l P ro

gr am

m in

C ha

pt er

3 6

In te

rn at

io na

liz at

io n

C ha

pt er

3 3

N et

w or

ki ng

C ha

pt er

3 4

Ja va

D at

ab as

ro gr

am m

in g

C ha

pt er

3 5

A dv

an ce

d D

at ab

as e

P ro

gr am

m in

C ha

pt er

3 7

Se rv

le ts

P ar

t V : A

dv an

ce d

Ja va

ro gr

am m

in g

C ha

pt er

1 4

Ja va

F X

B as

ic s

C ha

pt er

1 5

E ve

nt -D

ri ve

P ro

gr am

m in

g an

A ni

m at

io ns

C ha

pt er

2 0

L is

ts , S

ta ck

s, Q

ue ue

a nd

P ri

or it

y Q

ue ue

s C

ha pt

er 1

6 Ja

va F

X C

on tr

ol s

nd M

ul ti

m ed

C ha

pt er

3 1

A dv

an ce

d J

av aF

a nd

F X

M L

P ar

t I II

: G U

I P

ro gr

am m

in g

C ha

pt er

1 8

R ec

ur si

on C

h 7

C ha

pt er

1 9

G en

er ic

C ha

pt er

2 4

Im pl

em en

ti ng

L is

ts ,

ta ck

s, Q

ue ue

s, a

nd P

ri or

it y

ue ue

P ar

t I V

: D at

a St

ru ct

ur es

a nd

lg or

it hm

C h

C ha

pt er

3 0

A gg

re ga

te O

pe ra

tio ns

an d

C ol

le ct

io n

St re

am s

C ha

pt er

4 2

2- 4

T re

es a

nd B

T re

C ha

pt er

4 3

R ed

-B la

ck T

re es

C ha

pt er

2 7

H as

hi ng

C ha

pt er

2 3

So rt

in g

6 Preface

A01_LIAN1878_11_GE_FM.indd 6 1/2/18 11:57 PM

Organization of the Book The chapters can be grouped into five parts that, taken together, form a comprehensive introduc- tion to Java programming, data structures and algorithms, and database and Web programming. Because knowledge is cumulative, the early chapters provide the conceptual basis for under- standing programming and guide students through simple examples and exercises; subsequent chapters progressively present Java programming in detail, culminating with the development of comprehensive Java applications. The appendixes contain a mixed bag of topics, including an introduction to number systems, bitwise operations, regular expressions, and enumerated types.

Part I: Fundamentals of Programming (Chapters 1–8)

The first part of the book is a stepping stone, preparing you to embark on the journey of learning Java. You will begin to learn about Java (Chapter 1) and fundamental programming techniques with primitive data types, variables, constants, assignments, expressions, and operators ( Chapter 2), selection statements (Chapter 3), mathematical functions, characters, and strings (Chapter 4), loops (Chapter 5), methods (Chapter 6), and arrays (Chapters 7–8). After Chapter 7, you can jump to Chapter 18 to learn how to write recursive methods for solving inherently recursive problems.

Part II: Object-Oriented Programming (Chapters 9–13, and 17)

This part introduces object-oriented programming. Java is an object-oriented programming language that uses abstraction, encapsulation, inheritance, and polymorphism to provide great flexibility, modularity, and reusability in developing software. You will learn program- ming with objects and classes (Chapters 9–10), class inheritance (Chapter 11), polymorphism ( Chapter 11), exception handling (Chapter 12), abstract classes (Chapter 13), and interfaces (Chapter 13). Text I/O is introduced in Chapter 12 and binary I/O is discussed in Chapter 17.

Part III: GUI Programming (Chapters 14–16 and Bonus Chapter 31)

JavaFX is a new framework for developing Java GUI programs. It is not only useful for developing GUI programs, but also an excellent pedagogical tool for learning object-oriented programming. This part introduces Java GUI programming using JavaFX in Chapters 14–16. Major topics include GUI basics (Chapter 14), container panes (Chapter 14), drawing shapes (Chapter 14), event-driven programming (Chapter 15), animations (Chapter 15), and GUI controls (Chapter 16), and playing audio and video (Chapter 16). You will learn the architecture of JavaFX GUI programming and use the controls, shapes, panes, image, and video to develop useful applications. Chapter 31 covers advanced features in JavaFX.

Part IV: Data Structures and Algorithms (Chapters 18–30 and Bonus Chapters 42–43)

This part covers the main subjects in a typical data structures and algorithms course. Chapter 18 introduces recursion to write methods for solving inherently recursive problems. Chapter 19 presents how generics can improve software reliability. Chapters 20 and 21 introduce the Java Collection Framework, which defines a set of useful API for data structures. Chapter 22 discusses measur- ing algorithm efficiency in order to choose an appropriate algorithm for applications. Chapter 23 describes classic sorting algorithms. You will learn how to implement several classic data struc- tures lists, queues, and priority queues in Chapter 24. Chapters 25 and 26 introduce binary search trees and AVL trees. Chapter 27 presents hashing and implementing maps and sets using hashing. Chapters 28 and 29 introduce graph applications. Chapter 30 introduces aggregate operations for collection streams. The 2-4 trees, B-trees, and red-black trees are covered in Bonus Chapters 42–43.

Part V: Advanced Java Programming (Chapters 32-41, 44)

This part of the book is devoted to advanced Java programming. Chapter 32 treats the use of multithreading to make programs more responsive and interactive and introduces parallel pro- gramming. Chapter 33 discusses how to write programs that talk with each other from different

Preface 7

A01_LIAN1878_11_GE_FM.indd 7 1/2/18 11:57 PM

hosts over the Internet. Chapter 34 introduces the use of Java to develop database projects. Chapter 35 delves into advanced Java database programming. Chapter 36 covers the use of internationalization support to develop projects for international audiences. Chapters 37 and 38 introduce how to use Java servlets and JavaServer Pages to generate dynamic content from Web servers. Chapter 39 introduces modern Web application development using JavaServer Faces. Chapter 40 introduces remote method invocation and Chapter 41 discusses Web ser- vices. Chapter 44 introduces testing Java programs using JUnit.

Appendixes

This part of the book covers a mixed bag of topics. Appendix A lists Java keywords. Appendix B gives tables of ASCII characters and their associated codes in decimal and in hex. Appen- dix C shows the operator precedence. Appendix D summarizes Java modifiers and their usage. Appendix E discusses special floating-point values. Appendix F introduces number systems and conversions among binary, decimal, and hex numbers. Finally, Appendix G introduces bitwise operations. Appendix H introduces regular expressions. Appendix I covers enumerated types.

Java Development Tools You can use a text editor, such as the Windows Notepad or WordPad, to create Java programs and to compile and run the programs from the command window. You can also use a Java development tool, such as NetBeans or Eclipse. These tools support an integrated develop- ment environment (IDE) for developing Java programs quickly. Editing, compiling, building, executing, and debugging programs are integrated in one graphical user interface. Using these tools effectively can greatly increase your programming productivity. NetBeans and Eclipse are easy to use if you follow the tutorials. Tutorials on NetBeans and Eclipse can be found in the supplements on the Companion Website www.pearsonglobaleditions.com/Liang.

Student Resources The Companion Website (www.pearsonglobaleditions.com/Liang) contains the following resources:

■■ Answers to CheckPoint questions

■■ Solutions to majority of even-numbered programming exercises

■■ Source code for the examples in the book

■■ Interactive quiz (organized by sections for each chapter)

■■ Supplements

■■ Debugging tips

■■ Video notes

■■ Algorithm animations

Supplements The text covers the essential subjects. The supplements extend the text to introduce additional topics that might be of interest to readers. The supplements are available from the Companion Website.

IDE tutorials

8 Preface

A01_LIAN1878_11_GE_FM.indd 8 1/2/18 11:57 PM

Instructor Resources The Companion Website, accessible from www.pearsonglobaleditions.com/Liang, contains the following resources:

■■ Microsoft PowerPoint slides with interactive buttons to view full-color, syntax-highlighted source code and to run programs without leaving the slides.

■■ Solutions to a majority of odd-numbered programming exercises.

■■ More than 200 additional programming exercises and 300 quizzes organized by chapters. These exercises and quizzes are available only to the instructors. Solutions to these exercises and quizzes are provided.

■■ Web-based quiz generator. (Instructors can choose chapters to generate quizzes from a large database of more than two thousand questions.)

■■ Sample exams. Most exams have four parts:

■■ Multiple-choice questions or short-answer questions

■■ Correct programming errors

■■ Trace programs

■■ Write programs

■■ Sample exams with ABET course assessment.

■■ Projects. In general, each project gives a description and asks students to analyze, design, and implement the project.

Some readers have requested the materials from the Instructor Resource Center. Please understand that these are for instructors only. Such requests will not be answered.

Online Practice and Assessment with MyProgrammingLab MyProgrammingLab helps students fully grasp the logic, semantics, and syntax of program- ming. Through practice exercises and immediate, personalized feedback, MyProgrammingLab improves the programming competence of beginning students who often struggle with the basic concepts and paradigms of popular high-level programming languages.

A self-study and homework tool, a MyProgrammingLab course consists of hundreds of small practice problems organized around the structure of this textbook. For students, the system auto- matically detects errors in the logic and syntax of their code submissions and offers targeted hints that enable students to figure out what went wrong—and why. For instructors, a comprehensive gradebook tracks correct and incorrect answers and stores the code inputted by students for review.

MyProgrammingLab is offered to users of this book in partnership with Turing’s Craft, the makers of the CodeLab interactive programming exercise system. For a full demonstration, to see feedback from instructors and students, or to get started using MyProgrammingLab in your course, visit www.myprogramminglab.com.

Video Notes We are excited about the new Video Notes feature that is found in this new edition. These videos provide additional help by presenting examples of key topics and showing how to solve problems completely from design through coding. Video Notes are available from www.pearsonglobaleditions.com/Liang.

VideoNote

Preface 9

A01_LIAN1878_11_GE_FM.indd 9 04/01/18 6:28 pm

Algorithm Animations We have provided numerous animations for algorithms. These are valuable pedagogical tools to demonstrate how algorithms work. Algorithm animations can be accessed from the Com- panion Website.

Acknowledgments I would like to thank Armstrong State University for enabling me to teach what I write and for supporting me in writing what I teach. Teaching is the source of inspiration for continuing to improve the book. I am grateful to the instructors and students who have offered comments, suggestions, corrections, and praise. My special thanks go to Stefan Andrei of Lamar Univer- sity and William Bahn of University of Colorado Colorado Springs for their help to improve the data structures part of this book.

This book has been greatly enhanced thanks to outstanding reviews for this and previous edi- tions. The reviewers are: Elizabeth Adams (James Madison University), Syed Ahmed (North Georgia College and State University), Omar Aldawud (Illinois Institute of Technology), Ste- fan Andrei (Lamar University), Yang Ang (University of Wollongong, Australia), Kevin Bierre (Rochester Institute of Technology), Aaron Braskin (Mira Costa High School), David Champion (DeVry Institute), James Chegwidden (Tarrant County College), Anup Dargar (University of North Dakota), Daryl Detrick (Warren Hills Regional High School), Charles Dierbach (Towson Univer- sity), Frank Ducrest (University of Louisiana at Lafayette), Erica Eddy (University of Wisconsin at Parkside), Summer Ehresman (Center Grove High School), Deena Engel (New York University), Henry A. Etlinger (Rochester Institute of Technology), James Ten Eyck (Marist College), Myers Foreman (Lamar University), Olac Fuentes (University of Texas at El Paso), Edward F. Gehringer (North Carolina State University), Harold Grossman (Clemson University), Barbara Guillot (Loui- siana State University), Stuart Hansen (University of Wisconsin, Parkside), Dan Harvey (Southern Oregon University), Ron Hofman (Red River College, Canada), Stephen Hughes (Roanoke Col- lege), Vladan Jovanovic (Georgia Southern University), Deborah Kabura Kariuki (Stony Point High School), Edwin Kay (Lehigh University), Larry King (University of Texas at Dallas), Nana Kofi (Langara College, Canada), George Koutsogiannakis (Illinois Institute of Technology), Roger Kraft (Purdue University at Calumet), Norman Krumpe (Miami University), Hong Lin (DeVry Institute), Dan Lipsa (Armstrong State University), James Madison (Rensselaer Polytechnic Insti- tute), Frank Malinowski (Darton College), Tim Margush (University of Akron), Debbie Masada (Sun Microsystems), Blayne Mayfield (Oklahoma State University), John McGrath (J.P. McGrath Consulting), Hugh McGuire (Grand Valley State), Shyamal Mitra (University of Texas at Austin), Michel Mitri (James Madison University), Kenrick Mock (University of Alaska Anchorage), Frank Murgolo (California State University, Long Beach), Jun Ni (University of Iowa), Benjamin Nystuen (University of Colorado at Colorado Springs), Maureen Opkins (CA State University, Long Beach), Gavin Osborne (University of Saskatchewan), Kevin Parker (Idaho State University), Dale Par- son (Kutztown University), Mark Pendergast (Florida Gulf Coast University), Richard Povinelli (Marquette University), Roger Priebe (University of Texas at Austin), Mary Ann Pumphrey (De Anza Junior College), Pat Roth (Southern Polytechnic State University), Amr Sabry (Indiana Uni- versity), Ben Setzer (Kennesaw State University), Carolyn Schauble (Colorado State University), David Scuse (University of Manitoba), Ashraf Shirani (San Jose State University), Daniel Spiegel (Kutztown University), Joslyn A. Smith (Florida Atlantic University), Lixin Tao (Pace University), Ronald F. Taylor (Wright State University), Russ Tront (Simon Fraser University), Deborah Trytten (University of Oklahoma), Michael Verdicchio (Citadel), Kent Vidrine (George Washington Uni- versity), and Bahram Zartoshty (California State University at Northridge).

It is a great pleasure, honor, and privilege to work with Pearson. I would like to thank Tracy Johnson and her colleagues Marcia Horton, Demetrius Hall, Yvonne Vannatta, Kristy Alaura, Carole Snyder, Scott Disanno, Bob Engelhardt, Shylaja Gattupalli, and their colleagues for organizing, producing, and promoting this project.

As always, I am indebted to my wife, Samantha, for her love, support, and encouragement.

Animation

10 Preface

A01_LIAN1878_11_GE_FM.indd 10 1/2/18 11:57 PM

Acknowledgments for the Global Edition Pearson would like to thank and acknowledge Yvan Maillot (Univresite Haute-Alsace) and Steven Yuwono (National University of Singapore) for contributing to this Global Edition, and Arif Ahmed (National Institute of Technology, Silchar), Annette Bieniusa (University of Kaiserslautern), Shaligram Prajapat (Devi Ahilya Vishwavidyalaya, Indore), and Ram Gopal Raj (University of Malaya) for reviewing this Global Edition.

Preface 11

A01_LIAN1878_11_GE_FM.indd 11 1/2/18 11:57 PM

contentS chapter 1 Introduction to computers,

Programs, and Java™ 23 1.1 Introduction 24 1.2 What Is a Computer? 24 1.3 Programming Languages 29 1.4 Operating Systems 31 1.5 Java, the World Wide Web, and Beyond 32 1.6 The Java Language Specification, API, JDK,

JRE, and IDE 33 1.7 A Simple Java Program 34 1.8 Creating, Compiling, and Executing a Java Program 37 1.9 Programming Style and Documentation 40 1.10 Programming Errors 42 1.11 Developing Java Programs Using NetBeans 45 1.12 Developing Java Programs Using Eclipse 47

chapter 2 elementary Programming 55 2.1 Introduction 56 2.2 Writing a Simple Program 56 2.3 Reading Input from the Console 59 2.4 Identifiers 62 2.5 Variables 62 2.6 Assignment Statements and Assignment Expressions 64 2.7 Named Constants 65 2.8 Naming Conventions 66 2.9 Numeric Data Types and Operations 67 2.10 Numeric Literals 70 2.11 Evaluating Expressions and Operator Precedence 72 2.12 Case Study: Displaying the Current Time 74 2.13 Augmented Assignment Operators 76 2.14 Increment and Decrement Operators 77 2.15 Numeric Type Conversions 79 2.16 Software Development Process 81 2.17 Case Study: Counting Monetary Units 85 2.18 Common Errors and Pitfalls 87

chapter 3 Selections 97 3.1 Introduction 98 3.2 boolean Data Type 98 3.3 if Statements 100 3.4 Two-Way if-else Statements 102 3.5 Nested if and Multi-Way if-else Statements 103 3.6 Common Errors and Pitfalls 105 3.7 Generating Random Numbers 109 3.8 Case Study: Computing Body Mass Index 111 3.9 Case Study: Computing Taxes 112 3.10 Logical Operators 115 3.11 Case Study: Determining Leap Year 119 3.12 Case Study: Lottery 120 3.13 switch Statements 122

A01_LIAN1878_11_GE_FM.indd 12 1/2/18 11:57 PM

3.14 Conditional Operators 125 3.15 Operator Precedence and Associativity 126 3.16 Debugging 128

chapter 4 mathematical functions, characters, and Strings 141

4.1 Introduction 142 4.2 Common Mathematical Functions 142 4.3 Character Data Type and Operations 147 4.4 The String Type 152 4.5 Case Studies 161 4.6 Formatting Console Output 167

chapter 5 Loops 181 5.1 Introduction 182 5.2 The while Loop 182 5.3 Case Study: Guessing Numbers 185 5.4 Loop Design Strategies 188 5.5 Controlling a Loop with User Confirmation or a Sentinel Value 190 5.6 The do-while Loop 192 5.7 The for Loop 195 5.8 Which Loop to Use? 198 5.9 Nested Loops 200 5.10 Minimizing Numeric Errors 202 5.11 Case Studies 204 5.12 Keywords break and continue 208 5.13 Case Study: Checking Palindromes 211 5.14 Case Study: Displaying Prime Numbers 213

chapter 6 methods 227 6.1 Introduction 228 6.2 Defining a Method 228 6.3 Calling a Method 230 6.4 void vs. Value-Returning Methods 233 6.5 Passing Parameters by Values 236 6.6 Modularizing Code 239 6.7 Case Study: Converting Hexadecimals to Decimals 241 6.8 Overloading Methods 243 6.9 The Scope of Variables 246 6.10 Case Study: Generating Random Characters 247 6.11 Method Abstraction and Stepwise Refinement 249

chapter 7 Single-dimensional arrays 269 7.1 Introduction 270 7.2 Array Basics 270 7.3 Case Study: Analyzing Numbers 277 7.4 Case Study: Deck of Cards 278 7.5 Copying Arrays 280 7.6 Passing Arrays to Methods 281 7.7 Returning an Array from a Method 284 7.8 Case Study: Counting the Occurrences of Each Letter 285 7.9 Variable-Length Argument Lists 288 7.10 Searching Arrays 289 7.11 Sorting Arrays 293

Contents 13

A01_LIAN1878_11_GE_FM.indd 13 1/2/18 11:57 PM

7.12 The Arrays Class 294 7.13 Command-Line Arguments 296

chapter 8 multidimensional arrays 311 8.1 Introduction 312 8.2 Two-Dimensional Array Basics 312 8.3 Processing Two-Dimensional Arrays 315 8.4 Passing Two-Dimensional Arrays to Methods 317 8.5 Case Study: Grading a Multiple-Choice Test 318 8.6 Case Study: Finding the Closest Pair 320 8.7 Case Study: Sudoku 322 8.8 Multidimensional Arrays 325

chapter 9 objects and classes 345 9.1 Introduction 346 9.2 Defining Classes for Objects 346 9.3 Example: Defining Classes and Creating Objects 348 9.4 Constructing Objects Using Constructors 353 9.5 Accessing Objects via Reference Variables 354 9.6 Using Classes from the Java Library 358 9.7 Static Variables, Constants, and Methods 361 9.8 Visibility Modifiers 366 9.9 Data Field Encapsulation 368 9.10 Passing Objects to Methods 371 9.11 Array of Objects 375 9.12 Immutable Objects and Classes 377 9.13 The Scope of Variables 379 9.14 The this Reference 380

chapter 10 object-oriented thinking 389 10.1 Introduction 390 10.2 Class Abstraction and Encapsulation 390 10.3 Thinking in Objects 394 10.4 Class Relationships 397 10.5 Case Study: Designing the Course Class 400 10.6 Case Study: Designing a Class for Stacks 402 10.7 Processing Primitive Data Type Values as Objects 404 10.8 Automatic Conversion between Primitive Types

and Wrapper Class Types 407 10.9 The BigInteger and BigDecimal Classes 408 10.10 The String Class 410 10.11 The StringBuilder and StringBuffer Classes 416

chapter 11 Inheritance and Polymorphism 433

11.1 Introduction 434 11.2 Superclasses and Subclasses 434 11.3 Using the super Keyword 440 11.4 Overriding Methods 443 11.5 Overriding vs. Overloading 444 11.6 The Object Class and Its toString() Method 446 11.7 Polymorphism 447 11.8 Dynamic Binding 447 11.9 Casting Objects and the instanceof Operator 451 11.10 The Object’s equals Method 455

14 Contents

A01_LIAN1878_11_GE_FM.indd 14 1/2/18 11:57 PM

11.11 The ArrayList Class 456 11.12 Useful Methods for Lists 462 11.13 Case Study: A Custom Stack Class 463 11.14 The protected Data and Methods 464 11.15 Preventing Extending and Overriding 467

chapter 12 exception handling and text I/o 475

12.1 Introduction 476 12.2 Exception-Handling Overview 476 12.3 Exception Types 481 12.4 More on Exception Handling 484 12.5 The finally Clause 492 12.6 When to Use Exceptions 493 12.7 Rethrowing Exceptions 494 12.8 Chained Exceptions 495 12.9 Defining Custom Exception Classes 496 12.10 The File Class 499 12.11 File Input and Output 502 12.12 Reading Data from the Web 508 12.13 Case Study: Web Crawler 510

chapter 13 abstract classes and Interfaces 521 13.1 Introduction 522 13.2 Abstract Classes 522 13.3 Case Study: the Abstract Number Class 527 13.4 Case Study: Calendar and GregorianCalendar 529 13.5 Interfaces 532 13.6 The Comparable Interface 535 13.7 The Cloneable Interface 540 13.8 Interfaces vs. Abstract Classes 545 13.9 Case Study: The Rational Class 548 13.10 Class-Design Guidelines 553

chapter 14 JavafX Basics 563 14.1 Introduction 564 14.2 JavaFX vs Swing and AWT 564 14.3 The Basic Structure of a JavaFX Program 564 14.4 Panes, Groups, UI Controls, and Shapes 567 14.5 Property Binding 570 14.6 Common Properties and Methods for Nodes 573 14.7 The Color Class 575 14.8 The Font Class 576 14.9 The Image and ImageView Classes 578 14.10 Layout Panes and Groups 580 14.11 Shapes 589 14.12 Case Study: The ClockPane Class 602

chapter 15 event-driven Programming and animations 615

15.1 Introduction 616 15.2 Events and Event Sources 618 15.3 Registering Handlers and Handling Events 619 15.4 Inner Classes 623 15.5 Anonymous Inner Class Handlers 624

Contents 15

A01_LIAN1878_11_GE_FM.indd 15 1/2/18 11:57 PM

15.6 Simplifying Event Handling Using Lambda Expressions 627 15.7 Case Study: Loan Calculator 631 15.8 Mouse Events 633 15.9 Key Events 635 15.10 Listeners for Observable Objects 638 15.11 Animation 640 15.12 Case Study: Bouncing Ball 648 15.13 Case Study: US Map 652

chapter 16 JavafX uI controls and multimedia 665

16.1 Introduction 666 16.2 Labeled and Label 666 16.3 Button 668 16.4 CheckBox 670 16.5 RadioButton 673 16.6 TextField 676 16.7 TextArea 677 16.8 ComboBox 681 16.9 ListView 684 16.10 ScrollBar 687 16.11 Slider 690 16.12 Case Study: Developing a Tic-Tac-Toe Game 693 16.13 Video and Audio 698 16.14 Case Study: National Flags and Anthems 701

chapter 17 Binary I/o 713 17.1 Introduction 714 17.2 How Is Text I/O Handled in Java? 714 17.3 Text I/O vs. Binary I/O 715 17.4 Binary I/O Classes 716 17.5 Case Study: Copying Files 726 17.6 Object I/O 728 17.7 Random-Access Files 733

chapter 18 recursion 741 18.1 Introduction 742 18.2 Case Study: Computing Factorials 742 18.3 Case Study: Computing Fibonacci

Numbers 745 18.4 Problem Solving Using Recursion 748 18.5 Recursive Helper Methods 750 18.6 Case Study: Finding the Directory Size 753 18.7 Case Study: Tower of Hanoi 755 18.8 Case Study: Fractals 758 18.9 Recursion vs. Iteration 762 18.10 Tail Recursion 762

chapter 19 generics 773 19.1 Introduction 774 19.2 Motivations and Benefits 774 19.3 Defining Generic Classes and Interfaces 776 19.4 Generic Methods 778 19.5 Case Study: Sorting an Array of Objects 780

16 Contents

A01_LIAN1878_11_GE_FM.indd 16 1/2/18 11:57 PM

19.6 Raw Types and Backward Compatibility 782 19.7 Wildcard Generic Types 783 19.8 Erasure and Restrictions on Generics 786 19.9 Case Study: Generic Matrix Class 788

chapter 20 Lists, Stacks, Queues, and Priority Queues 797

20.1 Introduction 798 20.2 Collections 798 20.3 Iterators 802 20.4 Using the forEach Method 803 20.5 Lists 804 20.6 The Comparator Interface 809 20.7 Static Methods for Lists and Collections 813 20.8 Case Study: Bouncing Balls 816 20.9 Vector and Stack Classes 820 20.10 Queues and Priority Queues 821 20.11 Case Study: Evaluating Expressions 825

chapter 21 Sets and maps 837 21.1 Introduction 838 21.2 Sets 838 21.3 Comparing the Performance of Sets and Lists 846 21.4 Case Study: Counting Keywords 849 21.5 Maps 850 21.6 Case Study: Occurrences of Words 855 21.7 Singleton and Unmodifiable Collections and Maps 857

chapter 22 developing efficient algorithms 861

22.1 Introduction 862 22.2 Measuring Algorithm Efficiency Using Big O Notation 862 22.3 Examples: Determining Big O 864 22.4 Analyzing Algorithm Time Complexity 868 22.5 Finding Fibonacci Numbers Using Dynamic

Programming 871 22.6 Finding Greatest Common Divisors Using Euclid’s

Algorithm 873 22.7 Efficient Algorithms for Finding Prime Numbers 877 22.8 Finding the Closest Pair of Points Using

Divide-and-Conquer 883 22.9 Solving the Eight Queens Problem Using Backtracking 886 22.10 Computational Geometry: Finding a Convex Hull 889

chapter 23 Sorting 903 23.1 Introduction 904 23.2 Insertion Sort 904 23.3 Bubble Sort 906 23.4 Merge Sort 909 23.5 Quick Sort 912 23.6 Heap Sort 916 23.7 Bucket and Radix Sorts 923 23.8 External Sort 925

Contents 17

A01_LIAN1878_11_GE_FM.indd 17 1/2/18 11:57 PM

chapter 24 Implementing Lists, Stacks, Queues, and Priority Queues 939

24.1 Introduction 940 24.2 Common Operations for Lists 940 24.3 Array Lists 944 24.4 Linked Lists 951 24.5 Stacks and Queues 965 24.6 Priority Queues 969

chapter 25 Binary Search trees 975 25.1 Introduction 976 25.2 Binary Search Trees 976 25.3 Deleting Elements from a BST 989 25.4 Tree Visualization and MVC 995 25.5 Iterators 998 25.6 Case Study: Data Compression 1000

chapter 26 avL trees 1011 26.1 Introduction 1012 26.2 Rebalancing Trees 1012 26.3 Designing Classes for AVL Trees 1015 26.4 Overriding the insert Method 1016 26.5 Implementing Rotations 1017 26.6 Implementing the delete Method 1018 26.7 The AVLTree Class 1018 26.8 Testing the AVLTree Class 1024 26.9 AVL Tree Time Complexity Analysis 1027

chapter 27 hashing 1031 27.1 Introduction 1032 27.2 What Is Hashing? 1032 27.3 Hash Functions and Hash Codes 1033 27.4 Handling Collisions Using Open Addressing 1035 27.5 Handling Collisions Using Separate Chaining 1039 27.6 Load Factor and Rehashing 1039 27.7 Implementing a Map Using Hashing 1041 27.8 Implementing Set Using Hashing 1050

chapter 28 graphs and applications 1061 28.1 Introduction 1062 28.2 Basic Graph Terminologies 1063 28.3 Representing Graphs 1064 28.4 Modeling Graphs 1070 28.5 Graph Visualization 1080 28.6 Graph Traversals 1083 28.7 Depth-First Search (DFS) 1084 28.8 Case Study: The Connected Circles Problem 1088 28.9 Breadth-First Search (BFS) 1090 28.10 Case Study: The Nine Tails Problem 1093

18 Contents

A01_LIAN1878_11_GE_FM.indd 18 1/2/18 11:57 PM

chapter 29 Weighted graphs and applications 1107

29.1 Introduction 1108 29.2 Representing Weighted Graphs 1109 29.3 The WeightedGraph Class 1111 29.4 Minimum Spanning Trees 1119 29.5 Finding Shortest Paths 1125 29.6 Case Study: The Weighted Nine Tails Problem 1134

chapter 30 aggregate operations for collection Streams 1145

30.1 Introduction 1146 30.2 Stream Pipelines 1146 30.3 IntStream, LongStream, and DoubleStream 1152 30.4 Parallel Streams 1155 30.5 Stream Reduction Using the reduce Method 1157 30.6 Stream Reduction Using the collect Method 1160 30.7 Grouping Elements Using the groupingby Collector 1163 30.8 Case Studies 1166

chapter 31–44 are available from the companion Website at www.pearsonglobaleditions.com/Liang

chapter 31 advanced JavafX and fXmL chapter 32 multithreading and Parallel

Programming

chapter 33 networking chapter 34 Java database Programming chapter 35 advanced Java database Programming chapter 36 Internationalization chapter 37 Servlets chapter 38 JavaServer Pages chapter 39 JavaServer faces chapter 40 remote method Invocation chapter 41 Web Services chapter 42 2-4 trees and B-trees chapter 43 red-Black trees chapter 44 testing using Junit

Contents 19

A01_LIAN1878_11_GE_FM.indd 19 1/2/18 11:57 PM

Appendixes 1177

appendix a Java Keywords 1179 appendix B the aScII character Set 1180 appendix c operator Precedence chart 1182 appendix d Java modifiers 1184 appendix e Special floating-Point values 1186 appendix f number Systems 1187 appendix g Bitwise operations 1191 appendix h regular expressions 1192 appendix I enumerated types 1197

Quick Reference 1203 Index 1205

20 Contents

A01_LIAN1878_11_GE_FM.indd 20 1/2/18 11:57 PM

Chapter 1 Introduction to Computers, Programs, and Java™ 23 Your first Java program 34 Compile and run a Java program 39 NetBeans brief tutorial 45 Eclipse brief tutorial 47

Chapter 2 Elementary Programming 55 Obtain input 59 Use operators / and % 74 Software development process 81 Compute loan payments 82 Compute BMI 94

Chapter 3 Selections 97 Program addition quiz 99 Program subtraction quiz 109 Use multi-way if-else statements 112 Sort three integers 132 Check point location 134

Chapter 4 Mathematical Functions, Characters, and Strings 141 Introduce Math functions 142 Introduce strings and objects 152 Convert hex to decimal 165 Compute great circle distance 173 Convert hex to binary 176

Chapter 5 Loops 181 Use while loop 182 Guess a number 185 Multiple subtraction quiz 188 Use do-while loop 192 Minimize numeric errors 202 Display loan schedule 219 Sum a series 220

Chapter 6 Methods 227 Define/invoke max method 230 Use void method 233 Modularize code 239 Stepwise refinement 249 Reverse an integer 258 Estimate p 261

Chapter 7 Single-Dimensional Arrays 269 Random shuffling 274 Deck of cards 278 Selection sort 293 Command-line arguments 297

Coupon collector’s problem 304 Consecutive four 306

Chapter 8 Multidimensional Arrays 311 Find the row with the largest sum 316 Grade multiple-choice test 318 Sudoku 322 Multiply two matrices 331 Even number of 1s 338

Chapter 9 Objects and Classes 345 Define classes and objects 346 Use classes 358 Static vs. instance 361 Data field encapsulation 368 The this keyword 380 The Fan class 386

Chapter 10 Object-Oriented Thinking 389 The Loan class 391 The BMI class 394 The StackOfIntegers class 402 Process large numbers 408 The String class 410 The MyPoint class 424

Chapter 11 Inheritance and Polymorphism 433 Geometric class hierarchy 434 Polymorphism and dynamic binding demo 448 The ArrayList class 456 The MyStack class 463 New Account class 470

Chapter 12 Exception Handling and Text I/O 475 Exception-handling advantages 476 Create custom exception classes 496 Write and read data 502 HexFormatException 515

Chapter 13 Abstract Classes and Interfaces 521 Abstract GeometricObject class 522 Calendar and Gregorian Calendar classes 529 The concept of interface 532 Redesign the Rectangle class 558

Chapter 14 JavaFX Basics 563 Getting started with JavaFX 564 Understand property binding 570 Use Image and ImageView 578 Use layout panes 580 Use shapes 589

VideoNotes Locations of VideoNotes www.pearsonglobaleditions.com/Liang

VideoNote

A01_LIAN1878_11_GE_FM.indd 21 1/2/18 11:57 PM

22 VideoNotes

Display a tic-tac-toe board 608 Display a bar chart 610

Chapter 15 Event-Driven Programming and Animations 615 Handler and its registration 622 Anonymous handler 625 Move message using the mouse 634 Animate a rising flag 640 Flashing text 646 Simple calculator 656 Check mouse-point location 658 Display a running fan 661

Chapter 16 JavaFX UI Controls and Multimedia 665 Use ListView 684 Use Slider 690

Tic-Tac-Toe 693 Use Media, MediaPlayer, and MediaView 698 Use radio buttons and text fields 705 Set fonts 707

Chapter 17 Binary I/O 713 Copy file 726 Object I/O 728 Split a large file 738

Chapter 18 Recursion 741 Binary search 752 Directory size 753 Fractal (Sierpinski triangle) 758 Search a string in a directory 769 Recursive tree 772

Chapter 7 Single-Dimensional Arrays 269 linear search animation on Companion Website 290 binary search animation on Companion Website 290 selection sort animation on Companion Website 293

Chapter 8 Multidimensional Arrays 311 closest-pair animation on the Companion Website 320

Chapter 22 Developing Efficient Algorithms 861 binary search animation on the Companion Website 868 selection sort animation on the Companion Website 868 closest-pair animation on Companion Website 883 Eight Queens animation on the Companion Website 886 convex hull animation on the Companion Website 889

Chapter 23 Sorting 903 insertion-sort animation on Companion Website 904 bubble sort animation on the Companion Website 907 merge animation on Companion Website 911 partition animation on Companion Website 915 heap animation on Companion Website 917

radix sort on Companion Website 924

Chapter 24 Implementing Lists, Stacks, Queues, and Priority Queues 939 list animation on Companion Website 940 stack and queue animation on Companion Website 965

Chapter 25 Binary Search Trees 975 BST animation on Companion Website 976

Chapter 26 AVL Trees 1011 AVL tree animation on Companion Website 1012

Chapter 27 Hashing 1031 linear probing animation on Companion Website 1036 quadratic probing animation on Companion Website 1037 separate chaining animation on Companion Website 1040

Chapter 28 Graphs and Applications 1061 graph learning tool on Companion Website 1064 U.S. Map Search 1086

Chapter 29 Weighted Graphs and Applications 1107 weighted graph learning tool animation on Companion Website 1108

Animations

A01_LIAN1878_11_GE_FM.indd 22 1/2/18 11:57 PM

Chapter

1 Introduction to Computers, Programs, and Java™

Objectives ■■ To understand computer basics, programs, and operating systems

(§§1.2–1.4).

■■ To describe the relationship between Java and the World Wide Web (§1.5).

■■ To understand the meaning of Java language specification, API, JDK™, JRE™, and IDE (§1.6).

■■ To write a simple Java program (§1.7).

■■ To display output on the console (§1.7).

■■ To explain the basic syntax of a Java program (§1.7).

■■ To create, compile, and run Java programs (§1.8).

■■ To use sound Java programming style and document programs properly (§1.9).

■■ To explain the differences between syntax errors, runtime errors, and logic errors (§1.10).

■■ To develop Java programs using NetBeans™ (§1.11).

■■ To develop Java programs using Eclipse™ (§1.12).

M01_LIAN1878_11_GE_C01.indd 23 1/2/18 10:57 PM

24 Chapter 1 Introduction to Computers, Programs, and Java™

1.1 Introduction The central theme of this book is to learn how to solve problems by writing a program.

This book is about programming. So, what is programming? The term programming means to create (or develop) software, which is also called a program. In basic terms, software contains instructions that tell a computer—or a computerized device—what to do.

Software is all around you, even in devices you might not think would need it. Of course, you expect to find and use software on a personal computer, but software also plays a role in running airplanes, cars, cell phones, and even toasters. On a personal computer, you use word processors to write documents, web browsers to explore the Internet, and e-mail programs to send and receive messages. These programs are all examples of software. Software developers create software with the help of powerful tools called programming languages.

This book teaches you how to create programs by using the Java programming language. There are many programming languages, some of which are decades old. Each language was invented for a specific purpose—to build on the strengths of a previous language, for example, or to give the programmer a new and unique set of tools. Knowing there are so many program- ming languages available, it would be natural for you to wonder which one is best. However, in truth, there is no “best” language. Each one has its own strengths and weaknesses. Experienced programmers know one language might work well in some situations, whereas a different language may be more appropriate in other situations. For this reason, seasoned programmers try to master as many different programming languages as they can, giving them access to a vast arsenal of software-development tools.

If you learn to program using one language, you should find it easy to pick up other lan- guages. The key is to learn how to solve problems using a programming approach. That is the main theme of this book.

You are about to begin an exciting journey: learning how to program. At the outset, it is helpful to review computer basics, programs, and operating systems (OSs). If you are already familiar with such terms as central processing unit (CPU), memory, disks, operating systems, and programming languages, you may skip Sections 1.2–1.4.

1.2 What Is a Computer? A computer is an electronic device that stores and processes data.

A computer includes both hardware and software. In general, hardware comprises the visible, physical elements of the computer, and software provides the invisible instructions that control the hardware and make it perform specific tasks. Knowing computer hardware isn’t essential to learning a programming language, but it can help you better understand the effects that a program’s instructions have on the computer and its components. This section introduces computer hardware components and their functions.

A computer consists of the following major hardware components (see Figure 1.1):

■■ A central processing unit (CPU)

■■ Memory (main memory)

■■ Storage devices (such as disks and CDs)

■■ Input devices (such as the mouse and the keyboard)

■■ Output devices (such as monitors and printers)

■■ Communication devices (such as modems and network interface cards (NIC))

A computer’s components are interconnected by a subsystem called a bus. You can think of a bus as a sort of system of roads running among the computer’s components; data and power travel along the bus from one part of the computer to another. In personal computers,

Point Key

what is programming? programming program

Point Key

hardware software

bus

M01_LIAN1878_11_GE_C01.indd 24 1/2/18 10:57 PM

1.2 What Is a Computer? 25

the bus is built into the computer’s motherboard, which is a circuit case that connects all of the parts of a computer together.

1.2.1 Central Processing Unit The central processing unit (CPU) is the computer’s brain. It retrieves instructions from the memory and executes them. The CPU usually has two components: a control unit and an arithmetic/logic unit. The control unit controls and coordinates the actions of the other com- ponents. The arithmetic/logic unit performs numeric operations (addition, subtraction, multi- plication, and division) and logical operations (comparisons).

Today’s CPUs are built on small silicon semiconductor chips that contain millions of tiny electric switches, called transistors, for processing information.

Every computer has an internal clock that emits electronic pulses at a constant rate. These pulses are used to control and synchronize the pace of operations. A higher clock speed enables more instructions to be executed in a given period of time. The unit of measurement of clock speed is the hertz (Hz), with 1 Hz equaling 1 pulse per second. In the 1990s, computers meas- ured clock speed in megahertz (MHz), but CPU speed has been improving continuously; the clock speed of a computer is now usually stated in gigahertz (GHz). Intel’s newest processors run at about 3 GHz.

CPUs were originally developed with only one core. The core is the part of the processor that performs the reading and executing of instructions. In order to increase the CPU processing power, chip manufacturers are now producing CPUs that contain multiple cores. A multicore CPU is a single component with two or more independent cores. Today’s consumer comput- ers typically have two, three, and even four separate cores. Soon, CPUs with dozens or even hundreds of cores will be affordable.

1.2.2 Bits and Bytes Before we discuss memory, let’s look at how information (data and programs) are stored in a computer.

A computer is really nothing more than a series of switches. Each switch exists in two states: on or off. Storing information in a computer is simply a matter of setting a sequence of switches on or off. If the switch is on, its value is 1. If the switch is off, its value is 0. These 0s and 1s are interpreted as digits in the binary number system and are called bits (binary digits).

The minimum storage unit in a computer is a byte. A byte is composed of eight bits. A small number such as 3 can be stored as a single byte. To store a number that cannot fit into a single byte, the computer uses several bytes.

Data of various kinds, such as numbers and characters, are encoded as a series of bytes. As a programmer, you don’t need to worry about the encoding and decoding of data, which the computer system performs automatically, based on the encoding scheme. An encoding scheme is a set of rules that govern how a computer translates characters and numbers into data with which the computer can actually work. Most schemes translate each character into a

motherboard

CPU

speed

hertz megahertz

gigahertz

core

bits

byte

encoding scheme

Figure 1.1 A computer consists of a CPU, memory, storage devices, input devices, output devices, and communication devices.

Memory

e.g., Disk, CD, and Tape

e.g., Modem, and NIC

e.g., Keyboard, Mouse

e.g., Monitor, Printer

CPU

Bus

Storage Devices

Communication Devices

Input Devices

Output Devices

M01_LIAN1878_11_GE_C01.indd 25 1/2/18 10:57 PM

26 Chapter 1 Introduction to Computers, Programs, and Java™

predetermined string of bits. In the popular ASCII encoding scheme, for example, the character C is represented as 01000011 in 1 byte.

A computer’s storage capacity is measured in bytes and multiples of the byte, as follows:

■■ A kilobyte (KB) is about 1,000 bytes.

■■ A megabyte (MB) is about 1 million bytes.

■■ A gigabyte (GB) is about 1 billion bytes.

■■ A terabyte (TB) is about 1 trillion bytes.

A typical one-page word document might take 20 KB. Therefore, 1 MB can store 50 pages of documents, and 1 GB can store 50,000 pages of documents. A typical two-hour high- resolution movie might take 8 GB, so it would require 160 GB to store 20 movies.

1.2.3 Memory A computer’s memory consists of an ordered sequence of bytes for storing programs as well as data with which the program is working. You can think of memory as the computer’s work area for executing a program. A program and its data must be moved into the computer’s memory before they can be executed by the CPU.

Every byte in the memory has a unique address, as shown in Figure 1.2. The address is used to locate the byte for storing and retrieving the data. Since the bytes in the memory can be accessed in any order, the memory is also referred to as random-access memory (RAM).

kilobyte (KB)

megabyte (MB)

gigabyte (GB)

terabyte (TB)

memory

unique address

RAM

Figure 1.2 Memory stores data and program instructions in uniquely addressed memory locations.

01000011 01110010 01100101 01110111 00000011

Encoding for character ‘C’ Encoding for character ‘r’ Encoding for character ‘e’ Encoding for character ‘w’ Decimal number 3

2000 2001 2002 2003 2004

Memory address Memory content

Today’s personal computers usually have at least 4 GB of RAM, but they more commonly have 6 to 8 GB installed. Generally speaking, the more RAM a computer has, the faster it can operate, but there are limits to this simple rule of thumb.

A memory byte is never empty, but its initial content may be meaningless to your program. The current content of a memory byte is lost whenever new information is placed in it.

Like the CPU, memory is built on silicon semiconductor chips that have millions of transis- tors embedded on their surface. Compared to CPU chips, memory chips are less complicated, slower, and less expensive.

1.2.4 Storage Devices A computer’s memory (RAM) is a volatile form of data storage: Any information that has been saved in memory is lost when the system’s power is turned off. Programs and data are permanently stored on storage devices and are moved, when the computer actu- storage devices

M01_LIAN1878_11_GE_C01.indd 26 1/2/18 10:57 PM

1.2 What Is a Computer? 27

ally uses them, to memory, which operates at much faster speeds than permanent storage devices can.

There are three main types of storage devices:

■■ Magnetic disk drives

■■ Optical disc drives (CD and DVD)

■■ Universal serial bus (USB) flash drives

Drives are devices for operating a medium, such as disks and CDs. A storage medium physically stores data and program instructions. The drive reads data from the medium and writes data onto the medium.

Disks

A computer usually has at least one hard disk drive. Hard disks are used for permanently stor- ing data and programs. Newer computers have hard disks that can store from 500 GB to 1 TB of data. Hard disk drives are usually encased inside the computer, but removable hard disks are also available.

CDs and DVDs

CD stands for compact disc. There are three types of CDs: CD-ROM, CD-R, and CD-RW. A CD- ROM is a prepressed disc. It was popular for distributing software, music, and video. Software, music, and video are now increasingly distributed on the Internet without using CDs. A CD-R (CD-Recordable) is a write-once medium. It can be used to record data once and read any number of times. A CD-RW (CD-ReWritable) can be used like a hard disk; that is, you can write data onto the disc, then overwrite that data with new data. A single CD can hold up to 700 MB.

DVD stands for digital versatile disc or digital video disc. DVDs and CDs look alike, and you can use either to store data. A DVD can hold more information than a CD; a standard DVD’s storage capacity is 4.7 GB. There are two types of DVDs: DVD-R (Recordable) and DVD-RW (ReWritable).

USB Flash Drives

Universal serial bus (USB) connectors allow the user to attach many kinds of peripheral devices to the computer. You can use an USB to connect a printer, digital camera, mouse, external hard disk drive, and other devices to the computer.

An USB flash drive is a device for storing and transporting data. A flash drive is small—about the size of a pack of gum. It acts like a portable hard drive that can be plugged into your computer’s USB port. USB flash drives are currently available with up to 256 GB storage capacity.

1.2.5 Input and Output Devices Input and output devices let the user communicate with the computer. The most common input devices are the keyboard and mouse. The most common output devices are monitors and printers.

The Keyboard

A keyboard is a device for entering input. Compact keyboards are available without a numeric keypad.

Function keys are located across the top of the keyboard and are prefaced with the letter F. Their functions depend on the software currently being used.

A modifier key is a special key (such as the Shift, Alt, and Ctrl keys) that modifies the normal action of another key when the two are pressed simultaneously.

drive

hard disk

CD-ROM

CD-R

CD-RW

DVD

function key

modifier key

M01_LIAN1878_11_GE_C01.indd 27 1/2/18 10:57 PM

28 Chapter 1 Introduction to Computers, Programs, and Java™

The numeric keypad, located on the right side of most keyboards, is a separate set of keys styled like a calculator to use for quickly entering numbers.

Arrow keys, located between the main keypad and the numeric keypad, are used to move the mouse pointer up, down, left, and right on the screen in many kinds of programs.

The Insert, Delete, Page Up, and Page Down keys are used in word processing and other programs for inserting text and objects, deleting text and objects, and moving up or down through a document one screen at a time.

The Mouse

A mouse is a pointing device. It is used to move a graphical pointer (usually in the shape of an arrow) called a cursor around the screen, or to click on-screen objects (such as a button) to trigger them to perform an action.

The Monitor

The monitor displays information (text and graphics). The screen resolution and dot pitch determine the quality of the display.

The screen resolution specifies the number of pixels in horizontal and vertical dimensions of the display device. Pixels (short for “picture elements”) are tiny dots that form an image on the screen. A common resolution for a 17-inch screen, for example, is 1,024 pixels wide and 768 pixels high. The resolution can be set manually. The higher the resolution, the sharper and clearer the image is.

The dot pitch is the amount of space between pixels, measured in millimeters. The smaller the dot pitch, the sharper is the display.

1.2.6 Communication Devices Computers can be networked through communication devices, such as a dial-up modem (modulator/demodulator), a digital subscriber line (DSL) or cable modem, a wired network interface card, or a wireless adapter.

■■ A dial-up modem uses a phone line to dial a phone number to connect to the Internet and can transfer data at a speed up to 56,000 bps (bits per second).

■■ A digital subscriber line (DSL) connection also uses a standard phone line, but it can transfer data 20 times faster than a standard dial-up modem.

■■ A cable modem uses the cable line maintained by the cable company and is generally faster than DSL.

■■ A network interface card (NIC) is a device that connects a computer to a local area network (LAN). LANs are commonly used to connect computers within a limited area such as a school, a home, and an office. A high-speed NIC called 1000BaseT can transfer data at 1,000 million bits per second (mbps).

■■ Wireless networking is now extremely popular in homes, businesses, and schools. Every laptop computer sold today is equipped with a wireless adapter that enables the computer to connect to the LAN and the Internet.

Note Answers to the CheckPoint questions are available at www.pearsonglobaleditions .com/Liang. Choose this book and click Companion Website to select CheckPoint.

1.2.1 What are hardware and software? 1.2.2 List the five major hardware components of a computer.

numeric keypad

arrow keys

Insert key Delete key

Page Up key Page Down key

screen resolution pixels

dot pitch

dial-up modem

digital subscriber line (DSL)

cable modem

network interface card (NIC)

local area network (LAN) million bits per second

(mbps)

Point Check

M01_LIAN1878_11_GE_C01.indd 28 1/2/18 10:57 PM

1.3 Programming Languages 29

1.2.3 What does the acronym CPU stand for? What unit is used to measure CPU speed? 1.2.4 What is a bit? What is a byte? 1.2.5 What is memory for? What does RAM stand for? Why is memory called RAM? 1.2.6 What unit is used to measure memory size? What unit is used to measure disk size? 1.2.7 What is the primary difference between memory and a storage device?

1.3 Programming Languages Computer programs, known as software, are instructions that tell a computer what to do.

Computers do not understand human languages, so programs must be written in a language a computer can use. There are hundreds of programming languages, and they were developed to make the programming process easier for people. However, all programs must be converted into the instructions the computer can execute.

1.3.1 Machine Language A computer’s native language, which differs among different types of computers, is its machine language—a set of built-in primitive instructions. These instructions are in the form of binary code, so if you want to give a computer an instruction in its native language, you have to enter the instruction as binary code. For example, to add two numbers, you might have to write an instruction in binary code as follows:

1101101010011010

1.3.2 Assembly Language Programming in machine language is a tedious process. Moreover, programs written in machine language are very difficult to read and modify. For this reason, assembly language was created in the early days of computing as an alternative to machine languages. Assembly language uses a short descriptive word, known as a mnemonic, to represent each of the machine-language instructions. For example, the mnemonic add typically means to add num- bers, and sub means to subtract numbers. To add the numbers 2 and 3 and get the result, you might write an instruction in assembly code as follows:

add 2, 3, result

Assembly languages were developed to make programming easier. However, because the computer cannot execute assembly language, another program—called an assembler—is used to translate assembly-language programs into machine code, as shown in Figure 1.3.

Point Key

machine language

assembly language

assembler

Figure 1.3 An assembler translates assembly-language instructions into machine code.

Assembly Source File

... add 2, 3, result

...

Machine-Code File

... 1101101010011010

...

Assembler

Writing code in assembly language is easier than in machine language. However, it is still tedious to write code in assembly language. An instruction in assembly language essentially corresponds to an instruction in machine code. Writing in assembly language requires that you

M01_LIAN1878_11_GE_C01.indd 29 1/2/18 10:57 PM

30 Chapter 1 Introduction to Computers, Programs, and Java™

know how the CPU works. Assembly language is referred to as a low-level language, because assembly language is close in nature to machine language and is machine dependent.

1.3.3 High-Level Language In the 1950s, a new generation of programming languages known as high-level languages emerged. They are platform independent, which means that you can write a program in a high- level language and run it in different types of machines. High-level languages are similar to English and easy to learn and use. The instructions in a high-level programming language are called statements. Here, for example, is a high-level language statement that computes the area of a circle with a radius of 5:

area = 5 * 5 * 3.14159;

There are many high-level programming languages, and each was designed for a specific purpose. Table 1.1 lists some popular ones.

low-level language

high-level language

statement

Language Description

Ada Named for Ada Lovelace, who worked on mechanical general-purpose computers. Developed for the Department of Defense and used mainly in defense projects.

BASIC Beginner’s All-purpose Symbolic Instruction Code. Designed to be learned and used easily by beginners.

C Developed at Bell Laboratories. Combines the power of an assembly language with the ease of use and portability of a high-level language.

C++ An object-oriented language, based on C

C# Pronounced “C Sharp.” An object-oriented programming language developed by Microsoft.

COBOL COmmon Business Oriented Language. Used for business applications.

FORTRAN FORmula TRANslation. Popular for scientific and mathematical applications.

Java Developed by Sun Microsystems, now part of Oracle. An object-oriented programming language, widely used for developing platform-independent Internet applications.

JavaScript A Web programming language developed by Netscape

Pascal Named for Blaise Pascal, who pioneered calculating machines in the seventeenth century. A simple, structured, general-purpose language primarily for teaching programming.

Python A simple general-purpose scripting language good for writing short programs.

Visual Basic Visual Basic was developed by Microsoft. Enables the programmers to rapidly develop Windows-based applications.

Table 1.1 Popular High-Level Programming Languages

A program written in a high-level language is called a source program or source code. Because a computer cannot execute a source program, a source program must be translated into machine code for execution. The translation can be done using another programming tool called an interpreter or a compiler.

■■ An interpreter reads one statement from the source code, translates it to the machine code or virtual machine code, then executes it right away, as shown in Figure 1.4a. Note a statement from the source code may be translated into several machine instructions.

source program source code

interpreter

compiler

M01_LIAN1878_11_GE_C01.indd 30 1/2/18 10:57 PM

1.4 Operating Systems 31

Figure 1.4 (a) An interpreter translates and executes a program one statement at a time. (b) A compiler translates the entire source program into a machine-language file for execution.

Machine-Code File

... 0101100011011100 1111100011000100

...

High-Level Source File

... area = 5 * 5 * 3.1415;

...

(b)

Compiler Executor

High-Level Source File

... area = 5 * 5 * 3.1415;

...

(a)

Interpreter Output

Output

■■ A compiler translates the entire source code into a machine-code file, and the machine-code file is then executed, as shown in Figure 1.4b.

1.3.1 What language does the CPU understand? 1.3.2 What is an assembly language? What is an assembler? 1.3.3 What is a high-level programming language? What is a source program? 1.3.4 What is an interpreter? What is a compiler? 1.3.5 What is the difference between an interpreted language and a compiled language?

Point Check

1.4 Operating Systems The operating system (OS) is the most important program that runs on a computer. The OS manages and controls a computer’s activities.

The popular operating systems for general-purpose computers are Microsoft Windows, Mac OS, and Linux. Application programs, such as a web browser or a word processor, cannot run unless an operating system is installed and running on the computer. Figure 1.5 shows the interrelationship of hardware, operating system, application software, and the user.

Point Key

operating system (OS)

Figure 1.5 Users and applications access the computer’s hardware via the operating system.

User

Application Programs

Operating System

Hardware

M01_LIAN1878_11_GE_C01.indd 31 1/2/18 10:57 PM

32 Chapter 1 Introduction to Computers, Programs, and Java™

The major tasks of an operating system are as follows:

■■ Controlling and monitoring system activities

■■ Allocating and assigning system resources

■■ Scheduling operations

1.4.1 Controlling and Monitoring System Activities Operating systems perform basic tasks, such as recognizing input from the keyboard, sending output to the monitor, keeping track of files and folders on storage devices, and controlling peripheral devices such as disk drives and printers. An operating system must also ensure different programs and users working at the same time do not interfere with each other. In addition, the OS is responsible for security, ensuring unauthorized users and programs are not allowed to access the system.

1.4.2 Allocating and Assigning System Resources The operating system is responsible for determining what computer resources a program needs (such as CPU time, memory space, disks, and input and output devices) and for allocating and assigning them to run the program.

1.4.3 Scheduling Operations The OS is responsible for scheduling programs’ activities to make efficient use of system resources. Many of today’s operating systems support techniques such as multiprogramming, multithreading, and multiprocessing to increase system performance.

Multiprogramming allows multiple programs such as Microsoft Word, E-mail, and web browser to run simultaneously by sharing the same CPU. The CPU is much faster than the computer’s other components. As a result, it is idle most of the time—for example, while wait- ing for data to be transferred from a disk or waiting for other system resources to respond. A multiprogramming OS takes advantage of this situation by allowing multiple programs to use the CPU when it would otherwise be idle. For example, multiprogramming enables you to use a word processor to edit a file at the same time as your web browser is downloading a file.

Multithreading allows a single program to execute multiple tasks at the same time. For instance, a word-processing program allows users to simultaneously edit text and save it to a disk. In this example, editing and saving are two tasks within the same program. These two tasks may run concurrently.

Multiprocessing is similar to multithreading. The difference is that multithreading is for running multithreads concurrently within one program, but multiprocessing is for running multiple programs concurrently using multiple processors.

1.4.1 What is an operating system? List some popular operating systems. 1.4.2 What are the major responsibilities of an operating system? 1.4.3 What are multiprogramming, multithreading, and multiprocessing?

1.5 Java, the World Wide Web, and Beyond Java is a powerful and versatile programming language for developing software run- ning on mobile devices, desktop computers, and servers.

This book introduces Java programming. Java was developed by a team led by James Gosling at Sun Microsystems. Sun Microsystems was purchased by Oracle in 2010. Originally called Oak, Java was designed in 1991 for use in embedded chips in consumer electronic appliances.

multiprogramming multithreading multiprocessing

Point Check

Point Key

M01_LIAN1878_11_GE_C01.indd 32 1/2/18 10:57 PM

1.6 The Java Language Specification, API, JDK, JRE, and IDE 33

In 1995, renamed Java, it was redesigned for developing web applications. For the history of Java, see www.java.com/en/javahistory/index.jsp.

Java has become enormously popular. Its rapid rise and wide acceptance can be traced to its design characteristics, particularly its promise that you can write a program once and run it anywhere. As stated by its designer, Java is simple, object oriented, distributed, interpreted, robust, secure, architecture neutral, portable, high performance, multithreaded, and dynamic. For the anatomy of Java characteristics, see liveexample.pearsoncmg.com/etc/ JavaCharacteristics.pdf.

Java is a full-featured, general-purpose programming language that can be used to develop robust mission-critical applications. Today, it is employed not only for web programming but also for developing stand-alone applications across platforms on servers, desktop computers, and mobile devices. It was used to develop the code to communicate with and control the robotic rover on Mars. Many companies that once considered Java to be more hype than sub- stance are now using it to create distributed applications accessed by customers and partners across the Internet. For every new project being developed today, companies are asking how they can use Java to make their work easier.

The World Wide Web is an electronic information repository that can be accessed on the Internet from anywhere in the world. The Internet, the Web’s infrastructure, has been around for more than 40 years. The colorful World Wide Web and sophisticated web browsers are the major reason for the Internet’s popularity.

Java initially became attractive because Java programs can run from a web browser. Such programs are called applets. Today applets are no longer allowed to run from a Web browser in the latest version of Java due to security issues. Java, however, is now very popular for developing applications on web servers. These applications process data, perform computa- tions, and generate dynamic webpages. Many commercial Websites are developed using Java on the backend.

Java is a versatile programming language: You can use it to develop applications for desktop computers, servers, and small handheld devices. The software for Android cell phones is developed using Java.

1.5.1 Who invented Java? Which company owns Java now? 1.5.2 What is a Java applet? 1.5.3 What programming language does Android use?

1.6 The Java Language Specification, API, JDK, JRE, and IDE

Java syntax is defined in the Java language specification, and the Java library is defined in the Java application program interface (API). The JDK is the software for compiling and running Java programs. An IDE is an integrated development environ- ment for rapidly developing programs.

Computer languages have strict rules of usage. If you do not follow the rules when writing a program, the computer will not be able to understand it. The Java language specification and the Java API define the Java standards.

The Java language specification is a technical definition of the Java programming language’s syntax and semantics. You can find the complete Java language specification at docs.oracle.com/javase/specs/.

The application program interface (API), also known as library, contains predefined classes and interfaces for developing Java programs. The API is still expanding. You can view and download the latest version of the Java API at download.java.net/jdk8/docs/api/.

Point Check

Point Key

Java language specification

API library

M01_LIAN1878_11_GE_C01.indd 33 1/2/18 10:57 PM

34 Chapter 1 Introduction to Computers, Programs, and Java™

Java is a full-fledged and powerful language that can be used in many ways. It comes in three editions:

■■ Java Standard Edition (Java SE) to develop client-side applications. The applications can run on desktop.

■■ Java Enterprise Edition (Java EE) to develop server-side applications, such as Java servlets, JavaServer Pages (JSP), and JavaServer Faces (JSF).

■■ Java Micro Edition (Java ME) to develop applications for mobile devices, such as cell phones.

This book uses Java SE to introduce Java programming. Java SE is the foundation upon which all other Java technology is based. There are many versions of Java SE. The latest, Java SE 8, is used in this book. Oracle releases each version with a Java Development Toolkit (JDK). For Java SE 8, the Java Development Toolkit is called JDK 1.8 (also known as Java 8 or JDK 8).

The JDK consists of a set of separate programs, each invoked from a command line, for compiling, running, and testing Java programs. The program for running Java programs is known as JRE (Java Runtime Environment). Instead of using the JDK, you can use a Java development tool (e.g., NetBeans, Eclipse, and TextPad)—software that provides an integrated development environment (IDE) for developing Java programs quickly. Editing, compiling, building, debugging, and online help are integrated in one graphical user interface. You simply enter source code in one window or open an existing file in a window, and then click a button or menu item or press a function key to compile and run the program.

1.6.1 What is the Java language specification? 1.6.2 What does JDK stand for? What does JRE stand for? 1.6.3 What does IDE stand for? 1.6.4 Are tools like NetBeans and Eclipse different languages from Java, or are they dia-

lects or extensions of Java?

1.7 A Simple Java Program A Java program is executed from the main method in the class.

Let’s begin with a simple Java program that displays the message Welcome to Java! on the console. (The word console is an old computer term that refers to the text entry and display device of a computer. Console input means to receive input from the keyboard, and console output means to display output on the monitor.) The program is given in Listing 1.1.

lisTing 1.1 Welcome.java 1 public class Welcome { 2 public static void main(String[] args) { 3 // Display message Welcome to Java! on the console 4 System.out.println("Welcome to Java!"); 5 } 6 }

Java SE, EE, and ME

Java Development Toolkit (JDK)

JDK 1.8 = JDK 8

Java Runtime Environment (JRE)

Integrated development environment

Point Check

Point Key

what is a console? console input

console output

class main method

display message

VideoNote

Your first Java program

Welcome to Java!

Note the line numbers are for reference purposes only; they are not part of the program. So, don’t type line numbers in your program.

line numbers

M01_LIAN1878_11_GE_C01.indd 34 1/2/18 10:57 PM

1.7 A Simple Java Program 35

Line 1 defines a class. Every Java program must have at least one class. Each class has a name. By convention, class names start with an uppercase letter. In this example, the class name is Welcome.

Line 2 defines the main method. The program is executed from the main method. A class may contain several methods. The main method is the entry point where the program begins execution.

A method is a construct that contains statements. The main method in this program contains the System.out.println statement. This statement displays the string Welcome to Java! on the console (line 4). String is a programming term meaning a sequence of characters. A string must be enclosed in double quotation marks. Every statement in Java ends with a semi- colon (;), known as the statement terminator.

Reserved words, or keywords, have a specific meaning to the compiler and cannot be used for other purposes in the program. For example, when the compiler sees the word class, it understands that the word after class is the name for the class. Other reserved words in this program are public, static, and void.

Line 3 is a comment that documents what the program is and how it is constructed. Comments help programmers to communicate and understand the program. They are not programming statements, and thus are ignored by the compiler. In Java, comments are preceded by two slashes (//) on a line, called a line comment, or enclosed between /* and */ on one or several lines, called a block comment or paragraph comment. When the compiler sees //, it ignores all text after // on the same line. When it sees /*, it scans for the next */ and ignores any text between /* and */. Here are examples of comments:

// This application program displays Welcome to Java! /* This application program displays Welcome to Java! */ /* This application program

displays Welcome to Java! */

A pair of braces in a program forms a block that groups the program’s components. In Java, each block begins with an opening brace ({) and ends with a closing brace (}). Every class has a class block that groups the data and methods of the class. Similarly, every method has a method block that groups the statements in the method. Blocks can be nested, meaning that one block can be placed within another, as shown in the following code:

class name

main method

string

statement terminator reserved word keyword

comment

line comment block comment

block

match braces

public class Welcome { public static void main(String[] args) { System.out.println("Welcome to Java!"); } }

Method block Class block

Tip An opening brace must be matched by a closing brace. Whenever you type an opening brace, immediately type a closing brace to prevent the missing-brace error. Most Java IDEs automatically insert the closing brace for each opening brace.

Caution Java source programs are case sensitive. It would be wrong, for example, to replace main in the program with Main.

You have seen several special characters (e.g., { }, //, ;) in the program. They are used in almost every program. Table 1.2 summarizes their uses.

The most common errors you will make as you learn to program will be syntax errors. Like any programming language, Java has its own syntax, and you need to write code that conforms

case sensitive

special characters

common errors

M01_LIAN1878_11_GE_C01.indd 35 1/2/18 10:57 PM

36 Chapter 1 Introduction to Computers, Programs, and Java™

to the syntax rules. If your program violates a rule—for example, if the semicolon is missing, a brace is missing, a quotation mark is missing, or a word is misspelled—the Java compiler will report syntax errors. Try to compile the program with these errors and see what the com- piler reports.

Note You are probably wondering why the main method is defined this way and why System.out.println(...) is used to display a message on the console. For the time being, simply accept that this is how things are done. Your questions will be fully answered in subsequent chapters.

The program in Listing 1.1 displays one message. Once you understand the program, it is easy to extend it to display more messages. For example, you can rewrite the program to display three messages, as shown in Listing 1.2.

lisTing 1.2 WelcomeWithThreeMessages.java 1 public class WelcomeWithThreeMessages { 2 public static void main(String[] args) { 3 System.out.println("Programming is fun!"); 4 System.out.println("Fundamentals First"); 5 System.out.println("Problem Driven"); 6 } 7 }

syntax rules

class main method display message

Character Name Description

{} Opening and closing braces Denote a block to enclose statements.

() Opening and closing parentheses Used with methods.

[] Opening and closing brackets Denote an array.

// Double slashes Precede a comment line.

"" Opening and closing quotation marks Enclose a string (i.e., sequence of characters).

; Semicolon Mark the end of a statement.

Table 1.2 Special Characters

Programming is fun! Fundamentals First Problem Driven

Further, you can perform mathematical computations and display the result on the console.

Listing 1.3 gives an example of evaluating 10.5 + 2 * 3

45 - 3.5 .

lisTing 1.3 ComputeExpression.java 1 public class ComputeExpression { 2 public static void main(String[] args) { 3 System.out.print("(10.5 + 2 * 3) / (45 – 3.5) = "); 4 System.out.println((10.5 + 2 * 3) / (45 – 3.5)); 5 } 6 }

class main method

compute expression

(10.5 + 2 * 3) / (45 – 3.5) = 0.39759036144578314

M01_LIAN1878_11_GE_C01.indd 36 1/2/18 10:57 PM

1.8 Creating, Compiling, and Executing a Java Program 37

The print method in line 3

System.out.print("(10.5 + 2 * 3) / (45 – 3.5) = ");

is identical to the println method except that println moves to the beginning of the next line after displaying the string, but print does not advance to the next line when completed.

The multiplication operator in Java is *. As you can see, it is a straightforward process to translate an arithmetic expression to a Java expression. We will discuss Java expressions fur- ther in Chapter 2.

1.7.1 What is a keyword? List some Java keywords. 1.7.2 Is Java case sensitive? What is the case for Java keywords? 1.7.3 What is a comment? Is the comment ignored by the compiler? How do you denote a

comment line and a comment paragraph? 1.7.4 What is the statement to display a string on the console? 1.7.5 Show the output of the following code:

public class Test { public static void main(String[] args) { System.out.println("3.5 * 4 / 2 – 2.5 is "); System.out.println(3.5 * 4 / 2 – 2.5); } }

1.8 Creating, Compiling, and Executing a Java Program You save a Java program in a .java file and compile it into a .class file. The .class file is executed by the Java Virtual Machine (JVM).

You have to create your program and compile it before it can be executed. This process is repetitive, as shown in Figure 1.6. If your program has compile errors, you have to modify the program to fix them, then recompile it. If your program has runtime errors or does not produce the correct result, you have to modify the program, recompile it, and execute it again.

You can use any text editor or IDE to create and edit a Java source-code file. This section demonstrates how to create, compile, and run Java programs from a command window. Sec- tions 1.11 and 1.12 will introduce developing Java programs using NetBeans and Eclipse. From the command window, you can use a text editor such as Notepad to create the Java source-code file, as shown in Figure 1.7.

Note The source file must end with the extension .java and must have the same exact name as the public class name. For example, the file for the source code in Listing 1.1 should be named Welcome.java, since the public class name is Welcome.

A Java compiler translates a Java source file into a Java bytecode file. The following com- mand compiles Welcome.java:

javac Welcome.java

Note You must first install and configure the JDK before you can compile and run programs. See Supplement I.B, Installing and Configuring JDK 8, for how to install the JDK and set up the environment to compile and run Java programs. If you have trouble compiling and running programs, see Supplement I.C, Compiling and Running Java from the Command Window. This supplement also explains how to use basic DOS commands and how to use Windows Notepad to create and edit files. All the supplements are accessible from the Companion Website.

print vs. println

Point Check

Point Key

command window

file name Welcome.java,

compile

Supplement I.B

Supplement I.C

M01_LIAN1878_11_GE_C01.indd 37 1/2/18 10:57 PM

38 Chapter 1 Introduction to Computers, Programs, and Java™

Figure 1.6 The Java program-development process consists of repeatedly creating/modifying source code, compiling, and executing programs.

Create/Modify Source Code

Result

Compile Source Code e.g., javac Welcome.java

Saved on the disk

Stored on the disk If compile errors occur

If runtime errors or incorrect result

Source code (developed by the programmer)

Bytecode (generated by the compiler for JVM to read and interpret)

… Method Welcome() 0 aload_0 …

Method void main(java.lang.String[]) 0 getstatic #2 … 3 ldc #3 <String "Welcome to Java!"> 5 invokevirtual #4 … 8 return

public class Welcome { public static void main(String[] args) { System.out.println("Welcome to Java!"); } }

Run Bytecode e.g., java Welcome

Source Code

Bytecode

“Welcome to Java” is displayed on the console

Welcome to Java!

Figure 1.7 You can create a Java source file using Windows Notepad.

If there aren’t any syntax errors, the compiler generates a bytecode file with a .class extension. Thus, the preceding command generates a file named Welcome.class, as shown in Figure 1.8a. The Java language is a high-level language, but Java bytecode is a low-level language. The bytecode is similar to machine instructions but is architecture neutral and can run on any platform that has a Java Virtual Machine (JVM), as shown in Figure 1.8b. Rather than a physical machine, the virtual machine is a program that interprets Java bytecode. This is one of Java’s primary advantages: Java bytecode can run on a variety of hardware platforms and operating systems. Java source code is compiled into Java bytecode, and Java bytecode is interpreted by the JVM. Your Java code may use the code in the Java library. The JVM exe- cutes your code along with the code in the library.

To execute a Java program is to run the program’s bytecode. You can execute the bytecode on any platform with a JVM, which is an interpreter. It translates the individual instructions in the bytecode into the target machine language code one at a time, rather than the whole pro- gram as a single unit. Each step is executed immediately after it is translated.

.class bytecode file

bytecode

Java Virtual Machine (JVM)

interpret bytecode

M01_LIAN1878_11_GE_C01.indd 38 1/2/18 10:57 PM

1.8 Creating, Compiling, and Executing a Java Program 39

The following command runs the bytecode for Listing 1.1:

java Welcome

Figure 1.9 shows the javac command for compiling Welcome.java. The compiler generates the Welcome.class file, and this file is executed using the java command.

Note For simplicity and consistency, all source-code and class files used in this book are placed under c:\book unless specified otherwise.

run

javac command java command

c:\book

VideoNote

Compile and run a Java program

Figure 1.8 (a) Java source code is translated into bytecode. (b) Java bytecode can be executed on any computer with a Java Virtual Machine.

va Vir

tual Machine

Any Computer

Java Bytecode

Welcome.java (Java source-

code �le)

Welcome.class (Java bytecode executable �le)

Library Code

JVMJavaCompiler

compiled by generates

executed by

(a) (b)

Figure 1.9 The output of Listing 1.1 displays the message “Welcome to Java!”

Show �les

Run

Compile

Caution Do not use the extension .class in the command line when executing the program. Use java ClassName to run the program. If you use java ClassName.class in the command line, the system will attempt to fetch ClassName.class.class.

Tip If you execute a class file that does not exist, a NoClassDefFoundError will occur. If you execute a class file that does not have a main method or you mistype the main method (e.g., by typing Main instead of main), a NoSuchMethodError will occur.

java ClassName

NoClassDefFoundError

NoSuchMethodError

M01_LIAN1878_11_GE_C01.indd 39 1/2/18 10:57 PM

40 Chapter 1 Introduction to Computers, Programs, and Java™

Note When executing a Java program, the JVM first loads the bytecode of the class to memory using a program called the class loader. If your program uses other classes, the class loader dynamically loads them just before they are needed. After a class is loaded, the JVM uses a program called the bytecode verifier to check the validity of the bytecode and to ensure that the bytecode does not violate Java’s security restrictions. Java enforces strict security to make sure Java class files are not tampered with and do not harm your computer.

Pedagogical Note Your instructor may require you to use packages for organizing programs. For example, you may place all programs in this chapter in a package named chapter1. For instructions on how to use packages, see Supplement I.F, Using Packages to Organize the Classes in the Text.

1.8.1 What is the Java source filename extension, and what is the Java bytecode filename extension?

1.8.2 What are the input and output of a Java compiler? 1.8.3 What is the command to compile a Java program? 1.8.4 What is the command to run a Java program? 1.8.5 What is the JVM? 1.8.6 Can Java run on any machine? What is needed to run Java on a computer? 1.8.7 If a NoClassDefFoundError occurs when you run a program, what is the cause

of the error? 1.8.8 If a NoSuchMethodError occurs when you run a program, what is the cause of the

error?

1.9 Programming Style and Documentation Good programming style and proper documentation make a program easy to read and help programmers prevent errors.

Programming style deals with what programs look like. A program can compile and run properly even if written on only one line, but writing it all on one line would be bad pro- gramming style because it would be hard to read. Documentation is the body of explanatory remarks and comments pertaining to a program. Programming style and documentation are as important as coding. Good programming style and appropriate documentation reduce the chance of errors and make programs easy to read. This section gives several guidelines. For more detailed guidelines, see Supplement I.D, Java Coding Style Guidelines, on the Com- panion Website.

1.9.1 Appropriate Comments and Comment Styles Include a summary at the beginning of the program that explains what the program does, its key features, and any unique techniques it uses. In a long program, you should also include comments that introduce each major step and explain anything that is difficult to read. It is important to make comments concise so that they do not crowd the program or make it difficult to read.

In addition to line comments (beginning with //) and block comments (beginning with /*), Java supports comments of a special type, referred to as javadoc comments. javadoc comments begin with /** and end with */. They can be extracted into an HTML file using the JDK’s javadoc command. For more information, see Supplement III.Y, javadoc Comments, on the Companion Website.

class loader

bytecode verifier

use package

Point Check

Point Key

programming style documentation

javadoc comment

M01_LIAN1878_11_GE_C01.indd 40 1/2/18 10:57 PM

1.9 Programming Style and Documentation 41

Use javadoc comments (/** . . . */) for commenting on an entire class or an entire method. These comments must precede the class or the method header in order to be extracted into a javadoc HTML file. For commenting on steps inside a method, use line comments (//). To see an example of a javadoc HTML file, check out liveexample.pearsoncmg.com/javadoc/ Exercise1.html. Its corresponding Java code is shown in liveexample.pearsoncmg.com/java- doc/Exercise1.txt.

1.9.2 Proper Indentation and Spacing A consistent indentation style makes programs clear and easy to read, debug, and maintain. Indentation is used to illustrate the structural relationships between a program’s compo- nents or statements. Java can read the program even if all of the statements are on the same long line, but humans find it easier to read and maintain code that is aligned properly. Indent each subcomponent or statement at least two spaces more than the construct within which it is nested.

A single space should be added on both sides of a binary operator, as shown in (a), rather in (b).

System.out.println(3 + 4 * 4); System.out.println(3+4*4);

(a) Good style (b) Bad style

1.9.3 Block Styles A block is a group of statements surrounded by braces. There are two popular styles, next-line style and end-of-line style, as shown below.

public class Test { public static void main(String[] args) { System.out.println("Block Styles"); } }

Next-line style End-of-line style

The next-line style aligns braces vertically and makes programs easy to read, whereas the end-of-line style saves space and may help avoid some subtle programming errors. Both are acceptable block styles. The choice depends on personal or organizational preference. You should use a block style consistently—mixing styles is not recommended. This book uses the end-of-line style to be consistent with the Java API source code.

1.9.1 Reformat the following program according to the programming style and documen- tation guidelines. Use the end-of-line brace style.

public class Test { // Main method public static void main(String[] args) { /** Display output */ System.out.println("Welcome to Java"); } }

indent code

Point Check

M01_LIAN1878_11_GE_C01.indd 41 1/2/18 10:57 PM

42 Chapter 1 Introduction to Computers, Programs, and Java™

1.10 Programming Errors Programming errors can be categorized into three types: syntax errors, runtime errors, and logic errors.

1.10.1 Syntax Errors Errors that are detected by the compiler are called syntax errors or compile errors. Syntax errors result from errors in code construction, such as mistyping a keyword, omitting some necessary punctuation, or using an opening brace without a corresponding closing brace. These errors are usually easy to detect because the compiler tells you where they are and what caused them. For example, the program in Listing 1.4 has a syntax error, as shown in Figure 1.10.

lisTing 1.4 ShowSyntaxErrors.java 1 public class ShowSyntaxErrors { 2 public static main(String[] args) { 3 System.out.println("Welcome to Java); 4 } 5 }

Four errors are reported, but the program actually has two errors:

■■ The keyword void is missing before main in line 2.

■■ The string Welcome to Java should be closed with a closing quotation mark in line 3.

Since a single error will often display many lines of compile errors, it is a good practice to fix errors from the top line and work downward. Fixing errors that occur earlier in the program may also fix additional errors that occur later.

Point Key

syntax errors

compile errors

Figure 1.10 The compiler reports syntax errors.

Compile

Tip If you don’t know how to correct an error, compare your program closely, character by character, with similar examples in the text. In the first few weeks of this course, you will probably spend a lot of time fixing syntax errors. Soon you will be familiar with Java syntax, and can quickly fix syntax errors.

fix syntax errors

M01_LIAN1878_11_GE_C01.indd 42 1/2/18 10:57 PM

1.10 Programming Errors 43

1.10.2 Runtime Errors Runtime errors are errors that cause a program to terminate abnormally. They occur while a program is running if the environment detects an operation that is impossible to carry out. Input mistakes typically cause runtime errors. An input error occurs when the program is waiting for the user to enter a value, but the user enters a value that the program cannot handle. For instance, if the program expects to read in a number, but instead the user enters a string, this causes data-type errors to occur in the program.

Another example of runtime errors is division by zero. This happens when the divisor is zero for integer divisions. For instance, the program in Listing 1.5 would cause a runtime error, as shown in Figure 1.11.

lisTing 1.5 ShowRuntimeErrors.java 1 public class ShowRuntimeErrors { 2 public static void main(String[] args) { 3 System.out.println(1 / 0); 4 } 5 }

runtime errors

runtime error

Figure 1.11 The runtime error causes the program to terminate abnormally.

Run

1.10.3 Logic Errors Logic errors occur when a program does not perform the way it was intended to. Errors of this kind occur for many different reasons. For example, suppose you wrote the program in Listing 1.6 to convert Celsius 35 degrees to a Fahrenheit degree:

lisTing 1.6 ShowLogicErrors.java 1 public class ShowLogicErrors { 2 public static void main(String[] args) { 3 System.out.print("Celsius 35 is Fahrenheit degree "); 4 System.out.println((9 / 5) * 35 + 32); 5 } 6 }

logic errors

Celsius 35 is Fahrenheit degree 67

You will get Fahrenheit 67 degrees, which is wrong. It should be 95.0. In Java, the division for integers is the quotient—the fractional part is truncated—so in Java 9 / 5 is 1. To get the correct result, you need to use 9.0 / 5, which results in 1.8.

In general, syntax errors are easy to find and easy to correct because the compiler gives indications as to where the errors came from and why they are wrong. Runtime errors are not difficult to find, either, since the reasons and locations for the errors are displayed on the console when the program aborts. Finding logic errors, on the other hand, can be very challenging. In the upcoming chapters, you will learn the techniques of tracing programs and finding logic errors.

M01_LIAN1878_11_GE_C01.indd 43 1/2/18 10:57 PM

44 Chapter 1 Introduction to Computers, Programs, and Java™

1.10.4 Common Errors Missing a closing brace, missing a semicolon, missing quotation marks for strings, and mis- spelling names are common errors for new programmers.

Common Error 1: Missing Braces

The braces are used to denote a block in the program. Each opening brace must be matched by a closing brace. A common error is missing the closing brace. To avoid this error, type a closing brace whenever an opening brace is typed, as shown in the following example:

public class Welcome {

} Type this closing brace right away to match the opening brace.

If you use an IDE such as NetBeans and Eclipse, the IDE automatically inserts a closing brace for each opening brace typed.

Common Error 2: Missing Semicolons

Each statement ends with a statement terminator (;). Often, a new programmer forgets to place a statement terminator for the last statement in a block, as shown in the following example:

public static void main(String[] args) { System.out.println("Programming is fun!"); System.out.println("Fundamentals First"); System.out.println("Problem Driven") }

Missing a semicolon

Common Error 3: Missing Quotation Marks

A string must be placed inside the quotation marks. Often, a new programmer forgets to place a quotation mark at the end of a string, as shown in the following example:

System.out.println("Problem Driven);

Missing a quotation mark

If you use an IDE such as NetBeans and Eclipse, the IDE automatically inserts a closing quotation mark for each opening quotation mark typed.

Common Error 4: Misspelling Names

Java is case sensitive. Misspelling names is a common error for new programmers. For exam- ple, the word main is misspelled as Main and String is misspelled as string in the follow- ing code:

1 public class Test { 2 public static void Main(string[] args) { 3 System.out.println((10.5 + 2 * 3) / (45 – 3.5)); 4 } 5 }

1.10.1 What are syntax errors (compile errors), runtime errors, and logic errors? 1.10.2 Give examples of syntax errors, runtime errors, and logic errors. 1.10.3 If you forget to put a closing quotation mark on a string, what kind error of will be raised? 1.10.4 If your program needs to read integers, but the user entered strings, an error would

occur when running this program. What kind of error is this?

Point Check

M01_LIAN1878_11_GE_C01.indd 44 1/2/18 10:57 PM

1.11 Developing Java Programs Using NetBeans 45

1.10.5 Suppose you write a program for computing the perimeter of a rectangle and you mistak- enly write your program so it computes the area of a rectangle. What kind of error is this?

1.10.6 Identify and fix the errors in the following code:

1 public class Welcome { 2 public void Main(String[] args) { 3 System.out.println('Welcome to Java!); 4 } 5 )

Note Section 1.8 introduced developing programs from the command line. Many of our readers also use an IDE. The following two sections introduce two most popular Java IDEs: NetBeans and Eclipse. These two sections may be skipped.

1.11 Developing Java Programs Using NetBeans You can edit, compile, run, and debug Java Programs using NetBeans.

NetBeans and Eclipse are two free popular integrated development environments for develop- ing Java programs. They are easy to learn if you follow simple instructions. We recommend that you use either one for developing Java programs. This section gives the essential instruc- tions to guide new users to create a project, create a class, compile, and run a class in NetBeans. The use of Eclipse will be introduced in the next section. For instructions on downloading and installing latest version of NetBeans, see Supplement II.B.

1.11.1 Creating a Java Project Before you can create Java programs, you need to first create a project. A project is like a folder to hold Java programs and all supporting files. You need to create a project only once. Here are the steps to create a Java project:

1. Choose File, New Project to display the New Project dialog box, as shown in Figure 1.12.

2. Select Java in the Categories section and Java Application in the Projects section, and then click Next to display the New Java Application dialog box, as shown in Figure 1.13.

3. Type demo in the Project Name field and c:\michael in Project Location field. Uncheck Use Dedicated Folder for Storing Libraries and uncheck Create Main Class.

4. Click Finish to create the project, as shown in Figure 1.14.

Point Key

VideoNote

NetBeans brief tutorial

M01_LIAN1878_11_GE_C01.indd 45 1/2/18 10:57 PM

46 Chapter 1 Introduction to Computers, Programs, and Java™

1.11.2 Creating a Java Class After a project is created, you can create Java programs in the project using the following steps:

1. Right-click the demo node in the project pane to display a context menu. Choose New, Java Class to display the New Java Class dialog box, as shown in Figure 1.15.

2. Type Welcome in the Class Name field and select the Source Packages in the Location field. Leave the Package field blank. This will create a class in the default package.

3. Click Finish to create the Welcome class. The source-code file Welcome.java is placed under the <default package> node.

4. Modify the code in the Welcome class to match Listing 1.1 in the text, as shown in Figure 1.16.

1.11.3 Compiling and Running a Class To run Welcome.java, right-click Welcome.java to display a context menu and choose Run File, or simply press Shift + F6. The output is displayed in the Output pane, as shown in Figure 1.16. The Run File command automatically compiles the program if the program has been changed.

M01_LIAN1878_11_GE_C01.indd 46 1/2/18 10:57 PM

1.12 Developing Java Programs Using Eclipse 47

Edit pane

Output pane

1.12 Developing Java Programs Using Eclipse You can edit, compile, run, and debug Java Programs using Eclipse.

The preceding section introduced developing Java programs using NetBeans. You can also use Eclipse to develop Java programs. This section gives the essential instructions to guide new users to create a project, create a class, and compile/run a class in Eclipse. For instructions on downloading and installing latest version of Eclipse, see Supplement II.D.

1.12.1 Creating a Java Project Before creating Java programs in Eclipse, you need to first create a project to hold all files. Here are the steps to create a Java project in Eclipse:

1. Choose File, New, Java Project to display the New Project wizard, as shown in Figure 1.17.

2. Type demo in the Project name field. As you type, the Location field is automatically set by default. You may customize the location for your project.

Point Key

VideoNote

Eclipse brief tutorial

M01_LIAN1878_11_GE_C01.indd 47 1/2/18 10:57 PM

48 Chapter 1 Introduction to Computers, Programs, and Java™

3. Make sure you selected the options Use project folder as root for sources and class files so the .java and .class files are in the same folder for easy access.

4. Click Finish to create the project, as shown in Figure 1.18.

Figure 1.17 The New Java Project dialog is for specifying a project name and the properties. Source: Eclipse Foundation, Inc.

Figure 1.18 A New Java project named demo is created. Source: Eclipse Foundation, Inc.

M01_LIAN1878_11_GE_C01.indd 48 1/2/18 10:57 PM

1.12 Developing Java Programs Using Eclipse 49

1.12.2 Creating a Java Class After a project is created, you can create Java programs in the project using the following steps:

1. Choose File, New, Class to display the New Java Class wizard.

2. Type Welcome in the Name field.

3. Check the option public static void main(String[] args).

4. Click Finish to generate the template for the source code Welcome.java, as shown in Figure 1.19.

1.12.3 Compiling and Running a Class To run the program, right-click the class in the project to display a context menu. Choose Run, Java Application in the context menu to run the class. The output is displayed in the Console pane, as shown in Figure 1.20. The Run command automatically compiles the program if the program has been changed.

Figure 1.19 The New Java Class dialog box is used to create a new Java class. Source: Eclipse Foundation, Inc.

M01_LIAN1878_11_GE_C01.indd 49 1/2/18 10:57 PM

54 Chapter 1 Introduction to Computers, Programs, and Java™

*1.13 (Algebra: solve 2 * 2 linear equations) You can use Cramer’s rule to solve the following 2 * 2 system of linear equation provided that ad – bc is not 0:

ax + by = e cx + dy = f

x = ed - bf ad - bc

y = af - ec ad - bc

Write a program that solves the following equation and displays the value for x and y: (Hint: replace the symbols in the formula with numbers to compute x and y. This exercise can be done in Chapter 1 without using materials in later chapters.)

3.4x + 50.2y = 44.5 2.1x + .55y = 5.9

Note More than 200 additional programming exercises with solutions are provided to the instructors on the Instructor Resource Website.

M01_LIAN1878_11_GE_C01.indd 54 1/2/18 10:57 PM

50 Chapter 1 Introduction to Computers, Programs, and Java™

Figure 1.20 You can edit a program and run it in Eclipse. Source: Eclipse Foundation, Inc.

Edit pane

Output pane

Key Terms Application Program Interface (API) 33 assembler 29 assembly language 29 bit 25 block 35 block comment 35 bus 24 byte 25 bytecode 38 bytecode verifier 40 cable modem 28 central processing unit (CPU) 25 class loader 40 comment 35 compiler 30 console 34 dot pitch 28 DSL (digital subscriber line) 28 encoding scheme 25 hardware 24 high-level language 30 integrated development environment

(IDE) 34 interpreter 30 java command 39 Java Development Toolkit (JDK) 34 Java language specification 33

Java Runtime Environment (JRE) 34 Java Virtual Machine (JVM) 38 javac command 39 keyword (or reserved word) 35 library 33 line comment 35 logic error 43 low-level language 30 machine language 29 main method 35 memory 26 dial-up modem 28 motherboard 25 network interface card (NIC) 28 operating system (OS) 31 pixel 28 program 24 programming 24 runtime error 43 screen resolution 28 software 24 source code 30 source program 30 statement 30 statement terminator 35 storage devices 26 syntax error 42

Note The above terms are defined in this chapter. Supplement I.A, Glossary, lists all the key terms and descriptions in the book, organized by chapters.

Supplement I.A

M01_LIAN1878_11_GE_C01.indd 50 1/2/18 10:57 PM

Chapter Summary 51

ChapTer summary 1. A computer is an electronic device that stores and processes data.

2. A computer includes both hardware and software.

3. Hardware is the physical aspect of the computer that can be touched.

4. Computer programs, known as software, are the invisible instructions that control the hardware and make it perform tasks.

5. Computer programming is the writing of instructions (i.e., code) for computers to perform.

6. The central processing unit (CPU) is a computer’s brain. It retrieves instructions from memory and executes them.

7. Computers use zeros and ones because digital devices have two stable states, referred to by convention as zero and one.

8. A bit is a binary digit 0 or 1.

9. A byte is a sequence of 8 bits.

10. A kilobyte is about 1,000 bytes, a megabyte about 1 million bytes, a gigabyte about 1 billion bytes, and a terabyte about 1,000 gigabytes.

11. Memory stores data and program instructions for the CPU to execute.

12. A memory unit is an ordered sequence of bytes.

13. Memory is volatile, because information is lost when the power is turned off.

14. Programs and data are permanently stored on storage devices and are moved to memory when the computer actually uses them.

15. The machine language is a set of primitive instructions built into every computer.

16. Assembly language is a low-level programming language in which a mnemonic is used to represent each machine-language instruction.

17. High-level languages are English-like and easy to learn and program.

18. A program written in a high-level language is called a source program.

19. A compiler is a software program that translates the source program into a machine- language program.

20. The operating system (OS) is a program that manages and controls a computer’s activities.

21. Java is platform independent, meaning you can write a program once and run it on any computer.

22. The Java source file name must match the public class name in the program. Java source- code files must end with the .java extension.

23. Every class is compiled into a separate bytecode file that has the same name as the class and ends with the .class extension.

24. To compile a Java source-code file from the command line, use the javac command.

M01_LIAN1878_11_GE_C01.indd 51 1/2/18 10:57 PM

52 Chapter 1 Introduction to Computers, Programs, and Java™

25. To run a Java class from the command line, use the java command.

26. Every Java program is a set of class definitions. The keyword class introduces a class definition. The contents of the class are included in a block.

27. A block begins with an opening brace ({) and ends with a closing brace (}).

28. Methods are contained in a class. To run a Java program, the program must have a main method. The main method is the entry point where the program starts when it is executed.

29. Every statement in Java ends with a semicolon (;), known as the statement terminator.

30. Reserved words, or keywords, have a specific meaning to the compiler and cannot be used for other purposes in the program.

31. In Java, comments are preceded by two slashes (//) on a line, called a line comment, or enclosed between /* and */ on one or several lines, called a block comment or para- graph comment. Comments are ignored by the compiler.

32. Java source programs are case sensitive.

33. Programming errors can be categorized into three types: syntax errors, runtime errors, and logic errors. Errors reported by a compiler are called syntax errors or compile errors. Runtime errors are errors that cause a program to terminate abnormally. Logic errors occur when a program does not perform the way it was intended to.

Quiz Answer the quiz for this chapter at www.pearsonglobaleditions.com/Liang. Choose this book and click Companion Website to select Quiz.

programming exerCises

Pedagogical Note We cannot stress enough the importance of learning programming through exercises. For this reason, the book provides a large number of programming exercises at various levels of difficulty. The problems cover many application areas, including math, science, business, financial, gaming, animation, and multimedia. Solutions to most even- numbered programming exercises are on the Companion Website. Solutions to most odd-numbered programming exercises are on the Instructor Resource Website. The level of difficulty is rated easy (no star), moderate (*), hard (**), or challenging (***).

1.1 (Display three messages) Write a program that displays Welcome to Java, Learning Java Now, and Programming is fun.

1.2 (Display five messages) Write a program that displays I love Java five times. *1.3 (Display a pattern) Write a program that displays the following pattern:

J J aaa v vaaa J J aa v v a a J aaaa v aaaa

level of difficulty

M01_LIAN1878_11_GE_C01.indd 52 1/2/18 10:57 PM

Programming Exercises 53

1.4 (Print a table) Write a program that displays the following table:

a a^2 a^3 a^4 1 1 1 1 2 4 8 16 3 9 27 81 4 16 64 256

1.5 (Compute expressions) Write a program that displays the result of 7.5 * 6.5 - 4.5 * 3

47.5 - 5.5 .

1.6 (Summation of a series) Write a program that displays the result of 1 + 2 + 3 + 4 + 5 + 6 + 7 + 8 + 9 + 10.

1.7 (Approximate p) p can be computed using the following formula:

p = 4 * ¢1 - 1 3

+ 1 5

- 1 7

+ 1 9

- 1 11

+ c ≤ Write a program that displays the result of 4 * ¢1 - 1

3 +

1 5

- 1 7

+ 1 9

- 1 11

≤ and 4 * ¢1 - 1

3 +

1 5

- 1 7

+ 1 9

- 1 11

+ 1 13

≤. Use 1.0 instead of 1 in your program.

1.8 (Area and perimeter of a circle) Write a program that displays the area and perim- eter of a circle that has a radius of 6.5 using the following formula:

p = 3.14159 perimeter = 2 * radius * p

area = radius * radius * p

1.9 (Area and perimeter of a rectangle) Write a program that displays the area and perim- eter of a rectangle with a width of 5.3 and height of 8.6 using the following formula:

area = width * height

perimeter = 2 * (width + height)

1.10 (Average speed in miles) Assume that a runner runs 15 kilometers in 50 minutes and 30 seconds. Write a program that displays the average speed in miles per hour. (Note that 1 mile is 1.6 kilometers.)

*1.11 (Population projection) The U.S. Census Bureau projects population based on the following assumptions:

■■ One birth every 7 seconds ■■ One death every 13 seconds ■■ One new immigrant every 45 seconds

Write a program to display the population for each of the next five years. Assume that the current population is 312,032,486, and one year has 365 days. Hint: In Java, if two integers perform division, the result is an integer. The fractional part is truncated. For example, 5 / 4 is 1 (not 1.25) and 10 / 4 is 2 (not 2.5). To get an accurate result with the fractional part, one of the values involved in the division must be a number with a decimal point. For example, 5.0 / 4 is 1.25 and 10 / 4.0 is 2.5.

1.12 (Average speed in kilometers) Assume that a runner runs 24 miles in 1 hour, 40 minutes, and 35 seconds. Write a program that displays the average speed in kilometers per hour. (Note 1 mile is equal to 1.6 kilometers.)

M01_LIAN1878_11_GE_C01.indd 53 1/2/18 10:57 PM

Elementary Programming

Objectives ■■ To write Java programs to perform simple computations (§2.2).

■■ To obtain input from the console using the Scanner class (§2.3).

■■ To use identifiers to name variables, constants, methods, and classes (§2.4).

■■ To use variables to store data (§§2.5 and 2.6).

■■ To program with assignment statements and assignment expressions (§2.6).

■■ To use constants to store permanent data (§2.7).

■■ To name classes, methods, variables, and constants by following their naming conventions (§2.8).

■■ To explore Java numeric primitive data types: byte, short, int, long, float, and double (§2.9.1).

■■ To read a byte, short, int, long, float, or double value from the keyboard (§2.9.2).

■■ To perform operations using operators +, -, *, /, and % (§2.9.3).

■■ To perform exponent operations using Math.pow(a, b) (§2.9.4).

■■ To write integer literals, floating-point literals, and literals in scientific notation (§2.10).

■■ To write and evaluate numeric expressions (§2.11).

■■ To obtain the current system time using System.currentTimeMillis() (§2.12).

■■ To use augmented assignment operators (§2.13).

■■ To distinguish between postincrement and preincrement and between postdecrement and predecrement (§2.14).

■■ To cast the value of one type to another type (§2.15).

■■ To describe the software development process and apply it to develop the loan payment program (§2.16).

■■ To write a program that converts a large amount of money into smaller units (§2.17).

■■ To avoid common errors and pitfalls in elementary programming (§2.18).

Chapter

M02_LIAN1878_11_GE_C02.indd 55 12/29/17 3:47 PM

56 Chapter 2 Elementary Programming

2.1 Introduction The focus of this chapter is on learning elementary programming techniques to solve problems.

In Chapter 1, you learned how to create, compile, and run very basic Java programs. You will learn how to solve problems by writing programs. Through these problems, you will learn elementary programming using primitive data types, variables, constants, operators, expres- sions, and input and output.

Suppose, for example, you need to take out a student loan. Given the loan amount, loan term, and annual interest rate, can you write a program to compute the monthly payment and total payment? This chapter shows you how to write programs like this. Along the way, you will learn the basic steps that go into analyzing a problem, designing a solution, and implement- ing the solution by creating a program.

2.2 Writing a Simple Program Writing a program involves designing a strategy for solving the problem then using a programming language to implement that strategy.

Let’s first consider the simple problem of computing the area of a circle. How do we write a program for solving this problem?

Writing a program involves designing algorithms and translating algorithms into program- ming instructions, or code. An algorithm lists the steps you can follow to solve a problem. Algorithms can help the programmer plan a program before writing it in a programming language. Algorithms can be described in natural languages or in pseudocode (natural language mixed with some programming code). The algorithm for calculating the area of a circle can be described as follows:

1. Read in the circle’s radius.

2. Compute the area using the following formula:

area = radius * radius * p

3. Display the result.

Tip It’s always a good practice to outline your program (or its underlying problem) in the form of an algorithm before you begin coding.

When you code—that is, when you write a program—you translate an algorithm into a program. You already know every Java program begins with a class definition in which the keyword class is followed by the class name. Assume you have chosen ComputeArea as the class name. The outline of the program would look as follows:

public class ComputeArea { // Details to be given later }

As you know, every Java program must have a main method where program execution begins. The program is then expanded as follows:

public class ComputeArea { public static void main(String[] args) { // Step 1: Read in radius

// Step 2: Compute area

Point Key

problem

algorithm

pseudocode

M02_LIAN1878_11_GE_C02.indd 56 12/29/17 3:47 PM

2.2 Writing a Simple Program 57

// Step 3: Display the area } }

The program needs to read the radius entered by the user from the keyboard. This raises two important issues:

■■ Reading the radius

■■ Storing the radius in the program

Let’s address the second issue first. In order to store the radius, the program needs to declare a symbol called a variable. A variable represents a value stored in the computer’s memory.

Rather than using x and y as variable names, choose descriptive names: in this case, radius for radius and area for area. To let the compiler know what radius and area are, specify their data types. That is the kind of data stored in a variable, whether an integer, real number, or something else. This is known as declaring variables. Java provides simple data types for representing integers, real numbers, characters, and Boolean types. These types are known as primitive data types or fundamental types.

Real numbers (i.e., numbers with a decimal point) are represented using a method known as floating-point in computers. Therefore, the real numbers are also called floating-point numbers. In Java, you can use the keyword double to declare a floating-point variable. Declare radius and area as double. The program can be expanded as follows:

public class ComputeArea { public static void main(String[] args) { double radius; double area;

// Step 1: Read in radius

// Step 2: Compute area

// Step 3: Display the area } }

The program declares radius and area as variables. The reserved word double indicates that radius and area are floating-point values stored in the computer.

The first step is to prompt the user to designate the circle’s radius. You will soon learn how to prompt the user for information. For now, to learn how variables work, you can assign a fixed value to radius in the program as you write the code. Later, you’ll modify the program to prompt the user for this value.

The second step is to compute area by assigning the result of the expression radius * radius * 3.14159 to area.

In the final step, the program will display the value of area on the console by using the System.out.println method.

Listing 2.1 shows the complete program, and a sample run of the program is shown in Figure 2.1.

lisiTing 2.1 ComputeArea.java 1 public class ComputeArea { 2 public static void main(String[] args) { 3 double radius; // Declare radius 4 double area; // Declare area 5 6 // Assign a radius 7 radius = 20; // radius is now 20

variable descriptive names

declare variables data type

primitive data types

floating-point

M02_LIAN1878_11_GE_C02.indd 57 12/29/17 3:47 PM

58 Chapter 2 Elementary Programming

8 9 // Compute area 10 area = radius * radius * 3.14159; 11 12 // Display results 13 System.out.println("The area for the circle of radius " + 14 radius + " is " + area); 15 } 16 }

Figure 2.1 The program displays the area of a circle.

Compile

Run

Variables such as radius and area correspond to memory locations. Every variable has a name, a type, and a value. Line 3 declares that radius can store a double value. The value is not defined until you assign a value. Line 7 assigns 20 into the variable radius. Similarly, line 4 declares the variable area, and line 10 assigns a value into area. The following table shows the value in the memory for area and radius as the program is executed. Each row in the table shows the values of variables after the statement in the corresponding line in the program is executed. This method of reviewing how a program works is called tracing a pro- gram. Tracing programs are helpful for understanding how programs work, and they are useful tools for finding errors in programs.

line# radius area

3 no value

4 no value

7 20

10 1256.636

The plus sign (+) has two meanings: one for addition, and the other for concatenating (com- bining) strings. The plus sign (+) in lines 13–14 is called a string concatenation operator. It combines two strings into one. If a string is combined with a number, the number is converted into a string and concatenated with the other string. Therefore, the plus signs (+) in lines 13–14 concatenate strings into a longer string, which is then displayed in the output. Strings and string concatenation will be discussed further in Chapter 4.

Caution A string cannot cross lines in the source code. Thus, the following statement would result in a compile error:

System.out.println("Introduction to Java Programming, by Y. Daniel Liang");

To fix the error, break the string into separate substrings, and use the concatenation operator (+) to combine them:

System.out.println("Introduction to Java Programming, " + "by Y. Daniel Liang");

declare variable assign value

tracing program

concatenate strings with numbers

concatenate strings

break a long string

M02_LIAN1878_11_GE_C02.indd 58 12/29/17 3:47 PM

2.3 Reading Input from the Console 59

2.2.1 Identify and fix the errors in the following code:

1 public class Test { 2 public void main(string[] args) { 3 double i = 50.0; 4 double k = i + 50.0; 5 double j = k + 1; 6 7 System.out.println("j is " + j + " and 8 k is " + k); 9 } 10 }

2.3 Reading Input from the Console Reading input from the console enables the program to accept input from the user.

In Listing 2.1, the radius is fixed in the source code. To use a different radius, you have to modify the source code and recompile it. Obviously, this is not convenient, so instead you can use the Scanner class for console input.

Java uses System.out to refer to the standard output device, and System.in to the standard input device. By default, the output device is the display monitor, and the input device is the keyboard. To perform console output, you simply use the println method to display a primitive value or a string to the console. To perform console input, you need to use the Scanner class to create an object to read input from System.in, as follows:

Scanner input = new Scanner(System.in);

The syntax new Scanner(System.in) creates an object of the Scanner type. The syntax Scanner input declares that input is a variable whose type is Scanner. The whole line Scanner input = new Scanner(System.in) creates a Scanner object and assigns its reference to the variable input. An object may invoke its methods. To invoke a method on an object is to ask the object to perform a task. You can invoke the nextDouble() method to read a double value as follows:

double radius = input.nextDouble();

This statement reads a number from the keyboard and assigns the number to radius. Listing 2.2 rewrites Listing 2.1 to prompt the user to enter a radius.

Listing 2.2 ComputeAreaWithConsoleInput.java 1 import java.util.Scanner; // Scanner is in the java.util package 2 3 public class ComputeAreaWithConsoleInput { 4 public static void main(String[] args) { 5 // Create a Scanner object 6 Scanner input = new Scanner(System.in); 7 8 // Prompt the user to enter a radius 9 System.out.print("Enter a number for radius: "); 10 double radius = input.nextDouble(); 11 12 // Compute area 13 double area = radius * radius * 3.14159; 14 15 // Display results 16 System.out.println("The area for the circle of radius " +

Point Check

Point Key

VideoNote Obtain input

import class

read a double

create a Scanner

M02_LIAN1878_11_GE_C02.indd 59 12/29/17 3:47 PM

60 Chapter 2 Elementary Programming

17 radius + " is " + area); 18 } 19 }

Enter a number for radius: 2.5

The area for the circle of radius 2.5 is 19.6349375

Enter a number for radius: 23

The area for the circle of radius 23.0 is 1661.90111

The Scanner class is in the java.util package. It is imported in line 1. Line 6 creates a Scanner object. Note the import statement can be omitted if you replace Scanner by java. util.Scanner in line 6.

Line 9 displays a string "Enter a number for radius: " to the console. This is known as a prompt, because it directs the user to enter an input. Your program should always tell the user what to enter when expecting input from the keyboard.

Recall that the print method in line 9 is identical to the println method, except that println moves to the beginning of the next line after displaying the string, but print does not advance to the next line when completed.

Line 6 creates a Scanner object. The statement in line 10 reads input from the keyboard.

double radius = input.nextDouble();

After the user enters a number and presses the Enter key, the program reads the number and assigns it to radius.

More details on objects will be introduced in Chapter 9. For the time being, simply accept that this is how we obtain input from the console.

The Scanner class is in the java.util package. It is imported in line 1. There are two types of import statements: specific import and wildcard import. The specific import specifies a single class in the import statement. For example, the following statement imports Scanner from the package java.util.

import java.util.Scanner;

The wildcard import imports all the classes in a package by using the asterisk as the wildcard. For example, the following statement imports all the classes from the package java.util.

import java.util.*;

The information for the classes in an imported package is not read in at compile time or runtime unless the class is used in the program. The import statement simply tells the compiler where to locate the classes. There is no performance difference between a specific import and a wildcard import declaration.

Listing 2.3 gives an example of reading multiple inputs from the keyboard. The program reads three numbers and displays their average.

Listing 2.3 ComputeAverage.java 1 import java.util.Scanner; // Scanner is in the java.util package 2 3 public class ComputeAverage { 4 public static void main(String[] args) { 5 // Create a Scanner object 6 Scanner input = new Scanner(System.in); 7 8 // Prompt the user to enter three numbers 9 System.out.print("Enter three numbers: ");

prompt

specific import

wildcard import

no performance difference

import class

create a Scanner

M02_LIAN1878_11_GE_C02.indd 60 12/29/17 3:47 PM

2.3 Reading Input from the Console 61

10 double number1 = input.nextDouble(); 11 double number2 = input.nextDouble(); 12 double number3 = input.nextDouble(); 13 14 // Compute average 15 double average = (number1 + number2 + number3) / 3; 16 17 // Display results 18 System.out.println("The average of " + number1 + " " + number2 19 + " " + number3 + " is " + average); 20 } 21 }

read a double

enter input in one line

enter input in multiple lines

Enter three numbers: 10.5

11.5

The average of 10.5 11.0 11.5 is 11.0

Enter three numbers: 1 2 3

The average of 1.0 2.0 3.0 is 2.0

The codes for importing the Scanner class (line 1) and creating a Scanner object (line 6) are the same as in the preceding example, as well as in all new programs you will write for reading input from the keyboard.

Line 9 prompts the user to enter three numbers. The numbers are read in lines 10–12. You may enter three numbers separated by spaces, then press the Enter key, or enter each number followed by a press of the Enter key, as shown in the sample runs of this program.

If you entered an input other than a numeric value, a runtime error would occur. In Chapter 12, you will learn how to handle the exception so the program can continue to run.

Note Most of the programs in the early chapters of this book perform three steps— input, process, and output—called IPO. Input is receiving input from the user; process is pro- ducing results using the input; and output is displaying the results.

Note If you use an IDE such as Eclipse or NetBeans, you will get a warning to ask you to close the input for preventing a potential resource leak. Ignore the warning for the time being because the input is automatically closed when your program is terminated. In this case, there will be no resource leaking.

2.3.1 How do you write a statement to let the user enter a double value from the keyboard? What happens if you entered 5a when executing the following code?

double radius = input.nextDouble();

2.3.2 Are there any performance differences between the following two import statements?

import java.util.Scanner; import java.util.*;

runtime error

IPO

Warning in IDE

Point Check

M02_LIAN1878_11_GE_C02.indd 61 12/29/17 3:48 PM

62 Chapter 2 Elementary Programming

2.4 Identifiers Identifiers are the names that identify the elements such as classes, methods, and variables in a program.

As you see in Listing 2.3, ComputeAverage, main, input, number1, number2, number3, and so on are the names of things that appear in the program. In programming terminology, such names are called identifiers. All identifiers must obey the following rules:

■■ An identifier is a sequence of characters that consists of letters, digits, underscores (_), and dollar signs ($).

■■ An identifier must start with a letter, an underscore (_), or a dollar sign ($). It cannot start with a digit.

■■ An identifier cannot be a reserved word. (See Appendix A for a list of reserved words.)

■■ An identifier cannot be true, false, or null.

■■ An identifier can be of any length.

For example, $2, ComputeArea, area, radius, and print are legal identifiers, whereas 2A and d+4 are not because they do not follow the rules. The Java compiler detects illegal identifiers and reports syntax errors.

Note Since Java is case sensitive, area, Area, and AREA are all different identifiers.

Tip Identifiers are for naming variables, methods, classes, and other items in a program. Descriptive identifiers make programs easy to read. Avoid using abbreviations for identi- fiers. Using complete words is more descriptive. For example, numberOfStudents is better than numStuds, numOfStuds, or numOfStudents. We use descriptive names for complete programs in the text. However, we will occasionally use variable names such as i, j, k, x, and y in the code snippets for brevity. These names also provide a generic tone to the code snippets.

Tip Do not name identifiers with the $ character. By convention, the $ character should be used only in mechanically generated source code.

2.4.1 Which of the following identifiers are valid? Which are Java keywords?

miles, Test, a++, ––a, 4#R, $4, #44, apps class, public, int, x, y, radius

2.5 Variables Variables are used to represent values that may be changed in the program.

As you see from the programs in the preceding sections, variables are used to store values to be used later in a program. They are called variables because their values can be changed. In the program in Listing 2.2, radius and area are variables of the double type. You can assign any numerical value to radius and area, and the values of radius and area can be reassigned. For example, in the following code, radius is initially 1.0 (line 2) then changed to 2.0 (line 7), and area is set to 3.14159 (line 3) then reset to 12.56636 (line 8).

Point Key

identifiers

identifier naming rules

case sensitive

descriptive names

the $ character

Point Check

Point Key

why called variables?

M02_LIAN1878_11_GE_C02.indd 62 12/29/17 3:48 PM

2.5 Variables 63

1 // Compute the first area 2 radius = 1.0; radius: 1.0 3 area = radius * radius * 3.14159; area: 3.14159 4 System.out.println("The area is " + area + " for radius " + radius); 5 6 // Compute the second area 7 radius = 2.0; radius: 2.0 8 area = radius * radius * 3.14159; area: 12.56636 9 System.out.println("The area is " + area + " for radius " + radius);

Variables are for representing data of a certain type. To use a variable, you declare it by telling the compiler its name as well as what type of data it can store. The variable declaration tells the compiler to allocate appropriate memory space for the variable based on its data type. The syntax for declaring a variable is

datatype variableName;

Here are some examples of variable declarations:

int count; // Declare count to be an integer variable double radius; // Declare radius to be a double variable double interestRate; // Declare interestRate to be a double variable

These examples use the data types int and double. Later you will be introduced to additional data types, such as byte, short, long, float, char, and boolean.

If variables are of the same type, they can be declared together, as follows:

datatype variable1, variable2, ..., variablen;

The variables are separated by commas. For example,

int i, j, k; // Declare i, j, and k as int variables

Variables often have initial values. You can declare a variable and initialize it in one step. Consider, for instance, the following code:

int count = 1;

This is equivalent to the next two statements:

int count; count = 1;

You can also use a shorthand form to declare and initialize variables of the same type together. For example,

int i = 1, j = 2;

Tip A variable must be declared before it can be assigned a value. A variable declared in a method must be assigned a value before it can be used.

Whenever possible, declare a variable and assign its initial value in one step. This will make the program easy to read and avoid programming errors.

Every variable has a scope. The scope of a variable is the part of the program where the variable can be referenced. The rules that define the scope of a variable will be gradually introduced later in the book. For now, all you need to know is that a variable must be declared and initialized before it can be used.

declare variable

initialize variables

M02_LIAN1878_11_GE_C02.indd 63 12/29/17 3:48 PM

64 Chapter 2 Elementary Programming

2.5.1 Identify and fix the errors in the following code:

1 public class Test { 2 public static void main(String[] args) { 3 int i = k + 2; 4 System.out.println(i); 5 } 6 }

2.6 Assignment Statements and Assignment Expressions

An assignment statement designates a value for a variable. An assignment statement can be used as an expression in Java.

After a variable is declared, you can assign a value to it by using an assignment statement. In Java, the equal sign (=) is used as the assignment operator. The syntax for assignment state- ments is as follows:

variable = expression;

An expression represents a computation involving values, variables, and operators that, taking them together, evaluates to a value. For example, consider the following code:

int y = 1; // Assign 1 to variable y double radius = 1.0; // Assign 1.0 to variable radius int x = 5 * (3 / 2); // Assign the value of the expression to x x = y + 1; // Assign the addition of y and 1 to x double area = radius * radius * 3.14159; // Compute area

You can use a variable in an expression. A variable can also be used in both sides of the = operator. For example,

x = x + 1;

In this assignment statement, the result of x + 1 is assigned to x. If x is 1 before the statement is executed, then it becomes 2 after the statement is executed.

To assign a value to a variable, you must place the variable name to the left of the assign- ment operator. Thus, the following statement is wrong:

1 = x; // Wrong

Note In mathematics, x = 2 * x + 1 denotes an equation. However, in Java, x = 2 * x + 1 is an assignment statement that evaluates the expression 2 * x + 1 and assigns the result to x.

In Java, an assignment statement is essentially an expression that evaluates to the value to be assigned to the variable on the left side of the assignment operator. For this reason, an assignment statement is also known as an assignment expression. For example, the following statement is correct:

System.out.println(x = 1);

which is equivalent to

x = 1; System.out.println(x);

Point Check

Point Key

assignment statement assignment operator

expression

assignment expression

M02_LIAN1878_11_GE_C02.indd 64 12/29/17 3:48 PM

2.7 Named Constants 65

If a value is assigned to multiple variables, you can use the following syntax:

i = j = k = 1;

which is equivalent to

k = 1; j = k; i = j;

Note In an assignment statement, the data type of the variable on the left must be compatible with the data type of the value on the right. For example, int x = 1.0 would be ille- gal, because the data type of x is int. You cannot assign a double value (1.0) to an int variable without using type casting. Type casting will be introduced in Section 2.15.

2.6.1 Identify and fix the errors in the following code:

1 public class Test { 2 public static void main(String[] args) { 3 int i = j = k = 2; 4 System.out.println(i + " " + j + " " + k); 5 } 6 }

2.7 Named Constants A named constant is an identifier that represents a permanent value.

The value of a variable may change during the execution of a program, but a named constant, or simply constant, represents permanent data that never changes. A constant is also known as a final variable in Java. In our ComputeArea program, p is a constant. If you use it frequently, you don’t want to keep typing 3.14159; instead, you can declare a constant for p. Here is the syntax for declaring a constant:

final datatype CONSTANTNAME = value;

A constant must be declared and initialized in the same statement. The word final is a Java keyword for declaring a constant. By convention, all letters in a constant are in uppercase. For example, you can declare p as a constant and rewrite Listing 2.2, as in Listing 2.4.

Listing 2.4 ComputeAreaWithConstant.java

1 import java.util.Scanner; // Scanner is in the java.util package 2 3 public class ComputeAreaWithConstant { 4 public static void main(String[] args) { 5 final double PI = 3.14159; // Declare a constant 6 7 // Create a Scanner object 8 Scanner input = new Scanner(System.in); 9 10 // Prompt the user to enter a radius 11 System.out.print("Enter a number for radius: "); 12 double radius = input.nextDouble(); 13 14 // Compute area

Point Check

Point Key

constant

final keyword

M02_LIAN1878_11_GE_C02.indd 65 12/29/17 3:48 PM

66 Chapter 2 Elementary Programming

15 double area = radius * radius * PI; 16 17 // Display result 18 System.out.println("The area for the circle of radius " + 19 radius + " is " + area); 20 } 21 }

There are three benefits of using constants: (1) you don’t have to repeatedly type the same value if it is used multiple times; (2) if you have to change the constant value (e.g., from 3.14 to 3.14159 for PI), you need to change it only in a single location in the source code; and (3) a descriptive name for a constant makes the program easy to read.

2.7.1 What are the benefits of using constants? Declare an int constant SIZE with value 20.

2.8 Naming Conventions Sticking with the Java naming conventions makes your programs easy to read and avoids errors.

Make sure you choose descriptive names with straightforward meanings for the vari- ables, constants, classes, and methods in your program. As mentioned earlier, names are case sensitive. Listed below are the conventions for naming variables, methods, and classes.

■■ Use lowercase for variables and methods—for example, the variables radius and area, and the method print. If a name consists of several words, concatenate them into one, making the first word lowercase and capitalizing the first letter of each sub- sequent word—for example, the variable numberOfStudents. This naming style is known as the camelCase because the uppercase characters in the name resemble a camel’s humps.

■■ Capitalize the first letter of each word in a class name—for example, the class names ComputeArea and System.

■■ Capitalize every letter in a constant, and use underscores between words—for exam- ple, the constants PI and MAX_VALUE.

It is important to follow the naming conventions to make your programs easy to read.

Caution Do not choose class names that are already used in the Java library. For example, since the System class is defined in Java, you should not name your class System.

2.8.1 What are the naming conventions for class names, method names, constants, and variables? Which of the following items can be a constant, a method, a variable, or a class according to the Java naming conventions?

MAX_VALUE, Test, read, readDouble

2.8.2 Translate the following algorithm into Java code: Step 1: Declare a double variable named miles with an initial value 100.

Step 2: Declare a double constant named KILOMETERS_PER_MILE with value 1.609.

Step 3: Declare a double variable named kilometers, multiply miles and KILOMETERS_PER_MILE, and assign the result to kilometers.

benefits of constants

Point Check

Point Key

name variables and methods

name classes

name constants

name classes

Point Check

M02_LIAN1878_11_GE_C02.indd 66 12/29/17 3:48 PM

2.9 Numeric Data Types and Operations 67

Step 4: Display kilometers to the console.

What is kilometers after Step 4?

2.9 Numeric Data Types and Operations Java has six numeric types for integers and floating-point numbers with operators +, -, *, /, and %.

2.9.1 Numeric Types Every data type has a range of values. The compiler allocates memory space for each variable or constant according to its data type. Java provides eight primitive data types for numeric values, characters, and Boolean values. This section introduces numeric data types and operators.

Table 2.1 lists the six numeric data types, their ranges, and their storage sizes.

Point Key

byte type

short type

int type

long type

float type

double type

Name Range Storage Size

byte -27 to 27 -1 (-128 to 127) 8-bit signed short -215 to 215 -1 (-32768 to 32767) 16-bit signed int -231 to 231 -1 (-2147483648 to 2147483647) 32-bit signed long -263 to 263-1 64-bit signed

(i.e., -9223372036854775808 to 9223372036854775807) float Negative range: -3.4028235E + 38 to -1.4E -45 32-bit IEEE 754

Positive range: 1.4E -45 to 3.4028235E+38 double Negative range: -1.7976931348623157E+308 to -4.9E -324 64-bit IEEE 754

Positive range: 4.9E -324 to 1.7976931348623157E+308

tabLe 2.1 Numeric Data Types

Note IEEE 754 is a standard approved by the Institute of Electrical and Electronics Engineers for representing floating-point numbers on computers. The standard has been widely adopted. Java uses the 32-bit IEEE 754 for the float type and the 64-bit IEEE 754 for the double type. The IEEE 754 standard also defines special floating-point values, which are listed in Appendix E.

Java uses four types for integers: byte, short, int, and long. Choose the type that is most appropriate for your variable. For example, if you know an integer stored in a variable is within a range of a byte, declare the variable as a byte. For simplicity and consistency, we will use int for integers most of the time in this book.

Java uses two types for floating-point numbers: float and double. The double type is twice as big as float, so the double is known as double precision, and float as single preci- sion. Normally, you should use the double type, because it is more accurate than the float type.

2.9.2 Reading Numbers from the Keyboard You know how to use the nextDouble() method in the Scanner class to read a double value from the keyboard. You can also use the methods listed in Table 2.2 to read a number of the byte, short, int, long, and float type.

integer types

floating-point types

M02_LIAN1878_11_GE_C02.indd 67 12/29/17 3:48 PM

68 Chapter 2 Elementary Programming

Here are examples for reading values of various types from the keyboard:

1 Scanner input = new Scanner(System.in); 2 System.out.print("Enter a byte value: "); 3 byte byteValue = input.nextByte(); 4 5 System.out.print("Enter a short value: "); 6 short shortValue = input.nextShort(); 7 8 System.out.print("Enter an int value: "); 9 int intValue = input.nextInt(); 10 11 System.out.print("Enter a long value: "); 12 long longValue = input.nextLong(); 13 14 System.out.print("Enter a float value: "); 15 float floatValue = input.nextFloat();

If you enter a value with an incorrect range or format, a runtime error would occur. For example, if you enter a value 128 for line 3, an error would occur because 128 is out of range for a byte type integer.

2.9.3 Numeric Operators The operators for numeric data types include the standard arithmetic operators: addition (+), subtraction (–), multiplication (*), division (/), and remainder (%), as listed in Table 2.3. The operands are the values operated by an operator.operands

operators +, -, *, /, and %

Method Description

nextByte() reads an integer of the byte type.

nextShort() reads an integer of the short type.

nextInt() reads an integer of the int type.

nextLong() reads an integer of the long type.

nextFloat() reads a number of the float type.

nextDouble() reads a number of the double type.

tabLe 2.2 Methods for Scanner Objects

Name Meaning Example Result

+ Addition 34 + 1 35 - Subtraction 34.0 - 0.1 33.9 * Multiplication 300*30 9000

/ Division 1.0 / 2.0 0.5

% Remainder 20 % 3 2

tabLe 2.3 Numeric Operators

When both operands of a division are integers, the result of the division is the quotient and the fractional part is truncated. For example, 5 / 2 yields 2, not 2.5, and –5 / 2 yields –2, not –2.5. To perform a floating-point division, one of the operands must be a floating-point number. For example, 5.0 / 2 yields 2.5.

The % operator, known as remainder, yields the remainder after division. The operand on the left is the dividend, and the operand on the right is the divisor. Therefore, 7 % 3 yields 1, 3 % 7 yields 3, 12 % 4 yields 0, 26 % 8 yields 2, and 20 % 13 yields 7.

integer division

M02_LIAN1878_11_GE_C02.indd 68 12/29/17 3:48 PM

2.9 Numeric Data Types and Operations 69

Enter an integer for seconds: 500

500 seconds is 8 minutes and 20 seconds

4 12

12 0

8 26

24 2

Remainder

Quotient

Divisor Dividend13 20

13 7

7 3

0 3

3 7

6 1

The % operator is often used for positive integers, but it can also be used with negative inte- gers and floating-point values. The remainder is negative only if the dividend is negative. For example, -7 % 3 yields -1, -12 % 4 yields 0, -26 % -8 yields -2, and 20 % -13 yields 7.

Remainder is very useful in programming. For example, an even number % 2 is always 0 and a positive odd number % 2 is always 1. Thus, you can use this property to determine whether a number is even or odd. If today is Saturday, it will be Saturday again in 7 days. Suppose you and your friends are going to meet in 10 days. What will be the day in 10 days? You can find that the day is Tuesday using the following expression:

(6 + 10) % 7 is 2

After 10 days

Day 2 in a week is Tuesday Note: Day 0 in a week is Sunday

A week has 7 days Day 6 in a week is Saturday

The program in Listing 2.5 obtains minutes and remaining seconds from an amount of time in seconds. For example, 500 seconds contains 8 minutes and 20 seconds.

Listing 2.5 DisplayTime.java 1 import java.util.Scanner; 2 3 public class DisplayTime { 4 public static void main(String[] args) { 5 Scanner input = new Scanner(System.in); 6 // Prompt the user for input 7 System.out.print("Enter an integer for seconds: "); 8 int seconds = input.nextInt(); 9 10 int minutes = seconds / 60; // Find minutes in seconds 11 int remainingSeconds = seconds % 60; // Seconds remaining 12 System.out.println(seconds + " seconds is " + minutes + 13 " minutes and " + remainingSeconds + " seconds"); 14 } 15 }

import Scanner

create a Scanner

read an integer

divide remainder

line# seconds minutes remainingSeconds

8 500

10 8

11 20

The nextInt() method (line 8) reads an integer for seconds. Line 10 obtains the minutes using seconds / 60. Line 11 (seconds % 60) obtains the remaining seconds after taking away the minutes.

M02_LIAN1878_11_GE_C02.indd 69 12/29/17 3:48 PM

70 Chapter 2 Elementary Programming

The + and - operators can be both unary and binary. A unary operator has only one operand; a binary operator has two. For example, the – operator in –5 is a unary operator to negate number 5, whereas the – operator in 4 – 5 is a binary operator for subtracting 5 from 4.

2.9.4 Exponent Operations The Math.pow(a, b) method can be used to compute ab. The pow method is defined in the Math class in the Java API. You invoke the method using the syntax Math.pow(a, b) (e.g., Math.pow(2, 3)), which returns the result of ab (23). Here, a and b are parameters for the pow method and the numbers 2 and 3 are actual values used to invoke the method. For example,

System.out.println(Math.pow(2, 3)); // Displays 8.0 System.out.println(Math.pow(4, 0.5)); // Displays 2.0 System.out.println(Math.pow(2.5, 2)); // Displays 6.25 System.out.println(Math.pow(2.5, –2)); // Displays 0.16

Chapter 6 introduces more details on methods. For now, all you need to know is how to invoke the pow method to perform the exponent operation.

2.9.1 Find the largest and smallest byte, short, int, long, float, and double. Which of these data types requires the least amount of memory?

2.9.2 Show the result of the following remainders: 56 % 6 78 % -4 -34 % 5 -34 % -5 5 % 1 1 % 5

2.9.3 If today is Tuesday, what will be the day in 100 days? 2.9.4 What is the result of 25 / 4? How would you rewrite the expression if you wished

the result to be a floating-point number?

2.9.5 Show the result of the following code: System.out.println(2 * (5 / 2 + 5 / 2)); System.out.println(2 * 5 / 2 + 2 * 5 / 2); System.out.println(2 * (5 / 2)); System.out.println(2 * 5 / 2);

2.9.6 Are the following statements correct? If so, show the output. System.out.println("25 / 4 is " + 25 / 4); System.out.println("25 / 4.0 is " + 25 / 4.0); System.out.println("3 * 2 / 4 is " + 3 * 2 / 4); System.out.println("3.0 * 2 / 4 is " + 3.0 * 2 / 4);

2.9.7 Write a statement to display the result of 23.5. 2.9.8 Suppose m and r are integers. Write a Java expression for mr2 to obtain a floating-point

result.

2.10 Numeric Literals A literal is a constant value that appears directly in a program.

For example, 34 and 0.305 are literals in the following statements:

int numberOfYears = 34; double weight = 0.305;

unary operator

binary operator

Math.pow(a, b) method

Point Check

Point Key

literal

M02_LIAN1878_11_GE_C02.indd 70 12/29/17 3:48 PM

2.10 Numeric Literals 71

2.10.1 Integer Literals An integer literal can be assigned to an integer variable as long as it can fit into the variable. A compile error will occur if the literal is too large for the variable to hold. The statement byte b = 128, for example, will cause a compile error, because 128 cannot be stored in a variable of the byte type. (Note the range for a byte value is from –128 to 127.)

An integer literal is assumed to be of the int type, whose value is between -231 (-2147483648) and 231 -1 (2147483647). To denote an integer literal of the long type, append the letter L or l to it. For example, to write integer 2147483648 in a Java program, you have to write it as 2147483648L or 2147483648l, because 2147483648 exceeds the range for the int value. L is preferred because l (lowercase L) can easily be confused with 1 (the digit one).

Note By default, an integer literal is a decimal integer number. To denote a binary integer literal, use a leading 0b or 0B (zero B); to denote an octal integer literal, use a leading 0 (zero); and to denote a hexadecimal integer literal, use a leading 0x or 0X (zero X). For example,

System.out.println(0B1111); // Displays 15 System.out.println(07777); // Displays 4095 System.out.println(0XFFFF); // Displays 65535

Hexadecimal numbers, binary numbers, and octal numbers will be introduced in Appendix F.

Note To improve readability, Java allows you to use underscores between two digits in a number literal. For example, the following literals are correct.

long ssn = 232_45_4519; long creditCardNumber = 2324_4545_4519_3415L;

However, 45_ or _45 is incorrect. The underscore must be placed between two digits.

2.10.2 Floating-Point Literals Floating-point literals are written with a decimal point. By default, a floating-point literal is treated as a double type value. For example, 5.0 is considered a double value, not a float value. You can make a number a float by appending the letter f or F, and you can make a number a double by appending the letter d or D. For example, you can use 100.2f or 100.2F for a float number, and 100.2d or 100.2D for a double number.

Note The double type values are more accurate than the float type values. For example,

System.out.println("1.0 / 3.0 is " + 1.0 / 3.0);

displays 1.0 / 3.0 is 0 .3333333333333333

16 digits

System.out.println("1.0F / 3.0F is " + 1.0F / 3.0F);

displays 1.0F / 3.0F is 0.33333334

8 digits

A float value has 7–8 numbers of significant digits, and a double value has 15–17 numbers of significant digits.

binary, octal, and hex literals

underscores in numbers

suffix f or F suffix d or D

double vs. floatxs

M02_LIAN1878_11_GE_C02.indd 71 12/29/17 3:48 PM

72 Chapter 2 Elementary Programming

2.10.3 Scientific Notation Floating-point literals can be written in scientific notation in the form of a * 10b. For example, the scientific notation for 123.456 is 1.23456 * 102 and for 0.0123456 is 1.23456 * 10-2. A special syntax is used to write scientific notation numbers. For example, 1.23456 * 102 is written as 1.23456E2 or 1.23456E+2 and 1.23456 * 10-2 as 1.23456E-2. E (or e) repre- sents an exponent, and can be in either lowercase or uppercase.

Note The float and double types are used to represent numbers with a decimal point. Why are they called floating-point numbers? These numbers are stored in scientific notation internally. When a number such as 50.534 is converted into scientific notation, such as 5.0534E+1, its decimal point is moved (i.e., floated) to a new position.

2.10.1 How many accurate digits are stored in a float or double type variable?

2.10.2 Which of the following are correct literals for floating-point numbers? 12.3, 12.3e+2, 23.4e-2, –334.4, 20.5, 39F, 40D

2.10.3 Which of the following are the same as 52.534? 5.2534e+1, 0.52534e+2, 525.34e-1, 5.2534e+0

2.10.4 Which of the following are correct literals? 5_2534e+1, _2534, 5_2, 5_

2.11 Evaluating Expressions and Operator Precedence Java expressions are evaluated in the same way as arithmetic expressions.

Writing a numeric expression in Java involves a straightforward translation of an arithmetic expression using Java operators. For example, the arithmetic expression

3 + 4x 5

- 10(y - 5)(a + b + c)

x + 9¢ 4

x +

9 + x y

≤ can be translated into a Java expression as follows:

(3 + 4 * x) / 5 – 10 * (y - 5) * (a + b + c) / x + 9 * (4 / x + (9 + x) / y)

Although Java has its own way to evaluate an expression behind the scene, the result of a Java expression and its corresponding arithmetic expression is the same. Therefore, you can safely apply the arithmetic rule for evaluating a Java expression. Operators contained within pairs of parentheses are evaluated first. Parentheses can be nested, in which case the expression in the inner parentheses is evaluated first. When more than one operator is used in an expression, the following operator precedence rule is used to determine the order of evaluation:.

■■ Multiplication, division, and remainder operators are applied first. If an expression contains several multiplication, division, and remainder operators, they are applied from left to right.

■■ Addition and subtraction operators are applied last. If an expression contains several addition and subtraction operators, they are applied from left to right.

why called floating-point?

Point Check

Point Key

evaluating an expression

operator precedence rule

M02_LIAN1878_11_GE_C02.indd 72 12/29/17 3:48 PM

2.11 Evaluating Expressions and Operator Precedence 73

Here is an example of how an expression is evaluated:

3 + 4 * 4 + 5 * (4 + 3) - 1

3 + 4 * 4 + 5 * 7 – 1

3 + 16 + 5 * 7 – 1

3 + 16 + 35 – 1

19 + 35 – 1

54 – 1

(1) inside parentheses �rst

(2) multiplication

(3) multiplication

(4) addition

(5) addition

(6) subtraction

Listing 2.6 gives a program that converts a Fahrenheit degree to Celsius using the formula Celsius = (59 )(Fahrenheit - 32).

Listing 2.6 FahrenheitToCelsius.java 1 import java.util.Scanner; 2 3 public class FahrenheitToCelsius { 4 public static void main(String[] args) { 5 Scanner input = new Scanner(System.in); 6 7 System.out.print("Enter a degree in Fahrenheit: "); 8 double fahrenheit = input.nextDouble(); 9 10 // Convert Fahrenheit to Celsius 11 double celsius = (5.0 / 9) * (fahrenheit - 32); 12 System.out.println("Fahrenheit " + fahrenheit + " is " + 13 celsius + " in Celsius"); 14 } 15 }

divide

Enter a degree in Fahrenheit: 100

Fahrenheit 100.0 is 37.77777777777778 in Celsius

line# fahrenheit celsius 8 100

11 37.77777777777778

Be careful when applying division. Division of two integers yields an integer in Java. 59 is coded 5.0 / 9 instead of 5 / 9 in line 11, because 5 / 9 yields 0 in Java.

2.11.1 How would you write the following arithmetic expressions in Java?

a. 4

3(r + 34) - 9(a + bc) +

3 + d(2 + a) a + bd

b. 5.5 * (r + 2.5)2.5 + t

integer vs. floating-point division

Point Check

M02_LIAN1878_11_GE_C02.indd 73 12/29/17 3:48 PM

74 Chapter 2 Elementary Programming

2.12 Case Study: Displaying the Current Time You can invoke System.currentTimeMillis() to return the current time.

The problem is to develop a program that displays the current time in GMT (Greenwich Mean Time) in the format hour:minute:second, such as 13:19:8.

The currentTimeMillis method in the System class returns the current time in milli- seconds elapsed since the time midnight, January 1, 1970 GMT, as shown in Figure 2.2. This time is known as the UNIX epoch. The epoch is the point when time starts, and 1970 was the year when the UNIX operating system was formally introduced.

Point Key

VideoNote

Use operators / and %

currentTimeMillis UNIX epoch

Figure 2.2 The System.currentTimeMillis() returns the number of milliseconds since the UNIX epoch.

UNIX epoch 01-01-1970

00:00:00 GMT

Elapsed time

Current time System.currentTimeMillis()

Time

You can use this method to obtain the current time, then compute the current second, min- ute, and hour as follows:

1. Obtain the total milliseconds since midnight, January 1, 1970, in totalMilliseconds by invoking System.currentTimeMillis() (e.g., 1203183068328 milliseconds).

2. Obtain the total seconds totalSeconds by dividing totalMilliseconds by 1000 (e.g., 1203183068328 milliseconds / 1000 = 1203183068 seconds).

3. Compute the current second from totalSeconds % 60 (e.g., 1203183068 seconds % 60 = 8, which is the current second).

4. Obtain the total minutes totalMinutes by dividing totalSeconds by 60 (e.g., 1203183068 seconds / 60 = 20053051 minutes).

5. Compute the current minute from totalMinutes % 60 (e.g., 20053051 minutes % 60 = 31, which is the current minute).

6. Obtain the total hours totalHours by dividing totalMinutes by 60 (e.g., 20053051 minutes / 60 = 334217 hours).

7. Compute the current hour from totalHours % 24 (e.g., 334217 hours % 24 = 17, which is the current hour).

Listing 2.7 gives the complete program.

Listing 2.7 ShowCurrentTime.java 1 public class ShowCurrentTime { 2 public static void main(String[] args) { 3 // Obtain the total milliseconds since midnight, Jan 1, 1970 4 long totalMilliseconds = System.currentTimeMillis(); 5 6 // Obtain the total seconds since midnight, Jan 1, 1970 7 long totalSeconds = totalMilliseconds / 1000; 8 9 // Compute the current second in the minute in the hour 10 long currentSecond = totalSeconds % 60; 11

currentSecond

totalSeconds

totalMilliseconds

M02_LIAN1878_11_GE_C02.indd 74 12/29/17 3:48 PM

2.12 Case Study: Displaying the Current Time 75

12 // Obtain the total minutes 13 long totalMinutes = totalSeconds / 60; 14 15 // Compute the current minute in the hour 16 long currentMinute = totalMinutes % 60; 17 18 // Obtain the total hours 19 long totalHours = totalMinutes / 60; 20 21 // Compute the current hour 22 long currentHour = totalHours % 24; 23 24 // Display results 25 System.out.println("Current time is " + currentHour + ":" 26 + currentMinute + ":" + currentSecond + " GMT"); 27 } 28 }

totalMinutes

currentMinute

totalHours

currentHour

display output

Current time is 17:31:8 GMT

Line 4 invokes System.currentTimeMillis() to obtain the current time in milliseconds as a long value. Thus, all the variables are declared as the long type in this program. The seconds, minutes, and hours are extracted from the current time using the / and % operators (lines 6–22).

line#

variables

4 7 10 13 16 19 22

totalMilliseconds 1203183068328

totalSeconds 1203183068

currentSecond 8

totalMinutes 20053051

currentMinute 31

totalHours 334217

currentHour 17

In the sample run, a single digit 8 is displayed for the second. The desirable output would be 08. This can be fixed by using a method that formats a single digit with a prefix 0 (see Programming Exercise 6.37).

The hour displayed in this program is in GMT. Programming Exercise 2.8 enables to display the hour in any time zone.

Java also provides the System.nanoTime() method that returns the elapse time in nano- seconds. nanoTime() is more precise and accurate than currentTimeMillis().

2.12.1 How do you obtain the current second, minute, and hour?

nanoTime

Point Check

M02_LIAN1878_11_GE_C02.indd 75 12/29/17 3:48 PM

76 Chapter 2 Elementary Programming

2.13 Augmented Assignment Operators The operators +, -, *, /, and % can be combined with the assignment operator to form augmented operators.

Very often, the current value of a variable is used, modified, then reassigned back to the same variable. For example, the following statement increases the variable count by 1:

count = count + 1;

Java allows you to combine assignment and addition operators using an augmented (or compound) assignment operator. For example, the preceding statement can be written as

count += 1;

The += is called the addition assignment operator. Table 2.4 shows other augmented assign- ment operators.

Point Key

addition assignment operator

Operator Name Example Equivalent

+= Addition assignment i += 8 i = i + 8

-= Subtraction assignment i -= 8 i = i – 8

*= Multiplication assignment i *= 8 i = i * 8

/= Division assignment i /= 8 i = i / 8

%= Remainder assignment i %= 8 i = i % 8

tabLe 2.4 Augmented Assignment Operators

The augmented assignment operator is performed last after all the other operators in the expression are evaluated. For example,

x /= 4 + 5.5 * 1.5;

is same as

x = x / (4 + 5.5 * 1.5);

Caution There are no spaces in the augmented assignment operators. For example, + = should be +=.

Note Like the assignment operator (=), the operators (+=, -=, *=, /=, and %=) can be used to form an assignment statement as well as an expression. For example, in the follow- ing code, x += 2 is a statement in the first line, and an expression in the second line:

x += 2; // Statement System.out.println(x += 2); // Expression

2.13.1 Show the output of the following code:

double a = 6.5; a += a + 1;

Point Check

M02_LIAN1878_11_GE_C02.indd 76 12/29/17 3:48 PM

2.14 Increment and Decrement Operators 77

System.out.println(a); a = 6; a /= 2; System.out.println(a);

2.14 Increment and Decrement Operators The increment operator (+ +) and decrement operator (- -) are for incrementing and decrementing a variable by 1.

The ++ and — — are two shorthand operators for incrementing and decrementing a variable by 1. These are handy because that’s often how much the value needs to be changed in many program- ming tasks. For example, the following code increments i by 1 and decrements j by 1.

int i = 3, j = 3; i++; // i becomes 4 j— —; // j becomes 2

i++ is pronounced as "i plus plus" and i—— as "i minus minus." These operators are known as postfix increment (or postincrement) and postfix decrement (or postdecrement), because the operators ++ and —— are placed after the variable. These operators can also be placed before the variable. For example,

int i = 3, j = 3; ++i; // i becomes 4 — —j; // j becomes 2

++i increments i by 1 and ——j decrements j by 1. These operators are known as prefix increment (or preincrement) and prefix decrement (or predecrement).

As you see, the effect of i++ and ++i or i—— and ——i are the same in the preceding exam- ples. However, their effects are different when they are used in statements that do more than just increment and decrement. Table 2.5 describes their differences and gives examples.

Point Key

increment operator (+ +) decrement operator (- -)

postincrement

postdecrement

preincrement

predecrement

Operator Name Description Example (assume i = 1)

++var preincrement Increment var by 1, and use the new var value in the statement

int j = ++i; // j is 2, i is 2

var++ postincrement Increment var by 1, but use the original var value in the statement

int j = i++; // j is 1, i is 2

——var predecrement Decrement var by 1, and use the new var value in the statement

int j = — —i; // j is 0, i is 0

var—— postdecrement Decrement var by 1, and use the original var value in the statement

int j = i— —; // j is 1, i is 0

tabLe 2.5 Increment and Decrement Operators

Here are additional examples to illustrate the differences between the prefix form of ++ (or ——) and the postfix form of ++ (or ——). Consider the following code:

int i = 10; int newNum = 10 * i++;

Same effect as int newNum = 10 * i; i = i + 1;System.out.print("i is " + i

+ ", newNum is " + newNum);

i is 11, newNum is 100

M02_LIAN1878_11_GE_C02.indd 77 12/29/17 3:48 PM

78 Chapter 2 Elementary Programming

In this case, i is incremented by 1, then the old value of i is used in the multiplication. Thus, newNum becomes 100. If i++ is replaced by ++i, then it becomes as follows:

int i = 10; int newNum = 10 * (++i);

Same effect as i = i + 1; int newNum = 10 * i;

System.out.print("i is " + i + ", newNum is " + newNum);

i is 11, newNum is 110

i is incremented by 1, and the new value of i is used in the multiplication. Thus, newNum becomes 110.

Here is another example:

double x = 1.0; double y = 5.0; double z = x–– + (++y);

After all three lines are executed, y becomes 6.0, z becomes 7.0, and x becomes 0.0.

Operands are evaluated from left to right in Java. The left-hand operand of a binary operator is evaluated before any part of the right-hand operand is evaluated. This rule takes precedence over any other rules that govern expressions. Here is an example:

int i = 1; int k = ++i + i * 3;

++i is evaluated and returns 2. When evaluating i * 3, i is now 2. Therefore, k becomes 8.

Tip Using increment and decrement operators makes expressions short, but it also makes them complex and difficult to read. Avoid using these operators in expressions that modify multiple variables or the same variable multiple times, such as this one: int k = ++i + i * 3.

2.14.1 Which of these statements are true?

a. Any expression can be used as a statement.

b. The expression x++ can be used as a statement.

c. The statement x = x + 5 is also an expression.

d. The statement x = y = x = 0 is illegal.

2.14.2 Show the output of the following code:

int a = 6; int b = a++; System.out.println(a); System.out.println(b); a = 6; b = ++a; System.out.println(a); System.out.println(b);

Point Check

M02_LIAN1878_11_GE_C02.indd 78 12/29/17 3:48 PM

2.15 Numeric Type Conversions 79

2.15 Numeric Type Conversions Floating-point numbers can be converted into integers using explicit casting.

Can you perform binary operations with two operands of different types? Yes. If an integer and a floating-point number are involved in a binary operation, Java automatically converts the integer to a floating-point value. Therefore, 3 * 4.5 is the same as 3.0 * 4.5.

You can always assign a value to a numeric variable whose type supports a larger range of values; thus, for instance, you can assign a long value to a float variable. You cannot, how- ever, assign a value to a variable of a type with a smaller range unless you use type casting. Casting is an operation that converts a value of one data type into a value of another data type. Casting a type with a small range to a type with a larger range is known as widening a type. Casting a type with a large range to a type with a smaller range is known as narrowing a type. Java will automatically widen a type, but you must narrow a type explicitly.

The syntax for casting a type is to specify the target type in parentheses, followed by the variable’s name or the value to be cast. For example, the following statement

System.out.println((int)1.7);

displays 1. When a double value is cast into an int value, the fractional part is truncated. The following statement

System.out.println((double)1 / 2);

displays 0.5, because 1 is cast to 1.0 first, then 1.0 is divided by 2. However, the statement

System.out.println(1 / 2);

displays 0, because 1 and 2 are both integers and the resulting value should also be an integer.

Caution Casting is necessary if you are assigning a value to a variable of a smaller type range, such as assigning a double value to an int variable. A compile error will occur if casting is not used in situations of this kind. However, be careful when using casting, as loss of information might lead to inaccurate results.

Note Casting does not change the variable being cast. For example, d is not changed after casting in the following code:

double d = 4.5; int i = (int)d; // i becomes 4, but d is still 4.5

Note In Java, an augmented expression of the form x1 op= x2 is implemented as x1 = (T) (x1 op x2), where T is the type for x1. Therefore, the following code is correct:

int sum = 0; sum += 4.5; // sum becomes 4 after this statement sum += 4.5 is equivalent to sum = (int)(sum + 4.5).

Note To assign a variable of the int type to a variable of the short or byte type, explicit casting must be used. For example, the following statements have a compile error:

int i = 1; byte b = i; // Error because explicit casting is required

Point Key

casting

widening a type narrowing a type

possible loss of precision

casting in an augmented expression

M02_LIAN1878_11_GE_C02.indd 79 12/29/17 3:48 PM

80 Chapter 2 Elementary Programming

However, so long as the integer literal is within the permissible range of the target vari- able, explicit casting is not needed to assign an integer literal to a variable of the short or byte type (see Section 2.10, Numeric Literals).

The program in Listing 2.8 displays the sales tax with two digits after the decimal point.

Listing 2.8 SalesTax.java 1 import java.util.Scanner; 2 3 public class SalesTax { 4 public static void main(String[] args) { 5 Scanner input = new Scanner(System.in); 6 7 System.out.print("Enter purchase amount: "); 8 double purchaseAmount = input.nextDouble(); 9 10 double tax = purchaseAmount * 0.06; 11 System.out.println("Sales tax is $" + (int)(tax * 100) / 100.0); 12 } 13 }

casting

Enter purchase amount: 197.55

Sales tax is $11.85

line# purchaseAmount tax Output

8 197.55

10 11.853

11 11.85

Using the input in the sample run, the variable purchaseAmount is 197.55 (line 8). The sales tax is 6% of the purchase, so the tax is evaluated as 11.853 (line 10). Note

tax * 100 is 1185.3 (int)(tax * 100) is 1185 (int)(tax * 100) / 100.0 is 11.85

Thus, the statement in line 11 displays the tax 11.85 with two digits after the decimal point. Note the expression (int)(tax * 100) / 100.0 rounds down tax to two decimal places. If tax is 3.456, (int)(tax * 100) / 100.0 would be 3.45. Can it be rounded up to two decimal places? Note any double value x can be rounded up to an integer using (int)(x + 0.5). Thus, tax can be rounded up to two decimal places using (int)(tax * 100 + 0.5) / 100.0.

2.15.1 Can different types of numeric values be used together in a computation?

2.15.2 What does an explicit casting from a double to an int do with the fractional part of the double value? Does casting change the variable being cast?

2.15.3 Show the following output:

float f = 12.5F; int i = (int)f; System.out.println("f is " + f); System.out.println("i is " + i);

formatting numbers

Point Check

M02_LIAN1878_11_GE_C02.indd 80 12/29/17 3:48 PM

2.16 Software Development Process 81

2.15.4 If you change (int)(tax * 100) / 100.0 to (int)(tax * 100) / 100 in line 11 in Listing 2.8, what will be the output for the input purchase amount of 197.556?

2.15.5 Show the output of the following code:

double amount = 5; System.out.println(amount / 2); System.out.println(5 / 2);

2.15.6 Write an expression that rounds up a double value in variable d to an integer.

2.16 Software Development Process The software development life cycle is a multistage process that includes requirements specification, analysis, design, implementation, testing, deployment, and maintenance.

Developing a software product is an engineering process. Software products, no matter how large or how small, have the same life cycle: requirements specification, analysis, design, implementation, testing, deployment, and maintenance, as shown in Figure 2.3.

Point Key

VideoNote

Software development process

Figure 2.3 At any stage of the software development life cycle, it may be necessary to go back to a previous stage to correct errors or deal with other issues that might prevent the software from functioning as expected.

Requirements Speci�cation

System Analysis

System Design

Testing

Input, Process, Output IPO

Implementation

Maintenance

Deployment

Requirements specification is a formal process that seeks to understand the problem the software will address, and to document in detail what the software system needs to do. This phase involves close interaction between users and developers. Most of the examples in this book are simple, and their requirements are clearly stated. In the real world, however, problems are not always well defined. Developers need to work closely with their customers (the indi- viduals or organizations that will use the software) and study the problem carefully to identify what the software needs to do.

System analysis seeks to analyze the data flow and to identify the system’s input and output. When you perform analysis, it helps to identify what the output is first, then figure out what input data you need in order to produce the output.

requirements specification

system analysis

M02_LIAN1878_11_GE_C02.indd 81 12/29/17 3:48 PM

82 Chapter 2 Elementary Programming

System design is to design a process for obtaining the output from the input. This phase involves the use of many levels of abstraction to break down the problem into manageable components and design strategies for implementing each component. You can view each com- ponent as a subsystem that performs a specific function of the system. The essence of system analysis and design is input, process, and output (IPO).

Implementation involves translating the system design into programs. Separate programs are written for each component then integrated to work together. This phase requires the use of a programming language such as Java. The implementation involves coding, self-testing, and debugging (that is, finding errors, called bugs, in the code).

Testing ensures the code meets the requirements specification and weeds out bugs. An independent team of software engineers not involved in the design and implementation of the product usually conducts such testing.

Deployment makes the software available for use. Depending on the type of software, it may be installed on each user’s machine, or installed on a server accessible on the Internet.

Maintenance is concerned with updating and improving the product. A software product must continue to perform and improve in an ever-evolving environment. This requires periodic upgrades of the product to fix newly discovered bugs and incorporate changes.

To see the software development process in action, we will now create a program that computes loan payments. The loan can be a car loan, a student loan, or a home mortgage loan. For an introductory programming course, we focus on requirements specification, analysis, design, implementation, and testing.

Stage 1: Requirements Specification

The program must satisfy the following requirements:

■■ It must let the user enter the interest rate, the loan amount, and the number of years for which payments will be made.

■■ It must compute and display the monthly payment and total payment amounts.

Stage 2: System Analysis

The output is the monthly payment and total payment, which can be obtained using the fol- lowing formulas:

monthlyPayment = loanAmount * monthlyInterestRate

1 - 1

(1 + monthlyInterestRate)numberOfYears * 12

totalPayment = monthlyPayment * numberOfYears * 12

Therefore, the input needed for the program is the monthly interest rate, the length of the loan in years, and the loan amount.

Note The requirements specification says the user must enter the annual interest rate, the loan amount, and the number of years for which payments will be made. During analysis, however, it is possible you may discover that input is not sufficient or some values are unnecessary for the output. If this happens, you can go back and modify the require- ments specification.

Note In the real world, you will work with customers from all walks of life. You may develop software for chemists, physicists, engineers, economists, and psychologists, and of course you will not have (or need) complete knowledge of all these fields. Therefore,

system design

IPO implementation

testing

deployment

maintenance

VideoNote

Compute loan payments

M02_LIAN1878_11_GE_C02.indd 82 12/29/17 3:48 PM

2.16 Software Development Process 83

you don’t have to know how formulas are derived, but given the monthly interest rate, the number of years, and the loan amount, you can compute the monthly payment in this program. You will, however, need to communicate with customers and understand how a mathematical model works for the system.

Stage 3: System Design

During system design, you identify the steps in the program.

Step 3.1. Prompt the user to enter the annual interest rate, the number of years, and the loan amount.

(The interest rate is commonly expressed as a percentage of the principal for a period of one year. This is known as the annual interest rate.)

Step 3.2. The input for the annual interest rate is a number in percent format, such as 4.5%. The program needs to convert it into a decimal by dividing it by 100. To obtain the monthly interest rate from the annual interest rate, divide it by 12, since a year has 12 months. Thus, to obtain the monthly interest rate in decimal format, you need to divide the annual interest rate in percentage by 1200. For example, if the annual interest rate is 4.5%, then the monthly interest rate is 4.5/1200 = 0.00375.

Step 3.3. Compute the monthly payment using the preceding formula.

Step 3.4. Compute the total payment, which is the monthly payment multiplied by 12 and multiplied by the number of years.

Step 3.5. Display the monthly payment and total payment.

Stage 4: Implementation

Implementation is also known as coding (writing the code). In the formula, you have to com- pute (1 + monthlyInterestRate)numberOfYears*12, which can be obtained using Math. pow(1 + monthlyInterestRate, numberOfYears * 12).

Listing 2.9 gives the complete program.

Listing 2.9 ComputeLoan.java 1 import java.util.Scanner; 2 3 public class ComputeLoan { 4 public static void main(String[] args) { 5 // Create a Scanner 6 Scanner input = new Scanner(System.in); 7 8 // Enter annual interest rate in percentage, e.g., 7.25 9 System.out.print("Enter annual interest rate, e.g., 7.25: "); 10 double annualInterestRate = input.nextDouble(); 11 12 // Obtain monthly interest rate 13 double monthlyInterestRate = annualInterestRate / 1200; 14 15 // Enter number of years 16 System.out.print( 17 "Enter number of years as an integer, e.g., 5: "); 18 int numberOfYears = input.nextInt(); 19 20 // Enter loan amount 21 System.out.print("Enter loan amount, e.g., 120000.95: ");

Math.pow(a, b) method

enter years

enter interest rate

create a Scanner

import class

M02_LIAN1878_11_GE_C02.indd 83 12/29/17 3:48 PM

84 Chapter 2 Elementary Programming

22 double loanAmount = input.nextDouble(); 23 24 // Calculate payment 25 double monthlyPayment = loanAmount * monthlyInterestRate / (1 26 - 1 / Math.pow(1 + monthlyInterestRate, numberOfYears * 12)); 27 double totalPayment = monthlyPayment * numberOfYears * 12; 28 29 // Display results 30 System.out.println("The monthly payment is $" + 31 (int)(monthlyPayment * 100) / 100.0); 32 System.out.println("The total payment is $" + 33 (int)(totalPayment * 100) / 100.0); 34 } 35 }

monthlyPayment

enter loan amount

totalPayment

casting

Enter annual interest rate, for example, 7.25: 5.75

Enter number of years as an integer, for example, 5: 15

Enter loan amount, for example, 120000.95: 250000

The monthly payment is $2076.02

The total payment is $373684.53

line# 10 13 18 22 25 27

variables

annualInterestRate 5.75

monthlyInterestRate 0.0047916666666

numberOfYears 15

loanAmount 250000

monthlyPayment 2076.0252175

totalPayment 373684.539

Line 10 reads the annual interest rate, which is converted into the monthly interest rate in line 13.

Choose the most appropriate data type for the variable. For example, numberOfYears is best declared as an int (line 18), although it could be declared as a long, float, or double. Note byte might be the most appropriate for numberOfYears. For simplicity, however, the examples in this book will use int for integer and double for floating-point values.

The formula for computing the monthly payment is translated into Java code in lines 25–27. Casting is used in lines 31 and 33 to obtain a new monthlyPayment and totalPayment

with two digits after the decimal points. The program uses the Scanner class, imported in line 1. The program also uses the Math

class, and you might be wondering why that class isn’t imported into the program. The Math class is in the java.lang package, and all classes in the java.lang package are implicitly imported. Therefore, you don’t need to explicitly import the Math class.

Stage 5: Testing

After the program is implemented, test it with some sample input data and verify whether the output is correct. Some of the problems may involve many cases, as you will see in later chap- ters. For these types of problems, you need to design test data that cover all cases.

java.lang package

M02_LIAN1878_11_GE_C02.indd 84 12/29/17 3:48 PM

2.17 Case Study: Counting Monetary Units 85

Tip The system design phase in this example identified several steps. It is a good approach to code and test these steps incrementally by adding them one at a time. This approach makes it much easier to pinpoint problems and debug the program.

2.16.1 How would you write the following arithmetic expression?

-b + 2b2 - 4ac 2a

2.17 Case Study: Counting Monetary Units This section presents a program that breaks a large amount of money into smaller units.

Suppose you want to develop a program that changes a given amount of money into smaller monetary units. The program lets the user enter an amount as a double value representing a total in dollars and cents, and outputs a report listing the monetary equivalent in the maximum number of dollars, quarters, dimes, nickels, and pennies, in this order, to result in the minimum number of coins.

Here are the steps in developing the program:

1. Prompt the user to enter the amount as a decimal number, such as 11.56.

2. Convert the amount (e.g., 11.56) into cents (1156).

3. Divide the cents by 100 to find the number of dollars. Obtain the remaining cents using the cents remainder 100.

4. Divide the remaining cents by 25 to find the number of quarters. Obtain the remaining cents using the remaining cents remainder 25.

5. Divide the remaining cents by 10 to find the number of dimes. Obtain the remaining cents using the remaining cents remainder 10.

6. Divide the remaining cents by 5 to find the number of nickels. Obtain the remaining cents using the remaining cents remainder 5.

7. The remaining cents are the pennies.

8. Display the result.

The complete program is given in Listing 2.10.

Listing 2.10 ComputeChange.java 1 import java.util.Scanner; 2 3 public class ComputeChange { 4 public static void main(String[] args) { 5 // Create a Scanner 6 Scanner input = new Scanner(System.in); 7 8 // Receive the amount 9 System.out.print( 10 "Enter an amount in double, for example 11.56: "); 11 double amount = input.nextDouble(); 12 13 int remainingAmount = (int)(amount * 100); 14 15 // Find the number of one dollars

incremental code and test

Point Check

Point Key

enter input

import class

M02_LIAN1878_11_GE_C02.indd 85 12/29/17 3:48 PM

86 Chapter 2 Elementary Programming

16 int numberOfOneDollars = remainingAmount / 100; 17 remainingAmount = remainingAmount % 100; 18 19 // Find the number of quarters in the remaining amount 20 int numberOfQuarters = remainingAmount / 25; 21 remainingAmount = remainingAmount % 25; 22 23 // Find the number of dimes in the remaining amount 24 int numberOfDimes = remainingAmount / 10; 25 remainingAmount = remainingAmount % 10; 26 27 // Find the number of nickels in the remaining amount 28 int numberOfNickels = remainingAmount / 5; 29 remainingAmount = remainingAmount % 5; 30 31 // Find the number of pennies in the remaining amount 32 int numberOfPennies = remainingAmount; 33 34 // Display results 35 System.out.println("Your amount " + amount + " consists of"); 36 System.out.println(" " + numberOfOneDollars + " dollars"); 37 System.out.println(" " + numberOfQuarters + " quarters "); 38 System.out.println(" " + numberOfDimes + " dimes"); 39 System.out.println(" " + numberOfNickels + " nickels"); 40 System.out.println(" " + numberOfPennies + " pennies"); 41 } 42 }

pennies

nickels

dimes

quarters

dollars

output

Enter an amount in double, for example, 11.56: 11.56

Your amount 11.56 consists of

11 dollars 2 quarters 0 dimes 1 nickels 1 pennies

line#

variables

11 13 16 17 20 21 24 25 28 29 32

amount 11.56

remainingAmount 1156 56 6 6 1

numberOfOneDollars 11

numberOfQuarters 2

numberOfDimes 0

numberOfNickels 1

numberOfPennies 1

The variable amount stores the amount entered from the console (line 11). This variable is not changed, because the amount has to be used at the end of the program to display the results. The program introduces the variable remainingAmount (line 13) to store the chang- ing remaining amount.

M02_LIAN1878_11_GE_C02.indd 86 12/29/17 3:48 PM

2.18 Common Errors and Pitfalls 87

The variable amount is a double decimal representing dollars and cents. It is converted to an int variable remainingAmount, which represents all the cents. For instance, if amount is 11.56, then the initial remainingAmount is 1156. The division operator yields the integer part of the division, so 1156 / 100 is 11. The remainder operator obtains the remainder of the division, so 1156 % 100 is 56.

The program extracts the maximum number of singles from the remaining amount and obtains a new remaining amount in the variable remainingAmount (lines 16–17). It then extracts the maximum number of quarters from remainingAmount and obtains a new remainingAmount (lines 20–21). Continuing the same process, the program finds the maxi- mum number of dimes, nickels, and pennies in the remaining amount.

One serious problem with this example is the possible loss of precision when casting a double amount to an int remainingAmount. This could lead to an inaccurate result. If you try to enter the amount 10.03, 10.03 * 100 becomes 1002.9999999999999. You will find that the program displays 10 dollars and 2 pennies. To fix the problem, enter the amount as an integer value representing cents (see Programming Exercise 2.22).

2.17.1 Show the output of Listing 2.10 with the input value 1.99.

2.18 Common Errors and Pitfalls Common elementary programming errors often involve undeclared variables, uninitialized variables, integer overflow, unintended integer division, and round-off errors.

Common Error 1: Undeclared/Uninitialized Variables and Unused Variables

A variable must be declared with a type and assigned a value before using it. A common error is not declaring a variable or initializing a variable. Consider the following code:

double interestRate = 0.05; double interest = interestrate * 45;

This code is wrong, because interestRate is assigned a value 0.05; but interestrate has not been declared and initialized. Java is case sensitive, so it considers interestRate and interestrate to be two different variables.

If a variable is declared, but not used in the program, it might be a potential programming error. Therefore, you should remove the unused variable from your program. For example, in the following code, taxRate is never used. It should be removed from the code.

double interestRate = 0.05; double taxRate = 0.05; double interest = interestRate * 45; System.out.println("Interest is " + interest);

If you use an IDE such as Eclipse and NetBeans, you will receive a warning on unused variables.

Common Error 2: Integer Overflow

Numbers are stored with a limited numbers of digits. When a variable is assigned a value that is too large (in size) to be stored, it causes overflow. For example, executing the following statement causes overflow, because the largest value that can be stored in a variable of the int type is 2147483647. 2147483648 will be too large for an int value:

int value = 2147483647 + 1; // value will actually be -2147483648

loss of precision

Point Check

Point Key

what is overflow?

M02_LIAN1878_11_GE_C02.indd 87 12/29/17 3:48 PM

88 Chapter 2 Elementary Programming

Likewise, executing the following statement also causes overflow, because the smallest value that can be stored in a variable of the int type is -2147483648. -2147483649 is too large in size to be stored in an int variable.

int value = –2147483648 – 1; // value will actually be 2147483647

Java does not report warnings or errors on overflow, so be careful when working with integers close to the maximum or minimum range of a given type.

When a floating-point number is too small (i.e., too close to zero) to be stored, it causes underflow. Java approximates it to zero, so normally you don’t need to be concerned about underflow.

Common Error 3: Round-off Errors

A round-off error, also called a rounding error, is the difference between the calculated approximation of a number and its exact mathematical value. For example, 1/3 is approxi- mately 0.333 if you keep three decimal places, and is 0.3333333 if you keep seven decimal places. Since the number of digits that can be stored in a variable is limited, round-off errors are inevitable. Calculations involving floating-point numbers are approximated because these numbers are not stored with complete accuracy. For example,

System.out.println(1.0 - 0.1 - 0.1 - 0.1 - 0.1 - 0.1);

displays 0.5000000000000001, not 0.5, and

System.out.println(1.0 - 0.9);

displays 0.09999999999999998, not 0.1. Integers are stored precisely. Therefore, calcula- tions with integers yield a precise integer result.

Common Error 4: Unintended Integer Division

Java uses the same divide operator, namely /, to perform both integer and floating-point divi- sion. When two operands are integers, the / operator performs an integer division. The result of the operation is an integer. The fractional part is truncated. To force two integers to perform a floating-point division, make one of the integers into a floating-point number. For example, the code in (a) displays that average as 1 and the code in (b) displays that average as 1.5.

what is underflow?

floating-point approximation

int number1 = 1; int number2 = 2; double average = (number1 + number2) / 2; System.out.println(average);

int number1 = 1; int number2 = 2; double average = (number1 + number2) / 2.0; System.out.println(average);

(a) (b)

Common Pitfall 1: Redundant Input Objects

New programmers often write the code to create multiple input objects for each input. For example, the following code reads an integer and a double value:

Scanner input = new Scanner(System.in); System.out.print("Enter an integer: "); int v1 = input.nextInt();

Scanner input1 = new Scanner(System.in); BAD CODE System.out.print("Enter a double value: "); double v2 = input1.nextDouble();

M02_LIAN1878_11_GE_C02.indd 88 12/29/17 3:48 PM

Chapter Summary 89

The code is not good. It creates two input objects unnecessarily and may lead to some subtle errors. You should rewrite the code as follows:

Scanner input = new Scanner(System.in); GOOD CODE System.out.print("Enter an integer: "); int v1 = input.nextInt(); System.out.print("Enter a double value: "); double v2 = input.nextDouble();

2.18.1 Can you declare a variable as int and later redeclare it as double?

2.18.2 What is an integer overflow? Can floating-point operations cause overflow?

2.18.3 Will overflow cause a runtime error?

2.18.4 What is a round-off error? Can integer operations cause round-off errors? Can floating-point operations cause round-off errors?

Point Check

Key Terms algorithm 56 assignment operator (=) 64 assignment statement 64 byte type 67 casting 79 constant 65 data type 57 declare variables 57 decrement operator (– –) 77 double type 67 expression 64 final keyword 65 float type 67 floating-point number 57 identifier 62 increment operator (++) 77 incremental code and testing 85 int type 67 IPO 61 literal 70 long type 67

narrowing a type 79 operand 68 operator 68 overflow 87 postdecrement 77 postincrement 77 predecrement 77 preincrement 77 primitive data type 57 pseudocode 56 requirements specification 81 scope of a variable 63 short type 67 specific import 60 system analysis 81 system design 82 underflow 88 UNIX epoch 74 variable 57 widening a type 79 wildcard import 60

ChapTer summary 1. Identifiers are names for naming elements such as variables, constants, methods, classes,

and packages in a program.

2. An identifier is a sequence of characters that consists of letters, digits, underscores (_), and dollar signs ($). An identifier must start with a letter or an underscore. It cannot start with a digit. An identifier cannot be a reserved word. An identifier can be of any length.

3. Variables are used to store data in a program. To declare a variable is to tell the compiler what type of data a variable can hold.

M02_LIAN1878_11_GE_C02.indd 89 12/29/17 3:48 PM

90 Chapter 2 Elementary Programming

4. There are two types of import statements: specific import and wildcard import. The specific import specifies a single class in the import statement. The wildcard import imports all the classes in a package.

5. In Java, the equal sign (=) is used as the assignment operator.

6. A variable declared in a method must be assigned a value before it can be used.

7. A named constant (or simply a constant) represents permanent data that never changes.

8. A named constant is declared by using the keyword final.

9. Java provides four integer types (byte, short, int, and long) that represent integers of four different sizes.

10. Java provides two floating-point types (float and double) that represent floating-point numbers of two different precisions.

11. Java provides operators that perform numeric operations: + (addition), – (subtraction), * (multiplication), / (division), and % (remainder).

12. Integer arithmetic (/) yields an integer result.

13. The numeric operators in a Java expression are applied the same way as in an arithmetic expression.

14. Java provides the augmented assignment operators += (addition assignment), –= (sub- traction assignment), *= (multiplication assignment), /= (division assignment), and %= (remainder assignment).

15. The increment operator (++) and the decrement operator (––) increment or decrement a variable by 1.

16. When evaluating an expression with values of mixed types, Java automatically converts the operands to appropriate types.

17. You can explicitly convert a value from one type to another using the (type)value notation.

18. Casting a variable of a type with a small range to a type with a larger range is known as widening a type.

19. Casting a variable of a type with a large range to a type with a smaller range is known as narrowing a type.

20. Widening a type can be performed automatically without explicit casting. Narrowing a type must be performed explicitly.

21. In computer science, midnight of January 1, 1970, is known as the UNIX epoch.

Quiz Answer the quiz for this chapter online at the Companion Website.

M02_LIAN1878_11_GE_C02.indd 90 12/29/17 3:48 PM

Programming Exercises 91

programming exerCises

Debugging Tip The compiler usually gives a reason for a syntax error. If you don’t know how to correct it, compare your program closely, character by character, with similar examples in the text.

Pedagogical Note Instructors may ask you to document your analysis and design for selected exercises. Use your own words to analyze the problem, including the input, output, and what needs to be computed, and describe how to solve the problem in pseudocode.

Pedagogical Note The solution to most even-numbered programming exercises are provided to students. These exercises serve as additional examples for a variety of programs. To maximize the benefits of these solutions, students should first attempt to complete the even-numbered exercises and then compare their solutions with the solutions provided in the book. Since the book provides a large number of programming exercises, it is sufficient if you can complete all even-numbered programming exercises.

Sections 2.2–2.12 2.1 (Convert mile to kilometer) Write a program that reads a mile in a double value

from the console, converts it to kilometers, and displays the result. The formula for the conversion is as follows:

1 mile = 1.6 kilometers

Here is a sample run:

learn from examples

document analysis and design

even-numbered programming exercises

Enter miles: 96

96 miles is 153.6 kilometers

2.2 (Compute the volume of a triangle) Write a program that reads in the length of sides of an equilateral triangle and computes the area and volume using the fol- lowing formulas:

area = 23 4 (length of sides)2

volume = area * length

Here is a sample run:

Enter length of the sides and height of the Equilateral triangle: 3.5 The area is 3.89 The volume of the Triangular prism is 19.48

2.3 (Convert meters into feet) Write a program that reads a number in meters, converts it to feet, and displays the result. One meter is 3.2786 feet. Here is a sample run:

Enter a value for meter: 10 10.0 meters is 32.786 feet

M02_LIAN1878_11_GE_C02.indd 91 12/29/17 3:48 PM

92 Chapter 2 Elementary Programming

*2.5 (Financial application: calculate tips) Write a program that reads the subtotal and the gratuity rate and then computes the gratuity and total. For example, if the user enters 10 for subtotal and 12% for gratuity rate, the program displays $1.2 as gratuity and $11.2 as total. Here is a sample run:

Enter the subtotal and a gratuity rate: 10 12

The gratuity is $1.2 and total is $11.2

**2.6 (Multiply the digits in an integer) Write a program that reads an integer between 0 and 1000 and multiplies all the digits in the integer. For example, if an integer is 932, the multiplication of all its digits is 54.

Hint: Use the % operator to extract digits, and use the / operator to remove the extracted digit. For instance, 932 % 10 = 2 and 932 / 10 = 93.

Here is a sample run:

Enter a number between 0 and 1000: 999

The multiplication of all digits in 999 is 729

*2.7 (Find the number of years) Write a program that prompts the user to enter the minutes (e.g., 1 billion), and displays the number of years and remaining days for the minutes. For simplicity, assume that a year has 365 days. Here is a sample run:

Enter the number of minutes: 1000000000

1000000000 minutes is approximately 1902 years and 214 days

*2.8 (Current time) Listing 2.7, ShowCurrentTime.java, gives a program that displays the current time in GMT. Revise the program so it prompts the user to enter the time zone offset to GMT and displays the time in the specified time zone. Here is a sample run:

Enter the time zone offset to GMT: -5

The current time is 4:50:34

2.9 (Physics: acceleration) Average acceleration is defined as the change of velocity divided by the time taken to make the change, as given by the following formula:

a = v1 - v0

2.4 (Convert square meter into ping) Write a program that converts square meter into ping. The program prompts the user to enter a number in square meter, converts it to ping, and displays the result. One square meter is 0.3025 ping.

Here is a sample run:

Enter a number in square meters: 50

50.0 square meters is 15.125 pings

M02_LIAN1878_11_GE_C02.indd 92 12/29/17 3:48 PM

Programming Exercises 93

2.10 (Science: calculating energy) Write a program that calculates the energy needed to heat water from an initial temperature to a final temperature. Your program should prompt the user to enter the amount of water in kilograms and the initial and final temperatures of the water. The formula to compute the energy is

Q = M * (finalTemperature – initialTemperature) * 4184

where M is the weight of water in kilograms, initial and final temperatures are in degrees Celsius, and energy Q is measured in joules. Here is a sample run:

Enter v0, v1, and t: 5.5 50.9 4.5

The average acceleration is 10.0889

Enter the amount of water in kilograms: 55.5

Enter the initial temperature: 3.5

Enter the final temperature: 10.5

The energy needed is 1625484.0

2.11 (Population projection) Rewrite Programming Exercise 1.11 to prompt the user to enter the number of years and display the population after the number of years. Use the hint in Programming Exercise 1.11 for this program. Here is a sample run of the program:

Enter the number of years: 5

The population in 5 years is 325932969

2.12 (Physics: finding runway length) Given an airplane’s acceleration a and take-off speed v, you can compute the minimum runway length needed for an airplane to take off using the following formula:

length = v2

2a Write a program that prompts the user to enter v in meters/second (m/s) and the

acceleration a in meters/second squared (m/s2), then, displays the minimum run- way length. Here is a sample run:

Enter speed and acceleration: 60 3.5

The minimum runway length for this airplane is 514.286

**2.13 (Financial application: compound value) Suppose you save $100 each month in a savings account with annual interest rate 3.75%. Thus, the monthly interest rate is 0.0375/12 = 0.003125. After the first month, the value in the account becomes

Write a program that prompts the user to enter the starting velocity v0 in meters/ second, the ending velocity v1 in meters/second, and the time span t in seconds, then displays the average acceleration. Here is a sample run:

M02_LIAN1878_11_GE_C02.indd 93 12/29/17 3:48 PM

94 Chapter 2 Elementary Programming

Enter the monthly saving amount: 100 After the first month, the account value is 100.3125 After the second month, the account value is 200.9384765625 After the third month, the account value is 301.8789093017578 After the sixth month, the account value is 606.5967866995037

*2.14 (Health application: computing BMI) Body Mass Index (BMI) is a measure of health on weight. It can be calculated by taking your weight in kilograms and dividing, by the square of your height in meters. Write a program that prompts the user to enter a weight in pounds and height in inches and displays the BMI. Note one pound is 0.45359237 kilograms and one inch is 0.0254 meters. Here is a sample run:

VideoNote

Compute BMI

Enter weight in pounds: 95.5

Enter height in inches: 50

BMI is 26.8573

2.15 (Geometry: distance of two points) Write a program that prompts the user to enter two points (x1, y1) and (x2, y2) and displays their distance. The formula for computing the distance is 2(x2 - x1)2 + (y2 - y1)2. Note you can use Math. pow(a, 0.5) to compute 2a. Here is a sample run: Enter x1 and y1: 1.5 -3.4

Enter x2 and y2: 4 5

The distance between the two points is 8.764131445842194

2.16 (Geometry: area of a hexagon) Write a program that prompts the user to enter the side of a hexagon and displays its area. The formula for computing the area of a hexagon is

Area = 323

2 s2,

where s is the length of a side. Here is a sample run:

100 * (1 + 0.003125) = 100.3125

After the second month, the value in the account becomes

(100 + 100.3125) * (1 + 0.003125) = 200.938

After the third month, the value in the account becomes

(100 + 200.938) * (1 + 0.003125) = 301.878

and so on. Write a program that prompts the user to enter a monthly saving amount and

displays the account value after the sixth month. (In Exercise 5.30, you will use a loop to simplify the code and display the account value for any month.)

M02_LIAN1878_11_GE_C02.indd 94 12/29/17 3:48 PM

Programming Exercises 95

Enter the length of the side: 5.5

The area of the hexagon is 78.5918

*2.17 (Science: wind-chill temperature) How cold is it outside? The temperature alone is not enough to provide the answer. Other factors including wind speed, relative humidity, and sunshine play important roles in determining coldness outside. In 2001, the National Weather Service (NWS) implemented the new wind-chill temperature to measure the coldness using temperature and wind speed. The formula is

twc = 35.74 + 0.6215ta - 35.75v0.16 + 0.4275tav0.16

where ta is the outside temperature measured in degrees Fahrenheit, v is the speed measured in miles per hour, and twc is the wind-chill temperature. The formula cannot be used for wind speeds below 2 mph or temperatures below -58°F or above 41°F.

Write a program that prompts the user to enter a temperature between -58°F and 41°F and a wind speed greater than or equal to 2 then displays the wind-chill temperature. Use Math.pow(a, b) to compute v0.16. Here is a sample run:

Enter the temperature in Fahrenheit between -58°F and 41°F: 5.3

Enter the wind speed (7 = 2) in miles per hour: 6 The wind chill index is -5.56707

Enter the coordinates of three points separated by spaces like x1 y1 x2 y2 x3 y3: 1.5 -3.4 4.6 5 9.5 -3.4 The area of the triangle is 33.6

2.18 (Print a table) Write a program that displays the following table. Calculate the middle point of two points.

a b Middle Point (0, 0) (2, 1) (1.0, 0.5) (1, 4) (4, 2) (2.5, 3.0) (2, 7) (6, 3) (4.0, 5.0) (3, 9) (10, 5) (6.5, 7.0) (4, 11) (12, 7) (8.0, 9.0)

*2.19 (Geometry: area of a triangle) Write a program that prompts the user to enter three points, (x1, y1), (x2, y2), and (x3, y3), of a triangle then displays its area. The formula for computing the area of a triangle is

s = (side1 + side2 + side3)/2;

area = 2s(s - side1)(s - side2)(s - side3) Here is a sample run:

M02_LIAN1878_11_GE_C02.indd 95 12/29/17 3:48 PM

96 Chapter 2 Elementary Programming

Note More than 200 additional programming exercises with solutions are provided to the instructors on the Instructor Resource Website.

*2.21 (Financial application: calculate future investment value) Write a program that reads in investment amount, annual interest rate, and number of years and displays the future investment value using the following formula:

futureInvestmentValue =

investmentAmount * (1 + monthlyInterestRate)numberOfYears*12

For example, if you enter amount 1000, annual interest rate 3.25%, and number of years 1, the future investment value is 1032.98.

Here is a sample run:

Enter investment amount: 1000.56

Enter annual interest rate in percentage: 4.25

Enter number of years: 1

Future value is $1043.92

*2.22 (Financial application: monetary units) Rewrite Listing 2.10, ComputeChange. java, to fix the possible loss of accuracy when converting a double value to an int value. Enter the input as an integer whose last two digits represent the cents. For example, the input 1156 represents 11 dollars and 56 cents.

*2.23 (Cost of driving) Write a program that prompts the user to enter the distance to drive, the fuel efficiency of the car in miles per gallon, and the price per gallon then displays the cost of the trip. Here is a sample run:

Enter the driving distance: 900.5

Enter miles per gallon: 25.5

Enter price per gallon: 3.55

The cost of driving is $125.36

Enter balance and interest rate (e.g., 3 for 3%): 1000 3.5

The interest is 2.91667

Sections 2.13–2.17 *2.20 (Financial application: calculate interest) If you know the balance and the annual

percentage interest rate, you can compute the interest on the next monthly pay- ment using the following formula:

interest = balance * (annualInterestRate/1200)

Write a program that reads the balance and the annual percentage interest rate and displays the interest for the next month. Here is a sample run:

M02_LIAN1878_11_GE_C02.indd 96 12/29/17 3:49 PM

Selections

Objectives ■■ To declare boolean variables and write Boolean expressions using

relational operators (§3.2).

■■ To implement selection control using one-way if statements (§3.3).

■■ To implement selection control using two-way if-else statements (§3.4).

■■ To implement selection control using nested if and multi-way if statements (§3.5).

■■ To avoid common errors and pitfalls in if statements (§3.6).

■■ To generate random numbers using the Math.random() method (§3.7).

■■ To program using selection statements for a variety of examples (Sub- tractionQuiz, BMI, ComputeTax) (§§3.7–3.9).

■■ To combine conditions using logical operators (!, &&, ||, and ^) (§3.10).

■■ To program using selection statements with combined conditions (LeapYear, Lottery) (§§3.11 and 3.12).

■■ To implement selection control using switch statements (§3.13).

■■ To write expressions using the conditional operator (§3.14).

■■ To examine the rules governing operator precedence and associativity (§3.15).

■■ To apply common techniques to debug errors (§3.16).

Chapter

M03_LIAN1878_11_GE_C03.indd 97 12/29/17 3:54 PM

98 Chapter 3 Selections

3.1 Introduction The program can decide which statements to execute based on a condition.

If you enter a negative value for radius in Listing 2.2, ComputeAreaWithConsoleInput.java, the program displays an invalid result. If the radius is negative, you don’t want the program to compute the area. How can you deal with this situation?

Like all high-level programming languages, Java provides selection statements: statements that let you choose actions with alternative courses. You can use the following selection state- ment to replace lines 12–17 in Listing 2.2:

if (radius < 0) { System.out.println("Incorrect input"); } else { double area = radius * radius * 3.14159; System.out.println("Area is " + area); }

Selection statements use conditions that are Boolean expressions. A Boolean expression is an expression that evaluates to a Boolean value: true or false. We now introduce the boolean type and relational operators.

3.2 boolean Data Type The boolean data type declares a variable with the value either true or false.

How do you compare two values, such as whether a radius is greater than 0, equal to 0, or less than 0? Java provides six relational operators (also known as comparison opera- tors), shown in Table 3.1, which can be used to compare two values (assume radius is 5 in the table).

Point Keyproblem

selection statements

Boolean expression Boolean value

Point Key

boolean data type relational operators

Java Operator Mathematics Symbol Name Example (radius is 5) Result

< < Less than radius < 0 false

<= ≤ Less than or equal to radius <= 0 false

> > Greater than radius > 0 true

>= ≥ Greater than or equal to radius >= 0 true

== = Equal to radius == 0 false

!= ≠ Not equal to radius != 0 true

Table 3.1 Relational Operators

Caution The equality testing operator is two equal signs (==), not a single equal sign (=). The latter symbol is for assignment.

The result of the comparison is a Boolean value: true or false. For example, the follow- ing statement displays true:

double radius = 1; System.out.println(radius > 0);

A variable that holds a Boolean value is known as a Boolean variable. The boolean data type is used to declare Boolean variables. A boolean variable can hold one of the two

== vs. =

Boolean variable

M03_LIAN1878_11_GE_C03.indd 98 12/29/17 3:54 PM

3.2 boolean Data Type 99

values: true or false. For example, the following statement assigns true to the variable lightsOn:

boolean lightsOn = true;

true and false are literals, just like a number such as 10. They are treated as reserved words and cannot be used as identifiers in the program.

Suppose you want to develop a program to let a first-grader practice addition. The pro- gram randomly generates two single-digit integers, number1 and number2, and displays to the student a question such as “What is 1 + 7?, ” as shown in the sample run in Listing 3.1. After the student types the answer, the program displays a message to indicate whether it is true or false.

There are several ways to generate random numbers. For now, generate the first integer using System.currentTimeMillis() % 10 (i.e., the last digit in the current time) and the second using System.currentTimeMillis() / 10 % 10 (i.e., the second last digit in the current time). Listing 3.1 gives the program. Lines 5–6 generate two numbers, number1 and number2. Line 14 obtains an answer from the user. The answer is graded in line 18 using a Boolean expression number1 + number2 == answer.

lisTing 3.1 AdditionQuiz.java 1 import java.util.Scanner; 2 3 public class AdditionQuiz { 4 public static void main(String[] args) { 5 int number1 = (int)(System.currentTimeMillis() % 10); 6 int number2 = (int)(System.currentTimeMillis() / 10 % 10); 7 8 // Create a Scanner 9 Scanner input = new Scanner(System.in); 10 11 System.out.print( 12 "What is " + number1 + " + " + number2 + "? "); 13 14 int answer = input.nextInt(); 15 16 System.out.println( 17 number1 + " + " + number2 + " = " + answer + " is " + 18 (number1 + number2 == answer)); 19 } 20 }

Boolean literals

VideoNote

Program addition quiz

generate number1 generate number2

show question

receive answer

display result

What is 1 + 7? 8

1 + 7 = 8 is true

What is 4 + 8? 9

4 + 8 = 9 is false

line# number1 number2 answer output

5 4

6 8

14 9

16 4 + 8 = 9 is false

M03_LIAN1878_11_GE_C03.indd 99 12/29/17 3:54 PM

100 Chapter 3 Selections

3.2.1 List six relational operators. 3.2.2 Assuming x is 1, show the result of the following Boolean expressions:

(x > 0) (x < 0) (x != 0) (x >= 0) (x != 1)

3.2.3 Can the following conversions involving casting be allowed? Write a test program to verify it.

boolean b = true; i = (int)b;

int i = 1; boolean b = (boolean)i;

3.3 if Statements An if statement is a construct that enables a program to specify alternative paths of execution.

The preceding program displays a message such as “6 + 2 = 7 is false.” If you wish the message to be “6 + 2 = 7 is incorrect,” you have to use a selection statement to make this minor change.

Java has several types of selection statements: one-way if statements, two-way if-else statements, nested if statements, multi-way if-else statements, switch statements, and conditional operators.

A one-way if statement executes an action if and only if the condition is true. The syntax for a one-way if statement is as follows:

if (boolean-expression) { statement(s); }

The flowchart in Figure 3.1a illustrates how Java executes the syntax of an if statement. A flowchart is a diagram that describes an algorithm or process, showing the steps as boxes of various kinds, and their order by connecting these with arrows. Process operations are repre- sented in these boxes, and the arrows connecting them represent the flow of control. A diamond box denotes a Boolean condition, and a rectangle box represents statements.

Point Check

Point Key

why if statement?

if statement

flowchart

Figure 3.1 An if statement executes statements if the boolean-expression evaluates to true.

statement(s)

boolean- expression

true

false

(a)

area = radius * radius * PI; System.out.println("The area for the circle of" + " radius " + radius + " is " + area);

(radius >= 0)

true

false

(b)

M03_LIAN1878_11_GE_C03.indd 100 12/29/17 3:55 PM

3.3 if Statements 101

If the boolean-expression evaluates to true, the statements in the block are executed. As an example, see the following code:

if (radius >= 0) { area = radius * radius * PI; System.out.println("The area for the circle of radius " + radius + " is " + area); }

The flowchart of the preceding statement is shown in Figure 3.1b. If the value of radius is greater than or equal to 0, then the area is computed and the result is displayed; otherwise, the two statements in the block will not be executed.

The boolean-expression is enclosed in parentheses. For example, the code in (a) is wrong. It should be corrected, as shown in (b).

if i > 0 { System.out.println(“i is positive"); }

(a) Wrong

if (i > 0) { System.out.println(“i is positive"); }

(b) Correct

The block braces can be omitted if they enclose a single statement. For example, the fol- lowing statements are equivalent:

if (i > 0) { System.out.println(“i is positive"); }

(a)

if (i > 0) System.out.println(“i is positive");

(b)

Equivalent

Caution Omitting braces makes the code shorter, but it is prone to errors. It is a common mistake to forget the braces when you go back to modify the code that omits the braces.

Listing 3.2 gives a program that prompts the user to enter an integer. If the number is a mul- tiple of 5, the program displays HiFive. If the number is divisible by 2, it displays HiEven.

lisTing 3.2 SimpleIfDemo.java 1 import java.util.Scanner; 2 3 public class SimpleIfDemo { 4 public static void main(String[] args) { 5 Scanner input = new Scanner(System.in); 6 System.out.print("Enter an integer: "); 7 int number = input.nextInt(); 8 9 if (number % 5 == 0) 10 System.out.println("HiFive"); 11 12 if (number % 2 == 0) 13 System.out.println("HiEven"); 14 } 15 }

Omitting braces or not

enter input

check 5

check even

M03_LIAN1878_11_GE_C03.indd 101 12/29/17 3:55 PM

102 Chapter 3 Selections

The program prompts the user to enter an integer (lines 6–7) and displays HiFive if it is divisible by 5 (lines 9–10) and HiEven if it is divisible by 2 (lines 12–13).

3.3.1 Write an if statement that assigns 1 to x if y is greater than 0. 3.3.2 Write an if statement that increases pay by 3% if score is greater than 90. 3.3.3 What is wrong in the following code?

if radius >= 0 { area = radius * radius * PI; System.out.println("The area for the circle of " + " radius " + radius + " is " + area); }

3.4 Two-Way if-else Statements An if-else statement decides the execution path based on whether the condition is true or false.

A one-way if statement performs an action if the specified condition is true. If the condition is false, nothing is done. But what if you want to take alternative actions when the condition is false? You can use a two-way if-else statement. The actions that a two-way if-else statement specifies differ based on whether the condition is true or false.

Here is the syntax for a two-way if-else statement:

if (boolean-expression) { statement(s)-for-the-true-case; } else { statement(s)-for-the-false-case; }

The flowchart of the statement is shown in Figure 3.2.

Point Check

Point Key

Figure 3.2 An if-else statement executes statements for the true case if the boolean- expression evaluates to true; otherwise, statements for the false case are executed.

statement(s) for the true case statement(s) for the false case

boolean- expression

true false

Enter an integer: 30

HiFive HiEven

Enter an integer: 4

HiEven

M03_LIAN1878_11_GE_C03.indd 102 12/29/17 3:55 PM

3.5 Nested if and Multi-way if-else Statements 103

If the boolean-expression evaluates to true, the statement(s) for the true case are executed; otherwise, the statement(s) for the false case are executed. For example, consider the following code:

if (radius >= 0) { area = radius * radius * PI; System.out.println("The area for the circle of radius " + radius + " is " + area); } else { System.out.println("Negative input"); }

If radius >= 0 is true, area is computed and displayed; if it is false, the message "Negative input" is displayed.

As usual, the braces can be omitted if there is only one statement within them. The braces enclosing the System.out.println("Negative input") statement can therefore be omit- ted in the preceding example.

Here is another example of using the if-else statement. The example checks whether a number is even or odd, as follows:

if (number % 2 == 0) System.out.println(number + " is even."); else System.out.println(number + " is odd.");

3.4.1 Write an if statement that increases pay by 3% if score is greater than 90, other- wise increases pay by 1%.

3.4.2 What is the output of the code in (a) and (b) if number is 30? What if number is 35?

two-way if-else statement

Point Check

if (number % 2 == 0) System.out.println(number + " is even."); System.out.println(number + " is odd.");

(a)

if (number % 2 == 0) System.out.println(number + " is even."); else System.out.println(number + " is odd.");

(b)

3.5 Nested if and Multi-Way if-else Statements An if statement can be inside another if statement to form a nested if statement.

The statement in an if or if-else statement can be any legal Java statement, including another if or if-else statement. The inner if statement is said to be nested inside the outer if statement. The inner if statement can contain another if statement; in fact, there is no limit to the depth of the nesting. For example, the following is a nested if statement:

if (i > k) { if (j > k) System.out.println("i and j are greater than k"); } else System.out.println("i is less than or equal to k");

The if (j > k) statement is nested inside the if (i > k) statement. The nested if statement can be used to implement multiple alternatives. The statement

given in Figure 3.3a, for instance, prints a letter grade according to the score, with multiple alternatives.

Point Key

nested if statement

M03_LIAN1878_11_GE_C03.indd 103 12/29/17 3:55 PM

104 Chapter 3 Selections

The execution of this if statement proceeds as shown in Figure 3.4. The first condition (score >= 90) is tested. If it is true, the grade is A. If it is false, the second condition (score >= 80) is tested. If the second condition is true, the grade is B. If that condi- tion is false, the third condition and the rest of the conditions (if necessary) are tested until a condition is met or all of the conditions prove to be false. If all of the conditions are false, the grade is F. Note a condition is tested only when all of the conditions that come before it are false.

Figure 3.4 You can use a multi-way if-else statement to assign a grade.

grade is A

true

false

grade is B

score >= 80

true

grade is C

score >= 70

true

grade is D

score >= 60

true

grade is F

score >= 90

Figure 3.3 A preferred format for multiple alternatives is shown in (b) using a multi-way if-else statement.

if (score >= 90) System.out.print("A"); else if (score >= 80) System.out.print("B"); else if (score >= 70) System.out.print("C"); else if (score >= 60) System.out.print("D"); else System.out.print("F");

(a)

(b)

Equivalent

This is better

M03_LIAN1878_11_GE_C03.indd 104 12/29/17 3:55 PM

3.6 Common Errors and Pitfalls 105

The if statement in Figure 3.3a is equivalent to the if statement in Figure 3.3b. In fact, Figure 3.3b is the preferred coding style for multiple alternative if statements. This style, called multi-way if-else statements, avoids deep indentation and makes the program easy to read.

3.5.1 Suppose x = 3 and y = 2; show the output, if any, of the following code. What is the output if x = 3 and y = 4? What is the output if x = 2 and y = 2? Draw a flowchart of the code.

if (x > 2) { if (y > 2) { z = x + y; System.out.println("z is " + z); } } else System.out.println("x is " + x);

3.5.2 Suppose x = 2 and y = 3. Show the output, if any, of the following code. What is the output if x = 3 and y = 2? What is the output if x = 3 and y = 3?

if (x > 2) if (y > 2) { int z = x + y; System.out.println("z is " + z); } else System.out.println("x is " + x);

3.5.3 What is wrong in the following code? if (score >= 60) System.out.println("D"); else if (score >= 70) System.out.println("C"); else if (score >= 80) System.out.println("B"); else if (score >= 90) System.out.println("A"); else System.out.println("F");

3.6 Common Errors and Pitfalls Forgetting necessary braces, ending an if statement in the wrong place, mistaking == for =, and dangling else clauses are common errors in selection statements. Dupli- cated statements in if-else statements and testing equality of double values are common pitfalls.

The following errors are common among new programmers.

Common Error 1: Forgetting Necessary Braces

The braces can be omitted if the block contains a single statement. However, forgetting the braces when they are needed for grouping multiple statements is a common programming error. If you modify the code by adding new statements in an if statement without braces, you will have to insert the braces. For example, the following code in (a) is wrong. It should be written with braces to group multiple statements, as shown in (b).

multi-way if statement

Point Check

Point Key

M03_LIAN1878_11_GE_C03.indd 105 12/29/17 3:55 PM

106 Chapter 3 Selections

In (a), the console output statement is not part of the if statement. It is the same as the fol- lowing code:

if (radius >= 0) area = radius * radius * PI; System.out.println(“The area " + " is " + area);

(a) Wrong

if (radius >= 0) { area = radius * radius * PI; System.out.println("The area " + " is " + area); }

(b) Correct

if (radius >= 0) area = radius * radius * PI;

System.out.println(“The area " + “ is " + area);

Regardless of the condition in the if statement, the console output statement is always executed.

Common Error 2: Wrong Semicolon at the if Line

Adding a semicolon at the end of an if line, as shown in (a) below, is a common mistake.

Logic error

if (radius >= 0); { area = radius * radius * PI; System.out.println("The area " + " is " + area); }

(a)

if (radius >= 0) { }; { area = radius * radius * PI; System.out.println("The area " + " is " + area); }

(b)

Empty block

Equivalent

This mistake is hard to find, because it is neither a compile error nor a runtime error; it is a logic error. The code in (a) is equivalent to that in (b) with an empty block.

This error often occurs when you use the next-line block style. Using the end-of-line block style can help prevent this error.

Common Error 3: Redundant Testing of Boolean Values

To test whether a boolean variable is true or false in a test condition, it is redundant to use the equality testing operator like the code in (a):

(a)

if (even == true) System.out.println( "It is even.");

if (even) System.out.println( "It is even.");

Equivalent

This is better (b)

Instead, it is better to test the boolean variable directly, as shown in (b). Another good reason for doing this is to avoid errors that are difficult to detect. Using the = operator instead of the == operator to compare the equality of two items in a test condition is a common error. It could lead to the following erroneous statement:

if (even = true) System.out.println("It is even.");

This statement does not have compile errors. It assigns true to even, so even is always true.

M03_LIAN1878_11_GE_C03.indd 106 12/29/17 3:55 PM

3.6 Common Errors and Pitfalls 107

Common Error 4: Dangling else Ambiguity

The code in (a) below has two if clauses and one else clause. Which if clause is matched by the else clause? The indentation indicates that the else clause matches the first if clause. However, the else clause actually matches the second if clause. This situation is known as the dangling else ambiguity. The else clause always matches the most recent unmatched if clause in the same block. Therefore, the statement in (a) is equivalent to the code in (b).

dangling else ambiguity

int i = 1, j = 2, k = 3;

if (i > j) if (i > k) System.out.println("A"); else System.out.println("B");

(a)

Equivalent

This is better with correct indentation

int i = 1, j = 2, k = 3;

if (i > j) if (i > k) System.out.println("A"); else System.out.println("B");

(b)

Since (i > j) is false, nothing is displayed from the statements in (a) and (b). To force the else clause to match the first if clause, you must add a pair of braces:

int i = 1, j = 2, k = 3;

if (i > j) { if (i > k) System.out.println("A"); } else System.out.println("B");

This statement displays B.

Common Error 5: Equality Test of Two Floating-Point Values

As discussed in Common Error 3 in Section 2.8, floating-point numbers have a limited precision and calculations; involving floating-point numbers can introduce round-off errors. Therefore, equality test of two floating-point values is not reliable. For example, you expect the following code to display true, but surprisingly, it displays false:

double x = 1.0 - 0.1 - 0.1 - 0.1 - 0.1 - 0.1; System.out.println(x == 0.5);

Here, x is not exactly 0.5, but is 0.5000000000000001. You cannot reliably test equality of two floating-point values. However, you can compare whether they are close enough by testing whether the difference of the two numbers is less than some threshold. That is, two numbers x and y are very close if � x - y � 6 e, for a very small value, e. e, a Greek letter pronounced "epsilon", is commonly used to denote a very small value. Normally, you set e to 10-14 for comparing two values of the double type, and to 10-7 for comparing two values of the float type. For example, the following code

final double EPSILON = 1E-14; double x = 1.0 - 0.1 - 0.1 - 0.1 - 0.1 - 0.1; if (Math.abs(x - 0.5) < EPSILON) System.out.println(x + " is approximately 0.5");

will display

0.5000000000000001 is approximately 0.5.

The Math.abs(a) method can be used to return the absolute value of a.

M03_LIAN1878_11_GE_C03.indd 107 12/29/17 3:55 PM

108 Chapter 3 Selections

Common Pitfall 1: Simplifying Boolean Variable Assignment

Often, new programmers write the code that assigns a test condition to a boolean variable like the code in (a):

This is not an error, but it should be better written as shown in (b).

Common Pitfall 2: Avoiding Duplicate Code in Different Cases

Often, new programmers write the duplicate code in different cases that should be combined in one place. For example, the highlighted code in the following statement is duplicated:

if (inState) { tuition = 5000; System.out.println("The tuition is " + tuition); } else { tuition = 15000; System.out.println("The tuition is " + tuition); }

This is not an error, but it should be better written as follows:

if (inState) { tuition = 5000; } else { tuition = 15000; } System.out.println("The tuition is " + tuition);

The new code removes the duplication and makes the code easy to maintain, because you only need to change in one place if the print statement is modified.

3.6.1 Which of the following statements are equivalent? Which ones are correctly indented?Point

Check

if (number % 2 == 0) even = true; else even = false;

boolean even = number % 2 == 0;Equivalent

This is better

(a) (b)

3.6.2 Rewrite the following statement using a Boolean expression: if (count % 10 == 0) newLine = true; else newLine = false;

if (i > 0) if (j > 0) x = 0; else if (k > 0) y = 0; else z = 0;

(a)

if (i > 0) { if (j > 0) x = 0; else if (k > 0) y = 0; } else z = 0;

(b)

if (i > 0) if (j > 0) x = 0; else if (k > 0) y = 0; else z = 0;

(c)

if (i > 0) if (j > 0) x = 0; else if (k > 0) y = 0; else z = 0;

(d)

M03_LIAN1878_11_GE_C03.indd 108 12/29/17 3:55 PM

3.7 Generating Random Numbers 109

3.6.3 Are the following statements correct? Which one is better?

3.6.4 What is the output of the following code if number is 14, 15, or 30?

3.7 Generating Random Numbers You can use Math.random() to obtain a random double value between 0.0 and 1.0, excluding 1.0.

Suppose you want to develop a program for a first-grader to practice subtraction. The program randomly generates two single-digit integers, number1 and number2, with number1 >= number2, and it displays to the student a question such as “What is 9 - 2?” After the student enters the answer, the program displays a message indicating whether it is correct.

The previous programs generate random numbers using System.currentTimeMillis(). A better approach is to use the random() method in the Math class. Invoking this method returns a random double value d such that 0.0 … d 6 1.0. Thus, (int)(Math.random() * 10) returns a random single-digit integer (i.e., a number between 0 and 9).

The program can work as follows:

1. Generate two single-digit integers into number1 and number2.

2. If number1 < number2, swap number1 with number2.

3. Prompt the student to answer, "What is number1 − number2?"

4. Check the student’s answer and display whether the answer is correct.

The complete program is given in Listing 3.3.

lisTing 3.3 SubtractionQuiz.java 1 import java.util.Scanner; 2 3 public class SubtractionQuiz { 4 public static void main(String[] args) { 5 // 1. Generate two random single-digit integers 6 int number1 = (int)(Math.random() * 10); 7 int number2 = (int)(Math.random() * 10); 8 9 // 2. If number1 < number2, swap number1 with number2 10 if (number1 < number2) { 11 int temp = number1;

Point Key

VideoNote

Program subtraction quiz

random() method

random number

if (age < 16) System.out.println (“Cannot get a driver’s license"); if (age >= 16) System.out.println (“Can get a driver’s license");

if (age < 16) System.out.println (“Cannot get a driver’s license"); else System.out.println (“Can get a driver’s license");

(a) (b)

if (number % 2 == 0) System.out.println (number + “ is even"); if (number % 5 == 0) System.out.println (number + “ is multiple of 5");

if (number % 2 == 0) System.out.println (number + “ is even"); else if (number % 5 == 0) System.out.println (number + “ is multiple of 5");

(a) (b)

M03_LIAN1878_11_GE_C03.indd 109 12/29/17 3:55 PM

110 Chapter 3 Selections

12 number1 = number2; 13 number2 = temp; 14 } 15 16 // 3. Prompt the student to answer "What is number1 – number2?" 17 System.out.print 18 ("What is " + number1 + " − " + number2 + "? "); 19 Scanner input = new Scanner(System.in); 20 int answer = input.nextInt(); 21 22 // 4. Grade the answer and display the result 23 if (number1 − number2 == answer) 24 System.out.println("You are correct!"); 25 else { 26 System.out.println("Your answer is wrong."); 27 System.out.println(number1 + " − " + number2 + 28 " should be " + (number1 − number2)); 29 } 30 } 31 }

get answer

check the answer

What is 6 − 6? 0

You are correct!

What is 9 − 2? 5

Your answer is wrong

9 − 2 is 7

line# number1 number2 temp answer output

6 2

7 9

11 2

12 9

13 2

20 5

26 Your answer is wrong 9 − 2 should be 7

To swap two variables number1 and number2, a temporary variable temp (line 11) is used to first hold the value in number1. The value in number2 is assigned to number1 (line 12), and the value in temp is assigned to number2 (line 13).

3.7.1 Which of the following is a possible output from invoking Math.random()? 323.4, 0.5, 34, 1.0, 0.0, 0.234

3.7.2 a. How do you generate a random integer i such that 0 … i 6 20?

b. How do you generate a random integer i such that 10 … i 6 20?

c. How do you generate a random integer i such that 10 … i … 50?

d. Write an expression that returns 0 or 1 randomly.

Point Check

M03_LIAN1878_11_GE_C03.indd 110 12/29/17 3:55 PM

3.8 Case Study: Computing Body Mass Index 111

3.8 Case Study: Computing Body Mass Index You can use nested if statements to write a program that interprets body mass index.

Body mass index (BMI) is a measure of health based on height and weight. It can be calculated by taking your weight in kilograms and dividing it by the square of your height in meters. The interpretation of BMI for people 20 years or older is as follows:

BMI Interpretation BMI 6 18.5 Underweight 18.5 … BMI 6 25.0 Normal 25.0 … BMI 6 30.0 Overweight 30.0 … BMI Obese

Write a program that prompts the user to enter a weight in pounds and height in inches and displays the BMI. Note that one pound is 0.45359237 kilograms, and one inch is 0.0254 meters. Listing 3.4 gives the program.

lisTing 3.4 ComputeAndInterpretBMI.java 1 import java.util.Scanner; 2 3 public class ComputeAndInterpretBMI { 4 public static void main(String[] args) { 5 Scanner input = new Scanner(System.in); 6 7 // Prompt the user to enter weight in pounds 8 System.out.print("Enter weight in pounds: "); 9 double weight = input.nextDouble(); 10 11 // Prompt the user to enter height in inches 12 System.out.print("Enter height in inches: "); 13 double height = input.nextDouble(); 14 15 final double KILOGRAMS_PER_POUND = 0.45359237; // Constant 16 final double METERS_PER_INCH = 0.0254; // Constant 17 18 // Compute BMI 19 double weightInKilograms = weight * KILOGRAMS_PER_POUND; 20 double heightInMeters = height * METERS_PER_INCH; 21 double bmi = weightInKilograms / 22 (heightInMeters * heightInMeters); 23 24 // Display result 25 System.out.println("BMI is " + bmi); 26 if (bmi < 18.5) 27 System.out.println("Underweight"); 28 else if (bmi < 25) 29 System.out.println("Normal"); 30 else if (bmi < 30) 31 System.out.println("Overweight"); 32 else 33 System.out.println("Obese"); 34 } 35 }

Point Key

input weight

input height

compute bmi

display output

M03_LIAN1878_11_GE_C03.indd 111 12/29/17 3:55 PM

112 Chapter 3 Selections

line# weight height weightInKilograms heightInMeters bmi output

9 146

13 70

19 66.22448602

20 1.778

21 20.9486

25 BMI is 20.95

29 Normal

The constants KILOGRAMS_PER_POUND and METERS_PER_INCH are defined in lines 15–16. Using constants here makes programs easy to read.

You should test the input that covers all possible cases for BMI to ensure that the program works for all cases.

3.9 Case Study: Computing Taxes You can use nested if statements to write a program for computing taxes.

The U.S. federal personal income tax is calculated based on filing status and taxable income. There are four filing statuses: single filers, married filing jointly or qualified widow(er), mar- ried filing separately, and head of household. The tax rates vary every year. Table 3.2 shows the rates for 2009. If you are single with a taxable income of $10,000, for example, the first $8,350 is taxed at 10% and the other $1,650 is taxed at 15%, so your total tax is $1,082.50.

test all cases

Point Key

VideoNote

Use multi-way if-else statements

Enter weight in pounds: 146

Enter height in inches: 70

BMI is 20.948603801493316

Normal

Marginal Tax Rate Single

Married Filing Jointly or Qualifying Widow(er) Married Filing Separately Head of Household

10% $0–$8,350 $0–$16,700 $0–$8,350 $0–$11,950

15% $8,351–$33,950 $16,701–$67,900 $8,351–$33,950 $11,951–$45,500

25% $33,951–$82,250 $67,901–$137,050 $33,951–$68,525 $45,501–$117,450

28% $82,251–$171,550 $137,051–$208,850 $68,526–$104,425 $117,451–$190,200

33% $171,551–$372,950 $208,851–$372,950 $104,426–$186,475 $190,201–$372,950

35% +372,951+ +372,951+ +186,476+ +372,951+

Table 3.2 2009 U.S. Federal Personal Tax Rates

You are to write a program to compute personal income tax. Your program should prompt the user to enter the filing status and taxable income and compute the tax. Enter 0 for single filers, 1 for married filing jointly or qualified widow(er), 2 for married filing separately, and 3 for head of household.

M03_LIAN1878_11_GE_C03.indd 112 12/29/17 3:55 PM

3.9 Case Study: Computing Taxes 113

Your program computes the tax for the taxable income based on the filing status. The filing status can be determined using if statements outlined as follows:

if (status == 0) { // Compute tax for single filers } else if (status == 1) { // Compute tax for married filing jointly or qualifying widow(er) } else if (status == 2) { // Compute tax for married filing separately } else if (status == 3) { // Compute tax for head of household } else { // Display wrong status }

For each filing status there are six tax rates. Each rate is applied to a certain amount of tax- able income. For example, of a taxable income of $400,000 for single filers, $8,350 is taxed at 10%, (33,950 - 8,350) at 15%, (82,250 - 33,950) at 25%, (171,550 - 82,250) at 28%, (372,950 - 171,550) at 33%, and (400,000 - 372,950) at 35%.

Listing 3.5 gives the solution for computing taxes for single filers. The complete solution is left as an exercise.

lisTing 3.5 ComputeTax.java 1 import java.util.Scanner; 2 3 public class ComputeTax { 4 public static void main(String[] args) { 5 // Create a Scanner 6 Scanner input = new Scanner(System.in); 7 8 // Prompt the user to enter filing status 9 System.out.print("(0-single filer, 1-married jointly or " + 10 "qualifying widow(er), 2-married separately, 3-head of " + 11 "household) Enter the filing status: "); 12 13 int status = input.nextInt(); 14 15 // Prompt the user to enter taxable income 16 System.out.print("Enter the taxable income: "); 17 double income = input.nextDouble(); 18 19 // Compute tax 20 double tax = 0; 21 22 if (status == 0) { // Compute tax for single filers 23 if (income <= 8350) 24 tax = income * 0.10; 25 else if (income <= 33950) 26 tax = 8350 * 0.10 + (income − 8350) * 0.15; 27 else if (income <= 82250) 28 tax = 8350 * 0.10 + (33950 − 8350) * 0.15 + 29 (income − 33950) * 0.25; 30 else if (income <= 171550) 31 tax = 8350 * 0.10 + (33950 − 8350) * 0.15 + 32 (82250 - 33950) * 0.25 + (income − 82250) * 0.28;

input status

input income

compute tax

M03_LIAN1878_11_GE_C03.indd 113 12/29/17 3:55 PM

114 Chapter 3 Selections

33 else if (income <= 372950) 34 tax = 8350 * 0.10 + (33950 − 8350) * 0.15 + 35 (82250 - 33950) * 0.25 + (171550 − 82250) * 0.28 + 36 (income - 171550) * 0.33; 37 else 38 tax = 8350 * 0.10 + (33950 − 8350) * 0.15 + 39 (82250 − 33950) * 0.25 + (171550 − 82250) * 0.28 + 40 (372950 − 171550) * 0.33 + (income − 372950) * 0.35; 41 } 42 else if (status == 1) { // Left as an exercise 43 // Compute tax for married file jointly or qualifying widow(er) 44 } 45 else if (status == 2) { // Compute tax for married separately 46 // Left as an exercise in Programming Exercise 3.13 47 } 48 else if (status == 3) { // Compute tax for head of household 49 // Left as an exercise in Programming Exercise 3.13 50 } 51 else { 52 System.out.println("Error: invalid status"); 53 System.exit(1); 54 } 55 56 // Display the result 57 System.out.println("Tax is " + (int)(tax * 100) / 100.0); 58 } 59 }

exit program

display output

line# status income Tax output

13 0

17 400000

20 0

38 117683.5

57 Tax is 117683.5

(0-single filer, 1-married jointly or qualifying widow(er),

2-married separately, 3-head of household) Enter the filing status: 0 Enter the taxable income: 400000 Tax is 117683.5

The program receives the filing status and taxable income. The multi-way if-else state- ments (lines 22, 42, 45, 48, and 51) check the filing status and compute the tax based on the filing status.

System.exit(status) (line 53) is defined in the System class. Invoking this method terminates the program. The status 0 indicates that the program is terminated normally. A nonzero status code indicates abnormal termination.

An initial value of 0 is assigned to tax (line 20). A compile error would occur if it had no initial value, because all of the other statements that assign values to tax are within the if statement. The compiler thinks these statements may not be executed, and therefore reports a compile error.

System.exit(status)

M03_LIAN1878_11_GE_C03.indd 114 12/29/17 3:55 PM

3.10 Logical Operators 115

To test a program, you should provide the input that covers all cases. For this program, your input should cover all statuses (0, 1, 2, 3). For each status, test the tax for each of the six brackets. Thus, there are a total of 24 cases.

Tip For all programs, you should write a small amount of code and test it before moving on to add more code. This is called incremental development and testing. This approach makes testing easier, because the errors are likely in the new code you just added.

3.9.1 Are the following two statements equivalent?

test all cases

incremental development and testing

Point Check

Operator Name Description

! not Logical negation

&& and Logical conjunction

|| or Logical disjunction

^ exclusive or Logical exclusion

Table 3.3 Boolean Operators

p !p Example (assume age = 24, weight = 140)

true false !(age > 18) is false, because (age > 18) is true.

false true !(weight == 150) is true, because (weight == 150) is false.

Table 3.4 Truth Table for Operator !

if (income <= 10000) tax = income * 0.1; else if (income <= 20000) tax = 1000 + (income − 10000) * 0.15;

if (income <= 10000) tax = income * 0.1; else if (income > 10000 && income <= 20000) tax = 1000 + (income − 10000) * 0.15;

3.10 Logical Operators The logical operators !, &&, ||, and ^ can be used to create a compound Boolean expression.

Sometimes, whether a statement is executed is determined by a combination of several condi- tions. You can use logical operators to combine these conditions to form a compound Boolean expression. Logical operators, also known as Boolean operators, operate on Boolean values to create a new Boolean value. Table 3.3 lists the Boolean operators. Table 3.4 defines the not (!) operator, which negates true to false and false to true. Table 3.5 defines the and (&&) operator. The and (&&) of two Boolean operands is true if and only if both the operands are true. Table 3.6 defines the or (||) operator. The or (||) of two Boolean operands is true if at least one of the operands is true. Table 3.7 defines the exclusive or (^) operator. The exclusive or (^) of two Boolean operands is true if and only if the two operands have different Boolean values. Note p1 ^ p2 is the same as p1 != p2.

Point Key

M03_LIAN1878_11_GE_C03.indd 115 12/29/17 3:55 PM

116 Chapter 3 Selections

Listing 3.6 gives a program that checks whether a number is divisible by 2 and 3, by 2 or 3, and by 2 or 3 but not both.

lisTing 3.6 TestBooleanOperators.java 1 import java.util.Scanner; 2 3 public class TestBooleanOperators { 4 public static void main(String[] args) { 5 // Create a Scanner 6 Scanner input = new Scanner(System.in); 7 8 // Receive an input 9 System.out.print("Enter an integer: "); 10 int number = input.nextInt(); 11 12 if (number % 2 == 0 && number % 3 == 0) 13 System.out.println(number + " is divisible by 2 and 3."); 14

import class

input

and

p1 p2 p1 || p2 Example (assume age = 24, weight = 140)

false false false (age > 34) || (weight >= 150) is false, because (age > 34) and (weight >= 150) are both false.

false true true

true false true (age > 18) || (weight < 140) is true, because (age > 18) is true.

true true true

Table 3.6 Truth Table for Operator ||

p1 p2 p1 ^ p2 Example (assume age = 24, weight = 140)

false false false (age > 34) ^ (weight > 140) is false, because (age > 34) and (weight > 140) are both false.

false true true (age > 34) ^ (weight >= 140) is true, because (age > 34) is false but (weight >= 140) is true.

true false true

true true false

Table 3.7 Truth Table for Operator ^

p1 p2 p1 && p2 Example (assume age = 24, weight = 140)

false false false

false true false (age > 28) && (weight <= 140) is false, because (age > 28) is false.

true false false

true true true (age > 18) && (weight >= 140) is true, because (age > 18) and (weight >= 140) are both true.

Table 3.5 Truth Table for Operator &&

M03_LIAN1878_11_GE_C03.indd 116 12/29/17 3:55 PM

3.10 Logical Operators 117

15 if (number % 2 == 0 || number % 3 == 0) 16 System.out.println(number + " is divisible by 2 or 3."); 17 18 if (number % 2 == 0 ^ number % 3 == 0) 19 System.out.println(number + 20 " is divisible by 2 or 3, but not both."); 21 } 22 }

Enter an integer: 4 4 is divisible by 2 or 3. 4 is divisible by 2 or 3, but not both.

Enter an integer: 18 18 is divisible by 2 and 3. 18 is divisible by 2 or 3.

(number % 2 == 0 && number % 3 == 0) (line 12) checks whether the number is divisible by both 2 and 3. (number % 2 == 0 || number % 3 == 0) (line 15) checks whether the number is divisible by 2 or by 3. (number % 2 == 0 ^ number % 3 == 0) (line 18) checks whether the number is divisible by 2 or 3, but not both.

Caution In mathematics, the expression

28 <= numberOfDaysInAMonth <= 31 is correct. However, it is incorrect in Java, because 28 <= numberOfDaysInAMonth is evaluated to a boolean value, which cannot be compared with 31. Here, two oper- ands (a boolean value and a numeric value) are incompatible. The correct expression in Java is

28 <= numberOfDaysInAMonth && numberOfDaysInAMonth <= 31

Note De Morgan’s law, named after Indian-born British mathematician and logician Augustus De Morgan (1806–1871), can be used to simplify Boolean expressions. The law states the following:

!(condition1 && condition2) is the same as !condition1 || !condition2 !(condition1 || condition2) is the same as !condition1 && !condition2

For example,

!(number % 2 == 0 && number % 3 == 0)

can be simplified using an equivalent expression:

number % 2 != 0 || number % 3 != 0

As another example,

!(number == 2 || number == 3)

is better written as

number != 2 && number != 3

exclusive or

incompatible operands

De Morgan’s law

M03_LIAN1878_11_GE_C03.indd 117 12/29/17 3:55 PM

118 Chapter 3 Selections

If one of the operands of an && operator is false, the expression is false; if one of the operands of an || operator is true, the expression is true. Java uses these properties to improve the performance of these operators. When evaluating p1 && p2, Java first evaluates p1 then, if p1 is true, evaluates p2; if p1 is false, it does not evaluate p2. When evaluating p1 || p2, Java first evaluates p1 then, if p1 is false, evaluates p2; if p1 is true, it does not evaluate p2. In programming language terminology, && and || are known as the short- circuit or lazy operators. Java also provides the & and | operators, which are covered in Sup- plement III.C for advanced readers.

3.10.1 Assuming that x is 1, show the result of the following Boolean expressions: (true) && (3 > 4) !(x > 0) && (x > 0) (x > 0) || (x < 0) (x != 0) || (x == 0) (x >= 0) || (x < 0) (x != 1) == !(x == 1)

3.10.2 (a) Write a Boolean expression that evaluates to true if a number stored in variable num is between 1 and 100. (b) Write a Boolean expression that evaluates to true if a number stored in variable num is between 1 and 100 or the number is negative.

3.10.3 (a) Write a Boolean expression for � x - 5 � 6 4.5. (b) Write a Boolean expres- sion for � x - 5 � 7 4.5.

3.10.4 Assume x and y are int type. Which of the following are legal Java expressions? x > y > 0 x = y && y x /= y x or y x and y (x != 0) || (x = 0)

3.10.5 Are the following two expressions the same?

(a) x % 2 == 0 && x % 3 == 0

(b) x % 6 == 0

3.10.6 What is the value of the expression x >= 50 && x <= 100 if x is 45, 67, or 101? 3.10.7 Suppose, when you run the following program, you enter the input 2 3 6 from

the console. What is the output?

public class Test { public static void main(String[] args) { java.util.Scanner input = new java.util.Scanner(System.in); double x = input.nextDouble(); double y = input.nextDouble(); double z = input.nextDouble();

System.out.println("(x < y && y < z) is " + (x < y && y < z)); System.out.println("(x < y || y < z) is " + (x < y || y < z)); System.out.println("!(x < y) is " + !(x < y)); System.out.println("(x + y < z) is " + (x + y < z)); System.out.println("(x + y > z) is " + (x + y > z)); } }

3.10.8 Write a Boolean expression that evaluates to true if age is greater than 13 and less than 18.

short-circuit operator

lazy operator

Point Check

M03_LIAN1878_11_GE_C03.indd 118 12/29/17 3:55 PM

3.11 Case Study: Determining Leap Year 119

3.10.9 Write a Boolean expression that evaluates to true if weight is greater than 50 pounds or height is greater than 60 inches.

3.10.10 Write a Boolean expression that evaluates to true if weight is greater than 50 pounds and height is greater than 60 inches.

3.10.11 Write a Boolean expression that evaluates to true if either weight is greater than 50 pounds or height is greater than 60 inches, but not both.

3.11 Case Study: Determining Leap Year A year is a leap year if it is divisible by 4 but not by 100, or if it is divisible by 400.

A leap year has 366 days. The February of a leap year has 29 days. You can use the following Boolean expressions to check whether a year is a leap year:

// A leap year is divisible by 4 boolean isLeapYear = (year % 4 == 0);

// A leap year is divisible by 4 but not by 100 isLeapYear = isLeapYear && (year % 100 != 0);

// A leap year is divisible by 4 but not by 100 or divisible by 400 isLeapYear = isLeapYear || (year % 400 == 0);

Or you can combine all these expressions into one as follows:

isLeapYear = (year % 4 == 0 && year % 100 != 0) || (year % 400 == 0);

Listing 3.7 gives the program that lets the user enter a year and checks whether it is a leap year.

lisTing 3.7 LeapYear.java 1 import java.util.Scanner; 2 3 public class LeapYear { 4 public static void main(String[] args) { 5 // Create a Scanner 6 Scanner input = new Scanner(System.in); 7 System.out.print("Enter a year: "); 8 int year = input.nextInt(); 9 10 // Check if the year is a leap year 11 boolean isLeapYear = 12 (year % 4 == 0 && year % 100 != 0) || (year % 400 == 0); 13 14 // Display the result 15 System.out.println(year + " is a leap year? " + isLeapYear); 16 } 17 }

Point Key

input

leap year?

display result

Enter a year: 2008

2008 is a leap year? true

Enter a year: 1900

1900 is a leap year? false

M03_LIAN1878_11_GE_C03.indd 119 12/29/17 3:55 PM

120 Chapter 3 Selections

3.11.1 How many days in the February of a leap year? Which of the following is a leap year? 500, 1000, 2000, 2016, and 2020?

3.12 Case Study: Lottery The lottery program involves generating random numbers, comparing digits, and using Boolean operators.

Suppose you want to develop a program to play lottery. The program randomly generates a lottery of a two-digit number, prompts the user to enter a two-digit number, and determines whether the user wins according to the following rules:

1. If the user input matches the lottery number in the exact order, the award is $10,000.

2. If all digits in the user input match all digits in the lottery number, the award is $3,000.

3. If one digit in the user input matches a digit in the lottery number, the award is $1,000.

Note the digits of a two-digit number may be 0. If a number is less than 10, we assume that the number is preceded by a 0 to form a two-digit number. For example, number 8 is treated as 08, and number 0 is treated as 00 in the program. Listing 3.8 gives the complete program.

lisTing 3.8 Lottery.java 1 import java.util.Scanner; 2 3 public class Lottery { 4 public static void main(String[] args) { 5 // Generate a lottery number 6 int lottery = (int)(Math.random() * 100); 7 8 // Prompt the user to enter a guess 9 Scanner input = new Scanner(System.in); 10 System.out.print("Enter your lottery pick (two digits): "); 11 int guess = input.nextInt(); 12 13 // Get digits from lottery 14 int lotteryDigit1 = lottery / 10; 15 int lotteryDigit2 = lottery % 10; 16 17 // Get digits from guess 18 int guessDigit1 = guess / 10; 19 int guessDigit2 = guess % 10; 20 21 System.out.println("The lottery number is " + lottery); 22 23 // Check the guess 24 if (guess == lottery) 25 System.out.println("Exact match: you win $10,000"); 26 else if (guessDigit2 == lotteryDigit1 27 && guessDigit1 == lotteryDigit2) 28 System.out.println("Match all digits: you win $3,000"); 29 else if (guessDigit1 == lotteryDigit1 30 || guessDigit1 == lotteryDigit2 31 || guessDigit2 == lotteryDigit1

Point Check

Point Key

generate a lottery number

enter a guess

exact match?

match all digits?

match one digit?

Enter a year: 2002

2002 is a leap year? false

M03_LIAN1878_11_GE_C03.indd 120 12/29/17 3:55 PM

3.12 Case Study: Lottery 121

32 || guessDigit2 == lotteryDigit2) 33 System.out.println("Match one digit: you win $1,000"); 34 else 35 System.out.println("Sorry, no match"); 36 } 37 }

Enter your lottery pick: 23

The lottery number is 34 Match one digit: you win $1,000

Enter your lottery pick: 23

The lottery number is 14 Sorry: no match

Enter your lottery pick (two digits): 15

The lottery number is 15 Exact match: you win $10,000

Enter your lottery pick (two digits): 45

The lottery number is 54 Match all digits: you win $3,000

line#

variable

6 11 14 15 18 19 33

lottery 34

guess 23

lotteryDigit1 3

lotteryDigit2 4

guessDigit1 2

guessDigit2 3

Output Match one digit:

you win $1,000

The program generates a lottery using the random() method (line 6) and prompts the user to enter a guess (line 11). Note guess % 10 obtains the last digit from guess and guess /10 obtains the first digit from guess, since guess is a two-digit number (lines 18 and 19).

The program checks the guess against the lottery number in this order:

1. First, check whether the guess matches the lottery exactly (line 24).

2. If not, check whether the reversal of the guess matches the lottery (lines 26 and 27).

3. If not, check whether one digit is in the lottery (lines 29–32).

4. If not, nothing matches and display "Sorry, no match" (lines 34 and 35).

3.12.1 What happens if you enter an integer as 05? Point Check

M03_LIAN1878_11_GE_C03.indd 121 12/29/17 3:55 PM

122 Chapter 3 Selections

3.13 switch Statements A switch statement executes statements based on the value of a variable or an expression.

The if statement in Listing 3.5, ComputeTax.java, makes selections based on a single true or false condition. There are four cases for computing taxes, which depend on the value of status. To fully account for all the cases, nested if statements were used. Overuse of nested if statements makes a program difficult to read. Java provides a switch statement to simplify coding for multiple conditions. You can write the following switch statement to replace the nested if statement in Listing 3.5:

switch (status) { case 0: compute tax for single filers; break; case 1: compute tax for married jointly or qualifying widow(er); break; case 2: compute tax for married filing separately; break; case 3: compute tax for head of household; break; default: System.out.println("Error: invalid status"); System.exit(1); }

The flowchart of the preceding switch statement is shown in Figure 3.5.

Point Key

Figure 3.5 The switch statement checks all cases and executes the statements in the matched case.

Compute tax for single �lers

Compute tax for married jointly or qualifying widow(er)

Compute tax for head of household

Default actions

status is 0

status is 1

status is 2

status is 3

default

break

Compute tax for married �ling separately

This statement checks to see whether the status matches the value 0, 1, 2, or 3, in that order. If matched, the corresponding tax is computed; if not matched, a message is displayed. Here is the full syntax for the switch statement:

switch (switch-expression) { case value1: statement(s)1; break;

switch statement

M03_LIAN1878_11_GE_C03.indd 122 12/29/17 3:55 PM

3.13 switch Statements 123

case value2: statement(s)2; break; ... case valueN: statement(s)N; break; default: statement(s)-for-default; }

The switch statement observes the following rules:

■■ The switch-expression must yield a value of char, byte, short, int, or String type and must always be enclosed in parentheses. (The char and String types will be introduced in Chapter 4.)

■■ The value1, . . ., and valueN must have the same data type as the value of the switch-expression. Note that value1, . . ., and valueN are constant expressions, meaning they cannot contain variables, such as 1 + x.

■■ When the value in a case statement matches the value of the switch-expression, the statements starting from this case are executed until either a break statement or the end of the switch statement is reached.

■■ The default case, which is optional, can be used to perform actions when none of the specified cases matches the switch-expression.

■■ The keyword break is optional. The break statement immediately ends the switch statement.

Caution Do not forget to use a break statement when one is needed. Once a case is matched, the statements starting from the matched case are executed until a break statement or the end of the switch statement is reached. This is referred to as fall-through behavior. For example, the following code displays Weekday for days 1–5 and Weekend for day 0 and day 6.

without break

fall-through behavior

switch (day) { case 1: case 2: case 3: case 4: case 5: System.out.println(“Weekday"); break; case 0: case 6: System.out.println(“Weekend"); }

Tip To avoid programming errors and improve code maintainability, it is a good idea to put a comment in a case clause if break is purposely omitted.

Now let us write a program to find out the Chinese Zodiac sign for a given year. The Chi- nese Zodiac is based on a 12-year cycle, with each year represented by an animal—monkey, rooster, dog, pig, rat, ox, tiger, rabbit, dragon, snake, horse, or sheep—in this cycle, as shown in Figure 3.6.

Note year % 12 determines the Zodiac sign. 1900 is the year of the rat because 1900 % 12 is 4. Listing 3.9 gives a program that prompts the user to enter a year and displays the animal for the year.

M03_LIAN1878_11_GE_C03.indd 123 12/29/17 3:55 PM

124 Chapter 3 Selections

lisTing 3.9 ChineseZodiac.java 1 import java.util.Scanner; 2 3 public class ChineseZodiac { 4 public static void main(String[] args) { 5 Scanner input = new Scanner(System.in); 6 7 System.out.print("Enter a year: "); 8 int year = input.nextInt(); 9 10 switch (year % 12) { 11 case 0: System.out.println("monkey"); break; 12 case 1: System.out.println("rooster"); break; 13 case 2: System.out.println("dog"); break; 14 case 3: System.out.println("pig"); break; 15 case 4: System.out.println("rat"); break; 16 case 5: System.out.println("ox"); break; 17 case 6: System.out.println("tiger"); break; 18 case 7: System.out.println("rabbit"); break; 19 case 8: System.out.println("dragon"); break; 20 case 9: System.out.println("snake"); break; 21 case 10: System.out.println("horse"); break; 22 case 11: System.out.println("sheep"); 23 } 24 } 25 }

enter year

determine Zodiac sign

Figure 3.6 The Chinese Zodiac is based on a 12-year cycle.

rat 0: monkey 1: rooster 2: dog 3: pig 4: rat 5: ox 6: tiger 7: rabbit 8: dragon 9: snake 10: horse 11: sheep

tiger

rabbit

dragon

snakehorse

sheep

monkey

rooster

dog

pig

year % 12 =

Enter a year: 1963

rabbit

Enter a year: 1877

M03_LIAN1878_11_GE_C03.indd 124 12/29/17 3:55 PM

3.14 Conditional Operators 125

3.13.1 What data types are required for a switch variable? If the keyword break is not used after a case is processed, what is the next statement to be executed? Can you convert a switch statement to an equivalent if statement, or vice versa? What are the advantages of using a switch statement?

3.13.2 What is y after the following switch statement is executed? Rewrite the code using an if-else statement.

x = 3; y = 3; switch (x + 3) { case 6: y = 1; default: y += 1; }

3.13.3 What is x after the following if-else statement is executed? Use a switch statement to rewrite it and draw the flowchart for the new switch statement.

int x = 1, a = 3; if (a == 1) x += 5; else if (a == 2) x += 10; else if (a == 3) x += 16; else if (a == 4) x += 34;

3.13.4 Write a switch statement that displays Sunday, Monday, Tuesday, Wednesday, Thursday, Friday, Saturday, if day is 0, 1, 2, 3, 4, 5, 6, respectively.

3.13.5 Rewrite Listing 3.9 using an if-else statement.

3.14 Conditional Operators A conditional operator evaluates an expression based on a condition.

You might want to assign a value to a variable that is restricted by certain conditions. For example, the following statement assigns 1 to y if x is greater than 0 and -1 to y if x is less than or equal to 0:

if (x > 0) y = 1; else y = −1;

Alternatively, as in the following example, you can use a conditional operator to achieve the same result.

y = (x > 0) ? 1 : −1;

The symbols ? and : appearing together is called a conditional operator (also known as a ternary operator because it uses three operands. It is the only ternary operator in Java. The conditional operator is in a completely different style, with no explicit if in the statement. The syntax to use the operator is as follows:

boolean-expression ? expression1 : expression2

The result of this expression is expression1 if boolean-expression is true; otherwise the result is expression2.

Suppose you want to assign the larger number of variable num1 and num2 to max. You can simply write a statement using the conditional operator:

max = (num1 > num2) ? num1 : num2;

Point Check

Point Key

conditional operator

ternary operator

M03_LIAN1878_11_GE_C03.indd 125 12/29/17 3:55 PM

126 Chapter 3 Selections

For another example, the following statement displays the message “num is even” if num is even, and otherwise displays “num is odd.”

System.out.println((num % 2 == 0) ? "num is even" : "num is odd");

As you can see from these examples, the conditional operator enables you to write short and concise code.

3.14.1 Suppose when you run the following program, you enter the input 2 3 6 from the console. What is the output?

System.out.println((x < y && y < z) ? "sorted" : "not sorted"); } }

3.14.2 Rewrite the following if statements using the conditional operator.

Point Check

if (ages >= 16) ticketPrice = 20; else ticketPrice = 10;

3.14.3 Rewrite the following codes using if-else statements.

a. score = (x > 10) ? 3 * scale : 4 * scale; b. tax = (income > 10000) ? income * 0.2 : income * 0.17 + 1000; c. System.out.println((number % 3 == 0) ? i : j);

3.14.4 Write an expression using a conditional operator that returns randomly -1 or 1.

3.15 Operator Precedence and Associativity Operator precedence and associativity determine the order in which operators are evaluated.

Section 2.11 introduced operator precedence involving arithmetic operators. This section dis- cusses operator precedence in more detail. Suppose you have this expression:

3 + 4 * 4 > 5 * (4 + 3) – 1 && (4 – 3 > 5)

What is its value? What is the execution order of the operators? The expression within parentheses is evaluated first. (Parentheses can be nested, in which

case the expression within the inner parentheses is executed first.) When evaluating an expres- sion without parentheses, the operators are applied according to the precedence rule and the associativity rule.

The precedence rule defines precedence for operators, as shown in Table 3.8, which con- tains the operators you have learned so far. Operators are listed in decreasing order of prece- dence from top to bottom. The logical operators have lower precedence than the relational operators, and the relational operators have lower precedence than the arithmetic operators. Operators with the same precedence appear in the same group. (See Appendix C, Operator Precedence Chart, for a complete list of Java operators and their precedence.)

Point Key

operator precedence

M03_LIAN1878_11_GE_C03.indd 126 12/29/17 3:55 PM

3.15 Operator Precedence and Associativity 127

Assignment operators are right associative. Therefore, the expression

Precedence Operator

var++ and var−− (Postfix)

+, − (Unary plus and minus), ++var and −−var (Prefix)

(type) (Casting)

!(Not)

*, /, % (Multiplication, division, and remainder)

+, − (Binary addition and subtraction)

<, <=, >, >= (Relational)

==, != (Equality)

^ (Exclusive OR)

&& (AND)

|| (OR)

=, +=, −=, *=, /=, %= (Assignment operators)

Table 3.8 Operator Precedence Chart

a – b + c – d is equivalent to

((a – b) + c) - d

a = b += c = d is equivalent to

a = (b += (c = 5))

Suppose a, b, and c are 1 before the assignment; after the whole expression is evaluated, a becomes 6, b becomes 6, and c becomes 5. Note left associativity for the assignment operator would not make sense.

Note Java has its own way to evaluate an expression internally. The result of a Java evaluation is the same as that of its corresponding arithmetic evaluation. Advanced readers may refer to Sup- plement III.B for more discussions on how an expression is evaluated in Java behind the scenes.

3.15.1 List the precedence order of the Boolean operators. Evaluate the following expressions:

true || true && false true && true || false

3.15.2 True or false? All the binary operators except = are left associative. 3.15.3 Evaluate the following expressions:

2 * 2 – 3 > 2 && 4 – 2 > 5 2 * 2 – 3 > 2 || 4 – 2 > 5

3.15.4 Is (x > 0 && x < 10) the same as ((x > 0) && (x < 10))? Is (x > 0 || x < 10) the same as ((x > 0) || (x < 10))? Is (x > 0 || x < 10 && y < 0) the same as (x > 0 ||

(x < 10 && y < 0))?

behind the scenes

Point Check

If operators with the same precedence are next to each other, their associativity determines the order of evaluation. All binary operators except assignment operators are left associative. For example, since + and − are of the same precedence and are left associative, the expression operator associativity

M03_LIAN1878_11_GE_C03.indd 127 12/29/17 3:55 PM

128 Chapter 3 Selections

3.16 Debugging Debugging is the process of finding and fixing errors in a program.

As mentioned in Section 1.10, syntax errors are easy to find and easy to correct because the compiler gives indications as to where the errors came from and why they are there. Runtime errors are not difficult to find either, because the Java interpreter displays them on the console when the program aborts. Finding logic errors, on the other hand, can be very challenging.

Logic errors are called bugs. The process of finding and correcting errors is called debug- ging. A common approach to debugging is to use a combination of methods to help pinpoint the part of the program where the bug is located. You can hand-trace the program (i.e., catch errors by reading the program), or you can insert print statements in order to show the values of the variables or the execution flow of the program. These approaches might work for debug- ging a short, simple program, but for a large, complex program, the most effective approach is to use a debugger utility.

JDK includes a command-line debugger, jdb, which is invoked with a class name. jdb is itself a Java program, running its own copy of Java interpreter. All the Java IDE tools, such as Eclipse and NetBeans, include integrated debuggers. The debugger utilities let you follow the execution of a program. They vary from one system to another, but they all support most of the following helpful features.

■■ Executing a single statement at a time: The debugger allows you to execute one statement at a time so that you can see the effect of each statement.

■■ Tracing into or stepping over a method: If a method is being executed, you can ask the debugger to enter the method and execute one statement at a time in the method, or you can ask it to step over the entire method. You should step over the entire method if you know that the method works. For example, always step over system-supplied methods, such as System.out.println.

■■ Setting breakpoints: You can also set a breakpoint at a specific statement. Your program pauses when it reaches a breakpoint. You can set as many breakpoints as you want. Breakpoints are particularly useful when you know where your programming error starts. You can set a breakpoint at that statement, and have the program execute until it reaches the breakpoint.

■■ Displaying variables: The debugger lets you select several variables and display their values. As you trace through a program, the content of a variable is continu- ously updated.

■■ Displaying call stacks: The debugger lets you trace all of the method calls. This feature is helpful when you need to see a large picture of the program-execution flow.

■■ Modifying variables: Some debuggers enable you to modify the value of a variable when debugging. This is convenient when you want to test a program with different samples, but do not want to leave the debugger.

Tip If you use an IDE such as Eclipse or NetBeans, please refer to Learning Java Effectively with Eclipse/NetBeans in Supplements II.C and II.E on the Companion Website. The supplement shows you how to use a debugger to trace programs, and how debugging can help in learning Java effectively.

Point Key

bugs debugging hand-traces

debugging in IDE

M03_LIAN1878_11_GE_C03.indd 128 12/29/17 3:55 PM

Chapter Summary 129

ChapTer summary 1. A boolean-type variable can store a true or false value.

2. The relational operators (<, <=, ==, !=, >, and >=) yield a Boolean value.

3. Selection statements are used for programming with alternative courses of actions. There are several types of selection statements: one-way if statements, two-way if-else statements, nested if statements, multi-way if-else statements, switch statements, and conditional operators.

4. The various if statements all make control decisions based on a Boolean expression. Based on the true or false evaluation of the expression, these statements take one of the two possible courses.

5. The Boolean operators &&, ||, !, and ^ operate with Boolean values and variables.

6. When evaluating p1 && p2, Java first evaluates p1 then evaluates p2 if p1 is true; if p1 is false, it does not evaluate p2. When evaluating p1 || p2, Java first evaluates p1 then evaluates p2 if p1 is false; if p1 is true, it does not evaluate p2. Therefore, && is referred to as the short-circuit or lazy AND operator, and || is referred to as the short-circuit or lazy OR operator.

7. The switch statement makes control decisions based on a switch expression of type char, byte, short, int, or String.

8. The keyword break is optional in a switch statement, but it is normally used at the end of each case in order to skip the remainder of the switch statement. If the break statement is not present, the next case statement will be executed.

9. The operators in expressions are evaluated in the order determined by the rules of paren- theses, operator precedence, and operator associativity.

10. Parentheses can be used to force the order of evaluation to occur in any sequence.

11. Operators with higher precedence are evaluated earlier. For operators of the same prec- edence, their associativity determines the order of evaluation.

12. All binary operators except assignment operators are left associative; assignment opera- tors are right associative.

Key Terms Boolean expression 98 boolean data type 98 Boolean value 98 conditional operator 125 dangling else ambiguity 107 debugging 128 fall-through behavior 123

flowchart 100 lazy operator 118 operator associativity 127 operator precedence 126 selection statement 98 short-circuit operator 118

M03_LIAN1878_11_GE_C03.indd 129 12/29/17 3:55 PM

130 Chapter 3 Selections

Quiz Answer the quiz for this chapter online at the Companion Website.

programming exerCiSeS

Pedagogical Note For each exercise, carefully analyze the problem requirements and design strategies for solving the problem before coding.

Debugging Tip Before you ask for help, read and explain the program to yourself, and trace it using several representative inputs by hand or using an IDE debugger. You learn how to pro- gram by debugging your own mistakes.

Section 3.2 *3.1 (Algebra: solve quadratic equations) The two roots of a quadratic equation

ax2 + bx + c = 0 can be obtained using the following formula:

r1 = -b + 2b2 - 4ac

2a and r2 =

-b - 2b2 - 4ac 2a

b2 - 4ac is called the discriminant of the quadratic equation. If it is positive, the equation has two real roots. If it is zero, the equation has one root. If it is negative, the equation has no real roots.

Write a program that prompts the user to enter values for a, b, and c and displays the result based on the discriminant. If the discriminant is positive, display two roots. If the discriminant is 0, display one root. Otherwise, display “The equation has no real roots.”

Note you can use Math.pow(x, 0.5) to compute 2x. Here are some sample runs:

think before coding

learn from mistakes

Enter a, b, c: 1.0 3 1

The equation has two roots -0.381966 and -2.61803

Enter a, b, c: 1 2.0 1

The equation has one root -1.0

Enter a, b, c: 1 2 3

The equation has no real roots

3.2 (Game: multiply three numbers) The program in Listing 3.1, AdditionQuiz.java, generates two integers and prompts the user to enter the product of these two inte- gers. Revise the program to generate three single-digit integers and prompt the user to enter the multiplication of these three integers.

M03_LIAN1878_11_GE_C03.indd 130 12/29/17 3:55 PM

Programming Exercises 131

Sections 3.3–3.7 *3.3 (Algebra: solve 2 * 2 linear equations) A linear equation can be solved using

Cramer’s rule given in Programming Exercise 1.13. Write a program that prompts the user to enter a, b, c, d, e, and f and displays the result. If ad - bc is 0, report that “The equation has no solution.”

Enter a, b, c, d, e, f: 9.0 4.0 3.0 -5.0 -6.0 -21.0

x is -2.0 and y is 3.0

Enter a, b, c, d, e, f: 1.0 2.0 2.0 4.0 4.0 5.0

The equation has no solution

Enter today’s day: 1

Enter the number of days elapsed since today: 3 Today is Monday and the future day is Thursday

Enter today’s day: 0

Enter the number of days elapsed since today: 31

Today is Sunday and the future day is Wednesday

Enter weight in pounds: 140

Enter feet: 5

Enter inches: 10

BMI is 20.087702275404553

Normal

**3.4 (Random month) Write a program that randomly generates an integer between 1 and 12 and displays the English month names January, February, . . . , December for the numbers 1, 2, . . . , 12, accordingly.

*3.5 (Find future dates) Write a program that prompts the user to enter an integer for today’s day of the week (Sunday is 0, Monday is 1, . . . , and Saturday is 6). Also prompt the user to enter the number of days after today for a future day and display the future day of the week. Here is a sample run:

*3.6 (Health application: BMI) Revise Listing 3.4, ComputeAndInterpretBMI.java, to let the user enter weight, feet, and inches. For example, if a person is 5 feet and 10 inches, you will enter 5 for feet and 10 for inches. Here is a sample run:

3.7 (Financial application: monetary units) Modify Listing 2.10, ComputeChange. java, to display the nonzero denominations only, using singular words for single units such as 1 dollar and 1 penny, and plural words for more than one unit such as 2 dollars and 3 pennies.

M03_LIAN1878_11_GE_C03.indd 131 12/29/17 3:55 PM

132 Chapter 3 Selections

*3.8 (Sort three integers) Write a program that prompts the user to enter three integers and display the integers in non-decreasing order.

**3.9 (Business: check ISBN-10) An ISBN-10 (International Standard Book Number) consists of 10 digits: d1d2d3d4d5d6d7d8d9d10. The last digit, d10, is a checksum, which is calculated from the other 9 digits using the following formula:

(d1 * 1 + d2 * 2 + d3 * 3 + d4 * 4 + d5 * 5 +

d6 * 6 + d7 * 7 + d8 * 8 + d9 * 9),11

If the checksum is 10, the last digit is denoted as X according to the ISBN-10 convention. Write a program that prompts the user to enter the first 9 digits and displays the 10-digit ISBN (including leading zeros). Your program should read the input as an integer. Here are sample runs:

VideoNote

Sort three integers

Enter the first 9 digits of an ISBN as integer: 013601267

The ISBN-10 number is 0136012671

Enter the first 9 digits of an ISBN as integer: 013031997

The ISBN-10 number is 013031997X

Enter a three-digit integer: 121

121 is a palindrome

Enter a three-digit integer: 123

123 is not a palindrome

3.10 (Game: multiplication quiz) Listing 3.3, SubtractionQuiz.java, randomly generates a subtraction question. Revise the program to randomly generate a multiplication question with two integers less than 1000.

Sections 3.8–3.16 *3.11 (Find the number of days in a month) Write a program that prompts the user to enter

the month and year and displays the number of days in the month. For example, if the user entered month 2 and year 2012, the program should display that February 2012 has 29 days. If the user entered month 3 and year 2015, the program should display that March 2015 has 31 days.

3.12 (Palindrome integer) Write a program that prompts the user to enter a three-digit integer and determines whether it is a palindrome integer. An integer is palindrome if it reads the same from right to left and from left to right. A negative integer is treated the same as a positive integer. Here are sample runs of this program:

*3.13 (Financial application: compute taxes) Listing 3.5, ComputeTax.java, gives the source code to compute taxes for single filers. Complete this program to compute taxes for all filing statuses.

3.14 (Game: heads or tails) Write a program that lets the user guess whether the flip of a coin results in heads or tails. The program randomly generates an integer 0 or 1, which represents head or tail. The program prompts the user to enter a guess, and reports whether the guess is correct or incorrect.

M03_LIAN1878_11_GE_C03.indd 132 12/29/17 3:55 PM

Programming Exercises 133

**3.15 (Game: lottery) Revise Listing 3.8, Lottery.java, to generate a lottery of a three-digit number. The program prompts the user to enter a three-digit number and determines whether the user wins according to the following rules:

1. If the user input matches the lottery number in the exact order, the award is $12,000.

2. If all digits in the user input match all digits in the lottery number, the award is $5,000.

3. If one digit in the user input matches a digit in the lottery number, the award is $2,000.

3.16 (Random point) Write a program that displays a random coordinate in a rectangle. The rectangle is centred at (0, 0) with width 50 and height 150.

*3.17 (Game: scissor, rock, paper) Write a program that plays the popular scissor–rock– paper game. (A scissor can cut a paper, a rock can knock a scissor, and a paper can wrap a rock.) The program randomly generates a number 0, 1, or 2 representing scissor, rock, and paper. The program prompts the user to enter a number 0, 1, or 2 and displays a message indicating whether the user or the computer wins, loses, or draws. Here are sample runs:

scissor (0), rock (1), paper (2): 1

The computer is scissor. You are rock. You won

scissor (0), rock (1), paper (2): 2

The computer is paper. You are paper too. It is a draw

*3.18 (Cost of shipping) A shipping company uses the following function to calculate the cost (in dollars) of shipping based on the weight of the package (in pounds).

c(w) = d 2.5, if 0 6 w 6 = 24.5, if 2 6 w 6 = 4 7.5, if 4 6 w 6 = 10 10.5, if 10 6 w 6 = 20

Write a program that prompts the user to enter the weight of the package and display the shipping cost. If the weight is greater than 20, display a message “the package cannot be shipped.”

**3.19 (Compute the perimeter of a triangle) Write a program that reads three edges for a triangle and computes the perimeter if the input is valid. Otherwise, display that the input is invalid. The input is valid if the sum of every pair of two edges is greater than the remaining edge.

*3.20 (Science: wind-chill temperature) Programming Exercise 2.17 gives a formula to compute the wind-chill temperature. The formula is valid for temperatures in the range between -58°F and 41°F and wind speed greater than or equal to 2. Write a program that prompts the user to enter a temperature and a wind speed. The pro- gram displays the wind-chill temperature if the input is valid; otherwise, it displays a message indicating whether the temperature and/or wind speed is invalid.

M03_LIAN1878_11_GE_C03.indd 133 12/29/17 3:55 PM

134 Chapter 3 Selections

Comprehensive **3.21 (Science: day of the week) Zeller’s congruence is an algorithm developed by

Christian Zeller to calculate the day of the week. The formula is

h = aq + 26(m + 1) 10

+ k + k 4

+ j

4 + 5jb,7

where

■■ h is the day of the week (0: Saturday, 1: Sunday, 2: Monday, 3: Tuesday, 4: Wednesday, 5: Thursday, and 6: Friday).

■■ q is the day of the month. ■■ m is the month (3: March, 4: April, . . ., 12: December). January and February

are counted as months 13 and 14 of the previous year.

■■ j is year

100 .

■■ k is the year of the century (i.e., year % 100).

Note all divisions in this exercise perform an integer division. Write a program that prompts the user to enter a year, month, and day of the month, and displays the name of the day of the week. Here are some sample runs:

Enter year: (e.g., 2012): 2015

Enter month: 1-12: 1

Enter the day of the month: 1-31: 25

Day of the week is Sunday

Enter year: (e.g., 2012): 2012

Enter month: 1-12: 5

Enter the day of the month: 1-31: 12

Day of the week is Saturday

(Hint: January and February are counted as 13 and 14 in the formula, so you need to convert the user input 1 to 13 and 2 to 14 for the month and change the year to the previous year. For example, if the user enters 1 for m and 2015 for year, m will be 13 and year will be 2014 used in the formula.)

** 3.22 (Geometry: point in a circle?) Write a program that prompts the user to enter a point (x, y) and checks whether the point is within the circle centered at (0, 0) with radius 10. For example, (4, 5) is inside the circle and (9, 9) is outside the circle, as shown in Figure 3.7a.

(Hint: A point is in the circle if its distance to (0, 0) is less than or equal to 10. The formula for computing the distance is 2(x2 - x1)2 + (y2 - y1)2. Test your program to cover all cases.) Two sample runs are shown below:

VideoNote

Check point location

Enter a point with two coordinates: 4 5

Point (4.0, 5.0) is in the circle

Enter a point with two coordinates: 9 9

Point (9.0, 9.0) is not in the circle

M03_LIAN1878_11_GE_C03.indd 134 12/29/17 3:55 PM

** 3.23 (Geometry: point in a rectangle?) Write a program that prompts the user to enter a point (x, y) and checks whether the point is within the rectangle centred at (1, 1) with width 10 and height 5. For example, (2, 2) is inside the rectangle and (6, 4) is outside the rectangle, as shown in Figure 3.7b. (Hint: A point is in the rectangle if its horizontal distance to (0, 0) is less than or equal to 10 / 2 and its vertical distance to (0, 0) is less than or equal to 5.0 / 2. Test your program to cover all cases.) Here are two sample runs.

Figure 3.7 (a) Points inside and outside of the circle. (b) Points inside and outside of the rectangle.

x-axis(0, 0)

(a) (b)

y-axis

(4, 5)

(9, 9)

(2, 2) (6, 4)

x-axis

y-axis

(0, 0)

Enter a point with two coordinates: 2 2

Point (2.0, 2.0) is in the rectangle

Enter a point with two coordinates: 6 4

Point (6.0, 4.0) is not in the rectangle

**3.24 (Game: pick a card) Write a program that simulates picking a card from a deck of 52 cards. Your program should display the rank (Ace, 2, 3, 4, 5, 6, 7, 8, 9, 10, Jack, Queen, King) and suit (Clubs, Diamonds, Hearts, Spades) of the card. Here is a sample run of the program:

The card you picked is Jack of Hearts

*3.25 (Geometry: intersecting point) Two points on line 1 are given as (x1, y1) and (x2, y2) and on line 2 as (x3, y3) and (x4, y4), as shown in Figure 3.8a and b.

The intersecting point of the two lines can be found by solving the following linear equations:

(y1 - y2)x - (x1 - x2)y = (y1 - y2)x1 - (x1 - x2)y1 (y3 - y4)x - (x3 - x4)y = (y3 - y4)x3 - (x3 - x4)y3

This linear equation can be solved using Cramer’s rule (see Programming Exer- cise 3.3). If the equation has no solutions, the two lines are parallel (see Figure 3.8c).

Programming Exercises 135

M03_LIAN1878_11_GE_C03.indd 135 12/29/17 3:55 PM

136 Chapter 3 Selections

Write a program that prompts the user to enter four points and displays the intersect- ing point. Here are sample runs:

Figure 3.8 Two lines intersect in (a and b) and two lines are parallel in (c).

(x1, y1)

(x2, y2) (x3, y3)

(x4, y4)

(a) (b) (c)

(x1, y1)

(x2, y2)

(x3, y3)

(x4, y4) (x1, y1)

(x2, y2)

(x3, y3)

(x4, y4)

Enter x1, y1, x2, y2, x3, y3, x4, y4: 2 2 5 -1.0 4.0 2.0 -1.0 -2.0

The intersecting point is at (2.88889, 1.1111)

Enter x1, y1, x2, y2, x3, y3, x4, y4: 2 2 7 6.0 4.0 2.0 -1.0 -2.0

The two lines are parallel

Enter an integer: 10

Is 10 divisible by 4 and 5? false

Is 10 divisible by 4 or 5? true

Is 10 divisible by 4 or 5 but not both? true

3.26 (Use the &&, || and ^ operators) Write a program that prompts the user to enter an integer and determines whether it is divisible by 4 and 5, whether it is divisible by 4 or 5, and whether it is divisible by 4 or 5 but not both. Here is a sample run of this program:

** 3.27 (Geometry: points in triangle?) Suppose a right triangle is placed in a plane as shown below. The right-angle point is placed at (0, 0), and the other two points are placed at (200, 0) and (0, 100). Write a program that prompts the user to enter a point with x- and y-coordinates and determines whether the point is inside the triangle. Here are the sample runs:

(0, 100)

(0, 0) (200, 0)

Enter a point’s x- and y-coordinates: 100.5 25.5

The point is in the triangle

M03_LIAN1878_11_GE_C03.indd 136 12/29/17 3:55 PM

** 3.28 (Geometry: two rectangles) Write a program that prompts the user to enter the center x-, y-coordinates, width, and height of two rectangles and determines whether the second rectangle is inside the first or overlaps with the first, as shown in Figure 3.9. Test your program to cover all cases.

Enter a point’s x- and y-coordinates: 100.5 50.5

The point is not in the triangle

Figure 3.9 (a) A rectangle is inside another one. (b) A rectangle overlaps another one.

(a)

(x1, y1) (x2, y2)

h2h1

(b)

(x1, y1)

(x2, y2)

Here are the sample runs:

Enter r1’s center x-, y-coordinates, width, and height: 2.5 4 2.5 43

Enter r2’s center x-, y-coordinates, width, and height: 1.5 5 0.5 3

r2 is inside r1

Enter r1’s center x-, y-coordinates, width, and height: 1 2 3 5.5

Enter r2’s center x-, y-coordinates, width, and height: 3 4 4.5 5

r2 overlaps r1

Enter r1’s center x-, y-coordinates, width, and height: 1 2 3 3

Enter r2’s center x-, y-coordinates, width, and height: 40 45 3 2

r2 does not overlap r1

** 3.29 (Geometry: two circles) Write a program that prompts the user to enter the center coordinates and radii of two circles and determines whether the second circle is inside the first or overlaps with the first, as shown in Figure 3.10. (Hint: circle2 is inside circle1 if the distance between the two centers 6 = r1 - r2 and circle2 overlaps circle1 if the distance between the two centers 6 = r1 + r2. Test your program to cover all cases.)

Here are the sample runs:

Enter circle1’s center x-, y-coordinates, and radius: 0.5 5.1 13

Enter circle2’s center x-, y-coordinates, and radius: 1 1.7 4.5

circle2 is inside circle1

Programming Exercises 137

M03_LIAN1878_11_GE_C03.indd 137 12/29/17 3:55 PM

138 Chapter 3 Selections

*3.30 (Current time) Revise Programming Exercise 2.8 to display the hour using a 12-hour clock. Here is a sample run:

Figure 3.10 (a) A circle is inside another circle. (b) A circle overlaps another circle.

(a) (b)

(x1, y1)

(x2, y2)

(x1, y1)

(x2, y2)

Enter circle1’s center x-, y-coordinates, and radius: 3.4 5.7 5.5

Enter circle2’s center x-, y-coordinates, and radius: 6.7 3.5 3

circle2 overlaps circle1

Enter circle1’s center x-, y-coordinates, and radius: 3.4 5.5 1

Enter circle2’s center x-, y-coordinates, and radius: 5.5 7.2 1

circle2 does not overlap circle1

*3.31 (Financials: currency exchange) Write a program that prompts the user to enter the exchange rate from currency in U.S. dollars to Chinese RMB. Prompt the user to enter 0 to convert from U.S. dollars to Chinese RMB and 1 to convert from Chinese RMB to U.S. dollars. Prompt the user to enter the amount in U.S. dollars or Chinese RMB to convert it to Chinese RMB or U.S. dollars, respectively. Here are the sample runs:

Enter the time zone offset to GMT: -5

The current time is 4:50:34 AM

Enter the exchange rate from dollars to RMB: 6.81

Enter 0 to convert dollars to RMB and 1 vice versa: 1

Enter the RMB amount: 10000

10000.0 yuan is $1468.43

Enter the exchange rate from dollars to RMB: 6.81

Enter 0 to convert dollars to RMB and 1 vice versa: 0

Enter the dollar amount: 100

$100.0 is 681.0 yuan

M03_LIAN1878_11_GE_C03.indd 138 12/29/17 3:55 PM

*3.32 (Geometry: point position) Given a directed line from point p0(x0, y0) to p1(x1, y1), you can use the following condition to decide whether a point p2(x2, y2) is on the left of the line, on the right, or on the same line (see Figure 3.11):

(x1 - x0)*(y2 - y0) - (x2 - x0)*(y1 - y0) c 70 p2 is on the left side of the line=0 p2 is on the same line 60 p2 is on the right side of the line

Enter the exchange rate from dollars to RMB: 6.81

Enter 0 to convert dollars to RMB and 1 vice versa: 5

Incorrect input

Figure 3.11 (a) p2 is on the left of the line. (b) p2 is on the right of the line. (c) p2 is on the same line.

p2 p1

(a) (b) (c)

Write a program that prompts the user to enter the three points for p0, p1, and p2 and displays whether p2 is on the left of the line from p0 to p1, to the right, or on the same line. Here are some sample runs:

Enter three points for p0, p1, and p2: 4.4 2 6.5 9.5 -5 4

p2 is on the left side of the line

Enter three points for p0, p1, and p2: 1 1 5 5 2 2

p2 is on the same line

Enter three points for p0, p1, and p2: 3.4 2 6.5 9.5 5 2.5

p2 is on the right side of the line

*3.33 (Financial: compare costs) Suppose you shop for rice in two different packages. You would like to write a program to compare the cost. The program prompts the user to enter the weight and price of each package and displays the one with the better price. Here is a sample run:

Enter weight and price for package 1: 50 24.59

Enter weight and price for package 2: 25 11.99

Package 2 has a better price.

Programming Exercises 139

M03_LIAN1878_11_GE_C03.indd 139 12/29/17 3:55 PM

140 Chapter 3 Selections

*3.34 (Geometry: point on line segment) Exercise 3.32 shows how to test whether a point is on an unbounded line. Revise Exercise 3.32 to test whether a point is on a line segment. Write a program that prompts the user to enter the three points for p0, p1, and p2 and displays whether p2 is on the line segment from p0 to p1. Here are some sample runs:

Enter weight and price for package 1: 50 25

Enter weight and price for package 2: 25 12.5

Two packages have the same price.

Enter three points for p0, p1, and p2: 1 1 2.5 2.5 1.5 1.5

(1.5, 1.5) is on the line segment from (1.0, 1.0) to (2.5, 2.5)

Enter an integer: 200

200 is even number

Enter an integer: 211

211 is odd number

Enter three points for p0, p1, and p2: 1 1 2 2 3.5 3.5

(3.5, 3.5) is not on the line segment from (1.0, 1.0) to (2.0, 2.0)

3.35 (Even or odd number) Write a program that prompts the user to enter an integer and displays whether the integer is an odd number or not. Here are the sample runs:

Note More than 200 additional programming exercises with solutions are provided to the instructors on the Instructor Resource Website.

M03_LIAN1878_11_GE_C03.indd 140 12/29/17 3:55 PM

Mathematical Functions, Characters, and Strings

Objectives ■■ To solve mathematical problems by using the methods in the Math class (§4.2).

■■ To represent characters using the char type (§4.3).

■■ To encode characters using ASCII and Unicode (§4.3.1).

■■ To represent special characters using the escape sequences (§4.3.2).

■■ To cast a numeric value to a character and cast a character to an integer (§4.3.3).

■■ To compare and test characters using the static methods in the Character class (§4.3.4).

■■ To introduce objects and instance methods (§4.4).

■■ To represent strings using the String object (§4.4).

■■ To return the string length using the length() method (§4.4.1).

■■ To return a character in the string using the charAt(i) method (§4.4.2).

■■ To use the + operator to concatenate strings (§4.4.3).

■■ To return an uppercase string or a lowercase string and to trim a string (§4.4.4).

■■ To read strings from the console (§4.4.5).

■■ To read a character from the console (§4.4.6).

■■ To compare strings using the equals and the compareTo methods (§4.4.7).

■■ To obtain substrings (§4.4.8).

■■ To find a character or a substring in a string using the indexOf method (§4.4.9).

■■ To program using characters and strings (GuessBirthday) (§4.5.1).

■■ To convert a hexadecimal character to a decimal value (HexDigit2Dec) (§4.5.2).

■■ To revise the lottery program using strings (LotteryUsingStrings) (§4.5.3).

■■ To format output using the System.out.printf method (§4.6).

Chapter

M04_LIAN1878_11_GE_C04.indd 141 12/29/17 3:56 PM

142 Chapter 4 Mathematical Functions, Characters, and Strings

4.1 Introduction The focus of this chapter is to introduce mathematical functions, characters, string objects, and use them to develop programs.

The preceding chapters introduced fundamental programming techniques and taught you how to write simple programs to solve basic problems using selection statements. This chapter introduces methods for performing common mathematical operations. You will learn how to create custom methods in Chapter 6.

Suppose you need to estimate the area enclosed by four cities, given the GPS locations (lati- tude and longitude) of these cities, as shown in the following diagram. How would you write a program to solve this problem? You will be able to write such a program in this chapter.

Point Key

problem

Orlando (28.5383355, –81.3792365)

Savannah (32.0835407, –81.0998342)

Charlotte (35.2270869, –80.8431267)

Atlanta (33.7489954, –84.3879824)

Method Description

sin(radians) Returns the trigonometric sine of an angle in radians.

cos(radians) Returns the trigonometric cosine of an angle in radians.

tan(radians) Returns the trigonometric tangent of an angle in radians.

toRadians(degree) Returns the angle in radians for the angle in degrees.

toDegrees(radians) Returns the angle in degrees for the angle in radians.

asin(a) Returns the angle in radians for the inverse of sine.

acos(a) Returns the angle in radians for the inverse of cosine.

atan(a) Returns the angle in radians for the inverse of tangent.

Table 4.1 Trigonometric Methods in the Math Class

Because strings are frequently used in programming, it is beneficial to introduce strings early so that you can begin to use them to develop useful programs. This chapter also gives a brief introduction to string objects; you will learn more on objects and strings in Chapters 9 and 10.

4.2 Common Mathematical Functions Java provides many useful methods in the Math class for performing common math- ematical functions.

A method is a group of statements that performs a specific task. You have already used the pow(a, b) method to compute ab in Section 2.9.4, Exponent Operations and the random() method for generating a random number in Section 3.7. This section introduces other useful methods in the Math class. They can be categorized as trigonometric methods, exponent meth- ods, and service methods. Service methods include the rounding, min, max, absolute, and ran- dom methods. In addition to methods, the Math class provides two useful double constants, PI and E (the base of natural logarithms). You can use these constants as Math.PI and Math.E in any program.

4.2.1 Trigonometric Methods The Math class contains the following methods as listed in Table 4.1 for performing t rigonometric functions:

Point Key

VideoNote

Introduce Math functions

M04_LIAN1878_11_GE_C04.indd 142 12/29/17 3:56 PM

4.2 Common Mathematical Functions 143

The parameter for sin, cos, and tan is an angle in radians. The return value for asin and atan is an angle in radians in the range between -p/2 and p/2, and for acos is between 0 and p. One degree is equal to p/180 in radians, 90 degrees is equal to p/2 in radians, and 30 degrees is equal to p/6 in radians.

For example,

Math.toDegrees(Math.PI / 2) returns 90.0 Math.toRadians(30) returns 0.5236 (same as π/6) Math.sin(0) returns 0.0 Math.sin(Math.toRadians(270)) returns −1.0 Math.sin(Math.PI / 6) returns 0.5 Math.sin(Math.PI / 2) returns 1.0 Math.cos(0) returns 1.0 Math.cos(Math.PI / 6) returns 0.866 Math.cos(Math.PI / 2) returns 0 Math.asin(0.5) returns 0.523598333 (same as π/6) Math.acos(0.5) returns 1.0472 (same as π/3) Math.atan(1.0) returns 0.785398 (same as π/4)

4.2.2 Exponent Methods There are five methods related to exponents in the Math class as listed in Table 4.2.

Method Description

exp(x) Returns e raised to power of x (ex).

log(x) Returns the natural logarithm of x (ln(x) = loge(x)). log10(x) Returns the base 10 logarithm of x (log10(x)).

pow(a, b) Returns a raised to the power of b (ab). sqrt(x) Returns the square root of x (2x) for x 7 = 0.

Table 4.2 Exponent Methods in the Math Class

For example,

e3.5 is Math.exp(3.5), which returns 33.11545 ln(3.5) is Math.log(3.5), which returns 1.25276 log10 (3.5) is Math.log10(3.5), which returns 0.544 23 is Math.pow(2, 3), which returns 8.0 32 is Math.pow(3, 2), which returns 9.0 4.52.5 is Math.pow(4.5, 2.5), which returns 42.956724 is Math.sqrt(4), which returns 2.0210.5 is Math.sqrt(10.5), which returns 3.24

4.2.3 The Rounding Methods The Math class contains four rounding methods as listed in Table 4.3.

Method Description

ceil(x) x is rounded up to its nearest integer. This integer is returned as a double value.

floor(x) x is rounded down to its nearest integer. This integer is returned as a double value.

rint(x) x is rounded to its nearest integer. If x is equally close to two integers, the even one is returned as a double value.

round(x) Returns (int)Math.floor(x + 0.5) if x is a float and returns (long)Math.floor(x + 0.5) if x is a double.

Table 4.3 Rounding Methods in the Math Class

M04_LIAN1878_11_GE_C04.indd 143 12/29/17 3:56 PM

144 Chapter 4 Mathematical Functions, Characters, and Strings

For example,

Math.ceil(2.1) returns 3.0 Math.ceil(2.0) returns 2.0 Math.ceil(−2.0) returns −2.0 Math.ceil(−2.1) returns −2.0 Math.floor(2.1) returns 2.0 Math.floor(2.0) returns 2.0 Math.floor(−2.0) returns −2.0 Math.floor(−2.1) returns −3.0 Math.rint(2.1) returns 2.0 Math.rint(−2.0) returns −2.0 Math.rint(−2.1) returns −2.0 Math.rint(2.5) returns 2.0 Math.rint(4.5) returns 4.0 Math.rint(−2.5) returns −2.0 Math.round(2.6f) returns 3 // Returns int Math.round(2.0) returns 2 // Returns long Math.round(−2.0f) returns −2 // Returns int Math.round(−2.6) returns −3 // Returns long Math.round(−2.4) returns −2 // Returns long

4.2.4 The min, max, and abs Methods The min and max methods return the minimum and maximum numbers of two numbers (int, long, float, or double). For example, max(4.4, 5.0) returns 5.0, and min(3, 2) returns 2.

The abs method returns the absolute value of the number (int, long, float, or double). For example,

Math.max(2, 3) returns 3 Math.min(2.5, 4.6) returns 2.5 Math.max(Math.max(2.5, 4.6), Math.min(3, 5.6)) returns 4.6 Math.abs(−2) returns 2 Math.abs(−2.1) returns 2.1

4.2.5 The random Method You used the random() method in the preceding chapter. This method generates a random double value greater than or equal to 0.0 and less than 1.0 (0 <= Math.random() < 1.0). You can use it to write a simple expression to generate random numbers in any range. For example,

(int)(Math.random() * 10) Returns a random integer

between 0 and 9.

50 + (int)(Math.random() * 50) Returns a random integer

between 50 and 99.

In general,

a + Math.random() * b Returns a random number between a

and a + b, excluding a + b.

M04_LIAN1878_11_GE_C04.indd 144 12/29/17 3:56 PM

4.2 Common Mathematical Functions 145

4.2.6 Case Study: Computing Angles of a Triangle You can use the math methods to solve many computational problems. Given the three sides of a triangle, for example, you can compute the angles by using the following formulas:

x1, y1

x2, y2

x3, y3

A = acos((a * a − b * b − c * c) / (−2 * b * c)) B = acos((b * b − a * a − c * c) / (−2 * a * c)) C = acos((c * c − b * b − a * a) / (−2 * a * b))

Don’t be intimidated by the mathematical formula. As we discussed early in Listing 2.9, ComputeLoan.java, you don’t have to know how the mathematical formula is derived in order to write a program for computing the loan payments. Here, in this example, given the length of three sides, you can use this formula to write a program to compute the angles without having to know how the formula is derived. In order to compute the lengths of the sides, we need to know the coordinates of three corner points and compute the distances between the points.

Listing 4.1 is an example of a program that prompts the user to enter the x- and y-coordinates of the three corner points in a triangle then displays the three angles.

lisTing 4.1 ComputeAngles.java 1 import java.util.Scanner; 2 3 public class ComputeAngles { 4 public static void main(String[] args) { 5 Scanner input = new Scanner(System.in); 6 7 // Prompt the user to enter three points 8 System.out.print("Enter three points: "); 9 double x1 = input.nextDouble(); 10 double y1 = input.nextDouble(); 11 double x2 = input.nextDouble(); 12 double y2 = input.nextDouble(); 13 double x3 = input.nextDouble(); 14 double y3 = input.nextDouble(); 15 16 // Compute three sides 17 double a = Math.sqrt((x2 − x3) * (x2 − x3) 18 + (y2 − y3) * (y2 − y3)); 19 double b = Math.sqrt((x1 − x3) * (x1 − x3) 20 + (y1 − y3) * (y1 − y3)); 21 double c = Math.sqrt((x1 − x2) * (x1 − x2) 22 + (y1 − y2) * (y1 − y2)); 23 24 // Compute three angles 25 double A = Math.toDegrees(Math.acos((a * a − b * b − c * c) 26 / (−2 * b * c))); 27 double B = Math.toDegrees(Math.acos((b * b − a * a − c * c) 28 / (−2 * a * c))); 29 double C = Math.toDegrees(Math.acos((c * c − b * b − a * a) 30 / (−2 * a * b))); 31

compute sides

enter three points

M04_LIAN1878_11_GE_C04.indd 145 12/29/17 3:56 PM

146 Chapter 4 Mathematical Functions, Characters, and Strings

32 // Display results 33 System.out.println("The three angles are " + 34 Math.round(A * 100) / 100.0 + " " + 35 Math.round(B * 100) / 100.0 + " " + 36 Math.round(C * 100) / 100.0); 37 } 38 }

display result

Enter three points: 1 1 6.5 1 6.5 2.5

The three angles are 15.26 90.0 74.74

The program prompts the user to enter three points (line 8). This prompting message is not clear. You should give the user explicit instructions on how to enter these points as follows:

Sy stem.out.print("Enter the coordinates of three points separated " + "by spaces like x1 y1 x2 y2 x3 y3: ");

Note that the distance between two points (x1, y1) and (x2, y2) can be computed using the formula 2(x2 - x1)2 + (y2 - y1)2. The program computes the distances between two points (lines 17–22), and applies the formula to compute the angles (lines 25–30). The angles are rounded to display up to two digits after the decimal point (lines 34–36).

The Math class is used in the program, but not imported, because it is in the java.lang package. All the classes in the java.lang package are implicitly imported in a Java program.

4.2.1 Evaluate the following method calls: Point

Check

(a) Math.sqrt(4)

(b) Math.sin(2 * Math.PI)

(d) Math.pow(2, 2)

(e) Math.log(Math.E)

(f) Math.exp(1)

(g) Math.max(2, Math.min(3, 4))

(h) Math.rint(−2.5)

(i) Math.ceil(−2.5)

(j) Math.floor(−2.5)

(k) Math.round(−2.5f)

(l) Math.round(−2.5)

(m) Math.rint(2.5)

(n) Math.ceil(2.5)

(o) Math.floor(2.5)

(p) Math.round(2.5f)

(q) Math.round(2.5)

(r) Math.round(Math.abs(−2.5))

4.2.2 True or false? The argument for trigonometric methods is an angle in radians. 4.2.3 Write a statement that converts 47 degrees to radians and assigns the result to a

variable.

4.2.4 Write a statement that converts PI to an angle in degrees and assigns the result to a variable.

4.2.5 Write an expression that obtains a random integer between 34 and 55. Write an expression that obtains a random integer between 0 and 999. Write an expression that obtains a random number between 5.5 and 55.5.

4.2.6 Why does the Math class not need to be imported? 4.2.7 What is Math.log(Math.exp(5.5))? What is Math.exp(Math.log(5.5))? What is Math.asin(Math.sin(Math.PI / 6))? What is Math.sin(Math.asin(Math.PI / 6))?

M04_LIAN1878_11_GE_C04.indd 146 12/29/17 3:57 PM

4.3 Character Data Type and Operations 147

4.3 Character Data Type and Operations A character data type represents a single character.

In addition to processing numeric values, you can process characters in Java. The character data type, char, is used to represent a single character. A character literal is enclosed in single quotation marks. Consider the following code:

char letter = 'A'; char numChar = '4';

The first statement assigns character A to the char variable letter. The second statement assigns digit character 4 to the char variable numChar.

Caution A string literal must be enclosed in double quotation marks (" "). A character literal is a single character enclosed in single quotation marks (' '). Therefore, "A" is a string, but 'A' is a character.

4.3.1 Unicode and ASCII code Computers use binary numbers internally. A character is stored in a computer as a sequence of 0s and 1s. Mapping a character to its binary representation is called encoding. There are different ways to encode a character. How characters are encoded is defined by an encoding scheme.

Java supports Unicode, an encoding scheme established by the Unicode Consortium to support the interchange, processing, and display of written texts in the world’s diverse lan- guages. Unicode was originally designed as a 16-bit character encoding. The primitive data type char was intended to take advantage of this design by providing a simple data type that could hold any character. However, it turned out that the 65,536 characters possible in a 16-bit encoding are not sufficient to represent all the characters in the world. The Unicode standard therefore has been extended to allow up to 1,112,064 characters. Those characters that go beyond the original 16-bit limit are called supplementary characters. Java supports the supplementary characters. The processing and representing of supplementary characters are beyond the scope of this book. For simplicity, this book considers only the original 16-bit Unicode characters. These characters can be stored in a char type variable.

A 16-bit Unicode takes two bytes, preceded by \u, expressed in four hexadecimal digits that run from \u0000 to \uFFFF. Hexadecimal numbers are introduced in Appendix F, Number Systems. For example, the English word welcome is translated into Chinese using two characters, . The Unicodes of these two characters are \u6B22\u8FCE. The Unicodes for the Greek letters a b g are \u03b1 \u03b2 \u03b4 respectively.

Most computers use ASCII (American Standard Code for Information Interchange), an 8-bit encoding scheme, for representing all uppercase and lowercase letters, digits, punctuation marks, and control characters. Unicode includes ASCII code, with \u0000 to \u007F corre- sponding to the 128 ASCII characters. Table 4.4 shows the ASCII code for some commonly used characters. Appendix B, “The ASCII Character Set,” gives a complete list of ASCII characters and their decimal and hexadecimal codes.

Point Key

char type

char literal

encoding

Unicode

original Unicode

supplementary Unicode

ASCII

Characters Code Value in Decimal Unicode Value

'0' to '9' 48 to 57 \u0030 to \u0039

'A' to 'Z' 65 to 90 \u0041 to \u005A

'a' to 'z' 97 to 122 \u0061 to \u007A

Table 4.4 ASCII Code for Commonly Used Characters

M04_LIAN1878_11_GE_C04.indd 147 12/29/17 3:57 PM

148 Chapter 4 Mathematical Functions, Characters, and Strings

You can use ASCII characters such as 'X', '1', and '$' in a Java program as well as Unicodes. Thus, for example, the following statements are equivalent:

char letter = 'A'; char letter = '\u0041'; // Character A’s Unicode is 0041

Both statements assign character A to the char variable letter.

Note The increment and decrement operators can also be used on char variables to get the next or preceding Unicode character. For example, the following statements display character b:

char ch = 'a'; System.out.println(++ch);

4.3.2 Escape Sequences for Special Characters Suppose you want to print a message with quotation marks in the output. Can you write a statement like this?

System.out.println("He said "Java is fun"");

No, this statement has a compile error. The compiler thinks the second quotation character is the end of the string and does not know what to do with the rest of the characters.

To overcome this problem, Java uses a special notation to represent special characters, as listed in Table 4.5. This special notation, called an escape sequence, consists of a backslash (\) followed by a character or a combination of digits. For example, \t is an escape sequence for the Tab character, and an escape sequence such as \u03b1 is used to represent a Unicode. The symbols in an escape sequence are interpreted as a whole rather than individually. An escape sequence is considered as a single character.

So, now you can print the quoted message using the following statement:

System.out.println("He said \"Java is fun\"");

The output is

He said "Java is fun"

Note the symbols \ and " together represent one character.

char increment and decrement

escape sequence

Escape Sequence Name Unicode Code Decimal Value

\b Backspace \u0008 8

\t Tab \u0009 9

\n Linefeed \u000A 10

\f Formfeed \u000C 12

\r Carriage Return \u000D 13

\\ Backslash \u005C 92

\" Double Quote \u0022 34

Table 4.5 Escape Sequences

The backslash \ is called an escape character. It is a special character. To display this character, you have to use an escape sequence \\. For example, the following code

System.out.println("\\t is a tab character");

displays

\t is a tab character

escape character

M04_LIAN1878_11_GE_C04.indd 148 12/29/17 3:57 PM

4.3 Character Data Type and Operations 149

4.3.3 Casting between char and Numeric Types A char can be cast into any numeric type, and vice versa. When an integer is cast into a char, only its lower 16 bits of data are used; the other part is ignored. For example:

// Note a hex integer is written using prefix 0X char ch = (char)0XAB0041; // The lower 16 bits hex code 0041 is

// assigned to ch System.out.println(ch); // ch is character A

When a floating-point value is cast into a char, the floating-point value is first cast into an int, which is then cast into a char.

char ch = (char)65.25; // Decimal 65 is assigned to ch System.out.println(ch); // ch is character A

When a char is cast into a numeric type, the character’s Unicode is cast into the specified numeric type.

int i = (int)'A'; // The Unicode of character A is assigned to i System.out.println(i); // i is 65

Implicit casting can be used if the result of a casting fits into the target variable. Otherwise, explicit casting must be used. For example, since the Unicode of 'a' is 97, which is within the range of a byte, these implicit castings are fine:

byte b = 'a'; int i = 'a';

But the following statement is incorrect, because the Unicode \uFFF4 cannot fit into a byte:

byte b = '\uFFF4';

To force this assignment, use explicit casting, as follows:

byte b = (byte)'\uFFF4';

Any positive integer between 0 and FFFF in hexadecimal can be cast into a character implicitly. Any number not in this range must be cast into a char explicitly.

All numeric operators can be applied to char operands. A char operand is automatically cast into a number if the other operand is a number or a character. If the other operand is a string, the character is concatenated with the string. For example, the following statements

int i = '2' + '3'; // (int)'2' is 50 and (int)'3' is 51 System.out.println("i is " + i); // i is 101 int j = 2 + 'a'; // (int)'a' is 97 System.out.println("j is " + j); // j is 99 System.out.println(j + " is the Unicode for character ") + (char)j); // 99 is the Unicode for character c System.out.println("Chapter " + '2');

display

i is 101 j is 99 99 is the Unicode for character c Chapter 2

numeric operators on characters

M04_LIAN1878_11_GE_C04.indd 149 12/29/17 3:57 PM

150 Chapter 4 Mathematical Functions, Characters, and Strings

4.3.4 Comparing and Testing Characters Two characters can be compared using the relational operators just like comparing two numbers. This is done by comparing the Unicodes of the two characters. For example,

'a' < 'b' is true because the Unicode for 'a' (97) is less than the Unicode for 'b' (98).

'a' < 'A' is false because the Unicode for 'a' (97) is greater than the Unicode for 'A' (65).

'1' < '8' is true because the Unicode for '1' (49) is less than the Unicode for '8' (56).

Often in the program, you need to test whether a character is a number, a letter, an uppercase letter, or a lowercase letter. As given in Appendix B, the ASCII character set, that the Unicodes for lowercase letters are consecutive integers starting from the Unicode for 'a', then for 'b', 'c', . . . , and 'z'. The same is true for the uppercase letters and for numeric characters. This property can be used to write the code to test characters. For example, the following code tests whether a character ch is an uppercase letter, a lowercase letter, or a digital character:

if (ch >= 'A' && ch <= 'Z') System.out.println(ch + " is an uppercase letter"); else if (ch >= 'a' && ch <= 'z') System.out.println(ch + " is a lowercase letter"); else if (ch >= '0' && ch <= '9') System.out.println(ch + " is a numeric character");

For convenience, Java provides the following methods in the Character class for testing characters as listed in Table 4.6. The Character class is defined in the java.lang package.

Method Description

isDigit(ch) Returns true if the specified character is a digit.

isLetter(ch) Returns true if the specified character is a letter.

isLetterOrDigit(ch) Returns true if the specified character is a letter or digit.

isLowerCase(ch) Returns true if the specified character is a lowercase letter.

isUpperCase(ch) Returns true if the specified character is an uppercase letter.

toLowerCase(ch) Returns the lowercase of the specified character.

toUpperCase(ch) Returns the uppercase of the specified character.

Table 4.6 Methods in the Character Class

For example,

System.out.println("isDigit('a') is " + Character.isDigit('a')); System.out.println("isLetter('a') is " + Character.isLetter('a')); System.out.println("isLowerCase('a') is " + Character.isLowerCase('a')); System.out.println("isUpperCase('a') is " + Character.isUpperCase('a')); System.out.println("toLowerCase('T') is " + Character.toLowerCase('T')); System.out.println("toUpperCase('q') is " + Character.toUpperCase('q'));

displays

isDigit('a') is false isLetter('a') is true

M04_LIAN1878_11_GE_C04.indd 150 12/29/17 3:57 PM

4.3 Character Data Type and Operations 151

isLowerCase('a') is true isUpperCase('a') is false toLowerCase('T') is t toUpperCase('q') is Q

4.3.1 Use print statements to find out the ASCII code for '1', 'A', 'B', 'a', and 'b'. Use print statements to find out the character for the decimal codes 40, 59, 79, 85, and 90. Use print statements to find out the character for the hexadecimal code 40, 5A, 71, 72, and 7A.

4.3.2 Which of the following are correct literals for characters?

'1', '\u345dE', '\u3fFa', '\b', '\t'

4.3.3 How do you display the characters \ and "? 4.3.4 Evaluate the following:

int i = '1'; int j = '1' + '2' * ('4' − '3') + 'b' / 'a'; int k = 'a'; char c = 90;

4.3.5 Can the following conversions involving casting be allowed? If so, find the converted result.

char c = 'A'; int i = (int)c;

float f = 1000.34f; int i = (int)f;

double d = 1000.34; int i = (int)d;

int i = 97; char c = (char)i;

4.3.6 Show the output of the following program:

public class Test { public static void main(String[] args) { char x = 'a'; char y = 'c'; System.out.println(++x); System.out.println(y++); System.out.println(x − y); } }

4.3.7 Write the code that generates a random lowercase letter. 4.3.8 Show the output of the following statements:

System.out.println('a' < 'b'); System.out.println('a' <= 'A'); System.out.println('a' > 'b'); System.out.println('a' >= 'A'); System.out.println('a' == 'a'); System.out.println('a' != 'b');

Point Check

M04_LIAN1878_11_GE_C04.indd 151 12/29/17 3:57 PM

152 Chapter 4 Mathematical Functions, Characters, and Strings

4.4 The String Type A string is a sequence of characters.

The char type represents only one character. To represent a string of characters, use the data type called String. For example, the following code declares message to be a string with the value “Welcome to Java".

String message = "Welcome to Java";

String is a predefined class in the Java library, just like the classes System and Scanner. The String type is not a primitive type. It is known as a reference type. Any Java class can be used as a reference type for a variable. The variable declared by a reference type is known as a reference variable that references an object. Here, message is a reference variable that references a string object with contents Welcome to Java.

Reference data types will be discussed in detail in Chapter 9, Objects and Classes. For the time being, you need to know only how to declare a String variable, how to assign a string to the variable, and how to use the methods in the String class. More details on using strings will be covered in Chapter 10.

Table 4.7 lists the String methods for obtaining string length, for accessing characters in the string, for concatenating string, for converting string to uppercases or lowercases, and for trimming a string.

Point Key

VideoNote

Introduce strings and objects

Method Description

length() Returns the number of characters in this string.

charAt(index) Returns the character at the specified index from this string.

concat(s1) Returns a new string that concatenates this string with string s1.

toUpperCase() Returns a new string with all letters in uppercase.

toLowerCase() Returns a new string with all letters in lowercase.

trim() Returns a new string with whitespace characters trimmed on both sides.

Table 4.7 Simple Methods for String Objects

Strings are objects in Java. The methods listed in Table 4.7 can only be invoked from a specific string instance. For this reason, these methods are called instance methods. A noninstance method is called a static method. A static method can be invoked without using an object. All the methods defined in the Math class are static methods. They are not tied to a specific object instance. The syntax to invoke an instance method is referenceVariable.methodName(arguments). A method may have many arguments or no arguments. For example, the charAt(index)method has one argument, but the length() method has no arguments. Recall that the syntax to invoke a static method is ClassName.methodName(arguments). For example, the pow method in the Math class can be invoked using Math.pow(2, 2.5).

4.4.1 Getting String Length You can use the length() method to return the number of characters in a string. For example, the following code

String message = "Welcome to Java"; System.out.println("The length of " + message + " is " + message.length());

instance method static method

M04_LIAN1878_11_GE_C04.indd 152 12/29/17 3:57 PM

4.4 The String Type 153

displays

The length of Welcome to Java is 15

Note When you use a string, you often know its literal value. For convenience, Java allows you to use the string literal to refer directly to strings without creating new variables. Thus, "Welcome to Java".length() is correct and returns 15. Note that "" denotes an empty string and "".length() is 0.

4.4.2 Getting Characters from a String The s.charAt(index) method can be used to retrieve a specific character in a string s, where the index is between 0 and s.length()–1. For example, message.charAt(0) returns the character W, as shown in Figure 4.1. Note that the index for the first character in the string is 0.

string literal

empty string

charAt(index)

Figure 4.1 The characters in a String object can be accessed using its index.

Indices message

message.charAt(0) message.charAt(14)message.length() is 15

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14

W e l c o m e t o J a v a

Caution Attempting to access characters in a string s out of bounds is a common pro gramming error. To avoid it, make sure that you do not use an index beyond s.length()–1. For example, s.charAt(s.length()) would cause a StringIndexOutOfBoundsException.

4.4.3 Concatenating Strings You can use the concat method to concatenate two strings. The statement given below, for example, concatenates strings s1 and s2 into s3:

String s3 = s1.concat(s2);

Because string concatenation is heavily used in programming, Java provides a convenient way to accomplish it. You can use the plus (+) operator to concatenate two strings, so the previous statement is equivalent to

String s3 = s1 + s2;

The following code combines the strings message, " and ", and "HTML" into one string:

String myString = message + " and " + "HTML";

Recall that the + operator can also concatenate a number with a string. In this case, the number is converted into a string then concatenated. Note at least one of the operands must be a string in order for concatenation to take place. If one of the operands is a nonstring (e.g., a

string index range

s1.concat(s2)

s1 + s2

concatenate strings and numbers

M04_LIAN1878_11_GE_C04.indd 153 12/29/17 3:57 PM

154 Chapter 4 Mathematical Functions, Characters, and Strings

number), the nonstring value is converted into a string and concatenated with the other string. Here are some examples:

// Three strings are concatenated String message = "Welcome " + "to " + "Java";

// String Chapter is concatenated with number 2 String s = "Chapter" + 2; // s becomes Chapter2

// String Supplement is concatenated with character B String s1 = "Supplement" + 'B'; // s1 becomes SupplementB

If neither of the operands is a string, the plus sign (+) is the addition operator that adds two numbers.

The augmented += operator can also be used for string concatenation. For example, the following code appends the string " and Java is fun" with the string "Welcome to Java" in message.

message += " and Java is fun";

So the new message is "Welcome to Java and Java is fun." If i = 1 and j = 2, what is the output of the following statement?

System.out.println("i + j is " + i + j);

The output is "i + j is 12" because "i + j is" is concatenated with the value of i first. To force i + j to be executed first, enclose i + j in the parentheses, as follows:

System.out.println("i + j is " + (i + j));

4.4.4 Converting Strings The toLowerCase() method returns a new string with all lowercase letters, and the toUpperCase() method returns a new string with all uppercase letters. For example,

"Welcome".toLowerCase() returns a new string welcome. "Welcome".toUpperCase() returns a new string WELCOME.

The trim() method returns a new string by eliminating whitespace characters from both ends of the string. The characters ' ', \t, \f, \r, or \n are known as whitespace characters. For example,

"\t Good Night \n".trim() returns a new string Good Night.

4.4.5 Reading a String from the Console To read a string from the console, invoke the next() method on a Scanner object. For exam- ple, the following code reads three strings from the keyboard:

Scanner input = new Scanner(System.in); System.out.print("Enter three words separated by spaces: "); String s1 = input.next(); String s2 = input.next(); String s3 = input.next(); System.out.println("s1 is " + s1); System.out.println("s2 is " + s2); System.out.println("s3 is " + s3);

toLowerCase() toUpperCase()

whitespace character

trim()

read strings

M04_LIAN1878_11_GE_C04.indd 154 12/29/17 3:57 PM

4.4 The String Type 155

The next() method reads a string that ends with a whitespace character. You can use the nextLine() method to read an entire line of text. The nextLine() method reads a string that ends with the Enter key pressed. For example, the following statements read a line of text:

Scanner input = new Scanner(System.in); System.out.println("Enter a line: "); String s = input.nextLine(); System.out.println("The line entered is " + s);

Enter three words separated by spaces: Welcome to Java

s1 is Welcome s2 is to s3 is Java

Enter a line: Welcome to Java

The line entered is Welcome to Java

Method Description

equals(s1) Returns true if this string is equal to string s1.

equalsIgnoreCase(s1) Returns true if this string is equal to string s1; it is case insensitive.

compareTo(s1) Returns an integer greater than 0, equal to 0, or less than 0 to indicate whether this string is greater than, equal to, or less than s1.

compareToIgnoreCase(s1) Same as compareTo except that the comparison is case insensitive.

startsWith(prefix) Returns true if this string starts with the specified prefix.

endsWith(suffix) Returns true if this string ends with the specified suffix.

contains(s1) Returns true if s1 is a substring in this string.

Table 4.8 Comparison Methods for String Objects

For convenience, we call the input using the methods next(), nextByte(), nextShort(), nextInt(), nextLong(), nextFloat(), and nextDouble() the token-based input, because they read individual elements separated by whitespace characters rather than an entire line. The nextLine() method is called a line-based input.

Important Caution To avoid input errors, do not use a line-based input after a token-based input in the program. The reasons will be explained in Section 12.11.4, "How Does Scanner Work?"

4.4.6 Reading a Character from the Console To read a character from the console, use the nextLine() method to read a string and then invoke the charAt(0) method on the string to return a character. For example, the following code reads a character from the keyboard:

Scanner input = new Scanner(System.in); System.out.print("Enter a character: "); String s = input.nextLine(); char ch = s.charAt(0); System.out.println("The character entered is " + ch);

4.4.7 Comparing Strings The String class contains the methods, as listed in Table 4.8, for comparing two strings.

token-based input

line-based input

avoid input errors

M04_LIAN1878_11_GE_C04.indd 155 12/29/17 3:57 PM

156 Chapter 4 Mathematical Functions, Characters, and Strings

How do you compare the contents of two strings? You might attempt to use the == operator, as follows:

if (string1 == string2) System.out.println("string1 and string2 are the same object"); else System.out.println("string1 and string2 are different objects");

However, the == operator checks only whether string1 and string2 refer to the same object; it does not tell you whether they have the same contents. Therefore, you cannot use the == operator to find out whether two string variables have the same contents. Instead, you should use the equals method. The following code, for instance, can be used to compare two strings:

if (string1.equals(string2)) System.out.println("string1 and string2 have the same contents"); else System.out.println("string1 and string2 are not equal");

For example, the following statements display true then false:

String s1 = "Welcome to Java"; String s2 = "Welcome to Java"; String s3 = "Welcome to C++"; System.out.println(s1.equals(s2)); // true System.out.println(s1.equals(s3)); // false

The compareTo method can also be used to compare two strings. For example, consider the following code:

s1.compareTo(s2)

The method returns the value 0 if s1 is equal to s2, a value less than 0 if s1 is lexicographically (i.e., in terms of Unicode ordering) less than s2, and a value greater than 0 if s1 is lexicographically greater than s2.

The actual value returned from the compareTo method depends on the offset of the first two distinct characters in s1 and s2 from left to right. For example, suppose s1 is abc and s2 is abg, and s1.compareTo(s2) returns −4. The first two characters (a vs. a) from s1 and s2 are compared. Because they are equal, the second two characters (b vs. b) are compared. Because they are also equal, the third two characters (c vs. g) are compared. Since the character c is 4 less than g, the comparison returns −4.

Caution Syntax errors will occur if you compare strings by using relational operators >, >=, <, or <=. Instead, you have to use s1.compareTo(s2).

Note The equals method returns true if two strings are equal, and false if they are not. The compareTo method returns 0, a positive integer, or a negative integer, depending on whether one string is equal to, greater than, or less than the other string.

The String class also provides the equalsIgnoreCase and compareToIgnoreCase methods for comparing strings. The equalsIgnoreCase and compareToIgnoreCase methods ignore the case of the letters when comparing two strings. You can also use str.startsWith(prefix) to check whether string str starts with a specified prefix, str.endsWith(suffix) to check whether string str ends with a specified suffix, and str.contains(s1) to check whether string str contains string s1. For example,

"Welcome to Java".startsWith("We") returns true. "Welcome to Java".startsWith("we") returns false. "Welcome to Java".endsWith("va") returns true.

string1.equals(string2)

s1.compareTo(s2)

M04_LIAN1878_11_GE_C04.indd 156 12/29/17 3:57 PM

4.4 The String Type 157

"Welcome to Java".endsWith("v") returns false. "Welcome to Java".contains("to") returns true. "Welcome to Java".contains("To") returns false.

Listing 4.2 gives a program that prompts the user to enter two cities and displays them in alphabetical order.

lisTing 4.2 OrderTwoCities.java 1 import java.util.Scanner; 2 3 public class OrderTwoCities { 4 public static void main(String[] args) { 5 Scanner input = new Scanner(System.in); 6 7 // Prompt the user to enter two cities 8 System.out.print("Enter the first city: "); 9 String city1 = input.nextLine(); 10 System.out.print("Enter the second city: "); 11 String city2 = input.nextLine(); 12 13 if (city1.compareTo(city2) < 0) 14 System.out.println("The cities in alphabetical order are " + 15 city1 + " " + city2); 16 else 17 System.out.println("The cities in alphabetical order are " + 18 city2 + " " + city1); 19 } 20 }

Enter the first city: New York

Enter the second city: Boston

The cities in alphabetical order are Boston New York

The program reads two strings for two cities (lines 9 and 11). If input.nextLine() is replaced by input.next() (line 9), you cannot enter a string with spaces for city1. Since a city name may contain multiple words separated by spaces, the program uses the nextLine method to read a string (lines 9 and 11). Invoking city1.compareTo(city2) compares two strings city1 with city2 (line 13). A negative return value indicates that city1 is less than city2.

4.4.8 Obtaining Substrings You can obtain a single character from a string using the charAt method. You can also obtain a substring from a string using the substring method (see Figure 4.2) in the String class, as given in Table 4.9.

For example,

String message = "Welcome to Java"; String message = message.substring(0,11) + "HTML"; The string message now becomes Welcome to HTML.

input city1

input city2

compare two cities

Method Description

substring(beginIndex) Returns this string’s substring that begins with the character at the specified beginIndex and extends to the end of the string, as shown in Figure 4.2.

substring(beginIndex,

endIndex) Returns this string’s substring that begins at the specified beginIndex and extends to the character at index endIndex – 1, as shown in Figure 4.2. Note the character at endIndex is not part of the substring.

Table 4.9 The String Class Contains the Methods for Obtaining Substrings

M04_LIAN1878_11_GE_C04.indd 157 12/29/17 3:57 PM

158 Chapter 4 Mathematical Functions, Characters, and Strings

Note If beginIndex is endIndex, substring(beginIndex, endIndex) returns an empty string with length 0. If beginIndex > endIndex, it would be a runtime error.

4.4.9 Finding a Character or a Substring in a String The String class provides several versions of indexOf and lastIndexOf methods to find a character or a substring in a string, as listed in Table 4.10.

beginIndex <= endIndex

Figure 4.2 The substring method obtains a substring from a string.

Indices message

message.substring(0, 11) message.substring(11)

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14

W e l c o m e t o J a v a

Method Description

index Of (ch) Returns the index of the first occurrence of ch in the string. Returns −1 if not matched.

indexOf(ch, fromIndex) Returns the index of the first occurrence of ch after fromIndex in the string. Returns −1 if not matched.

indexOf(s) Returns the index of the first occurrence of string s in this string. Returns −1 if not matched.

indexOf(s, fromIndex) Returns the index of the first occurrence of string s in this string after fromIndex. Returns −1 if not matched.

lastIndexOf(ch) Returns the index of the last occurrence of ch in the string. Returns −1 if not matched.

lastIndexOf(ch, fromIndex) Returns the index of the last occurrence of ch before fromIndex in this string. Returns −1 if not matched.

lastIndexOf(s) Returns the index of the last occurrence of string s. Returns −1 if not matched.

lastIndexOf(s, fromIndex) Returns the index of the last occurrence of string s before fromIndex. Returns −1 if not matched.

Table 4.10 The String Class Contains the Methods for Finding Substrings

For example,

"Welcome to Java".indexOf('W') returns 0. "Welcome to Java".indexOf('o') returns 4. "Welcome to Java".indexOf('o', 5) returns 9. "Welcome to Java".indexOf("come") returns 3. "Welcome to Java".indexOf("Java", 5) returns 11. "Welcome to Java".indexOf("java", 5) returns −1.

"Welcome to Java".lastIndexOf('W') returns 0. "Welcome to Java".lastIndexOf('o') returns 9. "Welcome to Java".lastIndexOf('o', 5) returns 4. "Welcome to Java".lastIndexOf("come") returns 3. "Welcome to Java".lastIndexOf("Java", 5) returns −1. "Welcome to Java".lastIndexOf("Java") returns 11.

Suppose that a string s contains the first name and last name separated by a space. You can use the following code to extract the first name and last name from the string:

int k = s.indexOf(' '); String firstName = s.substring(0, k); String lastName = s.substring(k + 1);

indexOf

lastIndexOf

M04_LIAN1878_11_GE_C04.indd 158 12/29/17 3:57 PM

4.4 The String Type 159

For example, if s is Kim Jones, the following diagram illustrates how the first name and last name are extracted.

012345678

k is 3

s.substring(0, k) is Kim

s.substring(k + 1) is Jones

Kim Jones

4.4.10 Conversion between Strings and Numbers You can convert a numeric string into a number. To convert a string into an int value, use the Integer.parseInt method, as follows:

int intValue = Integer.parseInt(intString);

where intString is a numeric string such as "123". To convert a string into a double value, use the Double.parseDouble method, as

follows:

double doubleValue = Double.parseDouble(doubleString);

where doubleString is a numeric string such as "123.45". If the string is not a numeric string, the conversion would cause a runtime error. The

Integer and Double classes are both included in the java.lang package, and thus they are automatically imported.

You can convert a number into a string; simply use the string concatenating operator as follows:

String s = number + "";

4.4.1 Suppose s1, s2, and s3 are three strings, given as follows:

String s1 = "Welcome to Java"; String s2 = "Programming is fun"; String s3 = "Welcome to Java";

What are the results of the following expressions?

(a) s1 == s2

(b) s2 == s3

(d) s2.equals(s3)

(e) s1.compareTo(s2)

(f) s2.compareTo(s3)

(g) s2.compareTo(s2)

(h) s1.charAt(0)

(i) s1.indexOf('j')

(j) s1.indexOf("to")

(k) s1.lastIndexOf('a')

(l) s1.lastIndexOf("o", 15)

(m) s1.length()

(n) s1.substring(5)

(o) s1.substring(5, 11)

(p) s1.startsWith("Wel")

(q) s1.endsWith("Java")

(r) s1.toLowerCase()

(s) s1.toUpperCase()

(t) s1.concat(s2)

(u) s1.contain(s2)

(v) "\t Wel \t".trim()

Integer.parseInt method

Double.parseDouble method

number to string

Point Check

M04_LIAN1878_11_GE_C04.indd 159 12/29/17 3:57 PM

160 Chapter 4 Mathematical Functions, Characters, and Strings

4.4.2 Suppose s1 and s2 are two strings. Which of the following statements or expres- sions are incorrect?

String s = "Welcome to Java"; String s3 = s1 + s2; String s3 = s1 − s2; s1 == s2; s1 >= s2; s1.compareTo(s2); int i = s1.length(); char c = s1(0); char c = s1.charAt(s1.length());

4.4.3 Show the output of the following statements (write a program to verify your results):

System.out.println("1" + 1); System.out.println('1' + 1); System.out.println("1" + 1 + 1); System.out.println("1" + (1 + 1)); System.out.println('1' + 1 + 1);

4.4.4 Evaluate the following expressions (write a program to verify your results):

1 + "Welcome " + 1 + 1 1 + "Welcome " + (1 + 1) 1 + "Welcome " + ('\u0001' + 1) 1 + "Welcome " + 'a' + 1

4.4.5 Let s1 be " Welcome " and s2 be " welcome ". Write the code for the following statements:

(a) Check whether s1 is equal to s2 and assign the result to a Boolean variable isEqual.

(b) Check whether s1 is equal to s2, ignoring case, and assign the result to a Boolean variable isEqual.

(d) Compare s1 with s2, ignoring case, and assign the result to an int variable x.

(e) Check whether s1 has the prefix AAA and assign the result to a Boolean variable b.

(f) Check whether s1 has the suffix AAA and assign the result to a Boolean variable b.

(g) Assign the length of s1 to an int variable x.

(h) Assign the first character of s1 to a char variable x.

(i) Create a new string s3 that combines s1 with s2.

(j) Create a substring of s1 starting from index 1.

(k) Create a substring of s1 from index 1 to index 4.

(l) Create a new string s3 that converts s1 to lowercase.

(m) Create a new string s3 that converts s1 to uppercase.

(n) Create a new string s3 that trims whitespaces on both ends of s1.

M04_LIAN1878_11_GE_C04.indd 160 12/29/17 3:57 PM

4.5 Case Studies 161

(o) Assign the index of the first occurrence of the character e in s1 to an int variable x.

(p) Assign the index of the last occurrence of the string abc in s1 to an int variable x.

4.4.6 Write one statement to return the number of digits in an integer i. 4.4.7 Write one statement to return the number of digits in a double value d.

4.5 Case Studies Strings are fundamental in programming. The ability to write programs using strings is essential in learning Java programming.

You will frequently use strings to write useful programs. This section presents three examples of solving problems using strings.

4.5.1 Case Study: Guessing Birthdays You can find out the date of the month when your friend was born by asking five questions. Each question asks whether the day is in one of the five sets of numbers.

Point Key

16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

9 10 11 12 13 14 15 24 25 26 27 28 29 30 31

5 6 7 12 13 14 15 20 21 22 23 28 29 30 31

2 3 6 7 10 11 14 15 18 22 23 26 27 30 31

3 5 7 9 11 13 15

17 19 1921 23 25 27 29 31

Set1 Set2 Set3 Set4 Set5

= 19

841

The birthday is the sum of the first numbers in the sets where the day appears. For example, if the birthday is 19, it appears in Set1, Set2, and Set5. The first numbers in these three sets are 1, 2, and 16. Their sum is 19.

Listing 4.3 gives a program that prompts the user to answer whether the day is in Set1 (lines 41–44), in Set2 (lines 50–53), in Set3 (lines 59–62), in Set4 (lines 68–71), and in Set5 (lines 77–80). If the number is in the set, the program adds the first number in the set to day (lines 47, 56, 65, 74, and 83).

lisTing 4.3 GuessBirthday.java 1 import java.util.Scanner; 2 3 public class GuessBirthday { 4 public static void main(String[] args) { 5 String set1 = 6 " 1 3 5 7\n" + 7 " 9 11 13 15\n" + 8 "17 19 21 23\n" + 9 "25 27 29 31"; 10 11 String set2 = 12 " 2 3 6 7\n" +

M04_LIAN1878_11_GE_C04.indd 161 12/29/17 3:57 PM

162 Chapter 4 Mathematical Functions, Characters, and Strings

73 if (answer == 1) 74 day += 8; 75 76 // Prompt the user to answer questions 77 System.out.print("\nIs your birthday in Set5?\n"); 78 System.out.print(set5); 79 System.out.print("\nEnter 0 for No and 1 for Yes: "); 80 answer = input.nextInt(); 81 82 if (answer == 1) 83 day += 16; 84 85 System.out.println("\nYour birthday is " + day + "!"); 86 } 87 }

in Set3?

in Set2?

in Set1?

day to be determined

in Set4?

in Set5?

13 "10 11 14 15\n" + 14 "18 19 22 23\n" + 15 "26 27 30 31"; 16 17 String set3 = 18 " 4 5 6 7\n" + 19 "12 13 14 15\n" + 20 "20 21 22 23\n" + 21 "28 29 30 31"; 22 23 String set4 = 24 " 8 9 10 11\n" + 25 "12 13 14 15\n" + 26 "24 25 26 27\n" + 27 "28 29 30 31"; 28 29 String set5 = 30 "16 17 18 19\n" + 31 "20 21 22 23\n" + 32 "24 25 26 27\n" + 33 "28 29 30 31"; 34 35 int day = 0; 36 37 // Create a Scanner 38 Scanner input = new Scanner(System.in); 39 40 // Prompt the user to answer questions 41 System.out.print("Is your birthday in Set1?\n"); 42 System.out.print(set1); 43 System.out.print("\nEnter 0 for No and 1 for Yes: "); 44 int answer = input.nextInt(); 45 46 if (answer == 1) 47 day += 1; 48 49 // Prompt the user to answer questions 50 System.out.print("\nIs your birthday in Set2?\n"); 51 System.out.print(set2); 52 System.out.print("\nEnter 0 for No and 1 for Yes: "); 53 answer = input.nextInt(); 54 55 if (answer == 1) 56 day += 2; 57 58 // Prompt the user to answer questions 59 System.out.print("\nIs your birthday in Set3?\n"); 60 System.out.print(set3); 61 System.out.print("\nEnter 0 for No and 1 for Yes: "); 62 answer = input.nextInt(); 63 64 if (answer == 1) 65 day += 4; 66 67 // Prompt the user to answer questions 68 System.out.print("\nIs your birthday in Set4?\n"); 69 System.out.print(set4); 70 System.out.print("\nEnter 0 for No and 1 for Yes: "); 71 answer = input.nextInt(); 72

M04_LIAN1878_11_GE_C04.indd 162 12/29/17 3:57 PM

4.5 Case Studies 163

in Set3?

in Set2?

in Set1?

day to be determined

in Set4?

in Set5?

Is your birthday in Set1?

1 3 5 7

9 11 13 15

17 19 21 23

25 27 29 31

Enter 0 for No and 1 for Yes: 1

Is your birthday in Set2?

2 3 6 7

10 11 14 15

18 19 22 23

26 27 30 31

Enter 0 for No and 1 for Yes: 1

Is your birthday in Set3?

4 5 6 7

12 13 14 15

20 21 22 23

28 29 30 31

Enter 0 for No and 1 for Yes: 0

Is your birthday in Set4?

8 9 10 11

12 13 14 15

24 25 26 27

28 29 30 31

Enter 0 for No and 1 for Yes: 0

Is your birthday in Set5?

16 17 18 19

20 21 22 23

24 25 26 27

28 29 30 31

Enter 0 for No and 1 for Yes: 1

Your birthday is 19!

M04_LIAN1878_11_GE_C04.indd 163 12/29/17 3:57 PM

164 Chapter 4 Mathematical Functions, Characters, and Strings

line# day answer output

35 0

44 1

47 1

53 1

56 3

62 0

71 0

80 1

83 19

85 Your birthday is 19!

Figure 4.3 (a) A number between 1 and 31 can be represented using a five-digit binary number. (b) A five-digit binary number can be obtained by adding binary numbers 1, 10, 100, 1000, or 10000.

Decimal Binary

1 00001 2 00010

000113 ... 19 10011 ... 31 11111

10000 10 1+

10011

19 31

10000 1000 100 10

+ 1 11111

0 0 0

0 0

b5 b4

b3 b2

b5 b4 b3 b2 b1

0 00

(a) (b)

4.5.2 Case Study: Converting a Hexadecimal Digit to a Decimal Value

The hexadecimal number system has 16 digits: 0–9, A–F. The letters A, B, C, D, E, and F correspond to the decimal numbers 10, 11, 12, 13, 14, and 15. We now write a program that prompts the user to enter a hex digit and display its corresponding decimal value, as given in Listing 4.4.

This game is easy to program. You may wonder how the game was created. The mathemat- ics behind the game is actually quite simple. The numbers are not grouped together by accident—the way they are placed in the five sets is deliberate. The starting numbers in the five sets are 1, 2, 4, 8, and 16, which correspond to 1, 10, 100, 1000, and 10000 in binary (binary numbers are introduced in Appendix F, Number Systems). A binary number for deci- mal integers between 1 and 31 has at most five digits, as shown in Figure 4.3a. Let it be b5b4b3b2b1. Thus, b5b4b3b2b1 = b5 0000 + b4 000 + b3 00 + b2 0 + b1, as shown in Figure 4.3b. If a day’s binary number has a digit 1 in bk, the number should appear in Setk. For exam- ple, number 19 is binary 10011, so it appears in Set1, Set2, and Set5. It is binary 1 + 10 + 10000 = 10011 or decimal 1 + 2 + 16 = 19. Number 31 is binary 11111, so it appears in Set1, Set2, Set3, Set4, and Set5. It is binary 1 + 10 + 100 + 1000 + 10000 = 11111 or decimal 1 + 2 + 4 + 8 + 16 = 31.

mathematics behind the game

M04_LIAN1878_11_GE_C04.indd 164 12/29/17 3:57 PM

4.5 Case Studies 165

lisTing 4.4 HexDigit2Dec.java 1 import java.util.Scanner; 2 3 public class HexDigit2Dec { 4 public static void main(String[] args) { 5 Scanner input = new Scanner(System.in); 6 System.out.print("Enter a hex digit: "); 7 String hexString = input.nextLine(); 8 9 // Check if the hex string has exactly one character 10 if (hexString.length() != 1) { 11 System.out.println("You must enter exactly one character"); 12 System.exit(1); 13 } 14 15 // Display decimal value for the hex digit 16 char ch = Character.toUpperCase(hexString.charAt(0)); 17 if ('A' <= ch && ch <= 'F') { 18 int value = ch − 'A' + 10; 19 System.out.println("The decimal value for hex digit " 20 + ch + " is " + value); 21 } 22 else if (Character.isDigit(ch)) { 23 System.out.println("The decimal value for hex digit " 24 + ch + " is " + ch); 25 } 26 else { 27 System.out.println(ch + " is an invalid input"); 28 } 29 } 30 }

is 0–9?

is A–F?

check length

input string

VideoNote

Convert hex to decimal

Enter a hex digit: AB7C

You must enter exactly one character

Enter a hex digit: B

The decimal value for hex digit B is 11

Enter a hex digit: 8

The decimal value for hex digit 8 is 8

Enter a hex digit: T

T is an invalid input

The program reads a string from the console (line 7) and checks if the string contains a single character (line 10). If not, report an error and exit the program (line 12).

The program invokes the Character.toUpperCase method to obtain the character ch as an uppercase letter (line 16). If ch is between 'A' and 'F' (line 17), the corresponding decimal value is ch – 'A' + 10 (line 18). Note ch – 'A' is 0 if ch is 'A', ch – 'A' is

M04_LIAN1878_11_GE_C04.indd 165 12/29/17 3:57 PM

166 Chapter 4 Mathematical Functions, Characters, and Strings

1 if ch is 'B', and so on. When two characters perform a numerical operation, the characters' Unicodes are used in the computation.

The program invokes the Character.isDigit(ch) method to check if ch is between '0' and '9' (line 22). If so, the corresponding decimal digit is the same as ch (lines 23 and 24).

If ch is not between 'A' and 'F' nor a digit character, the program displays an error mes- sage (line 27).

4.5.3 Case Study: Revising the Lottery Program Using Strings The lottery program in Listing 3.8, Lottery.java, generates a random two-digit number, prompts the user to enter a two-digit number, and determines whether the user wins according to the following rule:

1. If the user input matches the lottery number in the exact order, the award is $10,000.

2. If all the digits in the user input match all the digits in the lottery number, the award is $3,000.

3. If one digit in the user input matches a digit in the lottery number, the award is $1,000.

The program in Listing 3.8 uses an integer to store the number. Listing 4.5 gives a new program that generates a random two-digit string instead of a number, and receives the user input as a string instead of a number.

lisTing 4.5 LotteryUsingStrings.java 1 import java.util.Scanner; 2 3 public class LotteryUsingStrings { 4 public static void main(String[] args) { 5 // Generate a lottery as a two-digit string 6 String lottery = "" + (int)(Math.random() * 10) 7 + (int)(Math.random() * 10); 8 9 // Prompt the user to enter a guess 10 Scanner input = new Scanner(System.in); 11 System.out.print("Enter your lottery pick (two digits): "); 12 String guess = input.nextLine(); 13 14 // Get digits from lottery 15 char lotteryDigit1 = lottery.charAt(0); 16 char lotteryDigit2 = lottery.charAt(1); 17 18 // Get digits from guess 19 char guessDigit1 = guess.charAt(0); 20 char guessDigit2 = guess.charAt(1); 21 22 System.out.println("The lottery number is " + lottery); 23 24 // Check the guess 25 if (guess.equals(lottery)) 26 System.out.println("Exact match: you win $10,000"); 27 else if (guessDigit2 == lotteryDigit1 28 && guessDigit1 == lotteryDigit2) 29 System.out.println("Match all digits: you win $3,000"); 30 else if (guessDigit1 == lotteryDigit1 31 || guessDigit1 == lotteryDigit2 32 || guessDigit2 == lotteryDigit1 33 || guessDigit2 == lotteryDigit2) 34 System.out.println("Match one digit: you win $1,000");

match one digit?

match all digits?

exact match?

enter a guess

generate a lottery

M04_LIAN1878_11_GE_C04.indd 166 12/29/17 3:57 PM

4.6 Formatting Console Output 167

35 else 36 System.out.println("Sorry, no match"); 37 } 38 }

Enter your lottery pick (two digits): 00 The lottery number is 00 Exact match: you win $10,000

Enter your lottery pick (two digits): 45

The lottery number is 54 Match all digits: you win $3,000

Enter your lottery pick: 23

The lottery number is 34 Match one digit: you win $1,000

Enter your lottery pick: 23

The lottery number is 14 Sorry: no match

The program generates two random digits and concatenates them into the string lottery (lines 6 and 7). After this, lottery contains two random digits.

The program prompts the user to enter a guess as a two-digit string (line 12) and checks the guess against the lottery number in this order:

■■ First, check whether the guess matches the lottery exactly (line 25).

■■ If not, check whether the reversal of the guess matches the lottery (line 27).

■■ If not, check whether one digit is in the lottery (lines 30–33).

■■ If not, nothing matches and display “Sorry, no match” (line 36).

4.5.1 If you run Listing 4.3 GuessBirthday.java with input 1 for Set1, Set3, and Set4 and 0 for Set2 and Set5, what will be the birthday?

4.5.2 If you enter a lowercase letter such as b, the program in Listing 4.4 displays B is 11. Revise the code as to display b is 11.

4.5.3 What would be wrong if lines 6 and 7 are in Listing 4.5 replaced by the following code? String lottery = "" + (int)(Math.random() * 100);

4.6 Formatting Console Output You can use the System.out.printf method to display formatted output on the console.

Often, it is desirable to display numbers in a certain format. For example, the following code computes interest, given the amount and the annual interest rate:

double amount = 12618.98; double interestRate = 0.0013; double interest = amount * interestRate; System.out.println("Interest is $" + interest);

Point Check

Point Key

M04_LIAN1878_11_GE_C04.indd 167 12/29/17 3:57 PM

168 Chapter 4 Mathematical Functions, Characters, and Strings

Because the interest amount is currency, it is desirable to display only two digits after the decimal point. To do this, you can write the code as follows:

double amount = 12618.98; double interestRate = 0.0013; double interest = amount * interestRate; System.out.println("Interest is $" + (int)(interest * 100) / 100.0);

Interest is $16.404674

Interest is $16.4

% 2 f4 .

�eld width

precision

conversion code

format speci�er

However, the format is still not correct. There should be two digits after the decimal point: 16.40 rather than 16.4. You can fix it by using the printf method, as follows:

double amount = 12618.98; double interestRate = 0.0013; double interest = amount * interestRate; System.out.printf("Interest is $%4.2f", interest);

printf

Interest is $16.40

The f in the printf stands for formatted, implying that the method prints an item in some format. The syntax to invoke this method is

System.out.printf(format, item1, item2, ..., itemk);

where format is a string that may consist of substrings and format specifiers. A format specifier specifies how an item should be formatted. An item may be a numeric value,

a character, a Boolean value, or a string. A simple format specifier consists of a percent sign (%) followed by a conversion code. Table 4.11 lists some frequently used simple format specifiers.

format specifier

Format Specifier Output Example

%b A Boolean value True or false

%c A character ‘a’

%d A decimal integer 200

%f A floating-point number 45.460000

%e A number in standard scientific notation 4.556000e+01 %s A string “Java is cool”

Table 4.11 Frequently Used Format Specifiers

Here is an example:

int count = 5; double amount = 45.56; System.out.printf("count is %d and amount is %f", count, amount);

display count is 5 and amount is 45.560000

items

M04_LIAN1878_11_GE_C04.indd 168 12/29/17 3:57 PM

4.6 Formatting Console Output 169

Items must match the format specifiers in order, in number, and in exact type. For example, the format specifier for count is %d and for amount is %f. By default, a floating-point value is displayed with six digits after the decimal point. You can specify the width and precision in a format specifier, as shown in the examples in Table 4.12.

Example Output

%5c Output the character and add four spaces before the character item, because the width is 5.

%6b Output the Boolean value and add one space before the false value and two spaces before the true value.

%5d Output the integer item with width 5. If the number of digits in the item is 6 5, add spaces before the number. If the number of digits in the item is 7 5, the width is auto- matically increased.

%10.2f Output the floating-point item with width 10 including a decimal point and two digits after the point. Thus, there are seven digits allocated before the decimal point. If the number of digits before the decimal point in the item is 6 7, add spaces before the number. If the number of digits before the decimal point in the item is 7 7, the width is automatically increased.

%10.2e Output the floating-point item with width 10 including a decimal point, two digits after the point and the exponent part. If the displayed number in scientific notation has width 6 10, add spaces before the number.

%12s Output the string with width 12 characters. If the string item has fewer than 12 char- acters, add spaces before the string. If the string item has more than 12 characters, the width is automatically increased.

Table 4.12 Examples of Specifying Width and Precision

If an item requires more spaces than the specified width, the width is automatically increased. For example, the following code

System.out.printf("%3d#%2s#%4.2f\n", 1234, "Java", 51.6653);

displays

1234#Java#51.67

The specified width for int item 1234 is 3, which is smaller than its actual size 4. The width is automatically increased to 4. The specified width for string item Java is 2, which is smaller than its actual size 4. The width is automatically increased to 4. The specified width for double item 51.6653 is 4, but it needs width 5 to display 51.67, so the width is automatically increased to 5.

You can display a number with comma separators by adding a comma in front of a number specifier. For example, the following code

System.out.printf("%,8d %,10.1f\n", 12345678, 12345678.263);

displays

12,345,678 12,345,678.3

You can pad a number with leading zeros rather than spaces by adding a 0 in front of a number specifier. For example, the following code

System.out.printf("%08d %08.1f\n", 1234, 5.63);

displays

00001234 000005.6

comma separators

leading zeros

M04_LIAN1878_11_GE_C04.indd 169 12/29/17 3:57 PM

170 Chapter 4 Mathematical Functions, Characters, and Strings

By default, the output is right justified. You can put the minus sign (−) in the format speci- fier to specify that the item is left justified in the output within the specified field. For example, the following statements

System.out.printf("%8d%8s%8.1f\n", 1234, "Java", 5.63); System.out.printf("%−8d%−8s%−8.1f \n", 1234, "Java", 5.63);

display

right justify left justify

8 1234 Java 5.6

1234 Java 5.6

8 8

where the square box (n) denotes a blank space.

Caution The items must match the format specifiers in exact type. The item for the format speci- fier %f or %e must be a floating-point type value such as 40.0, not 40. Thus, an int variable cannot match %f or %e. You can use %.2f to specify a floating-point value with two digits after the decimal point. However, %0.2f would be incorrect.

Tip The % sign denotes a format specifier. To output a literal % in the format string, use %%. For example, the following code

System.out.printf("%.2f%%\n", 75.234);

displays

75.23%

Listing 4.6 gives a program that uses printf to display a table.

lisTing 4.6 FormatDemo.java 1 public class FormatDemo { 2 public static void main(String[] args) { 3 // Display the header of the table 4 System.out.printf("%−10s%−10s%−10s%−10s%−10s\n", "Degrees", 5 "Radians", "Sine", "Cosine", "Tangent"); 6 7 // Display values for 30 degrees 8 int degrees = 30; 9 double radians = Math.toRadians(degrees); 10 System.out.printf("%−10d%−10.4f%−10.4f%−10.4f%−10.4f\n", degrees, 11 radians, Math.sin(radians), Math.cos(radians), 12 Math.tan(radians)); 13 14 // Display values for 60 degrees 15 degrees = 60; 16 radians = Math.toRadians(degrees); 17 System.out.printf("%−10d%−10.4f%−10.4f%−10.4f%−10.4f\n", degrees, 18 radians, Math.sin(radians), Math.cos(radians), 19 Math.tan(radians)); 20 } 21 }

values for 60 degrees

values for 30 degrees

display table header

M04_LIAN1878_11_GE_C04.indd 170 12/29/17 3:57 PM

Chapter Summary 171

The statements in lines 4 and 5 display the column names of the table. The column names are strings. Each string is displayed using the specifier %−10s, which left-justifies the string. The statements in lines 10–12 display the degrees as an integer and four float values. The integer is displayed using the specifier %−10d, and each float is displayed using the specifier %−10.4f, which specifies four digits after the decimal point.

4.6.1 What are the format specifiers for outputting a Boolean value, a character, a decimal integer, a floating-point number, and a string?

4.6.2 What is wrong in the following statements?

(a) System.out.printf("%5d %d", 1, 2, 3);

(b) System.out.printf("%5d %f", 1);

(d) System.out.printf("%.2f\n%0.3f\n", 1.23456, 2.34);

(e) System.out.printf("%08s\n", "Java");

4.6.3 Show the output of the following statements:

(a) System.out.printf("amount is %f %e\n", 32.32, 32.32);

(b) System.out.printf("amount is %5.2f%% %5.4e\n", 32.327, 32.32);

(d) System.out.printf("%6s\n", "Java");

(e) System.out.printf("%−6b%s\n", (1 > 2), "Java");

(f) System.out.printf("%6b%−8s\n", (1 > 2), "Java");

(g) System.out.printf("%,5d %,6.1f\n", 312342, 315562.932);

(h) System.out.printf("%05d %06.1f\n", 32, 32.32);

Point Check

Degrees Radians Sine Cosine Tangent

30 0.5236 0.5000 0.8660 0.5774 60 1.0472 0.8660 0.5000 1.7321

Key Terms char type 147 encoding 147 escape character 148 escape sequence 148 format specifier 168 instance method 152 line-based input 155

specific import 152 static method 152 supplementary Unicode 147 token-based input 155 Unicode 147 whitespace character 154

ChapTer summary 1. Java provides the mathematical methods sin, cos, tan, asin, acos, atan, toRadi-

ans, toDegrees, exp, log, log10, pow, sqrt, ceil, floor, rint, round, min, max, abs, and random in the Math class for performing mathematical functions.

2. The character type char represents a single character.

M04_LIAN1878_11_GE_C04.indd 171 12/29/17 3:57 PM

172 Chapter 4 Mathematical Functions, Characters, and Strings

3. An escape sequence consists of a backslash (\) followed by a character or a combina- tion of digits.

4. The character \ is called the escape character.

5. The characters ' ', \t, \f, \r, and \n are known as the whitespace characters.

6. Characters can be compared based on their Unicode using the relational operators.

7. The Character class contains the methods isDigit, isLetter, isLetterOrDigit, isLowerCase, and isUpperCase for testing whether a character is a digit, letter, low- ercase, or uppercase. It also contains the toLowerCase and toUpperCase methods for returning a lowercase or uppercase letter.

8. A string is a sequence of characters. A string value is enclosed in matching double quotes ("). A character value is enclosed in matching single quotes (').

9. Strings are objects in Java. A method that can only be invoked from a specific object is called an instance method. A noninstance method is called a static method, which can be invoked without using an object.

10. You can get the length of a string by invoking its length() method, retrieve a char- acter at the specified index in the string using the charAt(index) method, and use the indexOf and lastIndexOf methods to find a character or a substring in a string.

11. You can use the concat method to concatenate two strings or the plus (+) operator to concatenate two or more strings.

12. You can use the substring method to obtain a substring from the string.

13. You can use the equals and compareTo methods to compare strings. The equals method returns true if two strings are equal, and false if they are not equal. The compareTo method returns 0, a positive integer, or a negative integer, depending on whether one string is equal to, greater than, or less than the other string.

14. The printf method can be used to display a formatted output using format specifiers.

Quiz Answer the quiz for this chapter online at the Companion Website.

programming exerCises

Section 4.2 4.1 (Geometry: area of a pentagon) Write a program that prompts the user to enter

the length from the center of a pentagon to a vertex and computes the area of the pentagon, as shown in the following figure.

M04_LIAN1878_11_GE_C04.indd 172 12/29/17 3:57 PM

Programming Exercises 173

*4.2 (Geometry: great circle distance) The great circle distance is the distance between two points on the surface of a sphere. Let (x1, y1) and (x2, y2) be the geographical latitude and longitude of two points. The great circle distance between the two points can be computed using the following formula:

d = radius * arccos(sin (x1) * sin(x2) + cos(x1) * cos(x2) * cos(y1 - y2))

Write a program that prompts the user to enter the latitude and longitude of two points on the earth in degrees and displays its great circle distance. The average radius of the earth is 6,371.01 km. Note you need to convert the degrees into radi- ans using the Math.toRadians method since the Java trigonometric methods use radians. The latitude and longitude degrees in the formula are for north and west. Use negative to indicate south and east degrees. Here is a sample run:

VideoNote

Compute great circle distance

Enter the length from the center to a vertex: 5.5 The area of the pentagon is 71.92

The formula for computing the area of a pentagon is Area = 5 * s2

4 * tan ap 5 b

, where

s is the length of a side. The side can be computed using the formula s = 2r sin p

5 ,

where r is the length from the center of a pentagon to a vertex. Round up two digits after the decimal point. Here is a sample run:

Enter point 1 (latitude and longitude) in degrees: 39.55 −116.25 Enter point 2 (latitude and longitude) in degrees: 41.5 87.37 The distance between the two points is 10691.79183231593 km

*4.3 (Geography: estimate areas) Use the GPS locations for Atlanta, Georgia; Orlando, Florida; Savannah, Georgia; and Charlotte, North Carolina in the figure in Section 4.1 to compute the estimated area enclosed by these four cities. (Hint: Use the formula in Programming Exercise 4.2 to compute the distance between two cities. Divide the polygon into two triangles and use the formula in Programming Exercise 2.19 to compute the area of a triangle.)

4.4 (Geometry: area of a hexagon) The area of a hexagon can be computed using the following formula (s is the length of a side):

Area = 6 * s2

4 * tan( p

6 )

Write a program that prompts the user to enter the side of a hexagon and displays its area. Here is a sample run:

Enter the side: 5.5 The area of the hexagon is 78.59

M04_LIAN1878_11_GE_C04.indd 173 12/29/17 3:57 PM

174 Chapter 4 Mathematical Functions, Characters, and Strings

*4.5 (Geometry: area of a regular polygon) A regular polygon is an n-sided polygon in which all sides are of the same length and all angles have the same degree (i.e., the polygon is both equilateral and equiangular). The formula for computing the area of a regular polygon is

Area = n * s2

4 * tan( p

n )

Here, s is the length of a side. Write a program that prompts the user to enter the number of sides and their length of a regular polygon and displays its area. Here is a sample run:

Enter the number of sides: 5 Enter the side: 6.5 The area of the polygon is 72.69017017488385

*4.6 (Random points on a circle) Write a program that generates three random points on a circle centered at (0, 0) with radius 40 and displays three angles in a triangle formed by these three points, as shown in Figure 4.4a. (Hint: Generate a random angle a in radians between 0 and 2π, as shown in Figure 4.4b and the point deter- mined by this angle is (rxcos (a), rxsin (a)).)

Figure 4.4 (a) A triangle is formed from three random points on the circle. (b) A random point on the circle can be generated using a random angle a. (c) A pentagon is centered at (0, 0) with one point at the 0 o’clock position.

0 o’clock positionx = r × cos(a) and y = r ×sin(a)

(a) (b) (c)

(0, 0)

p5p4

(x, y)

r r

*4.7 (Corner point coordinates) Suppose a pentagon is centered at (0, 0) with one point at the 0 o’clock position, as shown in Figure 4.4c. Write a program that prompts the user to enter the radius of the bounding circle of a pentagon and displays the coordinates of the five corner points on the pentagon from p1 to p5 in this order.

M04_LIAN1878_11_GE_C04.indd 174 12/29/17 3:57 PM

Programming Exercises 175

Sections 4.3–4.6 *4.8 (Find the character of an ASCII code) Write a program that receives a character

and displays its ASCII code (an integer between 0 and 127). Here is a sample run:

Enter the radius of the bounding circle: 100.52 The coordinates of five points on the pentagon are (95.60, 31.06) (0.00, 100.52) (−95.60, 31.06) (−58.08, −81.32) (59.08, −81.32)

Use console format to display two digits after the decimal point. Here is a sample run:

Enter a character: E The ASCII code for character E is 69

*4.9 (Find the Unicode of a character) Write a program that receives a character and displays its Unicode. Here is a sample run:

Enter a character: E The Unicode for the character E is 69

Enter binary digits (0000 to 1111): 0111 The decimal value is 7

Enter binary digits (0000 to 1111): 1001 The decimal value is 9

Enter binary digits (0000 to 1111): 1100 The decimal value is 12

*4.10 (Guess birthday) Rewrite Listing 4.3, GuessBirthday.java, to prompt the user to enter the character Y for Yes and N for No, rather than entering 1 for Yes and 0 for No.

*4.11 (Binary to decimal) Write a program that prompts the user to enter binary digits and displays its corresponding decimal value. Here are some sample runs:

M04_LIAN1878_11_GE_C04.indd 175 12/29/17 3:57 PM

176 Chapter 4 Mathematical Functions, Characters, and Strings

4.12 (Hex to binary) Write a program that prompts the user to enter a hex digit and displays its corresponding binary number. For an incorrect input, display invalid input. Here is a sample run:

VideoNote

Convert hex to binary

Enter a hex digit: B

The binary value is 1011

Enter a hex digit: G

G is an invalid input

Enter a letter: B B is a consonant

Enter a letter: a a is a vowel

Enter a letter: #

# is an invalid input

*4.13 (Vowel or consonant?) Write a program that prompts the user to enter a letter and check whether the letter is a vowel or consonant. For a nonletter input, display invalid input. Here is a sample run:

*4.14 (Convert letter grade to number) Write a program that prompts the user to enter a letter grade A, B, C, D, or F and displays its corresponding numeric value 4, 3, 2, 1, or 0. For other input, display invalid grade. Here is a sample run:

Enter a letter grade: B The numeric value for grade B is 3

Enter a letter grade: T

T is an invalid grade

*4.15 (Phone key pads) The international standard letter/number mapping found on the telephone is shown below:

M04_LIAN1878_11_GE_C04.indd 176 12/29/17 3:57 PM

Programming Exercises 177

Write a program that prompts the user to enter a lowercase or uppercase letter and displays its corresponding number. For a nonletter input, display invalid input.

Enter a letter: A The corresponding number is 2

Enter a letter: a The corresponding number is 2

Enter a letter: +

+ is an invalid input

4.16 (Random character) Write a program that displays a random lowercase letter using the Math.random() method.

*4.17 (Days of a month) Write a program that prompts the user to enter the year and the first three letters of a month name (with the first letter in uppercase) and displays the number of days in the month. If the input for month is incorrect, display a mes- sage as presented in the following sample runs:

Enter a year: 2001 Enter a month: Jan Jan 2001 has 31 days

Enter a year: 2016 Enter a month: jan jan is not a correct month name

Enter two characters: I1 Information Management Freshman

*4.18 (Student major and status) Write a program that prompts the user to enter two characters and displays the major and status represented in the characters. The first character indicates the major and the second is number character 1, 2, 3, 4, which indicates whether a student is a freshman, sophomore, junior, or senior. Suppose the following characters are used to denote the majors:

I: Information Management C: Computer Science A: Accounting

Here is a sample run:

M04_LIAN1878_11_GE_C04.indd 177 12/29/17 3:57 PM

178 Chapter 4 Mathematical Functions, Characters, and Strings

4.19 (Business: check ISBN-10) Rewrite Programming Exercise 3.9 by entering the ISBN number as a string.

4.20 (Process a string) Write a program that prompts the user to enter a string and dis- plays its length and its last character.

*4.21 (Check SSN) Write a program that prompts the user to enter a Social Security number in the format DDD-DD-DDDD, where D is a digit. Your program should check whether the input is valid. Here are sample runs:

Enter two characters: T3

Invalid input

Enter a SSN: 232−23−5435

232−23−5435 is a valid social security number

Enter a SSN: 23−23−5435

23−23−5435 is an invalid social security number

Enter string s1: ABCD Enter string s2: BC BC is a substring of ABCD

4.22 (Check substring) Write a program that prompts the user to enter two strings, and reports whether the second string is a substring of the first string.

*4.23 (Financial application: payroll) Write a program that reads the following informa- tion and prints a payroll statement:

Employee’s name (e.g., Smith) Number of hours worked in a week (e.g., 10) Hourly pay rate (e.g., 9.75)

Enter string s1: ABCD

Enter string s2: BDC BDC is not a substring of ABCD

Enter two characters: A3

Accounting Junior

M04_LIAN1878_11_GE_C04.indd 178 12/29/17 3:57 PM

Programming Exercises 179

Federal tax withholding rate (e.g., 20%) State tax withholding rate (e.g., 9%)

A sample run is as follows:

Enter employee’s name: Smith

Enter number of hours worked in a week: 10 Enter hourly pay rate: 9.75

Enter federal tax withholding rate: 0.20 Enter state tax withholding rate: 0.09

Employee Name: Smith

Hours Worked: 10.0

Pay Rate: $9.75

Gross Pay: $97.5

Deductions:

Federal Withholding (20.0%): $19.5

State Withholding (9.0%): $8.77

Total Deduction: $28.27

Net Pay: $69.22

*4.24 (Enter three countries) Write a program that prompts the user to enter three coun- tries and displays them in descending order. Here is a sample run:

Enter the first country: Germany

Enter the second country: France

Enter the third country: Switzerland The three countries in descending order are Switzerland Germany France

*4.25 (Generate vehicle plate numbers) Assume that a vehicle plate number consists of three uppercase letters followed by four digits. Write a program to generate a plate number.

*4.26 (Financial application: monetary units) Rewrite Listing 2.10, ComputeChange.java, to fix the possible loss of accuracy when converting a float value to an int value. Read the input as a string such as "11.56". Your program should extract the dollar amount before the decimal point, and the cents after the decimal amount using the indexOf and substring methods.

Note More than 200 additional programming exercises with solutions are provided to the instructors on the Instructor Resource Website.

M04_LIAN1878_11_GE_C04.indd 179 12/29/17 3:57 PM

A01_GIPE3898_09_SE_FM.indd 3 12/20/17 5:56 PM

This page intentionally left blank

Loops

Objectives ■■ To write programs for executing statements repeatedly using a while

loop (§5.2).

■■ To write loops for the guessing number problem (§5.3).

■■ To follow the loop design strategy to develop loops (§5.4).

■■ To control a loop with the user confirmation or a sentinel value (§5.5).

■■ To obtain large input from a file using input redirection rather than typing from the keyboard (§5.5).

■■ To write loops using do-while statements (§5.6).

■■ To write loops using for statements (§5.7).

■■ To discover the similarities and differences of three types of loop statements (§5.8).

■■ To write nested loops (§5.9).

■■ To learn the techniques for minimizing numerical errors (§5.10).

■■ To learn loops from a variety of examples (GCD, FutureTuition, and Dec2Hex) (§5.11).

■■ To implement program control with break and continue (§5.12).

■■ To process characters in a string using a loop in a case study for checking palindrome (§5.13).

■■ To write a program that displays prime numbers (§5.14).

Chapter

M05_LIAN1878_11_GE_C05.indd 181 1/2/18 11:16 PM

182 Chapter 5 Loops

5.1 Introduction A loop can be used to tell a program to execute statements repeatedly.

Suppose you need to display a string (e.g., Welcome to Java!) a hundred times. It would be tedious to have to write the following statement a hundred times:Point

Keyproblem

100 times

System.out.println("Welcome to Java!"); System.out.println("Welcome to Java!"); ... System.out.println("Welcome to Java!");

So, how do you solve this problem? Java provides a powerful construct called a loop that controls how many times an operation

or a sequence of operations is performed in succession. Using a loop statement, you can simply tell the computer to display a string a hundred times without having to code the print statement a hundred times, as follows:

int count = 0; while (count < 100) { System.out.println("Welcome to Java!"); count++; }

The variable count is initially 0. The loop checks whether count < 100 is true. If so, it executes the loop body to display the message Welcome to Java! and increments count by 1. It repeatedly executes the loop body until count < 100 becomes false. When count < 100 is false (i.e., when count reaches 100), the loop terminates, and the next statement after the loop statement is executed.

Loops are constructs that control repeated executions of a block of statements. The concept of looping is fundamental to programming. Java provides three types of loop statements: while loops, do-while loops, and for loops.

5.2 The while Loop A while loop executes statements repeatedly while the condition is true.

The syntax for the while loop is as follows:

while (loop-continuation-condition) { // Loop body Statement(s); }

Figure 5.1a shows the while loop flowchart. The part of the loop that contains the state- ments to be repeated is called the loop body. A one-time execution of a loop body is referred to as an iteration (or repetition) of the loop. Each loop contains a loop-continuation- condition, a Boolean expression that controls the execution of the body. It is evaluated each time to determine if the loop body is executed. If its evaluation is true, the loop body is executed; if its evaluation is false, the entire loop terminates and the program control turns to the statement that follows the while loop.

The loop for displaying Welcome to Java! a hundred times introduced in the preceding section is an example of a while loop. Its flowchart is shown in Figure 5.1b.

loop

Point Key

while loop

VideoNote

Use while loop loop body iteration loop-continuation- condition

M05_LIAN1878_11_GE_C05.indd 182 1/2/18 11:16 PM

5.2 The while Loop 183

The loop-continuation-condition is count < 100 and the loop body contains two statements in the following code:

Figure 5.1 The while loop repeatedly executes the statements in the loop body when the loop-continuation-condition evaluates to true.

loop- continuation- condition?

true

false

(a)

(count < 100)?

true

false

(b)

count = 0;

System.out.println("Welcome to Java!"); count++;

Statement(s) (loop body)

Statement(s) Before loop

loop-continuation-condition

int count = 0; while (count < 100) { System.out.println("Welcome to Java!"); loop body count++; }

In this example, you know exactly how many times the loop body needs to be executed because the control variable count is used to count the number of iterations. This type of loop is known as a counter-controlled loop.

Note The loop-continuation-condition must always appear inside the parenthe- ses. The braces enclosing the loop body can be omitted only if the loop body contains one or no statement.

Here is another example to help understand how a loop works.

int sum = 0, i = 1; while (i < 10) { sum = sum + i; i++; } System.out.println("sum is " + sum); // sum is 45

If i < 10 is true, the program adds i to sum. Variable i is initially set to 1, then is incre- mented to 2, 3, and up to 10. When i is 10, i < 10 is false, so the loop exits. Therefore, the sum is 1 + 2 + 3 + ... + 9 = 45.

counter-controlled loop

M05_LIAN1878_11_GE_C05.indd 183 1/2/18 11:16 PM

184 Chapter 5 Loops

What happens if the loop is mistakenly written as follows?

int sum = 0, i = 1; while (i < 10) { sum = sum + i; }

This loop is infinite, because i is always 1 and i < 10 will always be true.

Note Make sure that the loop-continuation-condition eventually becomes false so that the loop will terminate. A common programming error involves infinite loops (i.e., the loop runs forever). If your program takes an unusually long time to run and does not stop, it may have an infinite loop. If you are running the program from the command window, press CTRL+C to stop it.

Caution Programmers often make the mistake of executing a loop one more or less time. This is commonly known as the off-by-one error. For example, the following loop displays Welcome to Java 101 times rather than 100 times. The error lies in the condition, which should be count < 100 rather than count <= 100.

int count = 0; while (count <= 100) { System.out.println("Welcome to Java!"); count++; }

Recall that Listing 3.1, AdditionQuiz.java, gives a program that prompts the user to enter an answer for a question on addition of two single digits. Using a loop, you can now rewrite the program to let the user repeatedly enter a new answer until it is correct, as given in Listing 5.1.

Listing 5.1 RepeatAdditionQuiz.java 1 import java.util.Scanner; 2 3 public class RepeatAdditionQuiz { 4 public static void main(String[] args) { 5 int number1 = (int)(Math.random() * 10); 6 int number2 = (int)(Math.random() * 10); 7 8 // Create a Scanner 9 Scanner input = new Scanner(System.in); 10 11 System.out.print( 12 "What is " + number1 + " + " + number2 + "? "); 13 int answer = input.nextInt(); 14 15 while (number1 + number2 != answer) { 16 System.out.print("Wrong answer. Try again. What is " 17 + number1 + " + " + number2 + "? "); 18 answer = input.nextInt(); 19 } 20 21 System.out.println("You got it!"); 22 } 23 }

infinite loop

off-by-one error

generate number1 generate number2

show question

get first answer

check answer

read an answer

What is 5 + 9? 12

Wrong answer. Try again. What is 5 + 9? 34

Wrong answer. Try again. What is 5 + 9? 14

You got it!

M05_LIAN1878_11_GE_C05.indd 184 1/2/18 11:16 PM

5.3 Case Study: Guessing Numbers 185

The loop in lines 15–19 repeatedly prompts the user to enter an answer when number1 + number2 != answer is true. Once number1 + number2 != answer is false, the loop exits.

5.2.1 Analyze the following code. Is count < 100 always true, always false, or sometimes true or sometimes false at Point A, Point B, and Point C?

int count = 0; while (count < 100) { // Point A System.out.println("Welcome to Java!"); count++; // Point B } // Point C

5.2.2 How many times are the following loop bodies repeated? What is the output of each loop?

Point Check

Guess a magic number between 0 and 100

Enter your guess: 50

Your guess is too high

Enter your guess: 25

Your guess is too low

Enter your guess: 42

Your guess is too high

Enter your guess: 39

Yes, the number is 39

(c)

int i = 1; while (i < 10) if (i % 2 == 0) System.out.println(i);

int i = 1; while (i < 10) if (i % 2 == 0) System.out.println(i++);

int i = 1; while (i < 10) if ((i++) % 2 == 0) System.out.println(i);

(a) (b)

5.2.3 What is the output of the following code? Explain the reason. int x = 80000000;

while (x > 0) x++;

System.out.println("x is " + x);

5.3 Case Study: Guessing Numbers This case study generates a random number and lets the user repeatedly guess a n umber until it is correct.

The problem is to guess what number a computer has in mind. You will write a program that randomly generates an integer between 0 and 100, inclusive. The program prompts the user to enter a number continuously until the number matches the randomly generated number. For each user input, the program tells the user whether the input is too low or too high, so the user can make the next guess intelligently. Here is a sample run:

Point Key

VideoNote

Guess a number

M05_LIAN1878_11_GE_C05.indd 185 1/2/18 11:16 PM

186 Chapter 5 Loops

The magic number is between 0 and 100. To minimize the number of guesses, enter 50 first. If your guess is too high, the magic number is between 0 and 49. If your guess is too low, the magic number is between 51 and 100. Thus, you can eliminate half of the numbers from further consideration after one guess.

How do you write this program? Do you immediately begin coding? No. It is important to think before coding. Think how you would solve the problem without writing a program. You need first to generate a random number between 0 and 100, inclusive, then to prompt the user to enter a guess, then to compare the guess with the random number.

It is a good practice to code incrementally one step at a time. For programs involving loops, if you don’t know how to write a loop right away, you may first write the code for executing the loop one time, then figure out how to repeatedly execute the code in a loop. For this pro- gram, you may create an initial draft, as given in Listing 5.2.

Listing 5.2 GuessNumberOneTime.java 1 import java.util.Scanner; 2 3 public class GuessNumberOneTime { 4 public static void main(String[] args) { 5 // Generate a random number to be guessed 6 int number = (int)(Math.random() * 101); 7 8 Scanner input = new Scanner(System.in); 9 System.out.println("Guess a magic number between 0 and 100"); 10 11 // Prompt the user to guess the number 12 System.out.print("\nEnter your guess: "); 13 int guess = input.nextInt(); 14 15 if (guess == number) 16 System.out.println("Yes, the number is " + number); 17 else if (guess > number) 18 System.out.println("Your guess is too high"); 19 else 20 System.out.println("Your guess is too low"); 21 } 22 }

When you run this program, it prompts the user to enter a guess only once. To let the user enter a guess repeatedly, you may wrap the code in lines 11–20 in a loop as follows:

while (true) { // Prompt the user to guess the number System.out.print("\nEnter your guess: "); guess = input.nextInt();

if (guess == number) System.out.println("Yes, the number is " + number); else if (guess > number) System.out.println("Your guess is too high"); else System.out.println("Your guess is too low"); } // End of loop

This loop repeatedly prompts the user to enter a guess. However, this loop is not correct, because it never terminates. When guess matches number, the loop should end. Thus, the loop can be revised as follows:

while (guess != number) { // Prompt the user to guess the number

intelligent guess

think before coding

code incrementally

generate a number

enter a guess

correct guess

too high

too low

M05_LIAN1878_11_GE_C05.indd 186 1/2/18 11:16 PM

5.3 Case Study: Guessing Numbers 187

System.out.print("\nEnter your guess: "); guess = input.nextInt();

The complete code is given in Listing 5.3.

Listing 5.3 GuessNumber.java 1 import java.util.Scanner; 2 3 public class GuessNumber { 4 public static void main(String[] args) { 5 // Generate a random number to be guessed 6 int number = (int)(Math.random() * 101); 7 8 Scanner input = new Scanner(System.in); 9 System.out.println("Guess a magic number between 0 and 100"); 10 11 int guess = –1; 12 while (guess != number) { 13 // Prompt the user to guess the number 14 System.out.print("\nEnter your guess: "); 15 guess = input.nextInt(); 16 17 if (guess == number) 18 System.out.println("Yes, the number is " + number); 19 else if (guess > number) 20 System.out.println("Your guess is too high"); 21 else 22 System.out.println("Your guess is too low"); 23 } // End of loop 24 } 25 }

generate a number

enter a guess

too high

too low

line# number guess output

6 39

11 −1

iteration 1 b 15 50 20 Your guess is too high

iteration 2 b 15 25 22 Your guess is too low

iteration 3 b 15 42 20 Your guess is too high

iteration 4 b 15 39 18 Yes, the number is 39

The program generates the magic number in line 6 and prompts the user to enter a guess continuously in a loop (lines 12–23). For each guess, the program checks whether the guess is

M05_LIAN1878_11_GE_C05.indd 187 1/2/18 11:16 PM

188 Chapter 5 Loops

correct, too high, or too low (lines 17–22). When the guess is correct, the program exits the loop (line 12). Note that guess is initialized to −1. Initializing it to a value between 0 and 100 would be wrong, because that could be the number to be guessed.

5.3.1 What is wrong if guess is initialized to 0 in line 11 in Listing 5.3?

5.4 Loop Design Strategies The key to designing a loop is to identify the code that needs to be repeated and write a condition for terminating the loop.

Writing a correct loop is not an easy task for novice programmers. Consider three steps when writing a loop.

Step 1: Identify the statements that need to be repeated.

Step 2: Wrap these statements in a loop as follows:

while (true) { Statements; }

Step 3: Code the loop-continuation-condition and add appropriate statements for controlling the loop.

while (loop-continuation-condition) { Statements; Additional statements for controlling the loop; }

The Math subtraction learning tool program in Listing 3.3, SubtractionQuiz.java, gener- ates just one question for each run. You can use a loop to generate questions repeatedly. How do you write the code to generate five questions? Follow the loop design strategy. First, identify the statements that need to be repeated. These are the statements for obtaining two random numbers, prompting the user with a subtraction question, and grading the question. Second, wrap the statements in a loop. Third, add a loop control variable and the loop- continuation-condition to execute the loop five times.

Listing 5.4 gives a program that generates five questions and, after a student answers all five, reports the number of correct answers. The program also displays the time spent on the test and lists all the questions.

Listing 5.4 SubtractionQuizLoop.java 1 import java.util.Scanner; 2 3 public class SubtractionQuizLoop { 4 public static void main(String[] args) { 5 final int NUMBER_OF_QUESTIONS = 5; // Number of questions 6 int correctCount = 0; // Count the number of correct answers 7 int count = 0; // Count the number of questions 8 long startTime = System.currentTimeMillis(); 9 String output = " "; // output string is initially empty 10 Scanner input = new Scanner(System.in); 11 12 while (count < NUMBER_OF_QUESTIONS) { 13 // 1. Generate two random single-digit integers 14 int number1 = (int)(Math.random() * 10); 15 int number2 = (int)(Math.random() * 10); 16 17 // 2. If number1 < number2, swap number1 with number2 18 if (number1 < number2) {

Point Check

Point Key

VideoNote Multiple subtraction quiz

loop

get start time

M05_LIAN1878_11_GE_C05.indd 188 1/2/18 11:16 PM

5.4 Loop Design Strategies 189

19 int temp = number1; 20 number1 = number2; 21 number2 = temp; 22 } 23 24 // 3. Prompt the student to answer "What is number1 – number2?" 25 System.out.print( 26 "What is " + number1 + " – " + number2 + "? "); 27 int answer = input.nextInt(); 28 29 // 4. Grade the answer and display the result 30 if (number1 – number2 == answer) { 31 System.out.println("You are correct!"); 32 correctCount++; // Increase the correct answer count 33 } 34 else 35 System.out.println("Your answer is wrong.\n" + number1 36 + " – " + number2 + " should be " + (number1 — number2)); 37 38 // Increase the question count 39 count++; 40 41 output += "\n" + number1 + "–" + number2 + "=" + answer + 42 ((number1 – number2 == answer) ? " correct": " wrong"); 43 } 44 45 long endTime = System.currentTimeMillis(); 46 long testTime = endTime – startTime; 47 48 System.out.println("Correct count is " + correctCount + 49 "\nTest time is " + testTime / 1000 + " seconds\n" + output); 50 } 51 }

display a question

grade an answer

increase correct count

increase control variable

prepare output

end loop

get end time test time

display result

What is 9 – 2? 7

You are correct!

What is 3 – 0? 3 You are correct!

What is 3 – 2? 1 You are correct!

What is 7 – 4? 4 Your answer is wrong. 7 – 4 should be 3

What is 7 – 5? 4 Your answer is wrong. 7 – 5 should be 2

Correct count is 3 Test time is 1021 seconds

9–2=7 correct 3–0=3 correct 3–2=1 correct 7–4=4 wrong 7–5=4 wrong

M05_LIAN1878_11_GE_C05.indd 189 1/2/18 11:16 PM

190 Chapter 5 Loops

The program uses the control variable count to control the execution of the loop. count is initially 0 (line 7) and is increased by 1 in each iteration (line 39). A subtraction question is displayed and processed in each iteration. The program obtains the time before the test starts in line 8 and the time after the test ends in line 45, then computes the test time in line 46. The test time is in milliseconds and is converted to seconds in line 49.

5.4.1 Revise the code using the System.nanoTime() to measure the time in nano seconds.

5.5 Controlling a Loop with User Confirmation or a Sentinel Value

It is a common practice to use a sentinel value to terminate the input.

The preceding example executes the loop five times. If you want the user to decide whether to continue, you can offer a user confirmation. The template of the program can be coded as follows:

char continueLoop = 'Y'; while (continueLoop == 'Y') { // Execute the loop body once ... // Prompt the user for confirmation System.out.print("Enter Y to continue and N to quit: "); continueLoop = input.getLine().charAt(0); }

You can rewrite the program given in Listing 5.4 with user confirmation to let the user decide whether to advance to the next question.

Another common technique for controlling a loop is to designate a special value when read- ing and processing a set of values. This special input value, known as a sentinel value, signifies the end of the input. A loop that uses a sentinel value to control its execution is called a sentinel-controlled loop.

Listing 5.5 gives a program that reads and calculates the sum of an unspecified number of integers. The input 0 signifies the end of the input. Do you need to declare a new variable for each input value? No. Just use one variable named data (line 12) to store the input value, and use a variable named sum (line 15) to store the total. Whenever a value is read, assign it to data and, if it is not zero, add it to sum (line 17).

Listing 5.5 SentinelValue.java 1 import java.util.Scanner; 2 3 public class SentinelValue { 4 /** Main method */ 5 public static void main(String[] args) { 6 // Create a Scanner 7 Scanner input = new Scanner(System.in); 8 9 // Read an initial data 10 System.out.print( 11 "Enter an integer (the input ends if it is 0): "); 12 int data = input.nextInt(); 13 14 // Keep reading data until the input is 0 15 int sum = 0; 16 while (data != 0) {

Point Check

Point Key

sentinel value

sentinel-controlled loop

loop

input

M05_LIAN1878_11_GE_C05.indd 190 1/2/18 11:16 PM

5.5 Controlling a Loop with User Confirmation or a Sentinel Value 191

17 sum += data; 18 19 // Read the next data 20 System.out.print( 21 "Enter an integer (the input ends if it is 0): "); 22 data = input.nextInt(); 23 } 24 25 System.out.println("The sum is " + sum); 26 } 27 }

end of loop

display result

Enter an integer (the input ends if it is 0): 2

Enter an integer (the input ends if it is 0): 3

Enter an integer (the input ends if it is 0): 4

Enter an integer (the input ends if it is 0): 0

The sum is 9

If data is not 0, it is added to sum (line 17) and the next item of input data is read (lines 20–22). If data is 0, the loop body is no longer executed and the while loop terminates. The input value 0 is the sentinel value for this loop. Note if the first input read is 0, the loop body never executes, and the resulting sum is 0.

Caution Don’t use floating-point values for equality checking in a loop control. Because floating-point values are approximations for some values, using them could result in imprecise counter values and inaccurate results.

Consider the following code for computing 1 + 0.9 + 0.8 + ... + 0.1:

double item = 1; double sum = 0; while (item != 0) { // No guarantee item will be 0 sum += item; item −= 0.1; } System.out.println(sum);

Variable item starts with 1 and is reduced by 0.1 every time the loop body is executed. The loop should terminate when item becomes 0. However, there is no guarantee that item will be exactly 0, because the floating-point arithmetic is approximated. This loop seems okay on the surface, but it is actually an infinite loop. numeric error

line# data sum output

12 2

15 0

iteration 1 b 17 2 22 3

iteration 2 b 17 5 22 4

iteration 3 b 17 9 22 0

25 The sum is 9

M05_LIAN1878_11_GE_C05.indd 191 1/2/18 11:16 PM

192 Chapter 5 Loops

In the preceding example, if you have a large number of data to enter, it would be cumbersome to type from the keyboard. You can store the data separated by whitespaces in a text file, say input.txt, and run the program using the following command:

java SentinelValue < input.txt

This command is called input redirection. The program takes the input from the file input. txt rather than having the user type the data from the keyboard at runtime. Suppose the contents of the file are as follows:

2 3 4 5 6 7 8 9 12 23 32 23 45 67 89 92 12 34 35 3 1 2 4 0

The program should get sum to be 518. Similarly, there is output redirection, which sends the output to a file rather than displaying

it on the console. The command for output redirection is

java ClassName > output.txt

Input and output redirections can be used in the same command. For example, the following command gets input from input.txt and sends output to output.txt:

java SentinelValue < input.txt > output.txt

Try running the program to see what contents are in output.txt.

5.5.1 Suppose the input is 2 3 4 5 0. What is the output of the following code? import java.util.Scanner;

public class Test { public static void main(String[] args) { Scanner input = new Scanner(System.in);

int number, max; number = input.nextInt(); max = number;

while (number != 0) { number = input.nextInt(); if (number > max) max = number; }

System.out.println("max is " + max); System.out.println("number " + number); } }

5.6 The do-while Loop A do-while loop is the same as a while loop except that it executes the loop body first then checks the loop continuation condition.

The do-while loop is a variation of the while loop. Its syntax is as follows:

do { // Loop body; Statement(s); } while (loop-continuation-condition);

Its execution flowchart is shown in Figure 5.2a.

input redirection

output redirection

Point Check

Point Key

VideoNote Use do-while loop do-while loop

M05_LIAN1878_11_GE_C05.indd 192 1/2/18 11:16 PM

The loop body is executed first, then the loop-continuation-condition is evaluated. If the evaluation is true, the loop body is executed again; if it is false, the do-while loop terminates. For example, the following while loop statement

int count = 0; while (count < 100) { System.out.println("Welcome to Java!"); count++; }

can be written using a do-while loop as follows:

int count = 0; do { System.out.println("Welcome to Java!"); count++; } while (count < 100);

The flowchart of this do-while loop is shown in Figure 5.2b. The difference between a while loop and a do-while loop is the order in which the loop-

continuation-condition is evaluated and the loop body is executed. In the case of a do- while loop, the loop body is executed at least once. You can write a loop using either the while loop or the do-while loop. Sometimes one is a more convenient choice than the other. For exam- ple, you can rewrite the while loop in Listing 5.5 using a do-while loop, as given in Listing 5.6.

Listing 5.6 TestDoWhile.java 1 import java.util.Scanner; 2 3 public class TestDoWhile { 4 /** Main method */ 5 public static void main(String[] args) { 6 int data; 7 int sum = 0;

Figure 5.2 The do-while loop executes the loop body first then checks the l oop- continuation-condition to determine whether to continue or terminate the loop.

loop- continuation-

condition?true

(a)

(count < 100)? true

falsefalse

(b)

count = 0;

System.out.println("Welcome to Java!"); count++;

Statement(s) (loop body)

Statement(s) Before loop

5.6 The do-while Loop 193

M05_LIAN1878_11_GE_C05.indd 193 1/2/18 11:16 PM

194 Chapter 5 Loops

8 9 // Create a Scanner 10 Scanner input = new Scanner(System.in); 11 12 // Keep reading data until the input is 0 13 do { 14 // Read the next data 15 System.out.print( 16 "Enter an integer (the input ends if it is 0): "); 17 data = input.nextInt(); 18 19 sum += data; 20 } while (data != 0); 21 22 System.out.println("The sum is " + sum); 23 } 24 }

loop

end loop

Enter an integer (the input ends if it is 0): 3

Enter an integer (the input ends if it is 0): 5

Enter an integer (the input ends if it is 0): 6

Enter an integer (the input ends if it is 0): 0

The sum is 14

Tip Use a do-while loop if you have statements inside the loop that must be executed at least once, as in the case of the do-while loop in the preceding TestDoWhile program. These statements must appear before the loop as well as inside it if you use a while loop.

5.6.1 Suppose the input is 2 3 4 5 0. What is the output of the following code?

import java.util.Scanner;

public class Test { public static void main(String[] args) { Scanner input = new Scanner(System.in);

int number, max; number = input.nextInt(); max = number;

do { number = input.nextInt(); if (number > max) max = number; } while (number != 0);

System.out.println("max is " + max); System.out.println("number " + number); } }

Point Check

M05_LIAN1878_11_GE_C05.indd 194 1/2/18 11:16 PM

5.7 The for Loop 195

5.6.2 What are the differences between a while loop and a do-while loop? Convert the following while loop into a do-while loop:

Scanner input = new Scanner(System.in); int sum = 0; System.out.println("Enter an integer " + "(the input ends if it is 0)"); int number = input.nextInt(); while (number != 0) { sum += number; System.out.println("Enter an integer " + "(the input ends if it is 0)"); number = input.nextInt(); }

5.7 The for Loop A for loop has a concise syntax for writing loops.

Often you write a loop in the following common form:

i = initialValue; // Initialize loop control variable while (i < endValue) { // Loop body ... i++; // Adjust loop control variable }

This loop is intuitive and easy for beginners to grasp. However, programmers often forget to adjust the control variable, which leads to an infinite loop. A for loop can be used to simplify the preceding loop as shown in (a), which is equivalent to (b)

Point Key

for (i = initialValue; i < endValue; i++) {

// Loop body ... }

(a)

i = initialValue;

while (i < endValue) {

// Loop body ... i++;

}

(b)

In general, the syntax of a for loop is as follows:

for (initial-action; loop-continuation-condition; action-after-each-iteration) { // Loop body; Statement(s); }

The flowchart of the for loop is shown in Figure 5.3a. The for loop statement starts with the keyword for, followed by a pair of parentheses

enclosing the control structure of the loop. This structure consists of initial-action, loop- continuation-condition, and action-after-each-iteration. The control structure is followed by the loop body enclosed inside braces. The initial-action, loop-contin- uation-condition, and action-after-each-iteration are separated by semicolons.

A for loop generally uses a variable to control how many times the loop body is executed and when the loop terminates. This variable is referred to as a control variable. The initial- action often initializes a control variable, the action-after-each-iteration usually increments or decrements the control variable, and the loop-continuation-condition

for loop

control variable

M05_LIAN1878_11_GE_C05.indd 195 1/2/18 11:16 PM

196 Chapter 5 Loops

tests whether the control variable has reached a termination value. For example, the following for loop prints Welcome to Java! a hundred times:

int i; for (i = 0; i < 100; i++) { System.out.println("Welcome to Java!"); }

The flowchart of the statement is shown in Figure 5.3b. The for loop initializes i to 0, then repeatedly executes the println statement and evaluates i++ while i is less than 100.

The initial-action, i = 0, initializes the control variable, i. The loop- continuation-condition, i < 100, is a Boolean expression. The expression is evaluated right after the initialization and at the beginning of each iteration. If this condition is true, the loop body is executed. If it is false, the loop terminates and the program control turns to the line following the loop.

The action-after-each-iteration, i++, is a statement that adjusts the control v ariable. This statement is executed after each iteration and increments the control variable. Eventually, the value of the control variable should force the loop-continuation-condi- tion to become false; otherwise, the loop is infinite.

The loop control variable can be declared and initialized in the for loop. Here is an example:

for (int i = 0; i < 100; i++) { System.out.println("Welcome to Java!"); }

If there is only one statement in the loop body, as in this example, the braces can be omitted.

Tip The control variable must be declared inside the control structure of the loop or before the loop. If the loop control variable is used only in the loop, and not elsewhere, it is a good programming practice to declare it in the initial-action of the for loop. If

initial-action

action-after-each-iteration

omitting braces

declare control variable

Figure 5.3 A for loop performs an initial action once, then repeatedly executes the statements in the loop body, and performs an action after an iteration when the loop-

continuation-condition evaluates to true.

Statement(s) (loop body)

(a)

Initial-action

action-after-each-iteration

true

false loop-

continuation- condition?

System.out.println( "Welcome to Java!");

(b)

i = 0

i++

true

false (i < 100)?

M05_LIAN1878_11_GE_C05.indd 196 1/2/18 11:16 PM

the variable is declared inside the loop control structure, it cannot be referenced outside the loop. In the preceding code, for example, you cannot reference i outside the for loop, because it is declared inside the for loop.

Note The initial-action in a for loop can be a list of zero or more comma-separated variable declaration statements or assignment expressions. For example:

for (int i = 0, j = 0; i + j < 10; i++, j++) { // Do something }

The action-after-each-iteration in a for loop can be a list of zero or more comma-separated statements. For example:

for (int i = 1; i < 100; System.out.println(i), i++) ;

This example is correct, but it is a bad example, because it makes the code difficult to read. Normally, you declare and initialize a control variable as an initial action, and increment or decrement the control variable as an action after each iteration.

Note If the loop-continuation-condition in a for loop is omitted, it is implicitly true. Thus, the statement given below in (a), which is an infinite loop, is the same as in (b). To avoid confusion, though, it is better to use the equivalent loop in (c).

for loop variations

5.7.1 Do the following two loops result in the same value in sum? Point Check

for ( ; ; ) { // Do something }

(a)

for ( ; true; ) { // Do something }

(b)

while (true) { // Do something }

Equivalent Equivalent

5.7.2 What are the three parts of a for loop control? Write a for loop that prints the numbers from 1 to 100.

5.7.3 Suppose the input is 2 3 4 5 0. What is the output of the following code? import java.util.Scanner;

public class Test { public static void main(String[] args) { Scanner input = new Scanner(System.in);

int number, sum = 0, count;

for (count = 0; count < 5; count++) { number = input.nextInt(); sum += number; }

System.out.println("sum is " + sum); System.out.println("count is " + count); } }

for (int i = 0; i < 10; ++i) { sum += i; }

(a)

for (int i = 0; i < 10; i++) { sum += i; }

(b)

5.7 The for Loop 197

M05_LIAN1878_11_GE_C05.indd 197 1/2/18 11:16 PM

198 Chapter 5 Loops

5.7.4 What does the following statement do? for ( ; ; ) { // Do something }

5.7.5 If a variable is declared in a for loop control, can it be used after the loop exits? 5.7.6 Convert the following for loop statement to a while loop and to a do-while loop:

long sum = 0; for (int i = 0; i <= 1000; i++) sum = sum + i;

5.7.7 Count the number of iterations in the following loops.

5.8 Which Loop to Use? You can use a for loop, a while loop, or a do-while loop, whichever is convenient.

The while loop and do-while loop are easier to learn than the for loop. However, you will learn the for loop quickly after some practice. A for loop places control variable initialization, loop continuation condition, and adjustment after each iteration all together. It is more concise and enables you to write the code with less errors than the other two loops.

The while loop and for loop are called pretest loops because the continuation condition is checked before the loop body is executed. The do-while loop is called a posttest loop because the condition is checked after the loop body is executed. The three forms of loop statements—while, do-while, and for—are expressively equivalent; that is, you can write a loop in any of these three forms. For example, a while loop in (a) in the following figure can always be converted into the for loop in (b).

Point Key

pretest loop

posttest loop

int count = 0; while (count < n) { count++; }

(a)

for (int count = 0; count <= n; count++) { }

(b)

int count = 5; while (count < n) { count++; }

(c)

int count = 5; while (count < n) { count = count + 3; }

(d)

A for loop in (a) in the next figure can generally be converted into the while loop in (b) except in certain special cases (see CheckPoint Question 5.12.2 in Section 5.12 for such a case).

while (loop-continuation-condition) { // Loop body }

(a)

for ( ; loop-continuation-condition; ) { // Loop body }

(b)

Equivalent

for (initial-action; loop-continuation-condition; action-after-each-iteration) { // Loop body; }

(a)

initial-action; while (loop-continuation-condition) { // Loop body; action-after-each-iteration; }

(b)

M05_LIAN1878_11_GE_C05.indd 198 1/2/18 11:16 PM

5.8 Which Loop to Use? 199

Use the loop statement that is most intuitive and comfortable for you. In general, a for loop may be used if the number of repetitions is known in advance, as, for example, when you need to display a message a hundred times. A while loop may be used if the number of repetitions is not fixed, as in the case of reading the numbers until the input is 0. A do-while loop can be used to replace a while loop if the loop body has to be executed before the continuation condition is tested.

Caution Adding a semicolon at the end of the for clause before the loop body is a common mistake, as shown below in (a). In (a), the semicolon signifies the end of the loop prematurely. The loop body is actually empty, as shown in (b). (a) and (b) are equivalent. Both are incorrect.

Similarly, the loop in (c) is also wrong. (c) is equivalent to (d). Both are incorrect.

for (int i = 0; i < 10; i++); { System.out.println("i is " + i); }

(a)

Error

for (int i = 0; i < 10; i++) { }; { System.out.println("i is " + i); }

(b)

Empty body

Error Empty body

These errors often occur when you use the next-line block style. Using the end-of-line block style can avoid errors of this type.

In the case of the do-while loop, the semicolon is needed to end the loop.

int i = 0; while (i < 10); { System.out.println("i is " + i); i++; }

int i = 0; while (i < 10) { }; { System.out.println("i is " + i); i++; }

int i = 0; do { System.out.println("i is " + i); i++; } while (i < 10);

Correct

5.8.1 Can you convert a for loop to a while loop? List the advantages of using for loops. 5.8.2 Can you always convert a while loop into a for loop? Convert the following

while loop into a for loop:

int i = 1; int sum = 0; while (sum < 10000) { sum = sum + i; i++; }

Point Check

M05_LIAN1878_11_GE_C05.indd 199 1/2/18 11:16 PM

200 Chapter 5 Loops

5.9 Nested Loops A loop can be nested inside another loop.

Nested loops consist of an outer loop and one or more inner loops. Each time the outer loop is repeated, the inner loops are reentered, and started anew.

Listing 5.7 presents a program that uses nested for loops to display a multiplication table.

Listing 5.7 MultiplicationTable.java 1 public class MultiplicationTable { 2 /** Main method */ 3 public static void main(String[] args) { 4 // Display the table heading 5 System.out.println(" Multiplication Table"); 6 7 // Display the number title 8 System.out.print(" "); 9 for (int j = 1; j <= 9; j++) 10 System.out.print(" " + j);

Point Key

table title

nested loop

5.8.4 What is wrong with the following programs?

5.8.3 Identify and fix the errors in the following code: 1 public class Test { 2 public void main(String[] args) { 3 for (int i = 0; i < 10; i++); 4 sum += i; 5 6 if (i < j); 7 System.out.println(i) 8 else 9 System.out.println(j); 10 11 while (j < 10); 12 { 13 j++; 14 } 15 16 do { 17 j++; 18 } while (j < 10) 19 } 20 }

1 public class ShowErrors { 2 public static void main(String[] args) { 3 int i = 0; 4 do { 5 System.out.println(i + 4); 6 i++; 7 } 8 while (i < 10) 8 } 9 }

1 public class ShowErrors { 2 public static void main(String[] args) { 3 for (int i = 0; i < 10; i++); 4 System.out.println(i + 4); 5 } 6 }

(b)(a)

M05_LIAN1878_11_GE_C05.indd 200 1/2/18 11:16 PM

5.9 Nested Loops 201

11 12 System.out.println("\n — — — — — — — — — — — — — — — —— — — — — — — — — — — — —"); 13 14 // Display table body 15 for (int i = 1; i <= 9; i++) { 16 System.out.print(i + " | "); 17 for (int j = 1; j <= 9; j++) { 18 // Display the product and align properly 19 System.out.printf("%4d", i * j); 20 } 21 System.out.println(); 22 } 23 } 24 }

outer loop

inner loop

Multiplication Table

1 2 3 4 5 6 7 8 9

1 | 1 2 3 4 5 6 7 8 9 2 | 2 4 6 8 10 12 14 16 18 3 | 3 6 9 12 15 18 21 24 27 4 | 4 8 12 16 20 24 28 32 36 5 | 5 10 15 20 25 30 35 40 45 6 | 6 12 18 24 30 36 42 48 54 7 | 7 14 21 28 35 42 49 56 63 8 | 8 16 24 32 40 48 56 64 72 9 | 9 18 27 36 45 54 63 72 81

The program displays a title (line 5) on the first line in the output. The first for loop (lines 9 and 10) displays the numbers 1–9 on the second line. A dashed (–) line is displayed on the third line (line 12).

The next loop (lines 15–22) is a nested for loop with the control variable i in the outer loop and j in the inner loop. For each i, the product i * j is displayed on a line in the inner loop, with j being 1, 2, 3, …, 9.

Note Be aware that a nested loop may take a long time to run. Consider the following loop nested in three levels:

for (int i = 0; i < 10000; i++) for (int j = 0; j < 10000; j++) for (int k = 0; k < 10000; k++) Perform an action

The action is performed one trillion times. If it takes 1 microsecond to perform the action, the total time to run the loop would be more than 277 hours. Note 1 microsecond is one-millionth (10-6) of a second.

5.9.1 How many times is the println statement executed? for (int i = 0; i < 10; i++) for (int j = 0; j < i; j++) System.out.println(i * j)

Point Check

M05_LIAN1878_11_GE_C05.indd 201 1/2/18 11:16 PM

202 Chapter 5 Loops

5.9.2 Show the output of the following programs. (Hint: Draw a table and list the vari- ables in the columns to trace these programs.)

public class Test { public static void main(String[] args) { for (int i = 1; i < 5; i++) { int j = 0; while (j < i) { System.out.print(j + " "); j++; } } } }

(a)

public class Test { public static void main(String[] args) { int i = 0; while (i < 5) { for (int j = i; j > 1; j––) System.out.print(j + " "); System.out.println("****"); i++; } } }

(b)

(c)

public class Test { public static void main(String[] args) { int i = 5; while (i >= 1) { int num = 1; for (int j = 1; j <= i; j++) { System.out.print(num + "xxx"); num *= 2; } System.out.println(); i--; } } }

public class Test { public static void main(String[] args) { int i = 1; do { int num = 1; for (int j = 1; j <= i; j++) { System.out.print(num + "G"); num += 2; }

System.out.println(); i++; } while (i <= 5); } }

(d)

5.10 Minimizing Numeric Errors Using floating-point numbers in the loop continuation condition may cause numeric errors.

Numeric errors involving floating-point numbers are inevitable, because floating-point num- bers are represented in approximation in computers by nature. This section discusses how to minimize such errors through an example.

Listing 5.8 presents an example summing a series that starts with 0.01 and ends with 1.0. The numbers in the series will increment by 0.01, as follows: 0.01 + 0.02 + 0.03, and so on.

Listing 5.8 TestSum.java 1 public class TestSum { 2 public static void main(String[] args) { 3 // Initialize sum 4 float sum = 0; 5 6 // Add 0.01, 0.02, ..., 0.99, 1 to sum 7 for (float i = 0.01f; i <= 1.0f; i = i + 0.01f) 8 sum += i; 9 10 // Display result

Point Key

VideoNote

Minimize numeric errors

loop

M05_LIAN1878_11_GE_C05.indd 202 1/2/18 11:16 PM

5.10 Minimizing Numeric Errors 203

The for loop (lines 7 and 8) repeatedly adds the control variable i to sum. This variable, which begins with 0.01, is incremented by 0.01 after each iteration. The loop terminates when i exceeds 1.0.

The for loop initial action can be any statement, but it is often used to initialize a control variable. From this example, you can see a control variable can be a float type. In fact, it can be any data type.

The exact sum should be 50.50, but the answer is 50.499985. The result is imprecise because computers use a fixed number of bits to represent floating-point numbers, and thus they cannot represent some floating-point numbers exactly. If you change float in the pro- gram to double, as follows, you should see a slight improvement in precision, because a double variable holds 64 bits, whereas a float variable holds 32 bits.

// Initialize sum double sum = 0;

// Add 0.01, 0.02, ..., 0.99, 1 to sum for (double i = 0.01; i <= 1.0; i = i + 0.01) sum += i;

However, you will be stunned to see the result is actually 49.50000000000003. What went wrong? If you display i for each iteration in the loop, you will see that the last i is slightly larger than 1 (not exactly 1). This causes the last i not to be added into sum. The fundamental problem is the floating-point numbers are represented by approximation. To fix the problem, use an integer count to ensure all the numbers are added to sum. Here is the new loop:

double currentValue = 0.01;

for (int count = 0; count < 100; count++) { sum += currentValue; currentValue += 0.01; }

After this loop, sum is 50.50000000000003. This loop adds the numbers from smallest to biggest. What happens if you add numbers from biggest to smallest (i.e., 1.0, 0.99, 0.98, . . . , 0.02, 0.01 in this order) is as follows:

double currentValue = 1.0;

for (int count = 0; count < 100; count++) { sum += currentValue; currentValue –= 0.01; }

After this loop, sum is 50.49999999999995. Adding from biggest to smallest is less accurate than adding from smallest to biggest. This phenomenon is an artifact of the finite-precision arithmetic. Adding a very small number to a very big number can have no effect if the result requires more precision than the variable can store. For example, the inaccurate result of 100000000.0 + 0.000000001 is 100000000.0. To obtain more accurate results, carefully select the order of computation. Adding smaller numbers before bigger numbers to sum is one way to minimize errors.

double precision

numeric error

avoiding numeric error

The sum is 50.499985

11 System.out.println("The sum is " + sum); 12 } 13 }

M05_LIAN1878_11_GE_C05.indd 203 1/2/18 11:16 PM

204 Chapter 5 Loops

5.11 Case Studies Loops are fundamental in programming. The ability to write loops is essential in learning Java programming.

If you can write programs using loops, you know how to program! For this reason, this section presents three additional examples of solving problems using loops.

5.11.1 Case Study: Finding the Greatest Common Divisor The greatest common divisor (gcd) of the two integers 4 and 2 is 2. The greatest common divisor of the two integers 16 and 24 is 8. How would you write this program to find the great- est common divisor? Would you immediately begin to write the code? No. It is important to think before you code. Thinking enables you to generate a logical solution for the problem without concern about how to write the code.

Let the two input integers be n1 and n2. You know that number 1 is a common divisor, but it may not be the greatest common divisor. Therefore, you can check whether k (for k = 2, 3, 4, and so on) is a common divisor for n1 and n2, until k is greater than n1 or n2. Store the common divisor in a variable named gcd. Initially, gcd is 1. Whenever a new common divisor is found, it becomes the new gcd. When you have checked all the possible common divisors from 2 up to n1 or n2, the value in variable gcd is the greatest common divisor.

Once you have a logical solution, type the code to translate the solution into a Java program as follows:

int gcd = 1; // Initial gcd is 1 int k = 2; // Possible gcd

while (k <= n1 && k <= n2) { if (n1 % k == 0 && n2 % k == 0) gcd = k; // Update gcd k++; // Next possible gcd }

// After the loop, gcd is the greatest common divisor for n1 and n2

Listing 5.9 presents the program that prompts the user to enter two positive integers and finds their greatest common divisor.

Listing 5.9 GreatestCommonDivisor.java 1 import java.util.Scanner; 2 3 public class GreatestCommonDivisor { 4 /** Main method */ 5 public static void main(String[] args) { 6 // Create a Scanner 7 Scanner input = new Scanner(System.in); 8 9 // Prompt the user to enter two integers 10 System.out.print("Enter first integer: "); 11 int n1 = input.nextInt(); 12 System.out.print("Enter second integer: "); 13 int n2 = input.nextInt(); 14 15 int gcd = 1; // Initial gcd is 1 16 int k = 2; // Possible gcd 17 while (k <= n1 && k <= n2) { 18 if (n1 % k == 0 && n2 % k == 0) 19 gcd = k; // Update gcd 20 k++;

Point Key

gcd

think before you code

logical solution

check divisor

gcd

input

M05_LIAN1878_11_GE_C05.indd 204 1/2/18 11:16 PM

5.11 Case Studies 205

21 } 22 23 System.out.println("The greatest common divisor for " + n1 + 24 " and " + n2 + " is " + gcd); 25 } 26 }

output

Enter first integer: 125

Enter second integer: 2525 The greatest common divisor for 125 and 2525 is 25

Translating a logical solution to Java code is not unique. For example, you could use a for loop to rewrite the code as follows:

for (int k = 2; k <= n1 && k <= n2; k++) { if (n1 % k == 0 && n2 % k == 0) gcd = k; }

A problem often has multiple solutions, and the gcd problem can be solved in many ways. Programming Exercise 5.14 suggests another solution. A more efficient solution is to use the classic Euclidean algorithm (see Section 22.6).

You might think that a divisor for a number n1 cannot be greater than n1 / 2 and would attempt to improve the program using the following loop:

for (int k = 2; k <= n1 / 2 && k <= n2 / 2; k++) { if (n1 % k == 0 && n2 % k == 0) gcd = k; }

This revision is wrong. Can you find the reason? See Checkpoint Question 5.11.1 for the answer.

5.11.2 Case Study: Predicting the Future Tuition Suppose the tuition for a university is $10,000 this year and tuition increases 7% every year. In how many years will the tuition be doubled?

Before you can write a program to solve this problem, first consider how to solve it by hand. The tuition for the second year is the tuition for the first year * 1.07. The tuition for a future year is the tuition of its preceding year * 1.07. Thus, the tuition for each year can be computed as follows:

double tuition = 10000; int year = 0; // Year 0 tuition = tuition * 1.07; year++; // Year 1 tuition = tuition * 1.07; year++; // Year 2 tuition = tuition * 1.07; year++; // Year 3 ...

Keep computing the tuition for a new year until it is at least 20000. By then, you will know how many years it will take for the tuition to be doubled. You can now translate the logic into the following loop:

double tuition = 10000; // Year 0 int year = 0; while (tuition < 20000) { tuition = tuition * 1.07; year++; }

The complete program is given in Listing 5.10.

think before you type

multiple solutions

erroneous solutions

think before you code

M05_LIAN1878_11_GE_C05.indd 205 1/2/18 11:16 PM

206 Chapter 5 Loops

Listing 5.10 FutureTuition.java 1 public class FutureTuition { 2 public static void main(String[] args) { 3 double tuition = 10000; // Year 0 4 int year = 0; 5 while (tuition < 20000) { 6 tuition = tuition * 1.07; 7 year++; 8 } 9 10 System.out.println("Tuition will be doubled in " 11 + year + " years"); 12 System.out.printf("Tuition will be $%.2f in %1d years", 13 tuition, year); 14 } 15 }

loop next year’s tuition

Tuition will be doubled in 11 years

Tuition will be $21048.52 in 11 years

Remainder

Quotient

16 123

112 11

16 7

0 7

The while loop (lines 5–8) is used to repeatedly compute the tuition for a new year. The loop terminates when the tuition is greater than or equal to 20000.

5.11.3 Case Study: Converting Decimals to Hexadecimals Hexadecimals are often used in computer systems programming (see Appendix F for an intro- duction to number systems). How do you convert a decimal number to a hexadecimal number? To convert a decimal number d to a hexadecimal number is to find the hexadecimal digits hn, hn - 1, hn - 2, c , h2, h1, and h0 such that

d = hn * 16n + hn - 1 * 16n - 1 + hn - 2 * 16n - 2 + g + h2 * 162 + h1 * 161 + h0 * 160

These hexadecimal digits can be found by successively dividing d by 16 until the quotient is 0. The remainders are h0, h1, h2, c , hn - 2, hn - 1, and hn. The hexadecimal digits include the decimal digits 0, 1, 2, 3, 4, 5, 6, 7, 8, and 9, plus A, which is the decimal value 10; B, which is the decimal value 11; C, which is 12; D, which is 13; E, which is 14; and F, which is 15.

For example, the decimal number 123 is 7B in hexadecimal. The conversion is done as follows. Divide 123 by 16. The remainder is 11 (B in hexadecimal) and the quotient is 7. Continue to divide 7 by 16. The remainder is 7 and the quotient is 0. Therefore, 7B is the hexadecimal number for 123.

Listing 5.11 gives a program that prompts the user to enter a decimal number and converts it into a hex number as a string.

M05_LIAN1878_11_GE_C05.indd 206 1/2/18 11:16 PM

5.11 Case Studies 207

Listing 5.11 Dec2Hex.java 1 import java.util.Scanner; 2 3 public class Dec2Hex { 4 /** Main method */ 5 public static void main(String[] args) { 6 // Create a Scanner 7 Scanner input = new Scanner(System.in); 8 9 // Prompt the user to enter a decimal integer 10 System.out.print("Enter a decimal number: "); 11 int decimal = input.nextInt(); 12 13 // Convert decimal to hex 14 String hex = ""; 15 16 while (decimal != 0) { 17 int hexValue = decimal % 16; 18 19 // Convert a decimal value to a hex digit 20 char hexDigit = (0 <= hexValue && hexValue <= 9) ? 21 (char)(hexValue + '0') : (char)(hexValue – 10 + 'A'); 22 23 hex = hexDigit + hex; 24 decimal = decimal / 16; 25 } 26 27 System.out.println("The hex number is " + hex); 28 } 29 }

get a hex char

decimal to hex

input decimal

add to hex string

Enter a decimal number: 1234 The hex number is 4D2

line# decimal hex hexValue hexDigit

14 1234 ""

iteration 1 c 17 223 "2" 2 24 77

iteration 2 c 17 1323 "D2" D 24 4

iteration 3 c 17 423 "4D2" 4 24 0

The program prompts the user to enter a decimal integer (line 11), converts it to a hex num- ber as a string (lines 14–25), and displays the result (line 27). To convert a decimal to a hex number, the program uses a loop to successively divide the decimal number by 16 and obtain its remainder (line 17). The remainder is converted into a hex character (lines 20 and 21). The character is then appended to the hex string (line 23). The hex string is initially empty (line 14). Divide the decimal number by 16 to remove a hex digit from the number (line 24). The loop ends when the remaining decimal number becomes 0.

M05_LIAN1878_11_GE_C05.indd 207 1/2/18 11:16 PM

208 Chapter 5 Loops

The program converts a hexValue between 0 and 15 into a hex character. If hexValue is between 0 and 9, it is converted to (char)(hexValue +'0') (line 21). Recall that when add- ing a character with an integer, the character’s Unicode is used in the evaluation. For example, if hexValue is 5, (char)(hexValue + '0') returns 5. Similarly, if hexValue is between 10 and 15, it is converted to (char)(hexValue – 10 + 'A') (line 21). For instance, if hexValue is 11, (char),(hexValue – 10 + 'A') returns B.

5.11.1 Will the program work if n1 and n2 are replaced by n1 / 2 and n2 / 2 in line 17 in Listing 5.9?

5.11.2 In Listing 5.11, why is it wrong if you change the code (char)(hexValue + '0') to hexValue + '0' in line 21?

5.11.3 In Listing 5.11, how many times the loop body is executed for a decimal number 245, and how many times the loop body is executed for a decimal number 3245?

5.11.4 What is the hex number after E? What is the hex number after F? 5.11.5 Revise line 27 in Listing 5.11 so the program displays hex number 0 if the input

decimal is 0.

5.12 Keywords break and continue The break and continue keywords provide additional controls in a loop.

Pedagogical Note Two keywords, break and continue, can be used in loop statements to provide additional controls. Using break and continue can simplify programming in some cases. Overusing or improperly using them, however, can make programs difficult to read and debug. (Note to instructors: You may skip this section without affecting students’ understanding of the rest of the book.)

You have used the keyword break in a switch statement. You can also use break in a loop to immediately terminate the loop. Listing 5.12 presents a program to demonstrate the effect of using break in a loop.

Listing 5.12 TestBreak.java 1 public class TestBreak { 2 public static void main(String[] args) { 3 int sum = 0; 4 int number = 0; 5 6 while (number < 20) { 7 number++; 8 sum += number; 9 if (sum >= 100) 10 break; 11 } 12 13 System.out.println("The number is " + number); 14 System.out.println("The sum is " + sum); 15 } 16 }

Point Check

Point Key

break statement

break

The number is 14

The sum is 105

M05_LIAN1878_11_GE_C05.indd 208 1/2/18 11:16 PM

5.12 Keywords break and continue 209

The program in Listing 5.12 adds integers from 1 to 20 in this order to sum until sum is greater than or equal to 100. Without the if statement (line 9), the program calculates the sum of the numbers from 1 to 20. However, with the if statement, the loop terminates when sum becomes greater than or equal to 100. Without the if statement, the output would be as follows:

The number is 20

The sum is 210

The sum is 189

The program in Listing 5.13 adds integers from 1 to 20 except 10 and 11 to sum. With the if statement in the program (line 8), the continue statement is executed when number becomes 10 or 11. The continue statement ends the current iteration so that the rest of the statement in the loop body is not executed; therefore, number is not added to sum when it is 10 or 11. Without the if statement in the program, the output would be as follows:

You can also use the continue keyword in a loop. When it is encountered, it ends the current iteration and program control goes to the end of the loop body. In other words, con- tinue breaks out of an iteration, while the break keyword breaks out of a loop. Listing 5.13 presents a program to demonstrate the effect of using continue in a loop.

Listing 5.13 TestContinue.java 1 public class TestContinue { 2 public static void main(String[] args) { 3 int sum = 0; 4 int number = 0; 5 6 while (number < 20) { 7 number++; 8 if (number == 10 || number == 11) 9 continue; 10 sum += number; 11 } 12 13 System.out.println("The sum is " + sum); 14 } 15 }

continue statement

continue

The sum is 210

In this case, all of the numbers are added to sum, even when number is 10 or 11. Therefore, the result is 210, which is 21 more than it was with the if statement.

Note The continue statement is always inside a loop. In the while and do-while loops, the loop-continuation-condition is evaluated immediately after the continue statement. In the for loop, the action-after-each-iteration is performed, then the loop-continuation-condition is evaluated immediately after the continue statement.

M05_LIAN1878_11_GE_C05.indd 209 1/2/18 11:16 PM

210 Chapter 5 Loops

Note Some programming languages have a goto statement. The goto statement indiscrimi- nately transfers control to any statement in the program and executes it. This makes your program vulnerable to errors. The break and continue statements in Java are different from goto statements. They operate only in a loop or a switch statement. The break statement breaks out of the loop, and the continue statement breaks out of the current iteration in the loop.

You can always write a program without using break or continue in a loop (see Check- Point Question 5.12.3). In general, though, using break and continue is appropriate if it simplifies coding and makes programs easier to read.

Suppose you need to write a program to find the smallest factor other than 1 for an integer n (assume n >= 2). You can write a simple and intuitive code using the break statement as follows:

int factor = 2; while (factor <= n) { if (n % factor == 0) break; factor++; } System.out.println("The smallest factor other than 1 for " + n + " is " + factor);

You may rewrite the code without using break as follows:

boolean found = false; int factor = 2; while (factor <= n && !found) { if (n % factor == 0) found = true; else factor++; } System.out.println("The smallest factor other than 1 for " + n + " is " + factor);

Obviously, the break statement makes this program simpler and easier to read in this case. However, you should use break and continue with caution. Too many break and continue statements will produce a loop with many exit points and make the program dif- ficult to read.

Note Programming is a creative endeavor. There are many different ways to write code. In fact, you can find a smallest factor using a rather simple code as follows:

int factor = 2; while (n % factor != 0) factor++; or for (int factor = 2; n % factor != 0; factor++);

The code here finds the smallest factor for an integer n. Programming Exercise 5.16 writes a program that finds all smallest factors in n.

5.12.1 What is the keyword break for? What is the keyword continue for? Will the fol- lowing programs terminate? If so, give the output.

goto

Point Check

M05_LIAN1878_11_GE_C05.indd 210 1/2/18 11:16 PM

5.13 Case Study: Checking Palindromes 211

5.12.2 The for loop on the left is converted into the while loop on the right. What is wrong? Correct it.

int balance = 10; while (true) { if (balance < 9) break; balance = balance – 9; }

System.out.println("Balance is " + balance);

(a)

int balance = 10; while (true) { if (balance < 9) continue; balance = balance – 9; }

System.out.println("Balance is " + balance);

(b)

int sum = 0; for (int i = 0; i < 4; i++) { if (i % 3 == 0) continue; sum += i; }

Converted

Wrong conversion

int i = 0, sum = 0; while (i < 4) { if (i % 3 == 0) continue; sum += i; i++; }

5.12.3 Rewrite the programs TestBreak and TestContinue in Listings 5.12 and 5.13 without using break and continue.

5.12.4 After the break statement in (a) is executed in the following loop, which state- ment is executed? Show the output. After the continue statement in (b) is exe- cuted in the following loop, which statement is executed? Show the output.

for (int i = 1; i < 4; i++) { for (int j = 1; j < 4; j++) { if (i * j > 2) break;

System.out.println(i * j); }

System.out.println(i); }

(a)

for (int i = 1; i < 4; i++) { for (int j = 1; j < 4; j++) { if (i * j > 2) continue;

System.out.println(i * j); }

System.out.println(i); }

(b)

5.13 Case Study: Checking Palindromes This section presents a program that checks whether a string is a palindrome.

A string is a palindrome if it reads the same forward and backward. The words “mom,” “dad,” and “noon,” for instance, are all palindromes.

The problem is to write a program that prompts the user to enter a string and reports whether the string is a palindrome. One solution is to check whether the first character in the string is the same as the last character. If so, check whether the second character is the same as the second-to-last

Point Key

M05_LIAN1878_11_GE_C05.indd 211 1/2/18 11:16 PM

212 Chapter 5 Loops

character. This process continues until a mismatch is found or all the characters in the string are checked, except for the middle character if the string has an odd number of characters.

Listing 5.14 gives the program.

Listing 5.14 Palindrome.java 1 import java.util.Scanner; 2 3 public class Palindrome { 4 /** Main method */ 5 public static void main(String[] args) { 6 // Create a Scanner 7 Scanner input = new Scanner(System.in); 8 9 // Prompt the user to enter a string 10 System.out.print("Enter a string: "); 11 String s = input.nextLine(); 12 13 // The index of the first character in the string 14 int low = 0; 15 16 // The index of the last character in the string 17 int high = s.length() – 1; 18 19 boolean isPalindrome = true; 20 while (low < high) { 21 if (s.charAt(low) != s.charAt(high)) { 22 isPalindrome = false; 23 break; 24 } 25 26 low++; 27 high––; 28 } 29 30 if (isPalindrome) 31 System.out.println(s + " is a palindrome"); 32 else 33 System.out.println(s + " is not a palindrome"); 34 } 35 }

think before you code

high index

low index

input string

update indices

Enter a string: noon

noon is a palindrome

Enter a string: abcdefgnhgfedcba

abcdefgnhgfedcba is not a palindrome

String s

low

a b c d e f g n h g f e d c b a

high

The program uses two variables, low and high, to denote the positions of the two characters at the beginning and the end in a string s (lines 14 and 17), as shown in the following figure.

M05_LIAN1878_11_GE_C05.indd 212 1/2/18 11:16 PM

5.14 Case Study: Displaying Prime Numbers 213

Initially, low is 0 and high is s.length() – 1. If the two characters at these positions match, increment low by 1 and decrement high by 1 (lines 26–27). This process continues until (low >= high) or a mismatch is found (line 21).

The program uses a boolean variable isPalindrome to denote whether the string s is a palindrome. Initially, it is set to true (line 19). When a mismatch is discovered (line 21), isPal- indrome is set to false (line 22) and the loop is terminated with a break statement (line 23).

5.13.1 What happens to the program if (low < high) in line 20 is changed to (low <= high)?

5.14 Case Study: Displaying Prime Numbers This section presents a program that displays the first 50 prime numbers in 5 lines, each containing 10 numbers.

An integer greater than 1 is prime if its only positive divisor is 1 or itself. For example, 2, 3, 5, and 7 are prime numbers, but 4, 6, 8, and 9 are not.

The problem is to display the first 50 prime numbers in 5 lines, each of which contains 10 numbers. The problem can be broken into the following tasks:

■■ Determine whether a given number is prime.

■■ For number = 2, 3, 4, 5, 6, …, test whether it is prime.

■■ Count the prime numbers.

■■ Display each prime number and display 10 numbers per line.

Obviously, you need to write a loop and repeatedly test whether a new number is prime. If the number is prime, increase the count by 1. The count is 0 initially. When it reaches 50, the loop terminates.

Here is the algorithm for the problem:

Set the number of prime numbers to be printed as a constant NUMBER_OF_PRIMES; Use count to track the number of prime numbers and set an initial count to 0; Set an initial number to 2;

while (count < NUMBER_OF_PRIMES) { Test whether number is prime; if number is prime { Display the prime number and increase the count; }

Increment number by 1; }

To test whether a number is prime, check whether it is divisible by 2, 3, 4, and so on up to number/2. If a divisor is found, the number is not a prime. The algorithm can be described as follows:

Use a boolean variable isPrime to denote whether the number is prime; Set isPrime to true initially;

for (int divisor = 2; divisor <= number / 2; divisor++) { if (number % divisor == 0) { Set isPrime to false Exit the loop; } }

Point Check

Point Key

M05_LIAN1878_11_GE_C05.indd 213 1/2/18 11:16 PM

214 Chapter 5 Loops

The complete program is given in Listing 5.15.

Listing 5.15 PrimeNumber.java 1 public class PrimeNumber { 2 public static void main(String[] args) { 3 final int NUMBER_OF_PRIMES = 50; // Number of primes to display 4 final int NUMBER_OF_PRIMES_PER_LINE = 10; // Display 10 per line 5 int count = 0; // Count the number of prime numbers 6 int number = 2; // A number to be tested for primeness 7 8 System.out.println("The first 50 prime numbers are \n"); 9 10 // Repeatedly find prime numbers 11 while (count < NUMBER_OF_PRIMES) { 12 // Assume the number is prime 13 boolean isPrime = true; // Is the current number prime? 14 15 // Test whether number is prime 16 for (int divisor = 2; divisor <= number / 2; divisor++) { 17 if (number % divisor == 0) { // If true, number is not prime 18 isPrime = false; // Set isPrime to false 19 break; // Exit the for loop 20 } 21 } 22 23 // Display the prime number and increase the count 24 if (isPrime) { 25 count++; // Increase the count 26 27 if (count % NUMBER_OF_PRIMES_PER_LINE == 0) { 28 // Display the number and advance to the new line 29 System.out.println(number); 30 } 31 else 32 System.out.print(number + " "); 33 } 34 35 // Check if the next number is prime 36 number++; 37 } 38 } 39 }

count prime numbers

check primeness

exit loop

display if prime

The first 50 prime numbers are

2 3 5 7 11 13 17 19 23 29 31 37 41 43 47 53 59 61 67 71 73 79 83 89 97 101 103 107 109 113 127 131 137 139 149 151 157 163 167 173 179 181 191 193 197 199 211 223 227 229

This is a complex program for novice programmers. The key to developing a programmatic solution for this problem, and for many other problems, is to break it into subproblems and develop solutions for each of them in turn. Do not attempt to develop a complete solution in the first trial. Instead, begin by writing the code to determine whether a given number is prime, then expand the program to test whether other numbers are prime in a loop.

To determine whether a number is prime, check whether it is divisible by a number between 2 and number/2 inclusive (lines 16–21). If so, it is not a prime number (line 18); otherwise, it is a prime number. For a prime number, display it (lines 27–33). If the count is divisible by 10,

subproblem

M05_LIAN1878_11_GE_C05.indd 214 1/2/18 11:16 PM

display the number followed by a newline (lines 27–30). The program ends when the count reaches 50.

The program uses the break statement in line 19 to exit the for loop as soon as the number is found to be a nonprime. You can rewrite the loop (lines 16–21) without using the break statement, as follows:

for (int divisor = 2; divisor <= number / 2 && isPrime; divisor++) { // If true, the number is not prime if (number % divisor == 0) { // Set isPrime to false, if the number is not prime isPrime = false; } }

However, using the break statement makes the program simpler and easier to read in this case. Prime numbers have many applications in computer science. Section 22.7 will study several

efficient algorithms for finding prime numbers.

5.14.1 Simplify the code in lines 27–32 using a conditional operator. Point Check

Key Terms break statement 208 continue statement 209 do-while loop 192 for loop 195 infinite loop 184 input redirection 192 iteration 182 loop 182

loop body 182 nested loop 200 off-by-one error 184 output redirection 192 posttest loop 198 pretest loop 198 sentinel value 190 while loop 182

ChapTer summary 1. There are three types of repetition statements: the while loop, the do-while loop, and

the for loop.

2. The part of the loop that contains the statements to be repeated is called the loop body.

3. A one-time execution of a loop body is referred to as an iteration of the loop.

4. An infinite loop is a loop statement that executes infinitely.

5. In designing loops, you need to consider both the loop control structure and the loop body.

6. The while loop checks the loop-continuation-condition first. If the condition is true, the loop body is executed; if it is false, the loop terminates.

7. The do-while loop is similar to the while loop, except the do-while loop executes the loop body first then checks the loop-continuation-condition to decide whether to continue or to terminate.

8. The while loop and the do-while loop often are used when the number of repetitions is not predetermined.

Chapter Summary 215

M05_LIAN1878_11_GE_C05.indd 215 1/2/18 11:16 PM

216 Chapter 5 Loops

9. A sentinel value is a special value that signifies the end of the loop.

10. The for loop generally is used to execute a loop body a fixed number of times.

11. The for loop control has three parts. The first part is an initial action that often initializes a control variable. The second part, the loop-continuation-condition, determines whether the loop body is to be executed. The third part is executed after each iteration and is often used to adjust the control variable. Usually, the loop control variables are initialized and changed in the control structure.

12. The while loop and for loop are called pretest loops because the continuation condition is checked before the loop body is executed.

13. The do-while loop is called a posttest loop because the condition is checked after the loop body is executed.

14. Two keywords break and continue can be used in a loop.

15. The break keyword immediately ends the innermost loop, which contains the break.

16. The continue keyword only ends the current iteration.

Quiz Answer the quiz for this chapter online at the Companion Website.

programming exerCises

Pedagogical Note Read each problem several times until you understand it. Think how to solve the problem before starting to write code. Translate your logic into a program.

A problem often can be solved in many different ways. Students are encouraged to explore various solutions.

Sections 5.2–5.7 *5.1 (Pass or fail) Write a program that prompts a student to enter a Java score. If the

score is greater or equal to 60, display “you pass the exam”; otherwise, display “you don’t pass the exam”. Your program ends with input -1. Here is a sample run:

read and think before coding

explore solutions

Enter your score: 80 You pass the exam.

Enter your score: 59 You don't pass the exam.

Enter your score: −1

No numbers are entered except 0

M05_LIAN1878_11_GE_C05.indd 216 1/2/18 11:16 PM

5.2 (Repeat multiplications) Listing 5.4, SubtractionQuizLoop.java, generates five random subtraction questions. Revise the program to generate ten random multiplication ques- tions for two integers between 1 and 12. Display the correct count and test time.

5.3 (Conversion from Cº to Fº) Write a program that displays the following table (note that farenheit = celsius * 9/5 + 32):

Celsius Fahrenheit 0 32.0 2 35.6 ... 98 208.4 100 212.0

5.4 (Conversion from inch to centimeter) Write a program that displays the following table (note that 1 inch is 2.54 centimeters):

Inches Centimetres 1 2.54 2 5.08 ... 9 22.86 10 25.4

5.5 (Conversion from Cº to Fº and Fº to Cº) Write a program that displays the following two tables side by side:

Celsius Fahrenheit | Fahrenheit Celsius 0 32.000 | 20 −6.667 2 35.600 | 25 −3.889 ... 98 208.400 | 265 129.444 100 212.000 | 270 132.222

5.6 (Conversion from square meter to ping) Write a program that displays the following two tables side by side (note that 1 ping = 3.305 square meters):

Ping Square meter | Square meter Ping 10 33.050 | 30 9.077 15 49.575 | 35 10.590 ... 75 247.875 | 95 28.744 80 264.400 | 100 30.257

**5.7 (Financial application: compute future tuition) Suppose that the tuition for a uni- versity is $10,000 this year and increases 6% every year. In one year, the tuition will be $10,600. Write a program that computes the tuition in ten years and the total cost of four years’ worth of tuition after the tenth year.

5.8 (Find the highest score) Write a program that prompts the user to enter the number of students and each student’s name and score, and finally displays the name of the student with the highest score. Use the next() method in the Scanner class to read a name, rather than using the nextLine() method.

*5.9 (Find the two lowest scores) Write a program that prompts the user to enter the number of students and each student’s name and score, and finally displays the names of the students with the lowest and second-lowest scores.

Programming Exercises 217

M05_LIAN1878_11_GE_C05.indd 217 1/2/18 11:16 PM

218 Chapter 5 Loops

5.10 (Find numbers divisible by 3 and 4) Write a program that displays all the numbers from 100 to 1,000, ten per line, that are divisible by 3 and 4. Numbers are separated by exactly one space.

5.11 (Find numbers divisible by 3 or 4, but not both) Write a program that displays all the numbers from 100 to 200, ten per line, that are divisible by 3 or 4, but not both. Numbers are separated by exactly one space.

5.12 (Find the smallest n such that n3 7 12,000) Use a while loop to find the smallest integer n such that n3 is greater than 12,000.

5.13 (Find the largest n such that n2 6 12,000) Use a while loop to find the largest integer n such that n2 is less than 12,000.

Sections 5.8–5.10 *5.14 (Compute the greatest common divisor) Another solution for Listing 5.9 to find the

greatest common divisor of two integers n1 and n2 is as follows: First find d to be the minimum of n1 and n2, then check whether d, d–1, d–2, …, 2, or 1 is a divi- sor for both n1 and n2 in this order. The first such common divisor is the greatest common divisor for n1 and n2. Write a program that prompts the user to enter two positive integers and displays the gcd.

*5.15 (Display the ASCII character table) Write a program that prints the characters in the ASCII character table from ! to ~. Display 10 characters per line. The ASCII table is given in Appendix B. Characters are separated by exactly one space.

*5.16 (Find the factors of an integer) Write a program that reads an integer and displays all its smallest factors in an increasing order. For example, if the input integer is 120, the output should be as follows: 2, 2, 2, 3, 5.

** 5.17 (Display pyramid) Write a program that prompts the user to enter an integer from 1 to 15 and displays a pyramid, as shown in the following sample run:

Enter the number of lines: 7

7 6 5 4 3 2 1 2 3 4 5 6 7

6 5 4 3 2 1 2 3 4 5 6

5 4 3 2 1 2 3 4 5

4 3 2 1 2 3 4

3 2 1 2 3

2 1 2

*5.18 (Display four patterns using loops) Use nested loops that display the following patterns in four separate programs:

Pattern A Pattern B Pattern C Pattern D

* * * * * * * * * * * * * *

M05_LIAN1878_11_GE_C05.indd 218 1/2/18 11:16 PM

**5.19 (Display numbers in a pyramid pattern) Write a nested for loop that prints the following output: 1

1 3 1

1 3 9 3 1

1 3 9 27 9 3 1

1 3 9 27 81 27 9 3 1

1 3 9 27 81 243 81 27 9 3 1

1 3 9 27 81 243 729 243 81 27 9 3 1

1 3 9 27 81 243 729 2187 729 243 81 27 9 3 1

*5.20 (Display prime numbers between 2 and 1,200) Modify Listing 5.15 to display all the prime numbers between 2 and 1,200, inclusive. Display eight prime numbers per line. Numbers are separated by exactly one space.

Comprehensive ** 5.21 (Financial application: compare loans with various interest rates) Write a program

that lets the user enter the loan amount and loan period in number of years and displays the monthly and total payments for each interest rate starting from 5% to 10%, with an increment of 1/4. Here is a sample run:

Loan Amount: 10000

Number of Years: 5

Interest Rate Monthly Payment Total Payment

5.000% 188.71 11322.74

5.250% 189.86 11391.59

5.500% 191.01 11460.70

...

9.750% 211.24 12674.55

10.000% 212.47 12748.23

For the formula to compute monthly payment, see Listing 2.9, ComputeLoan.java.

** 5.22 (Financial application: loan amortization schedule) The monthly payment for a given loan pays the principal and the interest. The monthly interest is computed by multiply- ing the monthly interest rate and the balance (the remaining principal). The principal paid for the month is therefore the monthly payment minus the monthly interest. Write a program that lets the user enter the loan amount, number of years, and interest rate then displays the amortization schedule for the loan. Here is a sample run:

VideoNote

Display loan schedule

Loan Amount: 10000

Number of Years: 1

Annual Interest Rate: 7

Monthly Payment: 865.26

Total Payment: 10383.21

Payment# Interest Principal Balance 1 58.33 806.93 9193.07 2 53.62 811.64 8381.43 ... 11 10.00 855.26 860.27 12 5.01 860.25 0.01

Programming Exercises 219

M05_LIAN1878_11_GE_C05.indd 219 1/2/18 11:16 PM

220 Chapter 5 Loops

Note The balance after the last payment may not be zero. If so, the last payment should be the normal monthly payment plus the final balance.

Hint: Write a loop to display the table. Since the monthly payment is the same for each month, it should be computed before the loop. The balance is initially the loan amount. For each iteration in the loop, compute the interest and principal, and update the balance. The loop may look as follows:

for (i = 1; i <= numberOfYears * 12; i++) { interest = monthlyInterestRate * balance; principal = monthlyPayment – interest; balance = balance – principal; System.out.println(i + "\t\t" + interest + "\t\t" + principal + "\t\t" + balance); }

*5.23 (Demonstrate cancellation errors) A cancellation error occurs when you are manipulating a very large number with a very small number. The large number may cancel out the smaller number. For example, the result of 100000000.0 + 0.000000001 is equal to 100000000.0. To avoid cancellation errors and obtain more accurate results, carefully select the order of computation. For example, in computing the following summation, you will obtain more accurate results by computing from right to left rather than from left to right:

1 + 1 2

+ 1 3

+ c + 1 n

Write a program that compares the results of the summation of the preceding series, computing from left to right and from right to left with n = 50000.

*5.24 (Sum a series) Write a program to compute the following summation:

1 3

+ 3 5

+ 5 7

+ 7 9

+ 9 11

+ 11 13

+ g + 95 97

+ 97 99

** 5.25 (Compute p) You can approximate p by using the following summation:

p = 4 a1 - 1 3

+ 1 5

- 1 7

+ 1 9

- 1 11

+ g + (-)i+ 1

2i - 1 b

Write a program that displays the p value for i = 10000, 20000, …, and 100000.

** 5.26 (Compute e) You can approximate e using the following summation:

e = 1 + 1 1!

+ 1 2!

+ 1 3!

+ 1 4!

+ g + 1 i!

Write a program that displays the e value for i = 10000, 20000, …, and 100000. (Hint: Because i! = i * (i - 1) * c * 2 * 1, then

1 i!

is 1

i(i - 1)!

Initialize e and item to be 1, and keep adding a new item to e. The new item is the previous item divided by i, for i >= 2.)

** 5.27 (Display leap years) Write a program that displays all the leap years, ten per line, from 2014 to 2114, separated by exactly one space. Also display the number of leap years in this period.

VideoNote

Sum a series

M05_LIAN1878_11_GE_C05.indd 220 1/2/18 11:16 PM

** 5.28 (Display the first days of each month) Write a program that prompts the user to enter the year and first day of the year, then displays the first day of each month in the year. For example, if the user entered the year 2013, and 2 for Tuesday, January 1, 2013, your program should display the following output:

January 1, 2013 is Tuesday ... December 1, 2013 is Sunday

** 5.29 (Display calendars) Write a program that prompts the user to enter the year and first day of the year and displays the calendar table for the year on the console. For example, if the user entered the year 2013, and 2 for Tuesday, January 1, 2013, your program should display the calendar for each month in the year, as follows:

January 2013

Sun Mon Tue Wed Thu Fri Sat

1 2 3 4 5

6 7 8 9 10 11 12

13 14 15 16 17 18 19

20 21 22 23 24 25 26

27 28 29 30 31

. . .

December 2013

Sun Mon Tue Wed Thu Fri Sat

1 2 3 4 5 6 7

8 9 10 11 12 13 14

15 16 17 18 19 20 21

22 23 24 25 26 27 28

29 30 31

* 5.30 (Financial application: compound value) Suppose you save $100 each month in a savings account with annual interest rate 3.75%. The monthly interest rate is 0.0375 / 12 = 0.003125. After the first month, the value in the account becomes

100 * (1 + 0.003125) = 100.3125

After the second month, the value in the account becomes

100 + 100.3125) * (1 + 0.003125) = 200.938

After the third month, the value in the account becomes

(100 + 200.938) * (1 + 0.003125) = 301.878

and so on.

Write a program that prompts the user to enter an amount (e.g., 100), the annual interest rate (e.g., 3.75), and the number of months (e.g., 6) and displays the amount in the savings account after the given month.

Programming Exercises 221

M05_LIAN1878_11_GE_C05.indd 221 1/2/18 11:16 PM

222 Chapter 5 Loops

*5.31 (Financial application: compute CD value) Suppose you put $10,000 into a CD with an annual percentage yield of 6.15%. After one month, the CD is worth

10000 + 10000 * 6.15 / 1200 = 10051.25

After two months, the CD is worth

10051.25 + 10051.25 * 6.15 / 1200 = 10102.76

After three months, the CD is worth

10102.76 + 10102.76 * 6.15 / 1200 = 10154.53

and so on.

Write a program that prompts the user to enter an amount (e.g., 10000), the annual percentage yield (e.g., 6.15), and the number of months (e.g., 18) and displays a table as shown in the sample run.

Enter the initial deposit amount: 10000

Enter annual percentage yield: 6.15

Enter maturity period (number of months): 18

Month CD Value 1 10051.25 2 10102.76 ... 17 10907.90 18 10963.81

** 5.32 (Game: lottery) Revise Listing 3.8, Lottery.java, to generate a lottery of a two-digit number. The two digits in the number are distinct. (Hint: Generate the first digit. Use a loop to continuously generate the second digit until it is different from the first digit.)

** 5.33 (Perfect number) A positive integer is called a perfect number if it is equal to the sum of all of its positive divisors, excluding itself. For example, 6 is the first perfect number because 6 = 3 + 2 + 1. The next is 28 = 14 + 7 + 4 + 2 + 1. There are four perfect numbers 6 10,000. Write a program to find all these four numbers.

*** 5.34 (Game: scissor, rock, paper) Programming Exercise 3.17 gives a program that plays the scissor-rock-paper game. Revise the program to let the user continuously play until either the user or the computer wins three times more than their opponent.

*5.35 (Summation) Write a program to compute the following summation:

1 + 22 + 122 + 23 + 123 + 24 + c + 12999 + 21000 ** 5.36 (Business application: checking ISBN) Use loops to simplify Programming

Exercise 3.9.

** 5.37 ( Decimal to binary) Write a program that prompts the user to enter a decimal integer then displays its corresponding binary value. Don’t use Java’s Integer .toBinaryString(int) in this program.

** 5.38 (Decimal to octal) Write a program that prompts the user to enter a decimal integer and displays its corresponding octal value. Don’t use Java’s Integer. toOctalString(int) in this program.

M05_LIAN1878_11_GE_C05.indd 222 1/2/18 11:16 PM

*5.39 (Financial application: find the sales amount) You have just started a sales job in a department store. Your pay consists of a base salary and a commission. The base salary is $5,000. The scheme shown below is used to determine the commission rate.

Sales Amount Commission Rate

$0.01–$5,000 6 percent

$5,000.01–$10,000 8 percent

$10,000.01 and above 10 percent

Note that this is a graduated rate. The rate for the first $5,000 is at 6%, the next $5000 is at 8%, and the rest is at 10%. If your sales amounts to $25,000, the com- mission is 5,000 * 6% + 5,000 * 8% + 15,000 * 10% = 2,200. Your goal is to earn $30,000 a year. Write a program to find the minimum sales you have to gener- ate in order to make $30,000.

5.40 (Simulation: heads or tails) Write a program that simulates flipping a coin two millions times and displays the number of heads and tails.

*5.41 (Occurrence of max numbers) Write a program that reads integers, finds the largest of them, and counts its occurrences. Assume the input ends with number 0. Sup- pose you entered 3 5 2 5 5 5 0; the program finds that the largest is 5 and the occurrence count for 5 is 4.

(Hint: Maintain two variables, max and count. max stores the current max number and count stores its occurrences. Initially, assign the first number to max and 1 to count. Compare each subsequent number with max. If the number is greater than max, assign it to max and reset count to 1. If the number is equal to max, increment count by 1.)

Enter numbers: 3 5 2 5 5 5 0

The largest number is 5

The occurrence count of the largest number is 4

*5.42 (Financial application: find the sales amount) Rewrite Programming Exercise 5.39 as follows:

■■ Use a while loop instead of a do-while loop. ■■ Let the user enter COMMISSION_SOUGHT instead of fixing it as a constant.

*5.43 (Math: combinations) Write a program that displays all possible combinations for picking two numbers from integers 1 to 7. Also display the total number of all combinations.

1 2

1 3 ... ...

The total number of all combinations is 21

Programming Exercises 223

M05_LIAN1878_11_GE_C05.indd 223 1/2/18 11:16 PM

224 Chapter 5 Loops

*5.44 (Computer architecture: bit-level operations) A byte value is stored in 8 bits. Write a program that prompts the user to enter a byte integer and displays the 8 bits for the integer. Here are sample runs:

Enter an integer: 5

The 8 bits are 00000101

Enter an integer: –5

The 8 bits are 11111011

(Hint: You need to use the bitwise right shift operator (>>) and the bitwise AND operator (&), which are covered in Appendix G, Bitwise Operations.)

** 5.45 (Statistics: compute mean and standard deviation) In business applications, you are often asked to compute the mean and standard deviation of data. The mean is simply the average of the numbers. The standard deviation is a statistic that tells you how tightly all the various data are clustered around the mean in a set of data. For example, what is the average age of the students in a class? How close are the ages? If all the students are the same age, the deviation is 0.

Write a program that prompts the user to enter 10 numbers and displays the mean and standard deviations of these numbers using the following formula:

mean = a

i = 1 xi

n =

x1 + x2 + g + xn n

deviation = cani = 1xi2 - aa n

i = 1 xib

n - 1 Here is a sample run:

Enter 10 numbers: 1 2 3 4.5 5.6 6 7 8 9 10

The mean is 5.61

The standard deviation is 2.99794

Enter a string: ABCD

The reversed string is DCBA

*5.46 (Reverse a string) Write a program that prompts the user to enter a string and displays the string in reverse order.

M05_LIAN1878_11_GE_C05.indd 224 1/2/18 11:16 PM

Enter the first 12 digits of an ISBN-13 as a string: 978013213080

The ISBN-13 number is 9780132130806

Enter the first 12 digits of an ISBN-13 as a string: 978013213079

The ISBN-13 number is 9780132130790

Enter the first 12 digits of an ISBN-13 as a string: 97801320

97801320 is an invalid input

*5.47 (Business: check ISBN-13) ISBN-13 is a new standard for identifying books. It uses 13 digits d1d2d3d4d5d6d7d8d9d10d11d12d13. The last digit d13 is a checksum, which is calculated from the other digits using the following formula:

10 - (d1 + 3d2 + d3 + 3d4 + d5 + 3d6 + d7 + 3d8 + d9 + 3d10 + d11 + 3d12),10

If the checksum is 10, replace it with 0. Your program should read the input as a string. Here are sample runs:

Enter a string: Beijing Chicago

ejnhcg

*5.48 (Process string) Write a program that prompts the user to enter a string and displays the characters at even positions. Here is a sample run:

Enter a string: Programming is fun

The number of vowels is 5

The number of consonants is 11

*5.49 (Count vowels and consonants) Assume that the letters A, E, I, O, and U are vowels. Write a program that prompts the user to enter a string, and displays the number of vowels and consonants in the string.

Programming Exercises 225

M05_LIAN1878_11_GE_C05.indd 225 1/2/18 11:16 PM

226 Chapter 5 Loops

Enter the first string: Welcome to C++

Enter the second string: Welcome to programming

The common prefix is Welcome to

*5.51 (Longest common prefix) Write a program that prompts the user to enter two strings and displays the largest common prefix of the two strings. Here are some sample runs:

Enter the first string: Atlanta

Enter the second string: Macon

Atlanta and Macon have no common prefix

*5.50 (Print multiplication table) Write a program that uses for, while, and do-while loop statements to display the multiplication table. Here is a sample run:

Enter a string: Welcome to Java

1*1= 1 2*1= 2 3*1= 3 4*1= 4 5*1= 5 6*1= 6 7*1= 7 8*1= 8 9*1= 9

1*2= 1 2*2= 2 3*2= 6 4*2= 8 5*2=10 6*2=12 7*2=14 8*2=16 9*2=18

1*3= 3 2*3= 6 3*3= 9 4*3=12 5*3=15 6*3=18 7*3=21 8*3=24 9*3=27

1*4= 4 2*4= 8 3*4=12 4*4=16 5*4=20 6*4=24 7*4=28 8*4=32 9*4=36

1*5= 5 2*5=10 3*5=15 4*5=20 5*5=25 6*5=30 7*5=35 8*5=40 9*5=45

1*6= 6 2*6=12 3*6=18 4*6=24 5*6=30 6*6=36 7*6=42 8*6=48 9*6=54

1*7= 7 2*7=14 3*7=21 4*7=28 5*7=35 6*7=42 7*7=49 8*7=56 9*7=63

1*8= 8 2*8=16 3*8=24 4*8=32 5*8=40 6*8=48 7*8=56 8*8=64 9*8=72

1*9= 9 2*9=18 3*9=27 4*9=36 5*9=45 6*9=54 7*9=63 8*9=72 9*9=81

M05_LIAN1878_11_GE_C05.indd 226 1/2/18 11:16 PM

Methods

Objectives ■■ To define methods with formal parameters (§6.2).

■■ To invoke methods with actual parameters (i.e., arguments) (§6.2).

■■ To define methods with a return value (§6.3).

■■ To define methods without a return value and distinguish the differ- ences between void methods and value-returning methods (§6.4).

■■ To pass arguments by value (§6.5).

■■ To develop reusable code that is modular, easy to read, easy to debug, and easy to maintain (§6.6).

■■ To write a method that converts hexadecimals to decimals (§6.7).

■■ To use method overloading and understand ambiguous overloading (§6.8).

■■ To determine the scope of variables (§6.9).

■■ To apply the concept of method abstraction in software development (§6.10).

■■ To design and implement methods using stepwise refinement (§6.11).

Chapter

M06_LIAN1878_11_GE_C06.indd 227 1/2/18 10:42 PM

228 Chapter 6 Methods

6.1 Introduction Methods can be used to define reusable code and organize and simplify coding.

Suppose you need to find the sum of integers from 1 to 10, 20 to 37, and 35 to 49, respectively. You may write the code as follows:

int sum = 0; for (int i = 1; i <= 10; i++) sum += i; System.out.println("Sum from 1 to 10 is " + sum);

sum = 0; for (int i = 20; i <= 37; i++) sum += i; System.out.println("Sum from 20 to 37 is " + sum);

sum = 0; for (int i = 35; i <= 49; i++) sum += i; System.out.println("Sum from 35 to 49 is " + sum);

You may have observed that computing these sums from 1 to 10, 20 to 37, and 35 to 49 are very similar, except that the starting and ending integers are different. Wouldn’t it be nice if we could write the common code once and reuse it? We can do so by defining a method and invoking it.

The preceding code can be simplified as follows:

1 public static int sum(int i1, int i2) { 2 int result = 0; 3 for (int i = i1; i <= i2; i++) 4 result += i; 5 6 return result; 7 } 8 9 public static void main(String[] args) { 10 System.out.println("Sum from 1 to 10 is " + sum(1, 10)); 11 System.out.println("Sum from 20 to 37 is " + sum(20, 37)); 12 System.out.println("Sum from 35 to 49 is " + sum(35, 49)); 13 }

Lines 1–7 define the method named sum with two parameters i1 and i2. The statements in the main method invoke sum(1, 10) to compute the sum from 1 to 10, sum(20, 37) to compute the sum from 20 to 37, and sum(35, 49) to compute the sum from 35 to 49.

A method is a collection of statements grouped together to perform an operation. In earlier chap- ters you have used predefined methods such as System.out.println, System.exit, Math .pow, and Math.random. These methods are defined in the Java library. In this chapter, you will learn how to define your own methods and apply method abstraction to solve complex problems.

6.2 Defining a Method A method definition consists of method name, parameters, return value type, and body.

The syntax for defining a method is as follows:

modifier returnValueType methodName(list of parameters) { // Method body; }

Point Key

problem

why methods?

define sum method

main method invoke sum

method

Point Key

M06_LIAN1878_11_GE_C06.indd 228 1/2/18 10:42 PM

6.2 Defining a Method 229

The method header specifies the modifiers, return value type, method name, and parameters of the method. The static modifier is used for all the methods in this chapter. The reason for using it will be discussed in Chapter 9, Objects and Classes.

A method may return a value. The returnValueType is the data type of the value the method returns. Some methods perform desired operations without returning a value. In this case, the returnValueType is the keyword void. For example, the returnValueType is void in the main method, as well as in System.exit, and System.out.println. If a method returns a value, it is called a value-returning method; otherwise, it is called a void method.

The variables defined in the method header are known as formal parameters or simply parameters. A parameter is like a placeholder: when a method is invoked, you pass a value to the parameter. This value is referred to as an actual parameter or argument. The parameter list refers to the method’s type, order, and the number of parameters. The method name and the parameter list together constitute the method signature. Parameters are optional; that is, a method may contain no parameters. For example, the Math.random() method has no parameters.

The method body contains a collection of statements that implement the method. The method body of the max method uses an if statement to determine which number is larger and return the value of that number. In order for a value-returning method to return a result, a return statement using the keyword return is required. The method terminates when a return statement is executed.

Note Some programming languages refer to methods as procedures and functions. In those languages, a value-returning method is called a function and a void method is called a procedure.

Caution In the method header, you need to declare each parameter separately. For instance, max(int num1, int num2) is correct, but max(int num1, num2) is wrong.

method header

modifier

value-returning method

void method formal parameter

parameter actual parameter argument

parameter list

method signature

Figure 6.1 A method definition consists of a method header and a method body.

De�ne a method Invoke a method

int z = max(x, y);

actual parameters (arguments)

public static int max(int num1, int num2) {

int result;

if (num1 > num2) result = num1;

else result = num2;

return result;

}

modi�er return value

type method name

formal parameters

return value

method body

method header

parameter list method signature

Let’s look at a method defined to find the larger between two integers. This method, named max, has two int parameters, num1 and num2, the larger of which is returned by the method. Figure 6.1 illustrates the components of this method.

M06_LIAN1878_11_GE_C06.indd 229 1/2/18 10:42 PM

230 Chapter 6 Methods

Note We say “define a method” and “declare a variable.” We are making a subtle distinction here. A definition defines what the defined item is, but a declaration usually involves allocating memory to store data for the declared item.

6.3 Calling a Method Calling a method executes the code in the method.

In a method definition, you define what the method is to do. To execute the method, you have to call or invoke it. The program that calls the function is called a caller. There are two ways to call a method, depending on whether the method returns a value or not.

If a method returns a value, a call to the method is usually treated as a value. For example,

int larger = max(3, 4);

calls max(3, 4) and assigns the result of the method to the variable larger. Another example of a call that is treated as a value is

System.out.println(max(3, 4));

which prints the return value of the method call max(3, 4). If a method returns void, a call to the method must be a statement. For example, the method

println returns void. The following call is a statement:

System.out.println("Welcome to Java!");

Note A value-returning method can also be invoked as a statement in Java. In this case, the caller simply ignores the return value. This is not often done, but it is permissible if the caller is not interested in the return value.

When a program calls a method, program control is transferred to the called method. A called method returns control to the caller when its return statement is executed or when its method- ending closing brace is reached.

Listing 6.1 presents a complete program that is used to test the max method.

Listing 6.1 TestMax.java 1 public class TestMax { 2 /** Main method */ 3 public static void main(String[] args) { 4 int i = 5; 5 int j = 2; 6 int k = max(i, j); 7 System.out.println("The maximum of " + i + 8 " and " + j + " is " + k); 9 } 10 11 /** Return the max of two numbers */ 12 public static int max(int num1, int num2) { 13 int result; 14 15 if (num1 > num2) 16 result = num1; 17 else 18 result = num2; 19 20 return result; 21 } 22 }

define vs. declare

Point Key

VideoNote

Define/invoke max method

main method

invoke max

define method

caller

M06_LIAN1878_11_GE_C06.indd 230 1/2/18 10:42 PM

6.3 Calling a Method 231

line# i j k num1 num2 result

4 5

Invoking max c 5 212 5 213 undefined 16 5

6 5

The maximum of 5 and 2 is 5

Figure 6.2 When the max method is invoked, the flow of control transfers to it. Once the max method is finished, it returns control back to the caller.

public static void main(String[] args) { int i = 5; int j = 2; int k = max(i, j); System.out.println( "The maximum of " + i + " and " + j + " is " + k); }

pass the value j

pass the value i

public static int max(int num1, int num2) { int result;

if (num1 > num2) result = num1; else result = num2;

return result; }

This program contains the main method and the max method. The main method is just like any other method, except that it is invoked by the JVM to start the program.

The main method’s header is always the same. Like the one in this example, it includes the modifiers public and static, return value type void, method name main, and a parameter of the String[] type. String[] indicates the parameter is an array of String, a subject addressed in Chapter 7.

The statements in main may invoke other methods that are defined in the class that contains the main method or in other classes. In this example, the main method invokes max(i, j), which is defined in the same class with the main method.

When the max method is invoked (line 6), variable i’s value 5 is passed to num1 and vari- able j’s value 2 is passed to num2 in the max method. The flow of control transfers to the max method and the max method is executed. When the return statement in the max method is executed, the max method returns the control to its caller (in this case, the caller is the main method). This process is illustrated in Figure 6.2.

main method

max method

Caution A return statement is required for a value-returning method. The method given in (a) is logically correct, but it has a compile error because the Java compiler thinks this method might not return a value.

M06_LIAN1878_11_GE_C06.indd 231 1/2/18 10:42 PM

232 Chapter 6 Methods

To fix this problem, delete if (n < 0) in (a), so the compiler will see a return statement to be reached regardless of how the if statement is evaluated.

Note Methods enable code sharing and reuse. The max method can be invoked from any class, not just TestMax. If you create a new class, you can invoke the max method using ClassName.methodName (i.e., TestMax.max).

Each time a method is invoked, the system creates an activation record (also called an activation frame) that stores parameters and variables for the method and places the activation record in an area of memory known as a call stack. A call stack is also known as an execution stack, runtime stack, or machine stack and it is often shortened to just “the stack.” When a method calls another method, the caller’s activation record is kept intact and a new activation record is created for the new method called. When a method finishes its work and returns to its caller, its activation record is removed from the call stack.

A call stack stores the activation records in a last-in, first-out fashion: The activation record for the method that is invoked last is removed first from the stack. For example, suppose method m1 calls method m2, and m2 calls method m3. The runtime system pushes m1’s activa- tion record into the stack, then m2’s, and then m3’s. After m3 is finished, its activation record is removed from the stack. After m2 is finished, its activation record is removed from the stack. After m1 is finished, its activation record is removed from the stack.

Understanding call stacks helps you to comprehend how methods are invoked. The vari- ables defined in the main method in Listing 6.1 are i, j, and k. The variables defined in the max method are num1, num2, and result. The variables num1 and num2 are defined in the method signature and are parameters of the max method. Their values are passed through method invocation. Figure 6.3 illustrates the activation records for method calls in the stack.

reusing method

activation record

call stack

Figure 6.3 When the max method is invoked, the flow of control transfers to the max method. Once the max method is finished, it returns control back to the caller.

Activation record for the main method

Activation record for the max method

k: j: 2 i: 5

(a) The main method is invoked.

Activation record for the main method

k: j: 2 i: 5

(d) The max method is �nished and the return value is sent to k.

Activation record for the main method

k: 5 j: 2 i: 5

Stack is empty

(b) The max method is invoked.

(e) The main method is �nished.

Activation record for the main method

Activation record for the max method

k: j: 2 i: 5

result: 5 num2: 2 num1: 5

result: num2: 2 num1: 5

(a)

public static int sign(int n) {

if (n > 0) return 1; else if (n == 0) return 0; else if (n < 0) return −1; }

public static int sign(int n) { if (n > 0) return 1; else if (n == 0) return 0; else return –1; }

(b)

Should be

M06_LIAN1878_11_GE_C06.indd 232 1/2/18 10:42 PM

6.4 void vs. Value-Returning Methods 233

6.4 void vs. Value-Returning Methods A void method does not return a value.

The preceding section gives an example of a value-returning method. This section shows how to define and invoke a void method. Listing 6.2 gives a program that defines a method named printGrade and invokes it to print the grade for a given score.

Listing 6.2 TestVoidMethod.java 1 public class TestVoidMethod { 2 public static void main(String[] args) { 3 System.out.print("The grade is "); 4 printGrade(78.5); 5 6 System.out.print("The grade is "); 7 printGrade(59.5); 8 } 9 10 public static void printGrade(double score) { 11 if (score >= 90.0) { 12 System.out.println('A'); 13 } 14 else if (score >= 80.0) { 15 System.out.println('B'); 16 } 17 else if (score >= 70.0) { 18 System.out.println('C'); 19 } 20 else if (score >= 60.0) { 21 System.out.println('D'); 22 } 23 else { 24 System.out.println('F'); 25 } 26 } 27 }

Point Key

VideoNote

Use void method

main method

invoke printGrade

printGrade method

The grade is C

The grade is F

The printGrade method is a void method because it does not return any value. A call to a void method must be a statement. Therefore, it is invoked as a statement in line 4 in the main method. Like any Java statement, it is terminated with a semicolon.

To see the differences between a void and value-returning method, let’s redesign the printGrade method to return a value. The new method, which we call getGrade, returns the grade as given in Listing 6.3.

Listing 6.3 TestReturnGradeMethod.java 1 public class TestReturnGradeMethod { 2 public static void main(String[] args) { 3 System.out.print("The grade is " + getGrade(78.5)); 4 System.out.print("\nThe grade is " + getGrade(59.5)); 5 } 6

invoke void method

void vs. value-returned

main method

invoke getGrade

M06_LIAN1878_11_GE_C06.indd 233 1/2/18 10:42 PM

234 Chapter 6 Methods

7 public static char getGrade(double score) { 8 if (score >= 90.0) 9 return 'A'; 10 else if (score >= 80.0) 11 return 'B'; 12 else if (score >= 70.0) 13 return 'C'; 14 else if (score >= 60.0) 15 return 'D'; 16 else 17 return 'F'; 18 } 19 }

getGrade method

The grade is C

The grade is F

The getGrade method defined in lines 7–18 returns a character grade based on the numeric score value. The caller invokes this method in lines 3 and 4.

The getGrade method can be invoked by a caller wherever a character may appear. The printGrade method does not return any value, so it must be invoked as a statement.

Note A return statement is not needed for a void method, but it can be used for terminat- ing the method and returning to the method’s caller. The syntax is simply

return;

This is not often done, but sometimes it is useful for circumventing the normal flow of control in a void method. For example, the following code has a return statement to terminate the method when the score is invalid:

public static void printGrade(double score) { if (score < 0 || score > 100) { System.out.println("Invalid score"); return; }

if (score >= 90.0) { System.out.println('A'); } else if (score >= 80.0) { System.out.println('B'); } else if (score >= 70.0) { System.out.println('C'); } else if (score >= 60.0) { System.out.println('D'); } else { System.out.println('F'); } }

return in void method

M06_LIAN1878_11_GE_C06.indd 234 1/2/18 10:42 PM

6.4.1 What are the benefits of using a method? 6.4.2 How do you define a method? How do you invoke a method? 6.4.3 How do you simplify the max method in Listing 6.1 using the conditional operator? 6.4.4 True or false? A call to a method with a void return type is always a statement

itself, but a call to a value-returning method cannot be a statement by itself.

6.4.5 What is the return type of a main method? 6.4.6 What would be wrong with not writing a return statement in a value-returning

method? Can you have a return statement in a void method? Does the return statement in the following method cause syntax errors?

public static void xMethod(double x, double y) { System.out.println(x + y); return x + y; }

6.4.7 Define the terms parameter, argument, and method signature. 6.4.8 Write method headers (not the bodies) for the following methods:

a. Return a sales commission, given the sales amount and the commission rate.

b. Display the calendar for a month, given the month and year.

c. Return a square root of a number.

d. Test whether a number is even, and returning true if it is.

e. Display a message a specified number of times.

f. Return the monthly payment, given the loan amount, number of years, and annual interest rate.

g. Return the corresponding uppercase letter, given a lowercase letter.

6.4.9 Identify and correct the errors in the following program: 1 public class Test { 2 public static method1(int n, m) { 3 n += m; 4 method2(3.4); 5 } 6 7 public static int method2(int n) { 8 if (n > 0) return 1; 9 else if (n == 0) return 0; 10 else if (n < 0) return −1; 11 } 12 }

6.4.10 Reformat the following program according to the programming style and docu- mentation guidelines proposed in Section 1.9, Programming Style and Documen- tation. Use the end-of-line brace style.

public class Test { public static double method(double i, double j) { while (i < j) { j––; }

return j; } }

Point Check

6.4 void vs. Value-Returning Methods 235

M06_LIAN1878_11_GE_C06.indd 235 1/2/18 10:42 PM

236 Chapter 6 Methods

6.5 Passing Parameters by Values The arguments are passed by value to parameters when invoking a method.

The power of a method is its ability to work with parameters. You can use println to print any string, and max to find the maximum of any two int values. When calling a method, you need to provide arguments, which must be given in the same order as their respective param- eters in the method signature. This is known as parameter order association. For example, the following method prints a message n times:

public static void nPrintln(String message, int n) { for (int i = 0; i < n; i++) System.out.println(message); }

You can use nPrintln("Hello", 3) to print Hello three times. The nPrintln("Hello", 3) statement passes the actual string parameter Hello to the parameter message, passes 3 to n, and prints Hello three times. However, the statement nPrintln(3, "Hello") would be wrong. The data type of 3 does not match the data type for the first parameter, message, nor does the second argument, Hello, match the second parameter, n.

Caution The arguments must match the parameters in order, number, and compatible type, as defined in the method signature. Compatible type means you can pass an argument to a parameter without explicit casting, such as passing an int value argument to a double value parameter.

When you invoke a method with an argument, the value of the argument is passed to the parameter. This is referred to as pass-by-value. If the argument is a variable rather than a literal value, the value of the variable is passed to the parameter. The variable is not affected, regardless of the changes made to the parameter inside the method. As given in Listing 6.4, the value of x (1) is passed to the parameter n to invoke the increment method (line 5). The parameter n is incremented by 1 in the method (line 10), but x is not changed no matter what the method does.

Listing 6.4 Increment.java 1 public class Increment { 2 public static void main(String[] args) { 3 int x = 1; 4 System.out.println("Before the call, x is " + x); 5 increment(x); 6 System.out.println("After the call, x is " + x); 7 } 8 9 public static void increment(int n) { 10 n++; 11 System.out.println("n inside the method is " + n); 12 } 13 }

Point Key

parameter order association

pass-by-value

invoke increment

increment n

Before the call, x is 1

n inside the method is 2

After the call, x is 1

M06_LIAN1878_11_GE_C06.indd 236 1/2/18 10:42 PM

6.5 Passing Parameters by Values 237

Listing 6.5 gives another program that demonstrates the effect of passing by value. The program creates a method for swapping two variables. The swap method is invoked by passing two arguments. Interestingly, the values of the arguments are not changed after the method is invoked.

Listing 6.5 TestPassByValue.java 1 public class TestPassByValue { 2 /** Main method */ 3 public static void main(String[] args) { 4 // Declare and initialize variables 5 int num1 = 1; 6 int num2 = 2; 7 8 System.out.println("Before invoking the swap method, num1 is " + 9 num1 + " and num2 is " + num2); 10 11 // Invoke the swap method to attempt to swap two variables 12 swap(num1, num2); 13 14 System.out.println("After invoking the swap method, num1 is " + 15 num1 + " and num2 is " + num2); 16 } 17 18 /** Swap two variables */ 19 public static void swap(int n1, int n2) { 20 System.out.println("\tInside the swap method"); 21 System.out.println("\t\tBefore swapping, n1 is " + n1 22 + " and n2 is " + n2); 23 24 // Swap n1 with n2 25 int temp = n1; 26 n1 = n2; 27 n2 = temp; 28 29 System.out.println("\t\tAfter swapping, n1 is " + n1 30 + " and n2 is " + n2); 31 } 32 }

false swap

Before invoking the swap method, num1 is 1 and num2 is 2 Inside the swap method Before swapping, n1 is 1 and n2 is 2 After swapping, n1 is 2 and n2 is 1 After invoking the swap method, num1 is 1 and num2 is 2

Before the swap method is invoked (line 12), num1 is 1 and num2 is 2. After the swap method is invoked, num1 is still 1 and num2 is still 2. Their values have not been swapped. As shown in Figure 6.4, the values of the arguments num1 and num2 are passed to n1 and n2, but n1 and n2 have their own memory locations independent of num1 and num2. Therefore, changes in n1 and n2 do not affect the contents of num1 and num2.

Another twist is to change the parameter name n1 in swap to num1. What effect does this have? No change occurs, because it makes no difference whether the parameter and the argu- ment have the same name. The parameter is a variable in the method with its own memory space. The variable is allocated when the method is invoked, and it disappears when the method is returned to its caller.

M06_LIAN1878_11_GE_C06.indd 237 1/2/18 10:42 PM

238 Chapter 6 Methods

Note For simplicity, Java programmers often say passing x to y, which actually means passing the value of argument x to parameter y.

6.5.1 How is an argument passed to a method? Can the argument have the same name as its parameter?

6.5.2 Identify and correct the errors in the following program:

1 public class Test { 2 public static void main(String[] args) { 3 nPrintln(5, "Welcome to Java!"); 4 } 5 6 public static void nPrintln(String message, int n) { 7 int n = 1; 8 for (int i = 0; i < n; i++) 9 System.out.println(message); 10 } 11 }

6.5.3 What is pass-by-value? Show the result of the following programs.

Point Check

Figure 6.4 The values of the variables are passed to the method’s parameters.

Space required for the main method

Space required for the swap method

The main method is invoked.

Space required for the main method

num2: 2 num1: 1

Space required for the swap method

num2: 2 num1: 1

The swap method is invoked.

The swap method is executed.

Space required for the main method

num2: 2 num1: 1

The swap method is �nished.

Stack is empty

The main method is �nished.

The values of num1 and num2 are passed to n1 and n2.

The values for n1 and n2 are swapped, but it does not affect num1 and num2.

n2: 2 n1: 1

temp: n2: 1 n1: 2

temp:

public class Test { public static void main(String[] args) { int max = 0; max(1, 2, max); System.out.println(max); }

public static void max( int value1, int value2, int max) { if (value1 > value2) max = value1; else max = value2; } }

(a)

public class Test { public static void main(String[] args) { int i = 1; while (i <= 6) { method1(i, 2); i++; } }

public static void method1( int i, int num) { for (int j = 1; j <= i; j++) { System.out.print(num + " "); num *= 2; }

System.out.println(); } }

(b)

M06_LIAN1878_11_GE_C06.indd 238 1/2/18 10:42 PM

6.6 Modularizing Code 239

6.5.4 For (a) in the preceding question, show the contents of the activation records in the call stack just before the method max is invoked, just as max is entered, just before max is returned, and right after max is returned.

6.6 Modularizing Code Modularizing makes the code easy to maintain and debug and enables the code to be reused.

Methods can be used to reduce redundant code and enable code reuse. Methods can also be used to modularize code and improve the quality of the program.

Listing 5.9 gives a program that prompts the user to enter two integers and displays their greatest common divisor. You can rewrite the program using a method, as given in Listing 6.6.

Listing 6.6 GreatestCommonDivisorMethod.java 1 import java.util.Scanner; 2 3 public class GreatestCommonDivisorMethod { 4 /** Main method */ 5 public static void main(String[] args) { 6 // Create a Scanner 7 Scanner input = new Scanner(System.in); 8 9 // Prompt the user to enter two integers 10 System.out.print("Enter first integer: "); 11 int n1 = input.nextInt(); 12 System.out.print("Enter second integer: "); 13 int n2 = input.nextInt(); 14

Point Key

VideoNote

Modularize code

public class Test { public static void main(String[] args) { // Initialize times int times = 3; System.out.println("Before the call," + " variable times is " + times);

// Invoke nPrintln and display times nPrintln("Welcome to Java!", times); System.out.println("After the call," + " variable times is " + times); }

// Print the message n times public static void nPrintln( String message, int n) { while (n > 0) { System.out.println("n = " + n); System.out.println(message); n––; } } }

(c)

public class Test { public static void main(String[] args) { int i = 0; while (i <= 4) { method1(i); i++; }

System.out.println("i is " + i); }

public static void method1(int i) { do { if (i % 3 != 0) System.out.print(i + " "); i––; } while (i >= 1);

System.out.println(); } }

(d)

M06_LIAN1878_11_GE_C06.indd 239 1/2/18 10:42 PM

240 Chapter 6 Methods

15 System.out.println("The greatest common divisor for " + n1 + 16 " and " + n2 + " is " + gcd(n1, n2)); 17 } 18 19 /** Return the gcd of two integers */ 20 public static int gcd(int n1,int n2) { 21 int gcd = 1; // Initial gcd is 1 22 int k = 2; // Possible gcd 23 24 while (k <= n1 && k <= n2) { 25 if (n1 % k == 0 && n2 % k == 0) 26 gcd = k; // Update gcd 27 k++; 28 } 29 30 return gcd; // Return gcd 31 } 32 }

invoke gcd

compute gcd

return gcd

Enter first integer: 45

Enter second integer: 75

The greatest common divisor for 45 and 75 is 15

By encapsulating the code for obtaining the gcd in a method, this program has several advantages:

1. It isolates the problem for computing the gcd from the rest of the code in the main method. Thus, the logic becomes clear, and the program is easier to read.

2. The errors on computing the gcd are confined in the gcd method, which narrows the scope of debugging.

3. The gcd method now can be reused by other programs.

Listing 6.7 applies the concept of code modularization to improve Listing 5.15, Prime- Number.java.

Listing 6.7 PrimeNumberMethod.java 1 public class PrimeNumberMethod { 2 public static void main(String[] args) { 3 System.out.println("The first 50 prime numbers are \n"); 4 printPrimeNumbers(50); 5 } 6 7 public static void printPrimeNumbers(int numberOfPrimes) { 8 final int NUMBER_OF_PRIMES_PER_LINE = 10; // Display 10 per line 9 int count = 0; // Count the number of prime numbers 10 int number = 2; // A number to be tested for primeness 11 12 // Repeatedly find prime numbers 13 while (count < numberOfPrimes) { 14 // Print the prime number and increase the count 15 if (isPrime(number)) { 16 count++; // Increase the count 17

invoke printPrimeNumbers

printPrimeNumbers method

invoke isPrime

M06_LIAN1878_11_GE_C06.indd 240 1/2/18 10:42 PM

6.7 Case Study: Converting Hexadecimals to Decimals 241

18 if (count % NUMBER_OF_PRIMES_PER_LINE == 0) { 19 // Print the number and advance to the new line 20 System.out.printf("%−5d\n", number); 21 } 22 else 23 System.out.printf("%−5d", number); 24 } 25 26 // Check whether the next number is prime 27 number++; 28 } 29 } 30 31 /** Check whether number is prime */ 32 public static boolean isPrime(int number) { 33 for (int divisor = 2; divisor <= number / 2; divisor++) { 34 if (number % divisor == 0) { // If true, number is not prime 35 return false; // Number is not a prime 36 } 37 } 38 39 return true; // Number is prime 40 } 41 }

isPrime method

The first 50 prime numbers are

2 3 5 7 11 13 17 19 23 29 31 37 41 43 47 53 59 61 67 71 73 79 83 89 97 101 103 107 109 113 127 131 137 139 149 151 157 163 167 173 179 181 191 193 197 199 211 223 227 229

We divided a large problem into two subproblems: determining whether a number is a prime, and printing the prime numbers. As a result, the new program is easier to read and easier to debug. Moreover, the methods printPrimeNumbers and isPrime can be reused by other programs.

6.6.1 Trace the gcd method to find the return value for gcd(4, 6). 6.6.2 Trace the isPrime method to find the return value for isPrime(25).

6.7 Case Study: Converting Hexadecimals to Decimals This section presents a program that converts a hexadecimal number into a decimal number.

Listing 5.11, Dec2Hex.java, gives a program that converts a decimal to a hexadecimal. How would you convert a hex number into a decimal?

Given a hexadecimal number hnhn - 1hn - 2 c h2h 1h 0 , the equivalent decimal value is

hn * 16n + hn - 1 * 16n - 1 + hn - 2 * 16n - 2 + g + h2 * 162 + h1 * 161 + h0 * 160

For example, the hex number AB8C is

10 * 163 + 11 * 162 + 8 * 161 + 12 * 160 = 43916

Our program will prompt the user to enter a hex number as a string and convert it into a decimal using the following method:

public static int hexToDecimal(String hex)

Point Check

Point Key

M06_LIAN1878_11_GE_C06.indd 241 1/2/18 10:42 PM

242 Chapter 6 Methods

A brute-force approach is to convert each hex character into a decimal number, multiply it by 16i for a hex digit at the i’s position, and then add all the items together to obtain the equivalent decimal value for the hex number.

Note that

hn * 16n + hn - 1 * 16n - 1 + hn - 2 * 16n - 2 + g + h1 * 161 + h0 * 160

= ( c ((hn * 16 + hn - 1) * 16 + hn - 2) * 16 + g + h1) * 16 + h0

This observation, known as the Horner’s algorithm, leads to the following efficient code for converting a hex string to a decimal number:

int decimalValue = 0; for (int i = 0; i < hex.length(); i++) { char hexChar = hex.charAt(i); decimalValue = decimalValue * 16 + hexCharToDecimal(hexChar); }

Here is a trace of the algorithm for hex number AB8C:

i hexChar hexCharToDecimal (hexChar) decimalValue

Before the loop 0

After the 1st iteration 0 A 10 10

After the 2nd iteration 1 B 11 10 * 16 + 11

After the 3rd iteration 2 8 8 (10 * 16 + 11) * 16 + 8

After the 4th iteration 3 C 12 ((10 * 16 + 11) * 16 + 8) * 16 + 12

Listing 6.8 gives the complete program.

Listing 6.8 Hex2Dec.java 1 import java.util.Scanner; 2 3 public class Hex2Dec { 4 /** Main method */ 5 public static void main(String[] args) { 6 // Create a Scanner 7 Scanner input = new Scanner(System.in); 8 9 // Prompt the user to enter a string 10 System.out.print("Enter a hex number: "); 11 String hex = input.nextLine(); 12 13 System.out.println("The decimal value for hex number " 14 + hex + " is " + hexToDecimal(hex.toUpperCase())); 15 } 16 17 public static int hexToDecimal(String hex) { 18 int decimalValue = 0; 19 for (int i = 0; i < hex.length(); i++) { 20 char hexChar = hex.charAt(i); 21 decimalValue = decimalValue * 16 + hexCharToDecimal(hexChar);

input string

hex to decimal

M06_LIAN1878_11_GE_C06.indd 242 1/2/18 10:42 PM

6.8 Overloading Methods 243

22 } 23 24 return decimalValue; 25 } 26 27 public static int hexCharToDecimal(char ch) { 28 if (ch >= 'A' && ch <= 'F') 29 return 10 + ch – 'A'; 30 else // ch is '0', '1', ..., or '9' 31 return ch − '0'; 32 } 33 }

hex char to decimal check uppercase

Enter a hex number: AB8C

The decimal value for hex number AB8C is 43916

Enter a hex number: af71

The decimal value for hex number af71 is 44913

The program reads a string from the console (line 11) and invokes the hexToDecimal method to convert a hex string to decimal number (line 14). The characters can be in either lowercase or uppercase. They are converted to uppercase before invoking the hexToDecimal method.

The hexToDecimal method is defined in lines 17–25 to return an integer. The length of the string is determined by invoking hex.length() in line 19.

The hexCharToDecimal method is defined in lines 27–32 to return a decimal value for a hex character. The character can be in either lowercase or uppercase. Recall that to subtract two characters is to subtract their Unicodes. For example, '5' – '0' is 5.

6.7.1 What is hexCharToDecimal('B'))? What is hexCharToDecimal('7'))? What is hexToDecimal("A9"))?

6.8 Overloading Methods Overloading methods enable you to define the methods with the same name as long as their parameter lists are different.

The max method used earlier works only with the int data type. But what if you need to determine which of the two floating-point numbers has the maximum value? The solution is to create another method with the same name but different parameters, as shown in the following code:

public static double max(double num1, double num2) { if (num1 > num2) return num1; else return num2; }

If you call max with int parameters, the max method that expects int parameters will be invoked; and if you call max with double parameters, the max method that expects double parameters will be invoked. This is referred to as method overloading; that is, two methods have the same name but different parameter lists within one class. The Java compiler determines which method to use based on the method signature.

Point Check

Point Key

method overloading

M06_LIAN1878_11_GE_C06.indd 243 1/2/18 10:42 PM

244 Chapter 6 Methods

Listing 6.9 is a program that creates three methods. The first finds the maximum integer, the second finds the maximum double, and the third finds the maximum among three double values. All three methods are named max.

Listing 6.9 TestMethodOverloading.java 1 public class TestMethodOverloading { 2 /** Main method */ 3 public static void main(String[] args) { 4 // Invoke the max method with int parameters 5 System.out.println("The maximum of 3 and 4 is " 6 + max(3, 4)); 7 8 // Invoke the max method with the double parameters 9 System.out.println("The maximum of 3.0 and 5.4 is " 10 + max(3.0, 5.4)); 11 12 // Invoke the max method with three double parameters 13 System.out.println("The maximum of 3.0, 5.4, and 10.14 is " 14 + max(3.0, 5.4, 10.14)); 15 } 16 17 /** Return the max of two int values */ 18 public static int max(int num1, int num2) { 19 if (num1 > num2) 20 return num1; 21 else 22 return num2; 23 } 24 25 /** Find the max of two double values */ 26 public static double max(double num1, double num2) { 27 if (num1 > num2) 28 return num1; 29 else 30 return num2; 31 } 32 33 /** Return the max of three double values */ 34 public static double max(double num1, double num2, double num3) { 35 return max(max(num1, num2), num3); 36 } 37 }

overloaded max

The maximum of 3 and 4 is 4 The maximum of 3.0 and 5.4 is 5.4 The maximum of 3.0, 5.4, and 10.14 is 10.14

When calling max(3, 4) (line 6), the max method for finding the maximum of two integers is invoked. When calling max(3.0, 5.4) (line 10), the max method for finding the maximum of two doubles is invoked. When calling max(3.0, 5.4, 10.14) (line 14), the max method for finding the maximum of three double values is invoked.

Can you invoke the max method with an int value and a double value, such as max(2, 2.5)? If so, which of the max methods is invoked? The answer to the first question is yes. The answer to the second question is that the max method for finding the maximum of two double values is invoked. The argument value 2 is automatically converted into a double value and passed to this method.

M06_LIAN1878_11_GE_C06.indd 244 1/2/18 10:42 PM

6.8 Overloading Methods 245

You may be wondering why the method max(double, double) is not invoked for the call max(3, 4). Both max(double, double) and max(int, int) are possible matches for max(3, 4). The Java compiler finds the method that best matches a method invocation. Since the method max(int, int) is a better match for max(3, 4) than max(double, double), max(int, int) is used to invoke max(3, 4).

Tip Overloading methods can make programs clearer and more readable. Methods that perform the same function with different types of parameters should be given the same name.

Note Overloaded methods must have different parameter lists. You cannot overload methods based on different modifiers or return types.

Note Sometimes there are two or more possible matches for the invocation of a method, but the compiler cannot determine the most specific match. This is referred to as ambiguous invocation. Ambiguous invocation causes a compile error. Consider the following code:

public class AmbiguousOverloading { public static void main(String[] args) { System.out.println(max(1, 2)); }

public static double max(int num1, double num2) { if (num1 > num2) return num1; else return num2; }

public static double max(double num1, int num2) { if (num1 > num2) return num1; else return num2; } }

Both max(int, double) and max(double, int) are possible candidates to match max(1, 2). Because neither is more specific than the other, the invocation is ambiguous, resulting in a compile error.

6.8.1 What is method overloading? Is it permissible to define two methods that have the same name but different parameter types? Is it permissible to define two methods in a class that have identical method names and parameter lists, but different return value types or different modifiers?

6.8.2 What is wrong in the following program?

public class Test { public static void method(int x) { }

public static int method(int y) {

ambiguous invocation

Point Check

M06_LIAN1878_11_GE_C06.indd 245 1/2/18 10:42 PM

246 Chapter 6 Methods

return y; } }

6.8.3 Given two method definitions,

public static double m(double x, double y)

public static double m(int x, double y)

tell which of the two methods is invoked for:

a. double z = m(4, 5);

b. double z = m(4, 5.4);

c. double z = m(4.5, 5.4);

6.9 The Scope of Variables The scope of a variable is the part of the program where the variable can be referenced.

Section 2.5 introduced the scope of a variable. This section discusses the scope of variables in detail. A variable defined inside a method is referred to as a local variable. The scope of a local variable starts from its declaration and continues to the end of the block that contains the variable. A local variable must be declared and assigned a value before it can be used.

A parameter is actually a local variable. The scope of a method parameter covers the entire method. A variable declared in the initial-action part of a for-loop header has its scope in the entire loop. However, a variable declared inside a for-loop body has its scope limited in the loop body from its declaration to the end of the block that contains the variable, as shown in Figure 6.5.

Point Key

scope of variables local variable

Figure 6.5 A variable declared in the initial-action part of a for-loop header has its scope in the entire loop.

The scope of j

The scope of i

public static void method1() {

for (int i = 1; i < 10; i++) {

int j;

} }

You can declare a local variable with the same name in different blocks in a method, but you cannot declare a local variable twice in the same block or in nested blocks, as shown in Figure 6.6.

M06_LIAN1878_11_GE_C06.indd 246 1/2/18 10:42 PM

6.10 Case Study: Generating Random Characters 247

Caution A common mistake is to declare a variable in a for loop and then attempt to use it outside the loop. As shown in the following code, i is declared in the for loop, but it is accessed from the outside of the for loop, which causes a syntax error.

for (int i = 0; i < 10; i++) { }

System.out.println(i); // Causes a syntax error on i

The last statement would cause a syntax error, because variable i is not defined outside of the for loop.

6.9.1 What is a local variable? 6.9.2 What is the scope of a local variable?

6.10 Case Study: Generating Random Characters A character is coded using an integer. Generating a random character is to generate an integer.

Computer programs process numerical data and characters. You have seen many examples that involve numerical data. It is also important to understand characters and how to process them. This section presents an example of generating random characters.

As introduced in Section 4.3, every character has a unique Unicode between 0 and FFFF in hexadecimal (65535 in decimal). To generate a random character is to generate a ran- dom integer between 0 and 65535 using the following expression (note since 0 <= Math .random() < 1.0, you have to add 1 to 65535):

(int)(Math.random() * (65535 + 1))

Now let’s consider how to generate a random lowercase letter. The Unicodes for lowercase letters are consecutive integers starting from the Unicode for a, then for b, c, . . . , and z. The Unicode for a is

(int)'a'

Thus, a random integer between (int)'a' and (int)'z' is

(int)((int)'a' + Math.random() * ((int)'z' – (int)'a' + 1))

Point Check

Point Key

Figure 6.6 A variable can be declared multiple times in nonnested blocks, but only once in nested blocks.

It is �ne to declare i in two nonnested blocks.

It is wrong to declare i in two nested blocks.

public static void method1() { int x = 1; int y = 1;

for (int i = 1; i < 10; i++) { x += i;

}

for (int i = 1; i < 10; i++) { y += i;

} }

public static void method2() {

int i = 1; int sum = 0;

for (int i = 1; i < 10; i++) { sum += i;

}

M06_LIAN1878_11_GE_C06.indd 247 1/2/18 10:42 PM

248 Chapter 6 Methods

As discussed in Section 4.3.3, all numeric operators can be applied to the char operands. The char operand is cast into a number if the other operand is a number or a character. Therefore, the preceding expression can be simplified as follows:

'a' + Math.random() * ('z' – 'a' + 1)

and a random lowercase letter is

(char)('a' + Math.random() * ('z' – 'a' + 1))

Hence, a random character between any two characters ch1 and ch2 with ch1 < ch2 can be generated as follows:

(char)(ch1 + Math.random() * (ch2 – ch1 + 1))

This is a simple but useful discovery. Listing 6.10 defines a class named RandomCharacter with overloaded methods to get a certain type of random character. You can use these methods in your future projects.

Listing 6.10 RandomCharacter.java 1 public class RandomCharacter { 2 /** Generate a random character between ch1 and ch2 */ 3 public static char getRandomCharacter(char ch1, char ch2) { 4 return (char)(ch1 + Math.random() * (ch2 – ch1 + 1)); 5 } 6 7 /** Generate a random lowercase letter */ 8 public static char getRandomLowerCaseLetter() { 9 return getRandomCharacter('a', 'z'); 10 } 11 12 /** Generate a random uppercase letter */ 13 public static char getRandomUpperCaseLetter() { 14 return getRandomCharacter('A', 'Z'); 15 } 16 17 /** Generate a random digit character */ 18 public static char getRandomDigitCharacter() { 19 return getRandomCharacter('0', '9'); 20 } 21 22 /** Generate a random character */ 23 public static char getRandomCharacter() { 24 return getRandomCharacter('\u0000', '\uFFFF'); 25 } 26 }

Listing 6.11 gives a test program that displays 175 random lowercase letters.

Listing 6.11 TestRandomCharacter.java 1 public class TestRandomCharacter { 2 /** Main method */ 3 public static void main(String[] args) { 4 final int NUMBER_OF_CHARS = 175; 5 final int CHARS_PER_LINE = 25; 6 7 // Print random characters between 'a' and 'z', 25 chars per line 8 for (int i = 0; i < NUMBER_OF_CHARS; i++) { 9 char ch = RandomCharacter.getRandomLowerCaseLetter(); 10 if ((i + 1) % CHARS_PER_LINE == 0) 11 System.out.println(ch);

getRandomCharacter()

getRandomLower CaseLetter()

getRandomUpper CaseLetter()

getRandomDigit Character()

getRandomCharacter

constants

lowercase letter

M06_LIAN1878_11_GE_C06.indd 248 1/2/18 10:42 PM

6.11 Method Abstraction and Stepwise Refinement 249

12 else 13 System.out.print(ch); 14 } 15 } 16 }

gmjsohezfkgtazqgmswfclrao pnrunulnwmaztlfjedmpchcif lalqdgivxkxpbzulrmqmbhikr lbnrjlsopfxahssqhwuuljvbe xbhdotzhpehbqmuwsfktwsoli cbuwkzgxpmtzihgatdslvbwbz bfesoklwbhnooygiigzdxuqni

Line 9 invokes getRandomLowerCaseLetter() defined in the RandomCharacter class. Note getRandomLowerCaseLetter() does not have any parameters, but you still have to use the parentheses when defining and invoking the method.

6.11 Method Abstraction and Stepwise Refinement The key to developing software is to apply the concept of abstraction.

You will learn many levels of abstraction from this book. Method abstraction is achieved by separating the use of a method from its implementation. The client can use a method without knowing how it is implemented. The details of the implementation are encapsulated in the method and hidden from the client who invokes the method. This is also known as information hiding or encapsulation. If you decide to change the implementation, the client program will not be affected, provided that you do not change the method signature. The implementation of the method is hidden from the client in a “black box,” as shown in Figure 6.7.

parentheses required

Point Key

method abstraction

information hiding

VideoNote

Stepwise refinement

Figure 6.7 The method body can be thought of as a black box that contains the detailed implementation for the method.

Method Header

Black box

Optional arguments for input

Optional return value

Method Body

You have already used the System.out.print method to display a string and the max method to find the maximum number. You know how to write the code to invoke these methods in your program, but as a user of these methods, you are not required to know how they are implemented.

The concept of method abstraction can be applied to the process of developing programs. When writing a large program, you can use the divide-and-conquer strategy, also known as stepwise refinement, to decompose it into subproblems. The subproblems can be further decomposed into smaller, more manageable problems.

Suppose that you write a program that displays the calendar for a given month of the year. The program prompts the user to enter the year and the month, and then displays the entire calendar for the month, as presented in the following sample run:

divide and conquer stepwise refinement

M06_LIAN1878_11_GE_C06.indd 249 1/2/18 10:42 PM

250 Chapter 6 Methods

Let us use this example to demonstrate the divide-and-conquer approach.

6.11.1 Top-Down Design How would you get started on such a program? Would you immediately start coding? Begin- ning programmers often start by trying to work out the solution to every detail. Although details are important in the final program, concern for detail in the early stages may block the problem-solving process. To make problem solving flow as smoothly as possible, this example begins by using method abstraction to isolate details from design and only later implements the details.

For this example, the problem is first broken into two subproblems: get input from the user, and print the calendar for the month. At this stage, you should be concerned with what the subproblems will achieve, not with how to get input and print the calendar for the month. You can draw a structure chart to help visualize the decomposition of the problem (see Figure 6.8a).

Enter full year (e.g., 2012): 2012

Enter month as number between 1 and 12: 3

March 2012

-----------------------------

Sun Mon Tue Wed Thu Fri Sat

1 2 3

4 5 6 7 8 9 10

11 12 13 14 15 16 17

18 19 20 21 22 23 24

25 26 27 28 29 30

Figure 6.8 The structure chart shows the printCalendar problem is divided into two subproblems, readInput and printMonth in (a), and printMonth is divided into two smaller subproblems, printMonthTitle and printMonth Body in (b).

printCalendar (main)

(a) (b)

printMonthBodyprintMonthTitlereadInput printMonth

printMonth

You can use Scanner to read input for the year and the month. The problem of printing the calendar for a given month can be broken into two subproblems: print the month title, and print the month body, as shown in Figure 6.8b. The month title consists of three lines: month and year, a dashed line, and the names of the seven days of the week. You need to get the month name (e.g., January) from the numeric month (e.g., 1). This is accomplished in getMonthName (see Figure 6.9a).

In order to print the month body, you need to know which day of the week is the first day of the month (getStartDay) and how many days the month has (getNumberOfDaysInMonth),

M06_LIAN1878_11_GE_C06.indd 250 1/2/18 10:42 PM

6.11 Method Abstraction and Stepwise Refinement 251

as shown in Figure 6.9b. For example, December 2013 has 31 days, and December 1, 2013 is a Sunday.

How would you get the start day for the first date in a month? There are several ways to do so. For now, we’ll use an alternative approach. Assume you know that the start day for January 1, 1800 was a Wednesday (START_DAY_FOR_JAN_1_1800 = 3). You could compute the total number of days (totalNumberOfDays) between January 1, 1800 and the first date of the calendar month. The start day for the calendar month is (totalNumberOfDays + START_ DAY_FOR_JAN_1_1800) % 7, since every week has seven days. Thus, the getStartDay problem can be further refined as getTotalNumberOfDays, as shown in Figure 6.10a.

Figure 6.9 (a) To printMonthTitle, you need getMonthName. (b) The printMonthBody problem is refined into several smaller problems.

(b)(a)

getNumberOfDaysInMonthgetStartDaygetMonthName

printMonthTitle

printMonthBody

Figure 6.10 (a) To getStartDay, you need getTotalNumberOfDays. (b) The getTotalNumberOfDays problem is refined into two smaller problems.

getTotalNumberOfDays

getStartDay

isLeapYear

(a) (b)

getNumberOfDaysInMonth

getTotalNumberOfDays

To get the total number of days, you need to know whether the year is a leap year and the number of days in each month. Thus, getTotalNumberOfDays can be further refined into two subproblems: isLeapYear and getNumberOfDaysInMonth, as shown in Figure 6.10b. The complete structure chart is shown in Figure 6.11.

6.11.2 Top-Down and/or Bottom-Up Implementation Now we turn our attention to implementation. In general, a subproblem corresponds to a method in the implementation, although some are so simple that this is unnecessary. You would need to decide which modules to implement as methods and which to combine with other methods. Decisions of this kind should be based on whether the overall program will be easier to read as a result of your choice. In this example, the subproblem readInput can be simply implemented in the main method.

You can use either a “top-down” or a “bottom-up” approach. The top-down approach imple- ments one method in the structure chart at a time from the top to the bottom. Stubs—a simple but incomplete version of a method—can be used for the methods waiting to be implemented. The use of stubs enables you to quickly build the framework of the program. Implement the main

stub

top-down approach

M06_LIAN1878_11_GE_C06.indd 251 1/2/18 10:42 PM

252 Chapter 6 Methods

method first then use a stub for the printMonth method. For example, let printMonth display the year and the month in the stub. Thus, your program may begin as follows:

public class PrintCalendar { /** Main method */ public static void main(String[] args) { Scanner input = new Scanner(System.in);

// Prompt the user to enter year System.out.print("Enter full year (e.g., 2012): "); int year = input.nextInt();

// Prompt the user to enter month System.out.print("Enter month as a number between 1 and 12: "); int month = input.nextInt();

// Print calendar for the month of the year printMonth(year, month); }

/** A stub for printMonth may look like this */ public static void printMonth(int year, int month) { System.out.print(month + " " + year); }

/** A stub for printMonthTitle may look like this */ public static void printMonthTitle(int year, int month) { }

/** A stub for printMonthBody may look like this */ public static void printMonthBody(int year, int month) { }

Figure 6.11 The structure chart shows the hierarchical relationship of the subproblems in the program.

printCalendar (main)

getTotalNumberOfDays

getNumberOfDaysInMonth

isLeapYear

printMonthreadInput

printMonthTitle

getMonthName

printMonthBody

getStartDay

M06_LIAN1878_11_GE_C06.indd 252 1/2/18 10:42 PM

6.11 Method Abstraction and Stepwise Refinement 253

/** A stub for getMonthName may look like this */ public static String getMonthName(int month) { return "January"; // A dummy value }

/** A stub for getStartDay may look like this */ public static int getStartDay(int year, int month) { return 1; // A dummy value }

/** A stub for getTotalNumberOfDays may look like this */ public static int getTotalNumberOfDays(int year, int month) { return 10000; // A dummy value }

/** A stub for getNumberOfDaysInMonth may look like this */ public static int getNumberOfDaysInMonth(int year, int month) { return 31; // A dummy value }

/** A stub for isLeapYear may look like this */ public static boolean isLeapYear(int year) { return true; // A dummy value } }

Compile and test the program, and fix any errors. You can now implement the printMonth method. For methods invoked from the printMonth method, you can again use stubs.

The bottom-up approach implements one method in the structure chart at a time from the bottom to the top. For each method implemented, write a test program, known as the driver, to test it. The top-down and bottom-up approaches are equally good: Both approaches imple- ment methods incrementally, help to isolate programming errors, and make debugging easy. They can be used together.

6.11.3 Implementation Details The isLeapYear(int year) method can be implemented using the following code from Section 3.11:

return year % 400 == 0 || (year % 4 == 0 && year % 100 != 0);

Use the following facts to implement getTotalNumberOfDaysInMonth(int year, int month):

■■ January, March, May, July, August, October, and December have 31 days.

■■ April, June, September, and November have 30 days.

■■ February has 28 days during a regular year, and 29 days during a leap year. A regular year, therefore, has 365 days, and a leap year has 366 days.

To implement getTotalNumberOfDays(int year, int month), you need to compute the total number of days (totalNumberOfDays) between January 1, 1800 and the first day of the calendar month. You could find the total number of days between the year 1800 and the calendar year then figure out the total number of days prior to the calendar month in the calendar year. The sum of these two totals is totalNumberOfDays.

To print a body, first pad some space before the start day then print the lines for every week. The complete program is given in Listing 6.12.

bottom-up approach driver

M06_LIAN1878_11_GE_C06.indd 253 1/2/18 10:42 PM

254 Chapter 6 Methods

Listing 6.12 PrintCalendar.java 1 import java.util.Scanner; 2 3 public class PrintCalendar { 4 /** Main method */ 5 public static void main(String[] args) { 6 Scanner input = new Scanner(System.in); 7 8 // Prompt the user to enter year 9 System.out.print("Enter full year (e.g., 2012): "); 10 int year = input.nextInt(); 11 12 // Prompt the user to enter month 13 System.out.print("Enter month as a number between 1 and 12: "); 14 int month = input.nextInt(); 15 16 // Print calendar for the month of the year 17 printMonth(year, month); 18 } 19 20 /** Print the calendar for a month in a year */ 21 public static void printMonth(int year, int month) { 22 // Print the headings of the calendar 23 printMonthTitle(year, month); 24 25 // Print the body of the calendar 26 printMonthBody(year, month); 27 } 28 29 /** Print the month title, e.g., March 2012 */ 30 public static void printMonthTitle(int year, int month) { 31 System.out.println(" " + getMonthName(month) 32 + " " + year); 33 System.out.println("−−−−−−−−−−−−−−−−−−−−−−−−−−−−−"); 34 System.out.println(" Sun Mon Tue Wed Thu Fri Sat"); 35 } 36 37 /** Get the English name for the month */ 38 public static String getMonthName(int month) { 39 String monthName = ""; 40 switch (month) { 41 case 1: monthName = "January"; break; 42 case 2: monthName = "February"; break; 43 case 3: monthName = "March"; break; 44 case 4: monthName = "April"; break; 45 case 5: monthName = "May"; break; 46 case 6: monthName = "June"; break; 47 case 7: monthName = "July"; break; 48 case 8: monthName = "August"; break; 49 case 9: monthName = "September"; break; 50 case 10: monthName = "October"; break; 51 case 11: monthName = "November"; break; 52 case 12: monthName = "December"; 53 } 54 55 return monthName; 56 } 57 58 /** Print month body */

printMonth

printMonthTitle

getMonthName

M06_LIAN1878_11_GE_C06.indd 254 1/2/18 10:42 PM

6.11 Method Abstraction and Stepwise Refinement 255

59 public static void printMonthBody(int year, int month) { 60 // Get start day of the week for the first date in the month 61 int startDay = getStartDay(year, month); 62 63 // Get number of days in the month 64 int numberOfDaysInMonth = getNumberOfDaysInMonth(year, month); 65 66 // Pad space before the first day of the month 67 int i = 0; 68 for (i = 0; i < startDay; i++) 69 System.out.print(" "); 70 71 for (i = 1; i <= numberOfDaysInMonth; i++) { 72 System.out.printf("%4d", i); 73 74 if ((i + startDay) % 7 == 0) 75 System.out.println(); 76 } 77 78 System.out.println(); 79 } 80 81 /** Get the start day of month/1/year */ 82 public static int getStartDay(int year, int month) { 83 final int START_DAY_FOR_JAN_1_1800 = 3; 84 // Get total number of days from 1/1/1800 to month/1/year 85 int totalNumberOfDays = getTotalNumberOfDays(year, month); 86 87 // Return the start day for month/1/year 88 return (totalNumberOfDays + START_DAY_FOR_JAN_1_1800) % 7; 89 } 90 91 /** Get the total number of days since January 1, 1800 */ 92 public static int getTotalNumberOfDays(int year, int month) { 93 int total = 0; 94 95 // Get the total days from 1800 to 1/1/year 96 for (int i = 1800; i < year; i++) 97 if (isLeapYear(i)) 98 total = total + 366; 99 else 100 total = total + 365; 101 102 // Add days from Jan to the month prior to the calendar month 103 for (int i = 1; i < month; i++) 104 total = total + getNumberOfDaysInMonth(year, i); 105 106 return total; 107 } 108 109 /** Get the number of days in a month */ 110 public static int getNumberOfDaysInMonth(int year, int month) { 111 if (month == 1 || month == 3 || month == 5 || month == 7 || 112 month == 8 || month == 10 || month == 12) 113 return 31; 114 115 if (month == 4 || month == 6 || month == 9 || month == 11) 116 return 30; 117 118 if (month == 2) return isLeapYear(year) ? 29 : 28;

getStartDay

printMonthBody

getTotalNumberOfDays

getNumberOfDaysInMonth

M06_LIAN1878_11_GE_C06.indd 255 1/2/18 10:42 PM

256 Chapter 6 Methods

119 120 return 0; // If month is incorrect 121 } 122 123 /** Determine if it is a leap year */ 124 public static boolean isLeapYear(int year) { 125 return year % 400 == 0 || (year % 4 == 0 && year % 100 != 0); 126 } 127 }

The program does not validate user input. For instance, if the user enters either a month not in the range between 1 and 12 or a year before 1800, the program displays an erroneous calen- dar. To avoid this error, add an if statement to check the input before printing the calendar.

This program prints calendars for a month, but could easily be modified to print calendars for a whole year. Although it can print months only after January 1800, it could be modified to print months before 1800.

6.11.4 Benefits of Stepwise Refinement Stepwise refinement breaks a large problem into smaller manageable subproblems. Each sub- problem can be implemented using a method. This approach makes the program easier to write, reuse, debug, test, modify, and maintain.

Simpler Program The print calendar program is long. Rather than writing a long sequence of statements in one method, stepwise refinement breaks it into smaller methods. This simplifies the program and makes the whole program easier to read and understand.

Reusing Methods Stepwise refinement promotes code reuse within a program. The isLeapYear method is defined once and invoked from the getTotalNumberOfDays and getNumberOfDaysInMonth methods. This reduces redundant code.

Easier Developing, Debugging, and Testing Since each subproblem is solved in a method, a method can be developed, debugged, and tested individually. This isolates the errors and makes developing, debugging, and testing easier.

When implementing a large program, use the top-down and/or bottom-up approach. Do not write the entire program at once. Using these approaches seems to take more development time (because you repeatedly compile and run the program), but it actually saves time and makes debugging easier.

Better Facilitating Teamwork When a large problem is divided into subprograms, subproblems can be assigned to different programmers. This makes it easier for programmers to work in teams.

isLeapYear

incremental development and testing

Key terms actual parameter 229 ambiguous invocation 245 argument 229 divide and conquer 249 formal parameter (i.e., parameter) 229 information hiding 249 method 228 method abstraction 249

method overloading 243 method signature 229 modifier 229 parameter 229 pass-by-value 236 scope of a variable 246 stepwise refinement 249 stub 251

M06_LIAN1878_11_GE_C06.indd 256 1/2/18 10:42 PM

ChapTer summary 1. Making programs modular and reusable is one of the central goals in software engineer-

ing. Java provides many powerful constructs that help to achieve this goal. Methods are one such construct.

2. The method header specifies the modifiers, return value type, method name, and param- eters of the method. The static modifier is used for all the methods in this chapter.

3. A method may return a value. The returnValueType is the data type of the value the method returns. If the method does not return a value, the returnValueType is the keyword void.

4. The parameter list refers to the type, order, and number of a method’s parameters. The method name and the parameter list together constitute the method signature. Parameters are optional; that is, a method doesn’t need to contain any parameters.

5. A return statement can also be used in a void method for terminating the method and returning to the method’s caller. This is useful occasionally for circumventing the normal flow of control in a method.

6. The arguments that are passed to a method should have the same number, type, and order as the parameters in the method signature.

7. When a program calls a method, program control is transferred to the called method. A called method returns control to the caller when its return statement is executed, or when its method-ending closing brace is reached.

8. A value-returning method can also be invoked as a statement in Java. In this case, the caller simply ignores the return value.

9. A method can be overloaded. This means that two methods can have the same name, as long as their method parameter lists differ.

10. A variable declared in a method is called a local variable. The scope of a local variable starts from its declaration and continues to the end of the block that contains the variable. A local variable must be declared and initialized before it is used.

11. Method abstraction is achieved by separating the use of a method from its implementa- tion. The client can use a method without knowing how it is implemented. The details of the implementation are encapsulated in the method and hidden from the client who invokes the method. This is known as information hiding or encapsulation.

12. Method abstraction modularizes programs in a neat, hierarchical manner. Programs written as collections of concise methods are easier to write, debug, maintain, and modify than would otherwise be the case. This writing style also promotes method reusability.

13. When implementing a large program, use the top-down and/or bottom-up coding approach. Do not write the entire program at once. This approach may seem to take more time for coding (because you are repeatedly compiling and running the program), but it actually saves time and makes debugging easier.

Chapter Summary 257

M06_LIAN1878_11_GE_C06.indd 257 1/2/18 10:42 PM

258 Chapter 6 Methods

Quiz Answer the quiz for this chapter online at the Companion Website.

programming exerCiSeS

Note A common error for the exercises in this chapter is that students don’t implement the methods to meet the requirements even though the output from the main program is correct. For an example of this type of error, see liveexample.pearsoncmg.com/etc/ CommonMethodErrorJava.pdf.

Sections 6.2–6.9 6.1 (Math: pentagonal numbers) A pentagonal number is defined as n(3n-1)/2 for

n = 1 , 2 , …, and so on. Therefore, the first few numbers are 1, 5, 12, 22, . . . . Write a method with the following header that returns a pentagonal number:

public static int getPentagonalNumber(int n)

For example, getPentagonalNumber(1) returns 1 and getPentagonal Number(2) returns 5. Write a test program that uses this method to display the first 100 pentagonal numbers with 10 numbers on each line. Use the %7d format to display each number.

*6.2 (Sum the digits in an integer) Write a method that computes the sum of the digits in an integer. Use the following method header:

public static int sumDigits(long n)

For example, sumDigits(234) returns 9 (= 2 + 3 + 4). (Hint: Use the % opera- tor to extract digits and the / operator to remove the extracted digit. For instance, to extract 4 from 234, use 234 % 10 (= 4 ). To remove 4 from 234, use 234 / 10 (= 2 3 ). Use a loop to repeatedly extract and remove the digit until all the digits are extracted. Write a test program that prompts the user to enter an integer then displays the sum of all its digits.

**6.3 (Palindrome integer) Write the methods with the following headers: // Return the reversal of an integer, e.g., reverse(456) returns 654 public static int reverse(int number)

// Return true if number is a palindrome public static boolean isPalindrome(int number)

Use the reverse method to implement isPalindrome. A number is a palin- drome if its reversal is the same as itself. Write a test program that prompts the user to enter an integer and reports whether the integer is a palindrome.

*6.4 (Display an integer reversed) Write a method with the following header to display an integer in reverse order:

public static void reverse(int number)

For example, reverse(3456) displays 6543. Write a test program that prompts the user to enter an integer then displays its reversal.

VideoNote

Reverse an integer

M06_LIAN1878_11_GE_C06.indd 258 1/2/18 10:42 PM

*6.5 (Sort three numbers) Write a method with the following header to display three numbers in decreasing order:

public static void displaySortedNumbers( double num1, double num2, double num3)

Write a test program that prompts the user to enter three numbers and invokes the method to display them in decreasing order.

*6.6 (Display patterns) Write a method to display a pattern as follows:

1 1 2 1 2 3 ... n n−1 ... 3 2 1

The method header is

public static void displayPattern(int n)

*6.7 (Financial application: compute the future investment value) Write a method that computes future investment value at a given interest rate for a specified number of years. The future investment is determined using the formula in Programming Exercise 2.21.

Use the following method header:

public static double futureInvestmentValue( double investmentAmount, double monthlyInterestRate,int years)

For example, futureInvestmentValue(10000, 0.05/12, 5) returns 12833.59.

Write a test program that prompts the user to enter the investment amount (e.g., 1,000) and the interest rate (e.g., 9%) and prints a table that displays future value for the years from 1 to 30, as shown below:

The amount invested: 1000

Annual interest rate: 9 Years Future Value 1 1093.80 2 1196.41 ... 29 13467.25 30 14730.57

6.8 (Conversions between mile and kilometer) Write a class that contains the follow- ing two methods:

/** Convert from Mile to Kilometer */ public static double mileToKilometer(double mile)

/** Convert from Kilometer to Mile */ public static double kilometerToMile(double kilometer)

Programming Exercises 259

M06_LIAN1878_11_GE_C06.indd 259 1/2/18 10:42 PM

260 Chapter 6 Methods

The formula for the conversion is:

1 mile = 1.6 kilometers

Write a test program that invokes these methods to display the following tables:

Miles Kilometers | Kilometers Miles

1 1.609 | 20 12.430

2 3.218 | 25 15.538

...

9 14.481 | 60 37.290

10 16.090 | 65 40.398

6.9 (Conversions between pounds and kilograms) Write a class that contains the following two methods:

/** Convert from pounds to kilograms */ public static double poundToKilogram(double pound)

/** Convert from kilograms to pounds */ public static double kilogramToPound(double kilogram)

The formula for the conversion is:

pound = 0.453 * kilogram kilogram = 2.204 * pound

Write a test program that invokes these methods to display the following tables:

Kilograms Pounds | Pounds Kilograms

1 2.2 | 20 9.09 3 6.6 | 25 11.36 ... 197 433.4 | 510 231.82 199 437.8 | 515 234.09

6.10 (Use the isPrime Method) Listing 6.7, PrimeNumberMethod.java, provides the isPrime(int number) method for testing whether a number is prime. Use this method to find the number of prime numbers less than 10000.

6.11 (Financial application: compute commissions) Write a method that computes the commission, using the scheme in Programming Exercise 5.39. The header of the method is as follows:

public static double computeCommission(double salesAmount)

Write a test program that displays the following table:

Sales Amount Commission

10000 900.0

15000 1500.0

...

95000 11100.0

100000 11700.0

M06_LIAN1878_11_GE_C06.indd 260 1/2/18 10:42 PM

6.12 (Display numbers) Write a method that prints numbers using the following header:

public static void printNumbers(int num1, int num2, int numberPerLine)

This method prints the characters between num1 and num2 with the specified numbers per line. Write a test program that prints ten characters per line from 1 to 100. Numbers are separated by exactly one space.

*6.13 (Sum series) Write a method to compute the following series:

m(i) = 1 3

+ 2 4

+ g + i

i + 2

Write a test program that displays the following table:

I m(i)

1 0.3333 2 0.8333 ... 19 14.7093 20 15.6184

*6.14 (Estimate p) p can be computed using the following summation:

m (i) = 4a1 - 1 3

+ 1 5

- 1 7

+ 1 9

- 1

1 1 + g +

(-1 )i + 1

2 i - 1 b

Write a method that returns m(i) for a given i and write a test program that displays the following table:

i m(i)

1 4.0000

101 3.1515

201 3.1466

301 3.1449

401 3.1441

501 3.1436

601 3.1433

701 3.1430

801 3.1428

901 3.1427

*6.15 (Financial application: print a tax table) Listing 3.5 gives a program to compute tax. Write a method for computing tax using the following header:

public static double computeTax(int status, double taxableIncome)

VideoNote

Estimate p

Programming Exercises 261

M06_LIAN1878_11_GE_C06.indd 261 1/2/18 10:42 PM

262 Chapter 6 Methods

Use this method to write a program that prints a tax table for taxable income from $50,000 to $70,000 with intervals of $100 for all the following statuses:

Taxable Income

Single Married Joint or Qualifying Widow(er)

Married Separate

Head of a House

50000 8688 6665 8688 7353 50100 8713 6680 8713 7378 ... 69900 13663 9850 12328 9840 70000 13688 9875 12353 9853

Hint: round the tax into integers using Math.round (i.e., Math .round(computeTax(status, taxableIncome))).

*6.16 (Number of days in a year) Write a method that returns the number of days in a year using the following header:

public static int numberOfDaysInAYear(int year)

Write a test program that displays the number of days in the years between 2014 and 2034.

Sections 6.10 and 6.11 *6.17 (Display matrix of 0s and 1s) Write a method that displays an n-by-n matrix using

the following header:

public static void printMatrix(int n)

Each element is 0 or 1, which is generated randomly. Write a test program that prompts the user to enter n and displays an n-by-n matrix. Here is a sample run:

Enter n: 3 0 1 0 0 0 0 1 1 1

**6.18 (Check password) Some websites impose certain rules for passwords. Write a method that checks whether a string is a valid password. Suppose the password rules are as follows:

■■ A password must have at least ten characters. ■■ A password consists of only letters and digits. ■■ A password must contain at least three digits.

Write a program that prompts the user to enter a password and displays Valid Password if the rules are followed or Invalid Password otherwise.

*6.19 (Triangles) Implement the following two methods:

/** Return true if the sum of every two sides is * greater than the third side. */ public static boolean isValid( double side1, double side2, double side3)

M06_LIAN1878_11_GE_C06.indd 262 1/2/18 10:42 PM

/** Return the area of the triangle. */ public static double area( double side1, double side2, double side3)

Write a test program that reads three sides for a triangle and uses the isValid method to test if the input is valid and uses the area method to obtain the area. The program displays the area if the input is valid. Otherwise, it displays that the input is invalid. The formula for computing the area of a triangle is given in Programming Exercise 2.19.

*6.20 (Count the letters in a string) Write a method that counts the number of letters in a string using the following header:

public static int countLetters(String s)

Write a test program that prompts the user to enter a string and displays the num- ber of letters in the string.

*6.21 (Phone keypads) The international standard letter/number mapping for tele- phones is given in Programming Exercise 4.15. Write a method that returns a number, given an uppercase letter, as follows:

public static int getNumber(char uppercaseLetter)

Write a test program that prompts the user to enter a phone number as a string. The input number may contain letters. The program translates a letter (uppercase or lowercase) to a digit and leaves all other characters intact. Here are sample runs of the program:

Enter a string: 1-800-Flowers 1-800-3569377

Enter a string: 1800flowers 18003569377

**6.22 (Math: approximate the square root) There are several techniques for implement- ing the sqrt method in the Math class. One such technique is known as the Babylonian method. It approximates the square root of a number, n, by repeat- edly performing the calculation using the following formula:

nextGuess = (lastGuess + n / lastGuess) / 2

When nextGuess and lastGuess are almost identical, nextGuess is the approximated square root. The initial guess can be any positive value (e.g., 1). This value will be the starting value for lastGuess. If the difference between nextGuess and lastGuess is less than a very small number, such as 0.0001, you can claim that nextGuess is the approximated square root of n. If not, next Guess becomes lastGuess and the approximation process continues. Implement the following method that returns the square root of n:

public static double sqrt(long n)

*6.23 (Occurrences of a specified character) Write a method that finds the number of occurrences of a specified character in a string using the following header:

public static int count(String str, char a)

Programming Exercises 263

M06_LIAN1878_11_GE_C06.indd 263 1/2/18 10:42 PM

264 Chapter 6 Methods

For example, count("Welcome", 'e') returns 2. Write a test program that prompts the user to enter a string followed by a character then displays the number of occurrences of the character in the string.

Sections 6.10–6.12 **6.24 (Display current date and time) Listing 2.7, ShowCurrentTime.java, displays the

current time. Revise this example to display the current date and time. The calen- dar example in Listing 6.12, PrintCalendar.java, should give you some ideas on how to find the year, month, and day.

**6.25 (Convert milliseconds to hours, minutes, and seconds) Write a method that con- verts milliseconds to hours, minutes, and seconds using the following header:

public static String convertMillis(long millis)

The method returns a string as hours:minutes:seconds. For example, convertMillis(5500) returns a string 0:0:5, convertMillis(100000) returns a string 0:1:40, and convertMillis(555550000) returns a string 154:19:10. Write a test program that prompts the user to enter a long integer for milliseconds and displays a string in the format of hours:minutes:seconds.

Comprehensive **6.26 (Palindromic prime) A palindromic prime is a prime number and also palindromic.

For example, 131 is a prime and also a palindromic prime, as are 313 and 757. Write a program that displays the first 120 palindromic prime numbers. Display 10 numbers per line, separated by exactly one space, as follows:

2 3 5 7 11 101 131 151 181 191 313 353 373 383 727 757 787 797 919 929

**6.27 (Emirp) An emirp (prime spelled backward) is a nonpalindromic prime number whose reversal is also a prime. For example, 17 is a prime and 71 is a prime, so 17 and 71 are emirps. Write a program that displays the first 120 emirps. Display 10 numbers per line, separated by exactly one space, as follows:

13 17 31 37 71 73 79 97 107 113 149 157 167 179 199 311 337 347 359 389 ...

**6.28 (Mersenne prime) A prime number is called a Mersenne prime if it can be written in the form 2 p - 1 for some positive integer p. Write a program that finds all Mersenne primes with p … 31 and displays the output as follows:

p 2^p – 1

2 3

3 7

5 31

...

**6.29 (Twin primes) Twin primes are a pair of prime numbers that differ by 2. For example, 3 and 5 are twin primes, 5 and 7 are twin primes, and 11 and 13 are twin primes. Write a program to find all twin primes less than 1,200. Display the output as follows:

(3, 5) (5, 7) ...

M06_LIAN1878_11_GE_C06.indd 264 1/2/18 10:42 PM

**6.30 (Game: craps) Craps is a popular dice game played in casinos. Write a program to play a variation of the game, as follows: Roll two dice. Each die has six faces representing values 1, 2, . . ., and 6, respec- tively. Check the sum of the two dice. If the sum is 2, 3, or 12 (called craps), you lose; if the sum is 7 or 11 (called natural), you win; if the sum is another value (i.e., 4, 5, 6, 8, 9, or 10), a point is established. Continue to roll the dice until either a 7 or the same point value is rolled. If 7 is rolled, you lose. Otherwise, you win.

Your program acts as a single player. Here are some sample runs.

You rolled 5 + 6 = 11 You win

You rolled 1 + 2 = 3 You lose

You rolled 4 + 4 = 8 point is 8 You rolled 6 + 2 = 8 You win

You rolled 3 + 2 = 5 point is 5 You rolled 2 + 5 = 7 You lose

**6.31 (Financial: credit card number validation) Credit card numbers follow certain patterns. A credit card number must have between 13 and 16 digits. It must start with

■■ 4 for Visa cards ■■ 5 for Master cards ■■ 37 for American Express cards ■■ 6 for Discover cards

In 1954, Hans Luhn of IBM proposed an algorithm for validating credit card numbers. The algorithm is useful to determine whether a card number is entered correctly, or whether a credit card is scanned correctly by a scanner. Credit card numbers are generated following this validity check, commonly known as the Luhn check or the Mod 10 check, which can be described as follows (for illustra- tion, consider the card number 4388576018402626):

1. Double every second digit from right to left. If doubling of a digit results in a two-digit number, add up the two digits to get a single-digit number.

4388576018402626

2 * 2 = 4 2 * 2 = 4 4 * 2 = 8 1 * 2 = 2 6 * 2 = 12 (1 + 2 = 3) 5 * 2 = 10 (1 + 0 = 1) 8 * 2 = 16 (1 + 6 = 7) 4 * 2 = 8

Programming Exercises 265

M06_LIAN1878_11_GE_C06.indd 265 1/2/18 10:42 PM

266 Chapter 6 Methods

2. Now add all single-digit numbers from Step 1.

4 + 4 + 8 + 2 + 3 + 1 + 7 + 8 = 3 7 3. Add all digits in the odd places from right to left in the card number.

6 + 6 + 0 + 8 + 0 + 7 + 8 + 3 = 3 8 4. Sum the results from Step 2 and Step 3.

3 7 + 3 8 = 7 5 5. If the result from Step 4 is divisible by 10, the card number is valid; otherwise,

it is invalid. For example, the number 4388576018402626 is invalid, but the number 4388576018410707 is valid.

Write a program that prompts the user to enter a credit card number as a long integer. Display whether the number is valid or invalid. Design your program to use the following methods:

/** Return true if the card number is valid */ public static boolean isValid(long number)

/** Get the result from Step 2 */ public static int sumOfDoubleEvenPlace(long number)

/** Return this number if it is a single digit, otherwise, * return the sum of the two digits */ public static int getDigit(int number)

/** Return sum of odd-place digits in number */ public static int sumOfOddPlace(long number)

/** Return true if the number d is a prefix for number */ public static boolean prefixMatched(long number, int d)

/** Return the number of digits in d */ public static int getSize(long d)

/** Return the first k number of digits from number. If the * number of digits in number is less than k, return number. */ public static long getPrefix(long number, int k)

Here are sample runs of the program: (You may also implement this program by reading the input as a string and processing the string to validate the credit card.)

Enter a credit card number as a long integer: 4388576018410707 4388576018410707 is valid

Enter a credit card number as a long integer: 4388576018402626 4388576018402626 is invalid

Current date and time is May 16, 2012 10:34:23

**6.32 (Game: chance of winning at craps) Revise Exercise 6.30 to run it 15,000 times and display the number of winning games.

**6.33 (Current date and time) Invoking System.currentTimeMillis() returns the elapsed time in milliseconds since midnight of January 1, 1970. Write a program that displays the date and time. Here is a sample run:

M06_LIAN1878_11_GE_C06.indd 266 1/2/18 10:42 PM

**6.34 (Print calendar) Programming Exercise 3.21 uses Zeller’s congruence to calcu- late the day of the week. Simplify Listing 6.12, PrintCalendar.java, using Zeller’s algorithm to get the start day of the month.

6.35 (Geometry: area of a pentagon) The area of a pentagon can be computed using the following formula:

Area = 5 * s 2

4 * tan ap 5 b

Write a method that returns the area of a pentagon using the following header:

public static double area(double side)

Write a main method that prompts the user to enter the side of a pentagon and displays its area. Here is a sample run:

Enter the side: 5.5 The area of the pentagon is 52.04444136781625

*6.36 (Geometry: area of a regular polygon) A regular polygon is an n-sided polygon in which all sides are of the same length and all angles have the same degree (i.e., the polygon is both equilateral and equiangular). The formula for computing the area of a regular polygon is

Area = n * s 2

4 * tan ap n b

Write a method that returns the area of a regular polygon using the following header:

public static double area(int n, double side)

Write a main method that prompts the user to enter the number of sides and the side of a regular polygon and displays its area. Here is a sample run:

Enter the number of sides: 5 Enter the side: 6.5 The area of the polygon is 72.69017017488385

6.37 (Format an integer) Write a method with the following header to format the inte- ger with the specified width.

public static String format(int number, int width)

The method returns a string for the number with one or more prefix 0s. The size of the string is the width. For example, format(34, 4) returns 0034 and format(34, 5) returns 00034. If the number is longer than the width, the method returns the string representation for the number. For example, format(34, 1) returns 34.

Write a test program that prompts the user to enter a number and its width, and displays a string returned by invoking format(number, width).

Programming Exercises 267

M06_LIAN1878_11_GE_C06.indd 267 1/2/18 10:42 PM

268 Chapter 6 Methods

*6.38 (Generate random characters) Use the methods in RandomCharacter in Listing 6.10 to print 200 uppercase letters and then 200 single digits, printing ten per line.

6.39 (Geometry: point position) Programming Exercise 3.32 shows how to test whether a point is on the left side of a directed line, on the right, or on the same line. Write the methods with the following headers:

/** Return true if point (x2, y2) is on the left side of the * directed line from (x0, y0) to (x1, y1) */ public static boolean leftOfTheLine(double x0, double y0, double x1, double y1, double x2, double y2)

/** Return true if point (x2, y2) is on the same * line from (x0, y0) to (x1, y1) */ public static boolean onTheSameLine(double x0, double y0, double x1, double y1, double x2, double y2)

/** Return true if point (x2, y2) is on the * line segment from (x0, y0) to (x1, y1) */ public static boolean onTheLineSegment(double x0, double y0, double x1, double y1, double x2, double y2)

Write a program that prompts the user to enter the three points for p0, p1, and p2 and displays whether p2 is on the left side of the line from p0 to p1, right side, the same line, or on the line segment. Here are some sample runs:

Enter three points for p0, p1, and p2: 1 1 2 2 1.5 1.5 (1.5, 1.5) is on the line segment from (1.0, 1.0) to (2.0, 2.0)

Enter three points for p0, p1, and p2: 1 1 2 2 3 3 (3.0, 3.0) is on the same line from (1.0, 1.0) to (2.0, 2.0)

Enter three points for p0, p1, and p2: 1 1 2 2 1 1.5 (1.0, 1.5) is on the left side of the line from (1.0, 1.0) to (2.0, 2.0)

Enter three points for p0, p1, and p2: 1 1 2 2 1 –1 (1.0, −1.0) is on the right side of the line from (1.0, 1.0) to (2.0, 2.0)

M06_LIAN1878_11_GE_C06.indd 268 1/2/18 10:42 PM

Single-Dimensional Arrays

Objectives ■■ To describe why arrays are necessary in programming (§7.1).

■■ To declare array reference variables and create arrays (§§7.2.1 and 7.2.2).

■■ To obtain array size using arrayRefVar.length and know default values in an array (§7.2.3).

■■ To access array elements using indexes (§7.2.4).

■■ To declare, create, and initialize an array using an array initializer (§7.2.5).

■■ To program common array operations (displaying arrays, summing all elements, finding the minimum and maximum elements, random shuf- fling, and shifting elements) (§7.2.6).

■■ To simplify programming using the foreach loops (§7.2.7).

■■ To apply arrays in application development (AnalyzeNumbers, and DeckOfCards) (§§7.3 and 7.4).

■■ To copy contents from one array to another (§7.5).

■■ To develop and invoke methods with array arguments and return values (§§7.6–7.8).

■■ To define a method with a variable-length argument list (§7.9).

■■ To search elements using the linear (§7.10.1) or binary (§7.10.2) search algorithm.

■■ To sort an array using the selection sort approach (§7.11).

■■ To use the methods in the java.util.Arrays class (§7.12).

■■ To pass arguments to the main method from the command line (§7.13).

Chapter

M07_LIAN1878_11_GE_C07.indd 269 1/3/18 12:10 AM

270 Chapter 7 Single-Dimensional Arrays

7.1 Introduction A single array variable can reference a large collection of data.

Often you will have to store a large number of values during the execution of a program. Suppose, for instance, that you need to read 100 numbers, compute their average, and find out how many numbers are above the average. Your program first reads the numbers and computes their average, then compares each number with the average to determine whether it is above the average. In order to accomplish this task, the numbers must all be stored in variables. You have to declare 100 variables and repeatedly write almost identical code 100 times. Writing a program this way would be impractical. So, how do you solve this problem?

An efficient, organized approach is needed. Java and most other high-level languages pro- vide a data structure, the array, which stores a fixed-size sequential collection of elements of the same type. In the present case, you can store all 100 numbers into an array and access them through a single array variable.

This chapter introduces single-dimensional arrays. The next chapter will introduce two- dimensional and multidimensional arrays.

7.2 Array Basics Once an array is created, its size is fixed. An array reference variable is used to access the elements in an array using an index.

An array is used to store a collection of data, but often we find it more useful to think of an array as a collection of variables of the same type. Instead of declaring individual variables, such as number0, number1, . . . , and number99, you declare one array variable such as numbers and use numbers[0], numbers[1], . . . , and numbers[99] to represent individual variables. This section introduces how to declare array variables, create arrays, and process arrays using indexes.

7.2.1 Declaring Array Variables To use an array in a program, you must declare a variable to reference the array and specify the array’s element type. Here is the syntax for declaring an array variable:

elementType[] arrayRefVar;

elementType arrayRefVar[]; // Allowed, but not preferred

The elementType can be any data type, and all elements in the array will have the same data type. For example, the following code declares a variable myList that references an array of double elements.

double[] myList;

double myList[]; // Allowed, but not preferred

Note You can use elementType arrayRefVar[] to declare an array variable. This style comes from the C/C+ + language and was adopted in Java to accommodate C/C+ + programmers. The style elementType[] arrayRefVar is preferred.

Point Key

problem why array?

Point Key

index

element type

preferred syntax

M07_LIAN1878_11_GE_C07.indd 270 1/3/18 12:10 AM

7.2 Array Basics 271

7.2.2 Creating Arrays Unlike declarations for primitive data type variables, the declaration of an array variable does not allocate any space in memory for the array. It creates only a storage location for the refer- ence to an array. If a variable does not contain a reference to an array, the value of the variable is null. You cannot assign elements to an array unless it has already been created. After an array variable is declared, you can create an array by using the new operator and assign its reference to the variable with the following syntax:

arrayRefVar = new elementType[arraySize];

This statement does two things: (1) it creates an array using new elementType[arraySize] and (2) it assigns the reference of the newly created array to the variable arrayRefVar.

Declaring an array variable, creating an array, and assigning the reference of the array to the variable can be combined in one statement as

elementType[] arrayRefVar = new elementType[arraySize];

elementType arrayRefVar[] = new elementType[arraySize];

Here is an example of such a statement:

double[] myList = new double[10];

This statement declares an array variable, myList, creates an array of 10 elements of double type, and assigns its reference to myList. To assign values to the elements, use the syntax

arrayRefVar[index] = value;

For example, the following code initializes the array:

myList[0] = 5.6; myList[1] = 4.5; myList[2] = 3.3; myList[3] = 13.2; myList[4] = 4.0; myList[5] = 34.33; myList[6] = 34.0; myList[7] = 45.45; myList[8] = 99.993; myList[9] = 11123;

This array is illustrated in Figure 7.1.

null

new operator

Figure 7.1 The array myList has 10 elements of double type and int indices from 0 to 9.

double[] myList = new double[10];

myList reference myList[0]

myList[1]

myList[2]

myList[3]

myList[4]

myList[6]

myList[5]

myList[7]

myList[8]

myList[9]

Array reference variable

Array element at index 5

5.6

4.5

3.3

13.2

4.0

34.33

34.0

45.45

99.993

11123

Element value

M07_LIAN1878_11_GE_C07.indd 271 1/3/18 12:10 AM

272 Chapter 7 Single-Dimensional Arrays

Note An array variable that appears to hold an array actually contains a reference to that array. Strictly speaking, an array variable and an array are different, but most of the time the distinction can be ignored. Thus, it is all right to say, for simplicity, that myList is an array, instead of stating, at greater length, that myList is a variable that contains a refer- ence to an array of double elements.

7.2.3 Array Size and Default Values When space for an array is allocated, the array size must be given, specifying the number of ele- ments that can be stored in it. The size of an array cannot be changed after the array is created. Size can be obtained using arrayRefVar.length. For example, myList.length is 10.

When an array is created, its elements are assigned the default value of 0 for the numeric primitive data types, \u0000 for char types, and false for boolean types.

7.2.4 Accessing Array Elements The array elements are accessed through the index. Array indices are 0 based; that is, they range from 0 to arrayRefVar.length − 1. In the example in Figure 7.1, myList holds 10 double values, and the indices are from 0 to 9.

Each element in the array is represented using the following syntax, known as an indexed variable:

arrayRefVar[index];

For example, myList[9] represents the last element in the array myList.

Caution Some programming languages use parentheses to reference an array element, as in myList(9), but Java uses brackets, as in myList[9].

An indexed variable can be used in the same way as a regular variable. For example, the following code adds the values in myList[0] and myList[1] to myList[2]:

myList[2] = myList[0] + myList[1];

The following loop assigns 0 to myList[0], 1 to myList[1], . . . , and 9 to myList[9]:

for (int i = 0; i < myList.length; i++) { myList[i] = i; }

7.2.5 Array Initializers Java has a shorthand notation, known as the array initializer, which combines the declaration, creation, and initialization of an array in one statement using the following syntax:

elementType[] arrayRefVar = {value0, value1, ..., valuek};

For example, the statement

double[] myList = {1.9, 2.9, 3.4, 3.5};

declares, creates, and initializes the array myList with four elements, which is equivalent to the following statements:

double[] myList = new double[4]; myList[0] = 1.9; myList[1] = 2.9;

array vs. array variable

array length

default values

0 based index

indexed variable

array initializer

M07_LIAN1878_11_GE_C07.indd 272 1/3/18 12:10 AM

7.2 Array Basics 273

myList[2] = 3.4; myList[3] = 3.5;

Caution The new operator is not used in the array-initializer syntax. Using an array initializer, you have to declare, create, and initialize the array all in one statement. Splitting it would cause a syntax error. Thus, the next statement is wrong:

double[] myList; myList = {1.9, 2.9, 3.4, 3.5}; // Wrong

7.2.6 Processing Arrays When processing array elements, you will often use a for loop for one of two reasons:

1. All of the elements in an array are of the same type. They are evenly processed in the same fashion repeatedly using a loop.

2. Since the size of the array is known, it is natural to use a for loop.

Assume that the array is created as follows:

double[] myList = new double[10];

The following are some examples of processing arrays:

1. Initializing arrays with input values: The following loop initializes the array myList with user input values:

java.util.Scanner input = new java.util.Scanner(System.in); System.out.print("Enter " + myList.length + " values: "); for (int i = 0; i < myList.length; i++) myList[i] = input.nextDouble();

2. Initializing arrays with random values: The following loop initializes the array myList with random values between 0.0 and 100.0, but less than 100.0:

for (int i = 0; i < myList.length; i++) { myList[i] = Math.random() * 100; }

3. Displaying arrays: To print an array, you have to print each element in the array using a loop such as the following:

for (int i = 0; i < myList.length; i++) { System.out.print(myList[i] + " "); }

Tip For an array of the char[] type, it can be printed using one print statement. For example, the following code displays Dallas:

char[] city = {'D', 'a', 'l', 'l', 'a', 's'}; System.out.println(city);

4. Summing all elements: Use a variable named total to store the sum. Initially total is 0. Add each element in the array to total using a loop such as the following:

double total = 0; for (int i = 0; i < myList.length; i++) { total += myList[i]; }

print character array

M07_LIAN1878_11_GE_C07.indd 273 1/3/18 12:10 AM

274 Chapter 7 Single-Dimensional Arrays

5. Finding the largest element: Use a variable named max to store the largest element. Initially max is myList[0]. To find the largest element in the array myList, compare each element with max, and update max if the element is greater than max.

double max = myList[0]; for (int i = 1; i < myList.length; i++) { if (myList[i] > max) max = myList[i]; }

6. Finding the smallest index of the largest element: Often you need to locate the largest element in an array. If an array has multiple elements with the same largest value, find the smallest index of such an element. Suppose that the array myList is {1, 5, 3, 4, 5, 5}. The largest element is 5, and the smallest index for 5 is 1. Use a variable named max to store the largest element, and a variable named indexOfMax to denote the index of the largest element. Initially max is myList[0] and indexOfMax is 0. Compare each element in myList with max and update max and indexOfMax if the element is greater than max.

double max = myList[0]; int indexOfMax = 0; for (int i = 1; i < myList.length; i++) { if (myList[i] > max) { max = myList[i]; indexOfMax = i; } }

7. Random shuffling: In many applications, you need to randomly reorder the elements in an array. This is called shuffling. To accomplish this, for each element myList[i], randomly generate an index j and swap myList[i] with myList[j], as follows:

double temp = myList[0]; // Retain the �rst element

// Shift elements left for (int i = 1; i < myList.length; i++) { myList[i - 1] = myList[i]; }

// Move the �rst element to �ll in the last position myList[myList.length - 1] = temp;

myList

8. Shifting elements: Sometimes you need to shift the elements left or right. Here is an example of shifting the elements one position to the left and filling the last element with the first element:

9. Simplifying coding: Arrays can be used to greatly simplify coding for certain tasks. For example, suppose you wish to obtain the English name of a given month by its number. If the month names are stored in an array, the month name for a given month can be

swap

myList i

[1]

[i]

for (int i = 0; i < myList.length – 1; i++) { // Generate an index j randomly int j = (int)(Math.random() * myList.length);

// Swap myList[i] with myList[j] double temp = myList[i]; myList[i] = myList[j]; myList[j] = temp; }

.[0]

A random index [j]

random shuffling

VideoNote

Random shuffling

M07_LIAN1878_11_GE_C07.indd 274 1/3/18 12:10 AM

7.2 Array Basics 275

accessed simply via the index. The following code prompts the user to enter a month number and displays its month name:

String[] months = {"January", "February",..., "December"}; System.out.print("Enter a month number (1 to 12): "); int monthNumber = input.nextInt(); System.out.println("The month is " + months[monthNumber − 1]);

If you didn’t use the months array, you would have to determine the month name using a lengthy multiway if−else statement as follows:

if (monthNumber == 1) System.out.println("The month is January"); else if (monthNumber == 2) System.out.println("The month is February"); ... else System.out.println("The month is December");

7.2.7 Foreach Loops Java supports a convenient for loop, known as a foreach loop, which enables you to traverse the array sequentially without using an index variable. For example, the following code dis- plays all the elements in the array myList:

for (double e: myList) { System.out.println(e); }

You can read the code as “for each element e in myList, do the following.” Note that the variable, e, must be declared as the same type as the elements in myList.

In general, the syntax for a foreach loop is

for (elementType element: arrayRefVar) { // Process the element }

You still have to use an index variable if you wish to traverse the array in a different order or change the elements in the array.

Caution Accessing an array out of bounds is a common programming error that throws a runtime ArrayIndexOutOfBoundsException. To avoid it, make sure you do not use an index beyond arrayRefVar.length − 1 or simply using a foreach loop if possible.

Programmers often mistakenly reference the first element in an array with index 1, but it should be 0. This is called the off-by-one error. Another common off-by-one error in a loop is using <= where < should be used. For example, the following loop is wrong:

for (int i = 0; i <= list.length; i++) System.out.print(list[i] + " ");

The <= should be replaced by <. Using a foreach loop can avoid the off-by-one error in this case.

7.2.1 How do you declare an array reference variable and how do you create an array? 7.2.2 When is the memory allocated for an array?

ArrayIndexOutOfBounds- Exception

off-by-one error

Point Check

M07_LIAN1878_11_GE_C07.indd 275 1/3/18 12:10 AM

276 Chapter 7 Single-Dimensional Arrays

7.2.3 What is the output of the following code? int x = 30; int[] numbers = new int[x]; x = 60; System.out.println("x is " + x); System.out.println("The size of numbers is " + numbers.length);

7.2.4 Indicate true or false for the following statements: a. Every element in an array has the same type. b. The array size is fixed after an array reference variable is declared. c. The array size is fixed after it is created. d. The elements in an array must be of a primitive data type.

7.2.5 Which of the following statements are valid? a. int i = new int(30); b. double d[] = new double[30]; c. char[] r = new char(1..30); d. int i[] = (3, 4, 3, 2); e. float f[] = {2.3, 4.5, 6.6}; f. char[] c = new char();

7.2.6 How do you access elements in an array? 7.2.7 What is the array index type? What is the lowest index? What is the representation

of the third element in an array named a?

7.2.8 Write statements to do the following:

a. Create an array to hold 10 double values.

b. Assign the value 5.5 to the last element in the array.

c. Display the sum of the first two elements.

d. Write a loop that computes the sum of all elements in the array.

e. Write a loop that finds the minimum element in the array.

f. Randomly generate an index and display the element of this index in the array.

g. Use an array initializer to create another array with the initial values 3.5, 5.5, 4.52, and 5.6.

7.2.9 What happens when your program attempts to access an array element with an invalid index?

7.2.10 Identify and fix the errors in the following code:

1 public class Test { 2 public static void main(String[] args) { 3 double[100] r; 4 5 for (int i = 0; i < r.length(); i++); 6 r(i) = Math.random * 100; 7 } 8 }

7.2.11 What is the output of the following code?

1 public class Test { 2 public static void main(String[] args) { 3 int list[] = {1, 2, 3, 4, 5, 6};

M07_LIAN1878_11_GE_C07.indd 276 1/3/18 12:10 AM

7.3 Case Study: Analyzing Numbers 277

4 for (int i = 1; i < list.length; i++) 5 list[i] = list[i − 1]; 6 7 for (int i = 0; i < list.length; i++) 8 System.out.print(list[i] + " "); 9 } 10 }

7.3 Case Study: Analyzing Numbers The problem is to write a program that finds the number of items above the average of all items.

Now you can write a program using arrays to solve the problem proposed at the beginning of this chapter. The problem is to read 100 numbers, get the average of these numbers, and find the number of the items greater than the average. To be flexible for handling any number of inputs, we will let the user enter the number of inputs, rather than fixing it to 100. Listing 7.1 gives a solution.

Point Key

Enter the number of items: 10

Enter the numbers: 3.4 5 6 1 6.5 7.8 3.5 8.5 6.3 9.5

Average is 5.75

Number of elements above the average is 6

Listing 7.1 AnalyzeNumbers.java 1 public class AnalyzeNumbers { 2 public static void main(String[] args) { numbers[0]: 3 java.util.Scanner input = new java.util.Scanner(System.in); numbers[1]: 4 System.out.print("Enter the number of items: "); numbers[2]: 5 int n = input.nextInt(); 6 double[] numbers = new double[n]; . 7 double sum = 0; . 8 numbers[i]: . 9 System.out.print("Enter the numbers: "); 10 for (int i = 0; i < n; i++) { numbers[n-3]: 11 numbers[i] = input.nextDouble(); numbers[n-2]: 12 sum += numbers[i]; numbers[n-1]: 13 } 14 15 double average = sum / n; 16 17 int count = 0; // The number of elements above average 18 for (int i = 0; i < n; i++) 19 if (numbers[i] > average) 20 count++; 21 22 System.out.println("Average is " + average); 23 System.out.println("Number of elements above the average is " 24 + count); 25 } 26 }

create array

store number in array

get average

above average?

The program prompts the user to enter the array size (line 5) and creates an array with the specified size (line 6). The program reads the input, stores numbers into the array (line 11), adds each number to sum in line 11, and obtains the average (line 15). It then compares

M07_LIAN1878_11_GE_C07.indd 277 1/3/18 12:10 AM

278 Chapter 7 Single-Dimensional Arrays

each number in the array with the average to count the number of values above the average (lines 7–20).

7.4 Case Study: Deck of Cards The problem is to create a program that will randomly select four cards from a deck of cards.

Say you want to write a program that will pick four cards at random from a deck of 52 cards. All the cards can be represented using an array named deck, filled with initial values 0–51, as follows:

int[] deck = new int[52];

// Initialize cards for (int i = 0; i < deck.length; i++) deck[i] = i;

Card numbers 0–12, 13–25, 26–38, and 39–51 represent 13 Spades, 13 Hearts, 13 Diamonds, and 13 Clubs, respectively, as shown in Figure 7.2. cardNumber / 13 determines the suit of the card, and cardNumber % 13 determines the rank of the card, as shown in Figure 7.3. After shuffling the array deck, pick the first four cards from deck. The program displays the cards from these four card numbers.

Point Key

VideoNote

Deck of cards

Figure 7.2 52 cards are stored in an array named deck.

0 . . .

12 13 . . .

25 26 . . .

38 39 . . .

13 Diamonds ( )

13 Clubs ( )

0 . . .

12 13 . . .

25 26 . . .

38 39 . . .

deck [0] . . .

[12] [13]

. [25] [26]

. [38] [39]

. [51]

Random shuf�e

6 48 11 24 . . . . . . . . . . . . . . . .

deck [0] [1] [2] [3] [4] [5] . . .

[25] [26]

. [38] [39]

. [51]

Card number 6 is the 7 (6 % 13 = 6) of Spades (6 / 13 is 0)

Card number 48 is the 10 (48 % 13 = 9) of Clubs (48 / 13 is 3)

Card number 11 is the Queen (11 % 13 = 11) of Spades (11 / 13 is 0)

Card number 24 is the Queen (24 % 13 = 11) of Hearts (24 / 13 is 1)

13 Hearts ( )

13 Spades ( )

Figure 7.3 cardNumber identifies a card’s suit and rank number.

cardNumber / 13 =

Spades

Hearts

Diamonds

Clubs

cardNumber % 13 =

Ace

1 2

Jack

Queen

King

M07_LIAN1878_11_GE_C07.indd 278 1/3/18 12:10 AM

7.4 Case Study: Deck of Cards 279

Listing 7.2 gives the solution to the problem.

Listing 7.2 DeckOfCards.java 1 public class DeckOfCards { 2 public static void main(String[] args) { 3 int[] deck = new int[52]; 4 String[] suits = {"Spades", "Hearts", "Diamonds", "Clubs"}; 5 String[] ranks = {"Ace", "2", "3", "4", "5", "6", "7", "8", "9", 6 "10", "Jack", "Queen", "King"}; 7 8 // Initialize the cards 9 for (int i = 0; i < deck.length; i++) 10 deck[i] = i; 11 12 // Shuffle the cards 13 for (int i = 0; i < deck.length; i++) { 14 // Generate an index randomly 15 int index = (int)(Math.random() * deck.length); 16 int temp = deck[i]; 17 deck[i] = deck[index]; 18 deck[index] = temp; 19 } 20 21 // Display the first four cards 22 for (int i = 0; i < 4; i++) { 23 String suit = suits[deck[i] / 13]; 24 String rank = ranks[deck[i] % 13]; 25 System.out.println("Card number " + deck[i] + ": " 26 + rank + " of " + suit); 27 } 28 } 29 }

create array deck array of strings array of strings

initialize deck

shuffle deck

suit of a card rank of a card

Card number 6: 7 of Spades Card number 48: 10 of Clubs Card number 11: Queen of Spades Card number 24: Queen of Hearts

The program creates an array suits for four suits (line 4) and an array ranks for 13 cards in a suit (lines 5 and 6). Each element in these arrays is a string.

The program initializes deck with values 0–51 in lines 9 and 10. The deck value 0 repre- sents the Ace of Spades, 1 represents the card 2 of Spades, 13 represents the Ace of Hearts, and 14 represents the 2 of Hearts.

Lines 13–19 randomly shuffle the deck. After a deck is shuffled, deck[i] contains an arbitrary value. deck[i] / 13 is 0, 1, 2, or 3, which determines the suit (line 23). deck[i] % 13 is a value between 0 and 12, which determines the rank (line 24). If the suits array is not defined, you would have to determine the suit using a lengthy multiway if−else state- ment as follows:

if (deck[i] / 13 == 0) System.out.print("suit is Spades"); else if (deck[i] / 13 == 1) System.out.print("suit is Hearts"); else if (deck[i] / 13 == 2) System.out.print("suit is Diamonds"); else System.out.print("suit is Clubs");

M07_LIAN1878_11_GE_C07.indd 279 1/3/18 12:10 AM

280 Chapter 7 Single-Dimensional Arrays

With suits = {"Spades", "Hearts", "Diamonds", "Clubs"} created in an array, suits[deck[i] / 13] gives the suit for deck[i]. Using arrays greatly simplifies the solu- tion for this program.

7.4.1 Will the program pick four random cards if you replace lines 22–27 in Listing 7.2, DeckOfCards.java, with the following code?

for (int i = 0; i < 4; i++) { int cardNumber = (int)(Math.random() * deck.length); String suit = suits[cardNumber / 13]; String rank = ranks[cardNumber % 13]; System.out.println("Card number " + cardNumber + ": " + rank + " of " + suit); }

7.5 Copying Arrays To copy the contents of one array into another, you have to copy the array’s individ- ual elements into the other array.

Often, in a program, you need to duplicate an array or a part of an array. In such cases you could attempt to use the assignment statement (=), as follows:

list2 = list1;

However, this statement does not copy the contents of the array referenced by list1 to list2, but instead merely copies the reference value from list1 to list2. After this statement, list1 and list2 reference the same array, as shown in Figure 7.4. The array previously referenced by list2 is no longer referenced; it becomes garbage, which will be automatically collected by the Java Virtual Machine. This process is called garbage collection.

Point Check

Point Key

copy reference

garbage collection

Figure 7.4 Before the assignment statement, list1 and list2 point to separate memory locations. After the assignment, the reference of the list1 array is passed to list2.

Contents of list1

Contents of list2

list1

list2

Before the assignment list2 = list1;

list1

list2

After the assignment list2 = list1;

In Java, you can use assignment statements to copy primitive data type variables, but not arrays. Assigning one array variable to another array variable actually copies one reference to another and makes both variables point to the same memory location.

There are three ways to copy arrays:

1. Use a loop to copy individual elements one by one.

2. Use the static arraycopy method in the System class.

3. Use the clone method to copy arrays; this will be introduced in Chapter 13, Abstract Classes and Interfaces.

M07_LIAN1878_11_GE_C07.indd 280 1/3/18 12:10 AM

7.6 Passing Arrays to Methods 281

You can write a loop to copy every element from the source array to the corresponding element in the target array. The following code, for instance, copies sourceArray to targetArray using a for loop:

int[] sourceArray = {2, 3, 1, 5, 10}; int[] targetArray = new int[sourceArray.length]; for (int i = 0; i < sourceArray.length; i++) { targetArray[i] = sourceArray[i]; }

Another approach is to use the arraycopy method in the java.lang.System class to copy arrays instead of using a loop. The syntax for arraycopy is:

arraycopy(sourceArray, srcPos, targetArray, tarPos, length);

The parameters srcPos and tarPos indicate the starting positions in sourceArray and targetArray, respectively. The number of elements copied from sourceArray to targetArray is indicated by length. For example, you can rewrite the loop using the following statement:

System.arraycopy(sourceArray, 0, targetArray, 0, sourceArray.length);

The arraycopy method does not allocate memory space for the target array. The target array must have already been created with its memory space allocated. After the copying takes place, targetArray and sourceArray have the same content but independent memory locations.

Note The arraycopy method violates the Java naming convention. By convention, this method should be named arrayCopy (i.e., with an uppercase C).

7.5.1 Use the arraycopy method to copy the following array to a target array t: int[] source = {3, 4, 5};

7.5.2 Once an array is created, its size cannot be changed. Does the following code resize the array?

int[] myList; myList = new int[10]; // Sometime later you want to assign a new array to myList myList = new int[20];

7.6 Passing Arrays to Methods When passing an array to a method, the reference of the array is passed to the method.

Just as you can pass primitive type values to methods, you can also pass arrays to methods. For example, the following method displays the elements in an int array:

public static void printArray(int[] array) { for (int i = 0; i < array.length; i++) { System.out.print(array[i] + " "); } }

You can invoke it by passing an array. For example, the following statement invokes the printArray method to display 3, 1, 2, 6, 4, and 2.

printArray(new int[]{3, 1, 2, 6, 4, 2});

arraycopy method

Point Check

Point Key

M07_LIAN1878_11_GE_C07.indd 281 1/3/18 12:10 AM

282 Chapter 7 Single-Dimensional Arrays

Note The preceding statement creates an array using the following syntax:

new elementType[]{value0, value1, ..., valuek};

There is no explicit reference variable for the array. Such array is called an anonymous array.

Java uses pass-by-value to pass arguments to a method. There are important differences between passing the values of variables of primitive data types and passing arrays.

■■ For an argument of a primitive type, the argument’s value is passed.

■■ For an argument of an array type, the value of the argument is a reference to an array; this reference value is passed to the method. Semantically, it can be best described as pass-by-sharing, that is, the array in the method is the same as the array being passed. Thus, if you change the array in the method, you will see the change outside the method.

Take the following code, for example:

public class TestArrayArguments { public static void main(String[] args) { int x = 1; // x represents an int value int[] y = new int[10]; // y represents an array of int values

m(x, y); // Invoke m with arguments x and y

System.out.println("x is " + x); System.out.println("y[0] is " + y[0]); }

public static void m(int number, int[] numbers) { number = 1001; // Assign a new value to number numbers[0] = 5555; // Assign a new value to numbers[0] } }

anonymous array

pass-by-value

pass-by-sharing

x is 1 y[0] is 5555

Figure 7.5 The primitive type value in x is passed to number, and the reference value in y is passed to numbers.

reference

Activation record for the main method int[] y: int x: 1

Stack

Activation record for method m int[] numbers: int number: 1 An array of

ten int values is stored here

Arrays are stored in a heap.

Heap

reference

You may wonder why after m is invoked, x remains 1, but y[0] becomes 5555. This is because y and numbers, although they are independent variables, reference the same array, as illustrated in Figure 7.5. When m(x, y) is invoked, the values of x and y are passed to number and numbers. Since y contains the reference value to the array, numbers now contains the same reference value to the same array.

M07_LIAN1878_11_GE_C07.indd 282 1/3/18 12:10 AM

7.6 Passing Arrays to Methods 283

Note Arrays are objects in Java (objects are introduced in Chapter 9). The JVM stores the objects in an area of memory called the heap, which is used for dynamic memory allocation.

Listing 7.3 gives another program that shows the difference between passing a primitive data type value and an array reference variable to a method.

The program contains two methods for swapping elements in an array. The first method, named swap, fails to swap two int arguments. The second method, named swapFirst TwoInArray, successfully swaps the first two elements in the array argument.

Listing 7.3 TestPassArray.java 1 public class TestPassArray { 2 /** Main method */ 3 public static void main(String[] args) { 4 int[] a = {1, 2}; 5 6 // Swap elements using the swap method 7 System.out.println("Before invoking swap"); 8 System.out.println("array is {" + a[0] + ", " + a[1] + "}"); 9 swap(a[0], a[1]); 10 System.out.println("After invoking swap"); 11 System.out.println("array is {" + a[0] + ", " + a[1] + "}"); 12 13 // Swap elements using the swapFirstTwoInArray method 14 System.out.println("Before invoking swapFirstTwoInArray"); 15 System.out.println("array is {" + a[0] + ", " + a[1] + "}"); 16 swapFirstTwoInArray(a); 17 System.out.println("After invoking swapFirstTwoInArray"); 18 System.out.println("array is {" + a[0] + ", " + a[1] + "}"); 19 } 20 21 /** Swap two variables */ 22 public static void swap(int n1, int n2) { 23 int temp = n1; 24 n1 = n2; 25 n2 = temp; 26 } 27 28 /** Swap the first two elements in the array */ 29 public static void swapFirstTwoInArray(int[] array) { 30 int temp = array[0]; 31 array[0] = array[1]; 32 array[1] = temp; 33 } 34 }

heap

false swap

swap array elements

Before invoking swap array is {1, 2} After invoking swap array is {1, 2} Before invoking swapFirstTwoInArray array is {1, 2} After invoking swapFirstTwoInArray array is {2, 1}

M07_LIAN1878_11_GE_C07.indd 283 1/3/18 12:10 AM

284 Chapter 7 Single-Dimensional Arrays

As shown in Figure 7.6, the two elements are not swapped using the swap method. However, they are swapped using the swapFirstTwoInArray method. Since the parameters in the swap method are primitive type, the values of a[0] and a[1] are passed to n1 and n2 inside the method when invoking swap(a[0], a[1]). The memory locations for n1 and n2 are independent of the ones for a[0] and a[1]. The contents of the array are not affected by this call.

Figure 7.6 When passing an array to a method, the reference of the array is passed to the method.

Invoke swap(int n1, int n2). The primitive type values in a[0] and a[1] are passed to the swap method.

Invoke swapFirstTwoInArray(int[] array). The reference value in a is passed to the swapFirstTwoInArray method.

The arrays are stored in a heap.

Stack Heap

Activation record for the swap method

Activation record for the main method

n2: 2 n1: 1

int[] a reference reference

reference

Stack

Activation record for the swapFirstTwoInArray method

Activation record for the main method

int[] a

int[] array

a[0]: 1 a[1]: 2

l public static int[] reverse(int[] list) { 2 int[] result = new int[list.length]; 3 4 for (int i = 0, j = result.length - 1; 5 i < list.length; i++, j--) { 6 result[j] = list[i]; 7 } 8 9 return result; 10 }

list

result

The parameter in the swapFirstTwoInArray method is an array. As shown in Figure 7.6, the reference of the array is passed to the method. Thus, the variables a (outside the method) and array (inside the method) both refer to the same array in the same memory location. Therefore, swapping array[0] with array[1] inside the method swapFirstTwoInArray is the same as swapping a[0] with a[1] outside of the method.

7.7 Returning an Array from a Method When a method returns an array, the reference of the array is returned.

You can pass arrays when invoking a method. A method may also return an array. For example, the following method returns an array that is the reversal of another array.Point

Key

create array

return array

Line 2 creates a new array result. Lines 4–7 copy elements from array list to array result. Line 9 returns the array. For example, the following statement returns a new array list2 with elements 6, 5, 4, 3, 2, 1:

int[] list1 = {1, 2, 3, 4, 5, 6}; int[] list2 = reverse(list1);

M07_LIAN1878_11_GE_C07.indd 284 1/3/18 12:10 AM

7.8 Case Study: Counting the Occurrences of Each Letter 285

7.7.1 Suppose the following code is written to reverse the contents in an array, explain why it is wrong. How do you fix it?

int[] list = {1, 2, 3, 5, 4};

for (int i = 0, j = list.length − 1; i < list.length; i++, j−−) { // Swap list[i] with list[j] int temp = list[i]; list[i] = list[j]; list[j] = temp; }

7.8 Case Study: Counting the Occurrences of Each Letter

This section presents a program to count the occurrences of each letter in an array of characters.

The program given in Listing 7.4 does the following:

1. Generates 100 lowercase letters randomly and assigns them to an array of characters, as shown in Figure 7.7a. You can obtain a random letter by using the getRandomLower CaseLetter() method in the RandomCharacter class in Listing 6.10.

2. Count the occurrences of each letter in the array. To do so, create an array, say counts, of 26 int values, each of which counts the occurrences of a letter, as shown in Figure 7.7b. That is, counts[0] counts the number of a’s, counts[1] counts the number of b’s, and so on.

Point Check

Point Key

Figure 7.7 The chars array stores 100 characters, and the counts array stores 26 counts, each of which counts the occurrences of a letter.

…

chars[0]

chars[1]

…

chars[98]

chars[99]

…

counts[0]

counts[1]

…

counts[24]

counts[25]

(a) (b)

Listing 7.4 CountLettersInArray.java 1 public class CountLettersInArray { 2 /** Main method */ 3 public static void main(String[] args) { 4 // Declare and create an array 5 char[] chars = createArray(); 6 7 // Display the array 8 System.out.println("The lowercase letters are:"); 9 displayArray(chars); 10

create array

pass array

M07_LIAN1878_11_GE_C07.indd 285 1/3/18 12:10 AM

286 Chapter 7 Single-Dimensional Arrays

11 // Count the occurrences of each letter 12 int[] counts = countLetters(chars); 13 14 // Display counts 15 System.out.println(); 16 System.out.println("The occurrences of each letter are:"); 17 displayCounts(counts); 18 } 19 20 /** Create an array of characters */ 21 public static char[] createArray() { 22 // Declare an array of characters and create it 23 char[] chars = new char[100]; 24 25 // Create lowercase letters randomly and assign 26 // them to the array 27 for (int i = 0; i < chars.length; i++) 28 chars[i] = RandomCharacter.getRandomLowerCaseLetter(); 29 30 // Return the array 31 return chars; 32 } 33 34 /** Display the array of characters */ 35 public static void displayArray(char[] chars) { 36 // Display the characters in the array 20 on each line 37 for (int i = 0; i < chars.length; i++) { 38 if ((i + 1) % 20 == 0) 39 System.out.println(chars[i]); 40 else 41 System.out.print(chars[i] + " "); 42 } 43 } 44 45 /** Count the occurrences of each letter */ 46 public static int[] countLetters(char[] chars) { 47 // Declare and create an array of 26 int 48 int[] counts = new int[26]; 49 50 // For each lowercase letter in the array, count it 51 for (int i = 0; i < chars.length; i++) 52 counts[chars[i] - 'a']++; 53 54 return counts; 55 } 56 57 /** Display counts */ 58 public static void displayCounts(int[] counts) { 59 for (int i = 0; i < counts.length; i++) { 60 if ((i + 1) % 10 == 0) 61 System.out.println(counts[i] + " " + (char)(i + 'a')); 62 else 63 System.out.print(counts[i] + " " + (char)(i + 'a') + " "); 64 } 65 } 66 }

return array

pass array

increase count

M07_LIAN1878_11_GE_C07.indd 286 1/3/18 12:10 AM

7.8 Case Study: Counting the Occurrences of Each Letter 287

The createArray method (lines 21–32) generates an array of 100 random lowercase letters. Line 5 invokes the method and assigns the array to chars. What would be wrong if you rewrote the code as follows?

char[] chars = new char[100]; chars = createArray();

You would be creating two arrays. The first line would create an array by using new char[100]. The second line would create an array by invoking createArray() and assign the reference of the array to chars. The array created in the first line would be garbage because it is no longer referenced, and as mentioned earlier, Java automatically collects garbage behind the scenes. Your program would compile and run correctly, but it would create an array unnecessarily.

Invoking getRandomLowerCaseLetter() (line 28) returns a random lowercase letter. This method is defined in the RandomCharacter class in Listing 6.10.

The countLetters method (lines 46–55) returns an array of 26 int values, each of which stores the number of occurrences of a letter. The method processes each letter in the array and increases its count by one. A brute-force approach to count the occurrences of each letter might be as follows:

for (int i = 0; i < chars.length; i++) if (chars[i] == 'a') counts[0]++; else if (chars[i] == 'b') counts[1]++; ...

However, a better solution is given in lines 51 and 52.

for (int i = 0; i < chars.length; i++) counts[chars[i] − 'a']++;

If the letter (chars[i]) is a, the corresponding count is counts['a' − 'a'] (i.e., counts[0]). If the letter is b, the corresponding count is counts['b' − 'a'] (i.e., counts[1]), since the Unicode of b is one more than that of a. If the letter is z, the corresponding count is counts['z' − 'a'] (i.e., counts[25]), since the Unicode of z is 25 more than that of a.

Figure 7.8 shows the call stack and heap during and after executing createArray. See CheckPoint Question 7.8.3 to show the call stack and heap for other methods in the program.

The lowercase letters are: e y l s r i b k j v j h a b z n w b t v s c c k r d w a m p w v u n q a m p l o a z g d e g f i n d x m z o u l o z j v h w i w n t g x w c d o t x h y v z y z q e a m f w p g u q t r e n n w f c r f

The occurrences of each letter are: 5 a 3 b 4 c 4 d 4 e 4 f 4 g 3 h 3 i 3 j 2 k 3 l 4 m 6 n 4 o 3 p 3 q 4 r 2 s 4 t 3 u 5 v 8 w 3 x 3 y 6 z

M07_LIAN1878_11_GE_C07.indd 287 1/3/18 12:10 AM

288 Chapter 7 Single-Dimensional Arrays

7.8.1 True or false? When an array is passed to a method, a new array is created and passed to the method.

7.8.2 Show the output of the following two programs: Point

Check

Figure 7.8 (a) An array of 100 characters is created when executing createArray. (b) This array is returned and assigned to the variable chars in the main method.

Array of 100 characters

Stack

(a) Executing createArray in line 5

(b) After exiting createArray in line 5

Heap

Activation record for the createArray method

Activation record for the main method

Array of 100 characters

Stack Heap

Activation record for the main method

char[] chars: refchar[] chars: ref

char[] chars: ref

public class Test { public static void main(String[] args) { int number = 0; int[] numbers = new int[1];

m(number, numbers);

System.out.println("number is " + number + " and numbers[0] is " + numbers[0]); }

public static void m(int x, int[] y) { x = 3; y[0] = 3; } }

(a)

public class Test { public static void main(String[] args) { int[] list = {1, 2, 3, 4, 5}; reverse(list); for (int i = 0; i < list.length; i++) System.out.print(list[i] + " "); }

public static void reverse(int[] list) { int[] newList = new int[list.length];

for (int i = 0; i < list.length; i++) newList[i] = list[list.length − 1 − i];

list = newList; } }

(b)

7.8.3 Where are the arrays stored during execution? Show the contents of the stack and heap during and after executing displayArray, countLetters, and display Counts in Listing 7.4.

7.9 Variable-Length Argument Lists A variable number of arguments of the same type can be passed to a method and treated as an array.

You can pass a variable number of arguments of the same type to a method. The parameter in the method is declared as follows:

typeName... parameterName

In the method declaration, you specify the type followed by an ellipsis (...). Only one variable-length parameter may be specified in a method, and this parameter must be the last parameter. Any regular parameters must precede it.

Point Key

M07_LIAN1878_11_GE_C07.indd 288 1/3/18 12:10 AM

7.10 Searching Arrays 289

Java treats a variable-length parameter as an array. You can pass an array or a variable number of arguments to a variable-length parameter. When invoking a method with a variable number of arguments, Java creates an array and passes the arguments to it. Listing 7.5 presents a method that prints the maximum value in a list of an unspecified number of values.

Listing 7.5 VarArgsDemo.java 1 public class VarArgsDemo { 2 public static void main(String[] args) { 3 printMax(34, 3, 3, 2, 56.5); 4 printMax(new double[]{1, 2, 3}); 5 } 6 7 public static void printMax(double... numbers) { 8 if (numbers.length == 0) { 9 System.out.println("No argument passed"); 10 return; 11 } 12 13 double result = numbers[0]; 14 15 for (int i = 1; i < numbers.length; i++) 16 if (numbers[i] > result) 17 result = numbers[i]; 18 19 System.out.println("The max value is " + result); 20 } 21 }

Line 3 invokes the printMax method with a variable-length argument list passed to the array numbers. If no arguments are passed, the length of the array is 0 (line 8).

Line 4 invokes the printMax method with an array.

7.9.1 What is wrong with each of the following method headers? a. public static void print(String... strings, double... numbers) b. public static void print(double... numbers, String name) c. public static double... print(double d1, double d2)

7.9.2 Can you invoke the printMax method in Listing 7.5 using the following statements? a. printMax(1, 2, 2, 1, 4); b. printMax(new double[]{1, 2, 3}); c. printMax(new int[]{1, 2, 3});

7.10 Searching Arrays If an array is sorted, binary search is more efficient than linear search for finding an element in the array.

Searching is the process of looking for a specific element in an array—for example, discover- ing whether a certain score is included in a list of scores. Searching is a common task in com- puter programming. Many algorithms and data structures are devoted to searching. This section discusses two commonly used approaches, linear search and binary search.

7.10.1 The Linear Search Approach The linear search approach compares the key element key sequentially with each element in the array. It continues to do so until the key matches an element in the array, or the array is exhausted without a match being found. If a match is made, the linear search returns the index

pass variable-length arg list pass an array arg

a variable-length arg parameter

Point Check

Point Key

linear search binary search

M07_LIAN1878_11_GE_C07.indd 289 1/3/18 12:10 AM

290 Chapter 7 Single-Dimensional Arrays

of the element in the array that matches the key. If no match is found, the search returns −1. The linearSearch method in Listing 7.6 gives the solution.

Listing 7.6 LinearSearch.java linear search animation on Companion Website

list

key Compare key with list[i] for i = 0, 1, …

[0] [1] [2] …

1 public class LinearSearch { 2 /** The method for �nding a key in the list */ 3 public static int linearSearch(int[] list, int key) { 4 for (int i = 0; i < list.length; i++) { 5 if (key == list[i]) 6 return i; 7 } 8 return -1; 9 } 10 }

To better understand this method, trace it with the following statements:

1 int[] list = {1, 4, 4, 2, 5, −3, 6, 2}; 2 int i = linearSearch(list, 4); // Returns 1 3 int j = linearSearch(list, −4); // Returns -1 4 int k = linearSearch(list, −3); // Returns 5

The linear search method compares the key with each element in the array. The elements can be in any order. On average, the algorithm will have to examine half of the elements in an array before finding the key, if it exists. Since the execution time of a linear search increases linearly as the number of array elements increases, linear search is inefficient for a large array.

7.10.2 The Binary Search Approach Binary search is the other common search approach for a list of values. For binary search to work, the elements in the array must already be ordered. Assume that the array is in ascending order. The binary search first compares the key with the element in the middle of the array. Consider the following three cases:

1. If the key is less than the middle element, you need to continue to search for the key only in the first half of the array.

2. If the key is equal to the middle element, the search ends with a match.

3. If the key is greater than the middle element, you need to continue to search for the key only in the second half of the array.

Clearly, the binary search method eliminates at least half of the array after each comparison. Sometimes you eliminate half of the elements, and sometimes you eliminate half plus one. Suppose the array has n elements. For convenience, let n be a power of 2. After the first com- parison, n/2 elements are left for further search; after the second comparison, (n/2)/2 ele- ments are left. After the kth comparison, n/2k elements are left for further search. When k = log2n, only one element is left in the array, and you need only one more comparison. There- fore, in the worst case when using the binary search approach, you need log2n+1 comparisons to find an element in the sorted array. In the worst case for a list of 1024 (210) elements, binary search requires only 11 comparisons, whereas a linear search requires 1023 comparisons in the worst case.

The portion of the array being searched shrinks by half after each comparison. Let low and high denote, respectively, the first index and last index of the array that is currently being searched. Initially, low is 0 and high is list.length − 1. Let mid denote the index of the middle element, so mid is (low + high)/2. Figure 7.9 shows how to find key 11 in the list {2, 4, 7, 10, 11, 45, 50, 59, 60, 66, 69, 70, 79} using binary search.

binary search animation on Companion Website

M07_LIAN1878_11_GE_C07.indd 290 1/3/18 12:10 AM

7.10 Searching Arrays 291

You now know how the binary search works. The next task is to implement it in Java. Don’t rush to give a complete implementation. Implement it incrementally, one step at a time. You may start with the first iteration of the search, as shown in Figure 7.10a. It compares the key with the middle element in the list whose low index is 0 and high index is list.length − 1. If key < list[mid], set the high index to mid − 1; if key == list[mid], a match is found and return mid; if key > list[mid], set the low index to mid + 1.

Next, consider implementing the method to perform the search repeatedly by adding a loop, as shown in Figure 7.10b. The search ends if the key is found, or if the key is not found when low > high.

why not -1?

Figure 7.9 Binary search eliminates half of the list from further consideration after each comparison.

key is 11 low

key , 50

key . 7

[0] [1] [2] [3] [4] [5] [7] [8] [9] [10] [11]

2list 4 7 10 11 45 50 59 60 66 69 70 79

mid

[6]

high

[12]

low

[0] [1] [2] [3] [4] [5]

2list 4 7 10 11 45

mid high

key 55 11

[3] [4] [5]

list 10 11 45

low mid high

Figure 7.10 Binary search is implemented incrementally.

(a) Version 1

public static int binarySearch( int[] list, int key) { int low = 0; int high = list.length − 1;

int mid = (low + high) / 2; if (key < list[mid]) high = mid − 1; else if (key == list[mid]) return mid; else low = mid + 1;

}

public static int binarySearch( int[] list, int key) { int low = 0; int high = list.length − 1;

while (high >= low) { int mid = (low + high) / 2; if (key < list[mid]) high = mid - 1; else if (key == list[mid]) return mid; else low = mid + 1; }

return 1; // Not found }

(b) Version 2

When the key is not found, low is the insertion point where a key would be inserted to maintain the order of the list. It is more useful to return the insertion point than −1. The method must return a negative value to indicate that the key is not in the list. Can it simply return −low? No. If the key is less than list[0], low would be 0. −0 is 0. This would indicate the key matches list[0]. A good choice is to let the method return −low − 1 if the key is not in the list. Returning −low − 1 indicates not only that the key is not in the list, but also where the key would be inserted.

M07_LIAN1878_11_GE_C07.indd 291 1/3/18 12:10 AM

292 Chapter 7 Single-Dimensional Arrays

The complete program is given in Listing 7.7.

Listing 7.7 BinarySearch.java 1 public class BinarySearch { 2 /** Use binary search to find the key in the list */ 3 public static int binarySearch(int[] list, int key) { 4 int low = 0; 5 int high = list.length − 1; 6 7 while (high >= low) { 8 int mid = (low + high) / 2; 9 if (key < list[mid]) 10 high = mid − 1; 11 else if (key == list[mid]) 12 return mid; 13 else 14 low = mid + 1; 15 } 16 17 return –low - 1; // Now high < low, key not found 18 } 19 }

The binary search returns the index of the search key if it is contained in the list (line 12). Otherwise, it returns −low − 1 (line 17).

What would happen if we replaced (high >= low) in line 7 with (high > low)? The search would miss a possible matching element. Consider a list with just one element. The search would miss the element.

Does the method still work if there are duplicate elements in the list? Yes, as long as the elements are sorted in increasing order. The method returns the index of one of the matching elements if the element is in the list.

The precondition for the binary search method is that the list must be sorted in increasing order. The postcondition is that the method returns the index of the element that matches the key if the key is in the list or a negative integer k such that −k 1 is the position for inserting the key. Precondition and postcondition are the terms often used to describe the properties of a method. Preconditions are the things that are true before the method is invoked, and postconditions are the things that are true after the method is returned:

To better understand this method, trace it with the following statements and identify low and high when the method returns.

int[] list = {2, 4, 7, 10, 11, 45, 50, 59, 60, 66, 69, 70, 79}; int i = BinarySearch.binarySearch(list, 2); // Returns 0 int j = BinarySearch.binarySearch(list, 11); // Returns 4 int k = BinarySearch.binarySearch(list, 12); // Returns –6 int l = BinarySearch.binarySearch(list, 1); // Returns –1 int m = BinarySearch.binarySearch(list, 3); // Returns –2

Here is the table that lists the low and high values when the method exits, and the value returned from invoking the method.

Method Low High Value Returned

binarySearch(list, 2) 0 1 0 (mid)

binarySearch(list, 11) 3 5 4 (mid)

binarySearch(list, 12) 5 4 −6

binarySearch(list, 1) 0 −1 −1

binarySearch(list, 3) 1 0 −2

first half

second half

precondition

postcondition

M07_LIAN1878_11_GE_C07.indd 292 1/3/18 12:10 AM

7.11 Sorting Arrays 293

Note Linear search is useful for finding an element in a small array or an unsorted array, but it is inefficient for large arrays. Binary search is more efficient, but it requires that the array be presorted.

7.10.1 If high is a very large integer such as the maximum int value 2147483647, (low + high) / 2 may cause overflow. How do you fix it to avoid overflow?

7.10.2 Use Figure 7.9 as an example to show how to apply the binary search approach to a search for key 10 and key 12 in list {2, 4, 7, 10, 11, 45, 50, 59, 60, 66, 69, 70, 79}.

7.10.3 If the binary search method returns −4, is the key in the list? Where should the key be inserted if you wish to insert the key into the list?

7.11 Sorting Arrays Sorting, like searching, is a common task in computer programming. Many different algorithms have been developed for sorting. This section introduces an intuitive sort- ing algorithm: selection sort.

Suppose you want to sort a list in ascending order. Selection sort finds the smallest number in the list and swaps it with the first element. It then finds the smallest number remaining and swaps it with the second element, and so on, until only a single number remains. Figure 7.11 shows how to sort the list {2, 9, 5, 4, 8, 1, 6} using selection sort.

binary search benefits

Point Check

Point Key

selection sort

VideoNote

Selection sort

selection sort animation on Companion Website

Figure 7.11 Selection sort repeatedly selects the smallest number and swaps it with the first number in the list.

Select 1 (the smallest) and swap it with 2 (the �rst) in the list.

The number 1 is now in the correct position, and thus no longer needs to be considered.

The number 2 is now in the correct position, and thus no longer needs to be considered.

The number 4 is now in the correct position, and thus no longer needs to be considered.

The number 5 is now in the correct position, and thus no longer needs to be considered.

The number 6 is now in the correct position, and thus no longer needs to be considered.

swap

9 The number 8 is now in the correct position, and thus no longer needs to be considered.

Select 2 (the smallest) and swap it with 9 (the �rst) in the remaining list.

Select 4 (the smallest) and swap it with 5 (the �rst) in the remaining list.

5 is the smallest and in the right position. No swap is necessary.

Select 6 (the smallest) and swap it with 8 (the �rst) in the remaining list.

Select 8 (the smallest) and swap it with 9 (the �rst) in the remaining list.

Since there is only one element remaining in the list, the sort is completed.

swap

You know how the selection-sort approach works. The task now is to implement it in Java. Beginners find it difficult to develop a complete solution on the first attempt. Start by writing the code for the first iteration to find the smallest element in the list and swap it with the first element, then observe what would be different for the second iteration, the third, and so on. The insight this gives will enable you to write a loop that generalizes all the iterations.

M07_LIAN1878_11_GE_C07.indd 293 1/3/18 12:10 AM

294 Chapter 7 Single-Dimensional Arrays

The solution can be described as follows:

for (int i = 0; i < list.length − 1; i++) { select the smallest element in list[i..list.length−1]; swap the smallest with list[i], if necessary; // list[i] is in its correct position. // The next iteration applies on list[i+1..list.length−1] }

Listing 7.8 implements the solution.

Listing 7.8 SelectionSort.java 1 public class SelectionSort { 2 /** The method for sorting the numbers */ 3 public static void selectionSort(double[] list) { 4 for (int i = 0; i < list.length - 1; i++) { 5 // Find the minimum in the list[i..list.length−1] 6 double currentMin = list[i]; 7 int currentMinIndex = i; 8 9 for (int j = i + 1; j < list.length; j++) { 10 if (currentMin > list[j]) { 11 currentMin = list[j]; 12 currentMinIndex = j; 13 } 14 } 15 16 // Swap list[i] with list[currentMinIndex] if necessary 17 if (currentMinIndex != i) { 18 list[currentMinIndex] = list[i]; 19 list[i] = currentMin; 20 } 21 } 22 } 23 }

The selectionSort(double[] list) method sorts any array of double elements. The method is implemented with a nested for loop. The outer loop (with the loop control vari- able i in line 4) is iterated in order to find the smallest element in the list, which ranges from list[i] to list[list.length−1], and exchanges it with list[i].

The variable i is initially 0. After each iteration of the outer loop, list[i] is in the right place. Eventually, all the elements are put in the right place; therefore, the whole list is sorted.

To understand this method better, trace it with the following statements:

double[] list = {1, 9, 4.5, 6.6, 5.7, −4.5}; SelectionSort.selectionSort(list);

7.11.1 Use Figure 7.11 as an example to show how to apply the selection-sort approach to sort {3.4, 5, 3, 3.5, 2.2, 1.9, 2}.

7.11.2 How do you modify the selectionSort method in Listing 7.8 to sort numbers in decreasing order?

7.12 The Arrays Class The java.util.Arrays class contains useful methods for common array operations such as sorting and searching.

select

swap

Point Check

Point Key

M07_LIAN1878_11_GE_C07.indd 294 1/3/18 12:10 AM

7.12 The Arrays Class 295

The java.util.Arrays class contains various static methods for sorting and searching arrays, comparing arrays, filling array elements, and returning a string representation of the array. These methods are overloaded for all primitive types.

You can use the sort or parallelSort method to sort a whole array or a partial array. For example, the following code sorts an array of numbers and an array of characters:

double[] numbers = {6.0, 4.4, 1.9, 2.9, 3.4, 3.5}; java.util.Arrays.sort(numbers); // Sort the whole array java.util.Arrays.parallelSort(numbers); // Sort the whole array

char[] chars = {'a', 'A', '4', 'F', 'D', 'P'}; java.util.Arrays.sort(chars, 1, 3); // Sort part of the array java.util.Arrays.parallelSort(chars, 1, 3); // Sort part of the array

Invoking sort(numbers) sorts the whole array numbers. Invoking sort(chars, 1, 3) sorts a partial array from chars[1] to chars[3−1]. parallelSort is more efficient if your computer has multiple processors.

You can use the binarySearch method to search for a key in an array. The array must be pre- sorted in increasing order. If the key is not in the array, the method returns −(insertionIndex + 1). For example, the following code searches the keys in an array of integers and an array of characters:

int[] list = {2, 4, 7, 10, 11, 45, 50, 59, 60, 66, 69, 70, 79}; System.out.println("1. Index is " + java.util.Arrays.binarySearch(list, 11)); System.out.println("2. Index is " + java.util.Arrays.binarySearch(list, 12));

char[] chars = {'a', 'c', 'g', 'x', 'y', 'z'}; System.out.println("3. Index is " + java.util.Arrays.binarySearch(chars, 'a')); System.out.println("4. Index is " + java.util.Arrays.binarySearch(chars, 't'));

The output of the preceding code is as follows:

Index is 4

Index is -6 Index is 0

Index is -4

You can use the equals method to check whether two arrays are strictly equal. Two arrays are strictly equal if their corresponding elements are the same. In the following code, list1 and list2 are equal, but list2 and list3 are not.

int[] list1 = {2, 4, 7, 10}; int[] list2 = {2, 4, 7, 10}; int[] list3 = {4, 2, 7, 10}; System.out.println(java.util.Arrays.equals(list1, list2)); // true System.out.println(java.util.Arrays.equals(list2, list3)); // false

You can use the fill method to fill in all or part of the array. For example, the following code fills list1 with 5 and fills 8 into elements list2[1] through list2[5−1].

int[] list1 = {2, 4, 7, 10}; int[] list2 = {2, 4, 7, 7, 7, 10}; java.util.Arrays.fill(list1, 5); // Fill 5 to the whole array java.util.Arrays.fill(list2, 1, 5, 8); // Fill 8 to a partial array

sort

parallelSort

binarySearch

equals

fill

M07_LIAN1878_11_GE_C07.indd 295 1/3/18 12:10 AM

296 Chapter 7 Single-Dimensional Arrays

You can also use the toString method to return a string that represents all elements in the array. This is a quick and simple way to display all elements in the array. For example, the following code:

int[] list = {2, 4, 7, 10}; System.out.println(java.util.Arrays.toString(list));

displays [2, 4, 7, 10].

7.12.1 What types of array can be sorted using the java.util.Arrays.sort method? Does this sort method create a new array?

7.12.2 To apply java.util.Arrays.binarySearch(array, key), should the array be sorted in increasing order, in decreasing order, or neither?

7.12.3 Show the output of the following code: int[] list1 = {2, 4, 7, 10}; java.util.Arrays.fill(list1, 7); System.out.println(java.util.Arrays.toString(list1));

int[] list2 = {2, 4, 7, 10}; System.out.println(java.util.Arrays.toString(list2)); System.out.print(java.util.Arrays.equals(list1, list2));

7.13 Command-Line Arguments The main method can receive string arguments from the command line.

Perhaps you have already noticed the unusual header for the main method, which has the parameter args of the String[] type. It is clear that args is an array of strings. The main method is just like a regular method with a parameter. You can call a regular method by passing actual parameters. Can you pass arguments to main? Yes, of course you can. In the following examples, the main method in class TestMain is invoked by a method in A:

toString

Point Check

Point Key

public class A { public static void main(String[] args) { String[] strings = {"New York", "Boston", "Atlanta"}; TestMain.main(strings); } }

public class TestMain { public static void main(String[] args) { for (int i = 0; i < args.length; i++) System.out.println(args[i]); } }

A main method is just like a regular method. Furthermore, you can pass arguments to a main method from the command line.

7.13.1 Passing Strings to the main Method You can pass strings to a main method from the command line when you run the program. The following command line, for example, starts the program TestMain with three strings: arg0, arg1, and arg2:

java TestMain arg0 arg1 arg2

arg0, arg1, and arg2 are strings, but they don’t have to appear in double quotes on the com- mand line. The strings are separated by a space. A string that contains a space must be enclosed in double quotes. Consider the following command line:

java TestMain "First num" alpha 53

M07_LIAN1878_11_GE_C07.indd 296 1/3/18 12:10 AM

7.13 Command-Line Arguments 297

It starts the program with three strings: First num, alpha, and 53. Since First num is a string, it is enclosed in double quotes. Note 53 is actually treated as a string. You can use "53" instead of 53 in the command line.

When the main method is invoked, the Java interpreter creates an array to hold the com- mand-line arguments and pass the array reference to args. For example, if you invoke a program with n arguments, the Java interpreter creates an array such as the one that follows:

args = new String[n];

The Java interpreter then passes args to invoke the main method.

Note If you run the program with no strings passed, the array is created with new String[0]. In this case, the array is empty with length 0. args references to this empty array. Therefore, args is not null, but args.length is 0.

7.13.2 Case Study: Calculator Suppose you are to develop a program that performs arithmetic operations on integers. The program receives an expression. The expression consists of an integer followed by an operator and another integer. For example, to add two integers, use this command:

java Calculator 2 + 3

The program will display the following output:

2 + 3 = 5

Figure 7.12 shows sample runs of the program. The strings passed to the main program are stored in args, which is an array of strings. The

first string is stored in args[0], and args.length is the number of strings passed. Here are the steps in the program:

1. Use args.length to determine whether the expression has been provided as three arguments in the command line. If not, terminate the program using System.exit(1).

2. Perform a binary arithmetic operation on the operands args[0] and args[2] using the operator in args[1].

VideoNote

Command-line arguments

Figure 7.12 The program takes three arguments (operand1 operator operand2) from the command line and displays the expression and the result of the arithmetic operation.

Add

Subtract

Multiply

Divide

M07_LIAN1878_11_GE_C07.indd 297 1/3/18 12:10 AM

298 Chapter 7 Single-Dimensional Arrays

The program is given in Listing 7.9.

Listing 7.9 Calculator.java 1 public class Calculator { 2 /** Main method */ 3 public static void main(String[] args) { 4 // Check number of strings passed 5 if (args.length != 3) { 6 System.out.println( 7 "Usage: java Calculator operand1 operator operand2"); 8 System.exit(1); 9 } 10 11 // The result of the operation 12 int result = 0; 13 14 // Determine the operator 15 switch (args[1].charAt(0)) { 16 case '+': result = Integer.parseInt(args[0]) + 17 Integer.parseInt(args[2]); 18 break; 19 case '−': result = Integer.parseInt(args[0]) − 20 Integer.parseInt(args[2]); 21 break; 22 case '.': result = Integer.parseInt(args[0]) * 23 Integer.parseInt(args[2]); 24 break; 25 case '/': result = Integer.parseInt(args[0]) / 26 Integer.parseInt(args[2]); 27 } 28 29 // Display result 30 System.out.println(args[0] + ' ' + args[1] + ' ' + args[2] 31 + " = " + result); 32 } 33 }

Integer.parseInt(args[0]) (line 16) converts a digital string into an integer. The string must consist of digits. If not, the program will terminate abnormally.

We used the . symbol for multiplication, not the common * symbol. The reason for this is the * symbol refers to all the files in the current directory when it is used on a command line. The following program displays all the files in the current directory when issuing the com- mand java Test *:

public class Test { public static void main(String[] args) { for (int i = 0; i < args.length; i++) System.out.println(args[i]); } }

To circumvent this problem, we will have to use a different symbol for the multiplication operator.

7.13.1 This book declares the main method as public static void main(String[] args)

Can it be replaced by one of the following lines?

a. public static void main(String args[]) b. public static void main(String[] x)

check argument

check operator

Point Check

M07_LIAN1878_11_GE_C07.indd 298 1/3/18 12:10 AM

Chapter Summary 299

c. public static void main(String x[]) d. static void main(String x[])

7.13.2 Show the output of the following program when invoked using

1. java Test I have a dream

2. java Test “1 2 3”

3. java Test

public class Test { public static void main(String[] args) { System.out.println("Number of strings is " + args.length); for (int i = 0; i < args.length; i++) System.out.println(args[i]); } }

Key Terms anonymous array 282 array 270 array initializer 272 binary search 289 garbage collection 280 index 270

indexed variable 272 linear search 289 off-by-one error 275 postcondition 292 precondition 292 selection sort 293

ChapTer summary 1. A variable is declared as an array type using the syntax elementType[] arrayRefVar

or elementType arrayRefVar[]. The style elementType[] arrayRefVar is preferred, although elementType arrayRefVar[] is legal.

2. Unlike declarations for primitive data type variables, the declaration of an array variable does not allocate any space in memory for the array. An array variable is not a primitive data type variable. An array variable contains a reference to an array.

3. You cannot assign elements to an array unless it has already been created. You can create an array by using the new operator with the following syntax: new elementType[arraySize].

4. Each element in the array is represented using the syntax arrayRefVar[index]. An index must be an integer or an integer expression.

5. After an array is created, its size becomes permanent and can be obtained using array RefVar.length. Since the index of an array always begins with 0, the last index is always arrayRefVar.length − 1. An out-of-bounds error will occur if you attempt to reference elements beyond the bounds of an array.

6. Programmers often mistakenly reference the first element in an array with index 1, but it should be 0. This is called the index off-by-one error.

M07_LIAN1878_11_GE_C07.indd 299 1/3/18 12:10 AM

300 Chapter 7 Single-Dimensional Arrays

7. When an array is created, its elements are assigned the default value of 0 for the numeric primitive data types, \u0000 for char types, and false for boolean types.

8. Java has a shorthand notation, known as the array initializer, which combines declaring an array, creating an array, and initializing an array in one statement, using the syntax elementType[] arrayRefVar = {value0, value1, . . . , valuek}.

9. When you pass an array argument to a method, you are actually passing the reference of the array; that is, the called method can modify the elements in the caller’s original array.

10. If an array is sorted, binary search is more efficient than linear search for finding an element in the array.

11. Selection sort finds the smallest number in the list and swaps it with the first element. It then finds the smallest number remaining and swaps it with the first element in the remaining list, and so on, until only a single number remains.

Quiz Answer the quiz for this chapter online at the Companion Website.

programming exerCises

Sections 7.2–7.5 *7.1 (Assign grades) Write a program that reads student scores, gets the best score, and

then assigns grades based on the following scheme:

Grade is A if score is Ú best -5 Grade is B if score is Ú best -10; Grade is C if score is Ú best -15; Grade is D if score is Ú best -20; Grade is F otherwise.

The program prompts the user to enter the total number of students, and then prompts the user to enter all of the scores, and concludes by displaying the grades. Here is a sample run:

Enter the number of students: 4

Enter 4 scores: 40 55 70 58 Student 0 score is 40 and grade is F Student 1 score is 55 and grade is C Student 2 score is 70 and grade is A Student 3 score is 58 and grade is C

7.2 (Comparing numbers) Write a program that reads 11 integers, compares each integer with the 11th integer, and displays whether the integers are “greater”, “smaller”, or “equal” to the 11th integer.

M07_LIAN1878_11_GE_C07.indd 300 1/3/18 12:10 AM

Programming Exercises 301

**7.3 (Count occurrence of numbers) Write a program that reads the integers between 1 and 50 and counts the occurrences of each. Assume the input ends with 0. Here is a sample run of the program:

Enter the integers between 1 and 50: 2 5 6 5 4 3 23 43 2 0 2 occurs 2 times 3 occurs 1 time 4 occurs 1 time 5 occurs 2 times 6 occurs 1 time 23 occurs 1 time 43 occurs 1 time

Enter numbers: 1 2 3 2 1 6 3 4 5 2 3 6 8 9 9 0 The number of odd numbers: 8 The number of even numbers: 7

Note if a number occurs more than one time, the plural word “times” is used in the output.

7.4 (Analyze scores) Write a program that reads an unspecified number of scores and determines how many scores are above or equal to the average, and how many scores are below the average. Enter a negative number to signify the end of the input. Assume the maximum number of scores is 100.

**7.5 (The number of even numbers and odd numbers) Write a program that reads ten integers, and then display the number of even numbers and odd numbers. Assume that the input ends with 0. Here is the sample run of the program.

7.6 (The Sieve of Eratosthenes) The Sieve of Eratosthenes is an algorithm that finds all prime numbers up to a given limit, n. It works by creating an array of Booleans, flag, of size n+1, initializing the array to true (assuming every number is a prime number in the beginning), and iteratively traversing the numbers from 2 to n, setting the values of indices that are multiples of other numbers to false. Write a program using the algorithm above, and display all the prime numbers up to 50.

*7.7 (Count single digits) Write a program that generates 200 random integers between 0 and 9 and displays the count for each number. (Hint: Use an array of ten integers, say counts, to store the counts for the number of 0s, 1s, . . . , 9s.)

Sections 7.6–7.8 7.8 (Average an array) Write two overloaded methods that return the average of an

array with the following headers:

public static int average(int[] array) public static double average(double[] array)

Write a test program that prompts the user to enter 10 double values, invokes this method, then displays the average value.

M07_LIAN1878_11_GE_C07.indd 301 1/3/18 12:10 AM

302 Chapter 7 Single-Dimensional Arrays

7.9 (Find the largest element) Write a method that finds the largest element in an array of double values using the following header:

public static double max(double[] array)

Write a test program that prompts the user to enter ten numbers, invokes this method to return the maximum value, and displays the maximum value. Here is a sample run of the program:

Enter ten numbers: 1.9 2.5 3.7 2 1.5 6 3 4 5 2 The minimum number is: 6

Enter 10 numbers: 1.9 2.5 3.7 2 1 6 3 4 5 2 The mean is 3.11 The standard deviation is 1.55738

7.10 (Find the index of the largest element) Write a method that returns the index of the largest element in an array of integers. If there is more than one element, return the largest index. Use the following header:

public static int indexOfLargestElement(double[] array)

Write a test program that prompts the user to enter 10 numbers, invokes this method to return the index of the largest element, and displays the index and the element itself.

*7.11 (Statistics: compute deviation) Programming Exercise 5.45 computes the standard deviation of numbers. This exercise uses a different but equivalent formula to compute the standard deviation of n numbers.

mean = a

i = 1 xi

n =

x1 + x2 + g + xn n

deviation = H ani = 1(xi - mean)2n - 1 To compute the standard deviation with this formula, you have to store the indi-

vidual numbers using an array, so they can be used after the mean is obtained.

Your program should contain the following methods:

/** Compute the deviation of double values */ public static double deviation(double[] x)

/** Compute the mean of an array of double values */ public static double mean(double[] x)

Write a test program that prompts the user to enter 10 numbers and displays the mean and standard deviation, as presented in the following sample run:

*7.12 (Reverse an array) The reverse method in Section 7.7 reverses an array by copying it to a new array. Rewrite the method that reverses the array passed in the argument and returns this array. Write a test program that prompts the user to enter 10 numbers, invokes the method to reverse the numbers, and displays the numbers.

M07_LIAN1878_11_GE_C07.indd 302 1/3/18 12:10 AM

Programming Exercises 303

Section 7.9 *7.13 (Random number selector) Write a method that returns a random number from a

list of numbers passed in the argument. The method header is specified as follows:

public static int getRandom(int... numbers)

7.14 (Computing lcm) Write a method that returns the lcm (least common multiple) of an unspecified number of integers. The method header is specified as follows:

public static int lcm(int... numbers)

Write a test program that prompts the user to enter five numbers, invokes the method to find the lcm of these numbers, and displays the lcm.

Sections 7.10–7.12 7.15 (Eliminate duplicates) Write a method that returns a new array by eliminating the

duplicate values in the array using the following method header:

public static int[] eliminateDuplicates(int[] list)

Write a test program that reads in 10 integers, invokes the method, and displays the distinct numbers separated by exactly one space. Here is a sample run of the program:

Enter 10 numbers: 1 2 3 2 1 6 3 4 5 2 The distinct numbers are: 1 2 3 6 4 5

7.16 (Execution time) Write a program that creates an array of numbers from 1 to 100,000,000 in ascending order. Display the execution time of invoking the linearSearch method and the binarySearch method in Listings 7.6 and 7.7 respectively. Display the execution time of invoking both searches for the fol- lowing numbers: 1; 25,000,000; 50,000,000; 75,000,000; 100,000,000. You can use the following code template to obtain the execution time:

long startTime = System.nanoTime(); perform the task; long endTime = System.nanoTime(); long executionTime = endTime − startTime;

**7.17 (Sort students) Write a program that prompts the user to enter the number of stu- dents, the students’ names, and their scores and prints student names in decreasing order of their scores. Assume the name is a string without spaces, use the Scan ner’s next() method to read a name.

**7.18 (Bubble sort) Write a sort method that uses the bubble-sort algorithm. The bub- ble-sort algorithm makes several passes through the array. On each pass, suc- cessive neighboring pairs are compared. If a pair is not in order, its values are swapped; otherwise, the values remain unchanged. The technique is called a bubble sort or sinking sort because the smaller values gradually “bubble” their way to the top, and the larger values “sink” to the bottom. Write a test program that reads in 10 double numbers, invokes the method, and displays the sorted numbers.

M07_LIAN1878_11_GE_C07.indd 303 1/3/18 12:10 AM

304 Chapter 7 Single-Dimensional Arrays

**7.19 (Sorted?) Write the following method that returns true if the list is already sorted in decreasing order.

public static boolean isSorted(int[] list)

Write a test program that prompts the user to enter a list and displays whether the list is sorted or not. Here is a sample run. Note that the first number in the input indicates the number of the elements in the list. This number is not part of the list.

Enter list: 8 10 1 5 16 61 9 11 1 The list is not sorted

Enter list: 10 21 11 9 7 5 4 4 3 1 1 The list is already sorted

*7.20 (Descending Bubble Sort) Modify the bubble sort method of Programming Exercise 7.18, and sort the elements in the array in descending order instead of ascending order. Write a test program that reads 10 double numbers, invokes the method, and displays the sorted numbers in descending order.

Section 7.13 *7.21 (Normalizing integers) Write a program that prompts the user to enter 10 integers,

and normalizes the integers to be in the range of 0 to 1 by dividing every integer with the largest integer. Display the 10 normalized real-valued numbers.

*7.22 (Count the number of vowels in a string) Write a program that prompts the user to enter a string, and counts and displays the number of both lowercase and uppercase vowels in the string.

Comprehensive **7.23 (Game: locker puzzle) A school has 100 lockers and 100 students. All lockers are

closed on the first day of school. As the students enter, the first student, denoted as S1, opens every locker. Then the second student, S2, begins with the second locker, denoted as L2, and closes every other locker. Student S3 begins with the third locker and changes every third locker (closes it if it was open and opens it if it was closed). Student S4 begins with locker L4 and changes every fourth locker. Student S5 starts with L5 and changes every fifth locker, and so on, until student S100 changes L100.

After all the students have passed through the building and changed the lockers, which lockers are open? Write a program to find your answer and display all open locker numbers separated by exactly one space.

(Hint: Use an array of 100 Boolean elements, each of which indicates whether a locker is open (true) or closed (false). Initially, all lockers are closed.)

**7.24 (Simulation: coupon collector’s problem) Coupon collector is a classic statistics problem with many practical applications. The problem is to repeatedly pick objects from a set of objects and find out how many picks are needed for all the objects to be picked at least once. A variation of the problem is to pick cards from a shuffled deck of 52 cards repeatedly, and find out how many picks are needed before you see one of each suit. Assume a picked card is placed back in the deck

VideoNote

Coupon collector’s problem

M07_LIAN1878_11_GE_C07.indd 304 1/3/18 12:10 AM

Programming Exercises 305

before picking another. Write a program to simulate the number of picks needed to get four cards from each suit and display the four cards picked (it is possible a card may be picked twice). Here is a sample run of the program:

Queen of Spades 5 of Clubs Queen of Hearts 4 of Diamonds Number of picks: 12

Enter list: 5 2 5 6 9 10

The list is sorted

7.25 (Algebra: solve quadratic equations) Write a method for solving a quadratic equa- tion using the following header:

public static int solveQuadratic(double[] eqn, double[] roots)

The coefficients of a quadratic equation ax2 + bx + c = 0 are passed to the array eqn and the real roots are stored in roots. The method returns the number of real roots. See Programming Exercise 3.1 on how to solve a quadratic equation.

Write a program that prompts the user to enter values for a, b, and c and displays the number of real roots and all real roots.

7.26 (Check whether array is sorted) An array list is required to be sorted in ascending order. Write a method that returns true if list is sorted, using the following header:

public static Bboolean isSorted(int[] list)

Write a test program that prompts the user to enter a list of integers. Note the first number in the input indicates the number of elements in the list. This number is not part of the list.

Enter list: 5 2 5 6 1 6

The list is not sorted.

7.27 (Check whether the array is sorted with constant interval) An array list is sorted with a constant interval if its elements are arranged in an ascending order and there is a constant difference between adjacent elements. Write a method that returns true if list is sorted with a constant interval, using the following header:

public static boolean isSortedConstantInterval(int[] list)

Write a test program that prompts the user to enter a list of integers. Note the first number in the input indicates the number of elements in the list. This number is not part of the list.

M07_LIAN1878_11_GE_C07.indd 305 1/3/18 12:10 AM

306 Chapter 7 Single-Dimensional Arrays

*7.28 (Math: permutations) Write a program that prompts the user to enter four integers and then displays all possible ways of arranging the four integers.

*7.29 (Game: dice) Suppose three dice are thrown at random. Write a program that shows all possible permutations or configurations of the three dice that yield the sum of nine.

*7.30 (Pattern recognition: consecutive four equal numbers) Write the following method that tests whether the array has four consecutive numbers with the same value:

public static boolean isConsecutiveFour(int[] values)

Write a test program that prompts the user to enter a series of integers and dis- plays it if the series contains four consecutive numbers with the same value. Your program should first prompt the user to enter the input size—i.e., the number of values in the series. Here are sample runs:

VideoNote

Consecutive four

Enter list: 5 2 5 6 9 10

The list is not sorted with a constant interval

Enter list: 5 2 4 6 8 10 The list is sorted with a constant interval.

Enter the number of values: 8

Enter the values: 3 4 5 5 5 5 4 5

The list has consecutive fours

Enter the number of values: 9

Enter the values: 3 4 5 5 6 5 5 4 5

The list has no consecutive fours

Enter list1 size and contents: 5 1 5 16 61 111

Enter list2 size and contents: 4 2 4 5 6

list1 is 1 5 16 61 111

list2 is 2 4 5 6

The merged list is 1 2 4 5 5 6 16 61 111

**7.31 (Merge two sorted lists) Write the following method that merges two sorted lists into a new sorted list:

public static int[] merge(int[] list1, int[] list2)

Implement the method in a way that takes at most list1.length + list2. length comparisons. See liveexample.pearsoncmg.com/dsanimation/ MergeSortNeweBook.html for an animation of the implementation. Write a test program that prompts the user to enter two sorted lists and displays the merged list. Here is a sample run. Note the first number in the input indicates the number of the elements in the list. This number is not part of the list.

M07_LIAN1878_11_GE_C07.indd 306 1/3/18 12:10 AM

Programming Exercises 307

** 7.32 (Partition of a list) Write the following method that partitions the list using the first element, called a pivot:

public static int partition(int[] list)

After the partition, the elements in the list are rearranged so all the elements before the pivot are less than or equal to the pivot, and the elements after the pivot are greater than the pivot. The method returns the index where the pivot is located in the new list. For example, suppose the list is {5, 2, 9, 3, 6, 8}. After the partition, the list becomes {3, 2, 5, 9, 6, 8}. Implement the method in a way that takes at most list.length comparisons. See liveexample.pearsoncmg.com/dsanima- tion/QuickSortNeweBook.html for an animation of the implementation. Write a test program that prompts the user to enter the size of the list and the contents of the list and displays the list after the partition. Here is a sample run.

Enter list size: 8

Enter list content: 10 1 5 16 61 9 11 1

After the partition, the list is 9 1 5 1 10 61 11 16

*7.33 (Month name) Write a program that prompts the user to enter an integer between 1 and 12 and then displays the name of the month that corresponds to the integer entered by the user. For example, if the user enters three, the program should display March.

**7.34 (Sort characters in a string) Write a method that returns a sorted string using the following header:

public static String sort(String s)

For example, sort("acb") returns abc.

Write a test program that prompts the user to enter a string and displays the sorted string.

***7.35 (Game: hangman) Write a hangman game that randomly generates a word and prompts the user to guess one letter at a time, as presented in the sample run. Each letter in the word is displayed as an asterisk. When the user makes a correct guess, the actual letter is then displayed. When the user finishes a word, display the number of misses and ask the user whether to continue to play with another word. Declare an array to store words, as follows: // Add any words you wish in this array String[] words = {"write", "that",...};

(Guess) Enter a letter in word ******* > p

(Guess) Enter a letter in word p****** > r

(Guess) Enter a letter in word pr**r** > p

p is already in the word

(Guess) Enter a letter in word pr**r** > o

(Guess) Enter a letter in word pro*r** > g

(Guess) Enter a letter in word progr** > n

n is not in the word

(Guess) Enter a letter in word progr** > m

(Guess) Enter a letter in word progr*m > a

The word is program. You missed 1 time

Do you want to guess another word? Enter y or n>

M07_LIAN1878_11_GE_C07.indd 307 1/3/18 12:10 AM

308 Chapter 7 Single-Dimensional Arrays

***7.36 (Game: Eight Queens) The classic Eight Queens puzzle is to place eight queens on a chessboard such that no two queens can attack each other (i.e., no two queens are on the same row, same column, or same diagonal). There are many possible solutions. Write a program that displays one such solution. A sample output is shown below:

Figure 7.13 Each ball takes a random path and falls into a slot.

(a) (b) (c)

Balls are dropped from the opening of the board. Every time a ball hits a nail, it has a 50% chance of falling to the left or to the right. The piles of balls are accu- mulated in the slots at the bottom of the board.

Write a program that simulates the bean machine. Your program should prompt the user to enter the number of the balls and the number of the slots in the machine. Simulate the falling of each ball by printing its path. For example, the path for the ball in Figure 7.13b is LLRRLLR and the path for the ball in Figure 7.13c is RLRRLRR. Display the final buildup of the balls in the slots in a histogram. Here is a sample run of the program:

(Hint: Create an array named slots. Each element in slots stores the number of balls in a slot. Each ball falls into a slot via a path. The number of Rs in a path is the position of the slot where the ball falls. For example, for the path LRLR- LRR, the ball falls into slots[4], and for the path RRLLLLL, the ball falls into slots[2].)

|Q| | | | | | | | | | | | |Q| | | | | | | | | | | |Q| | | | | | |Q| | | | | |Q| | | | | | | | | | | | |Q| | | |Q| | | | | | | | | | |Q| | | | |

*** 7.37 (Game: bean machine) The bean machine, also known as a quincunx or the Gal- ton box, is a device for statistics experiments named after English scientist Sir Francis Galton. It consists of an upright board with evenly spaced nails (or pegs) in a triangular form, as shown in Figure 7.13.

M07_LIAN1878_11_GE_C07.indd 308 1/3/18 12:10 AM

Programming Exercises 309

Enter the number of balls to drop: 5

Enter the number of slots in the bean machine: 8

LRLRLRR

RRLLLRR

LLRLLRR

RRLLLLL

LRLRRLR

OOO

M07_LIAN1878_11_GE_C07.indd 309 1/3/18 12:10 AM

A01_GIPE3898_09_SE_FM.indd 3 12/20/17 5:56 PM

This page intentionally left blank

Objectives ■■ To give examples of representing data using two-dimensional arrays

(§8.1).

■■ To declare variables for two-dimensional arrays, create arrays, and access array elements in a two-dimensional array using row and column indices (§8.2).

■■ To program common operations for two-dimensional arrays ( displaying arrays, summing all elements, finding the minimum and maximum elements, and random shuffling) (§8.3).

■■ To pass two-dimensional arrays to methods (§8.4).

■■ To write a program for grading multiple-choice questions using two-dimensional arrays (§8.5).

■■ To solve the closest pair problem using two-dimensional arrays (§8.6).

■■ To check a Sudoku solution using two-dimensional arrays (§8.7).

■■ To use multidimensional arrays (§8.8).

Multidimensional Arrays

Chapter

M08_LIAN1878_11_GE_C08.indd 311 12/29/17 4:06 PM

312 Chapter 8 Multidimensional Arrays

8.1 Introduction Data in a table or a matrix can be represented using a two-dimensional array.

The preceding chapter introduced how to use one-dimensional arrays to store linear collections of elements. You can use a two-dimensional array to store a matrix or a table. For example, the following table that lists the distances between cities can be stored using a two-dimensional array named distances.

Point Key

problem

Distance Table (in miles)

Chicago Boston New York Atlanta Miami Dallas Houston

Chicago 0 983 787 714 1375 967 1087

Boston 983 0 214 1102 1763 1723 1842

New York 787 214 0 888 1549 1548 1627

Atlanta 714 1102 888 0 661 781 810

Miami 1375 1763 1549 661 0 1426 1187

Dallas 967 1723 1548 781 1426 0 239

Houston 1087 1842 1627 810 1187 239 0

double[][] distances = { {0, 983, 787, 714, 1375, 967, 1087}, {983, 0, 214, 1102, 1763, 1723, 1842}, {787, 214, 0, 888, 1549, 1548, 1627}, {714, 1102, 888, 0, 661, 781, 810}, {1375, 1763, 1549, 661, 0, 1426, 1187}, {967, 1723, 1548, 781, 1426, 0, 239}, {1087, 1842, 1627, 810, 1187, 239, 0}, };

8.2 Two-Dimensional Array Basics An element in a two-dimensional array is accessed through a row and a column index.

How do you declare a variable for two-dimensional arrays? How do you create a two- dimensional array? How do you access elements in a two-dimensional array? This section will address these issues.

8.2.1 Declaring Variables of Two-Dimensional Arrays and Creating Two-Dimensional Arrays

The syntax for declaring a two-dimensional array is as follows:

elementType[][] arrayRefVar;

elementType arrayRefVar[][]; // Allowed, but not preferred

As an example, here is how you would declare a two-dimensional array variable matrix of int values:

int[][] matrix;

Point Key

M08_LIAN1878_11_GE_C08.indd 312 12/29/17 4:06 PM

8.2 Two-Dimensional Array Basics 313

int matrix[][]; // This style is allowed, but not preferred

You can create a two-dimensional array of 5-by-5 int values and assign it to matrix using this syntax:

matrix = new int[5][5];

Two subscripts are used in a two-dimensional array: one for the row, and the other for the column. As in a one-dimensional array, the index for each subscript is of the int type and starts from 0, as shown in Figure 8.1a.

Figure 8.1 The index of each subscript of a two-dimensional array is an int value, starting from 0.

[4][3][2][1][0]

[4]

[3]

[2]

[1]

[0]

[4][3][2][1][0]

[4]

[3]

[2]

[1]

[0]

[2][1][0]

[3]

[2]

[1]

[0] 1

matrix = new int[5][5]; matrix[2][1] = 7;

2 3

4 5 6

7 8 9

10 11 12

0000 0

000 0

0000 0

0000 0 int[][] array = { {1, 2, 3}, {4, 5, 6}, {7, 8, 9}, {10, 11, 12} };

(a) (b) (c)

To assign the value 7 to a specific element at row index 2 and column index 1, as shown in Figure 8.1b, you can use the following syntax:

matrix[2][1] = 7;

Caution It is a common mistake to use matrix[2, 1] to access the element at row 2 and column 1. In Java, each subscript must be enclosed in a pair of square brackets.

You can also use an array initializer to declare, create, and initialize a two-dimensional array. For example, the following code in (a) creates an array with the specified initial values, as shown in Figure 8.1c. This is equivalent to the code in (b).

int[][] array = { {1, 2, 3}, {4, 5, 6}, {7, 8, 9}, {10, 11, 12} };

Equivalent

int[][] array = new int[4][3]; array[0][0] = 1; array[0][1] = 2; array[0][2] = 3; array[1][0] = 4; array[1][1] = 5; array[1][2] = 6; array[2][0] = 7; array[2][1] = 8; array[2][2] = 9; array[3][0] = 10; array[3][1] = 11; array[3][2] = 12;

(a) (b)

8.2.2 Obtaining the Lengths of Two-Dimensional Arrays A two-dimensional array is actually an array in which each element is a one-dimensional array. The length of an array x is the number of elements in the array, which can be obtained using x. length. x[0], x[1], . . . , and x[x.length − 1] are arrays. Their lengths can be obtained using x[0].length, x[1].length, . . . , and x[x.length − 1].length.

M08_LIAN1878_11_GE_C08.indd 313 12/29/17 4:06 PM

314 Chapter 8 Multidimensional Arrays

For example, suppose that x = new int[3][4], x[0], x[1], and x[2] are one- dimensional arrays and each contains four elements, as shown in Figure 8.2. x.length is 3, and x[0].length, x[1].length, and x[2].length are 4.

Figure 8.2 A two-dimensional array is a one-dimensional array in which each element is another one-dimensional array.

x[0]

x[1]

x[2]

x[0][0] x[0][1] x[0][3]

x[1][0] x[1][1] x[1][2] x[1][3]

x[2][0] x[2][1] x[2][2] x[2][3] x.length is 3

x[0].length is 4

x[1].length is 4

x[2].length is 4

x[0][2]

int[][] triangleArray = { {1, 2, 3, 4, 5}, {2, 3, 4, 5}, {3, 4, 5}, {4, 5}, {5} };

1 2 3 4 5

2 3 4 5

4 5

3 4 5

8.2.3 Ragged Arrays Each row in a two-dimensional array is itself an array. Thus, the rows can have different lengths. An array of this kind is known as a ragged array. Here is an example of creating a ragged array:

ragged array

As you can see, triangleArray[0].length is 5, triangleArray[1].length is 4, triangleArray[2].length is 3, triangleArray[3].length is 2, and triangle− Array[4].length is 1.

If you don’t know the values in a ragged array in advance, but do know the sizes—say, the same as in the preceding figure—you can create a ragged array using the following syntax:

int[][] triangleArray = new int[5][]; triangleArray[0] = new int[5]; triangleArray[1] = new int[4]; triangleArray[2] = new int[3]; triangleArray[3] = new int[2]; triangleArray[4] = new int[1];

You can now assign values to the array. For example,

triangleArray[0][3] = 4; triangleArray[4][0] = 5;

Note The syntax new int[5][] for creating an array requires the first index to be specified. The syntax new int[][] would be wrong.

M08_LIAN1878_11_GE_C08.indd 314 12/29/17 4:06 PM

8.3 Processing Two-Dimensional Arrays 315

8.2.1 Declare an array reference variable for a two-dimensional array of int values, create a 4-by-5 int matrix, and assign it to the variable.

8.2.2 Can the rows in a two-dimensional array have different lengths? 8.2.3 What is the output of the following code?

int[][] array = new int[5][6]; int[] x = {1, 2}; array[0] = x; System.out.println("array[0][1] is " + array[0][1]);

8.2.4 Which of the following statements are valid? int[][] r = new int[2]; int[] x = new int[]; int[][] y = new int[3][]; int[][] z = {{1, 2}}; int[][] m = {{1, 2}, {2, 3}}; int[][] n = {{1, 2}, {2, 3}, };

8.3 Processing Two-Dimensional Arrays Nested for loops are often used to process a two-dimensional array.

Suppose an array matrix is created as follows:

int[][] matrix = new int[10][10];

The following are some examples of processing two-dimensional arrays.

1. Initializing arrays with input values. The following loop initializes the array with user input values:

java.util.Scanner input = new java.util.Scanner(System.in); System.out.println("Enter " + matrix.length + " rows and " + matrix[0].length + " columns: "); for (int row = 0; row < matrix.length; row++) { for (int column = 0; column < matrix[row].length; column++) { matrix[row][column] = input.nextInt(); } }

2. Initializing arrays with random values. The following loop initializes the array with random values between 0 and 99:

for (int row = 0; row < matrix.length; row++) { for (int column = 0; column < matrix[row].length; column++) { matrix[row][column] = (int)(Math.random() * 100); } }

3. Printing arrays. To print a two-dimensional array, you have to print each element in the array using a loop like the following loop:

for (int row = 0; row < matrix.length; row++) { for (int column = 0; column < matrix[row].length; column++) { System.out.print(matrix[row][column] + " "); }

System.out.println(); }

Point Check

Point Key

M08_LIAN1878_11_GE_C08.indd 315 12/29/17 4:06 PM

316 Chapter 8 Multidimensional Arrays

4. Summing all elements. Use a variable named total to store the sum. Initially total is 0. Add each element in the array to total using a loop like this:

int total = 0; for (int row = 0; row < matrix.length; row++) { for (int column = 0; column < matrix[row].length; column++) { total += matrix[row][column]; } }

5. Summing elements by column. For each column, use a variable named total to store its sum. Add each element in the column to total using a loop like this:

for (int column = 0; column < matrix[0].length; column++) { int total = 0; for (int row = 0; row < matrix.length; row++) total += matrix[row][column]; System.out.println("Sum for column " + column + " is " + total); }

6. Which row has the largest sum? Use variables maxRow and indexOfMaxRow to track the largest sum and index of the row. For each row, compute its sum and update maxRow and indexOfMaxRow if the new sum is greater.

int maxRow = 0; int indexOfMaxRow = 0;

// Get sum of the first row in maxRow for (int column = 0; column < matrix[0].length; column++) { maxRow += matrix[0][column]; }

for (int row = 1; row < matrix.length; row++) { int totalOfThisRow = 0; for (int column = 0; column < matrix[row].length; column++) totalOfThisRow += matrix[row][column];

if (totalOfThisRow > maxRow) { maxRow = totalOfThisRow; indexOfMaxRow = row; } }

System.out.println("Row " + indexOfMaxRow + " has the maximum sum of " + maxRow);

7. Random shuffling. Shuffling the elements in a one-dimensional array was introduced in Section 7.2.6. How do you shuffle all the elements in a two-dimensional array? To accomplish this, for each element matrix[i][j], randomly generate indices i1 and j1 and swap matrix[i][j] with matrix[i1][j1], as follows:

for (int i = 0; i < matrix.length; i++) { for (int j = 0; j < matrix[i].length; j++) { int i1 = (int)(Math.random() * matrix.length); int j1 = (int)(Math.random() * matrix[i].length);

// Swap matrix[i][j] with matrix[i1][j1]

VideoNote

Find the row with the largest sum

M08_LIAN1878_11_GE_C08.indd 316 12/29/17 4:06 PM

8.4 Passing Two-Dimensional Arrays to Methods 317

int temp = matrix[i][j]; matrix[i][j] = matrix[i1][j1]; matrix[i1][j1] = temp; } }

8.3.1 Show the output of the following code: int[][] array = {{1, 2}, {3, 4}, {5, 6}}; for (int i = array.length − 1; i >= 0; i——) { for (int j = array[i].length − 1; j >= 0; j——) System.out.print(array[i][j] + " "); System.out.println(); }

8.3.2 Show the output of the following code: int[][] array = {{1, 2}, {3, 4}, {5, 6}}; int sum = 0; for (int i = 0; i < array.length; i++) sum += array[i][0]; System.out.println(sum);

8.4 Passing Two-Dimensional Arrays to Methods When passing a two-dimensional array to a method, the reference of the array is passed to the method.

You can pass a two-dimensional array to a method just as you pass a one-dimensional array. You can also return an array from a method. Listing 8.1 gives an example with two meth- ods. The first method, getArray(), returns a two-dimensional array and the second method, sum(int[][] m), returns the sum of all the elements in a matrix.

Listing 8.1 PassTwoDimensionalArray.java

1 import java.util.Scanner; 2 3 public class PassTwoDimensionalArray { 4 public static void main(String[] args) { 5 int[][] m = getArray(); // Get an array 6 7 // Display sum of elements 8 System.out.println("\nSum of all elements is " + sum(m)); 9 } 10 11 public static int[][] getArray() { 12 // Create a Scanner 13 Scanner input = new Scanner(System.in); 14 15 // Enter array values 16 int[][] m = new int[3][4]; 17 System.out.println("Enter " + m.length + " rows and " 18 + m[0].length + " columns: "); 19 for (int i = 0; i < m.length; i++) 20 for (int j = 0; j < m[i].length; j++) 21 m[i][j] = input.nextInt(); 22

Point Check

Point Key

get array

pass array

getArray method

M08_LIAN1878_11_GE_C08.indd 317 12/29/17 4:06 PM

318 Chapter 8 Multidimensional Arrays

23 return m; 24 } 25 26 public static int sum(int[][] m) { 27 int total = 0; 28 for (int row = 0; row < m.length; row++) { 29 for (int column = 0; column < m[row].length; column++) { 30 total += m[row][column]; 31 } 32 } 33 34 return total; 35 } 36 }

return array

sum method

Enter 3 rows and 4 columns:

1 2 3 4 5 6 7 8 9 10 11 12 Sum of all elements is 78

The method getArray prompts the user to enter values for the array (lines 11–24) and returns the array (line 23).

The method sum (lines 26–35) has a two-dimensional array argument. You can obtain the number of rows using m.length (line 28), and the number of columns in a specified row using m[row].length (line 29).

8.4.1 Show the output of the following code: public class Test { public static void main(String[] args) { int[][] array = {{1, 2, 3, 4}, {5, 6, 7, 8}}; System.out.println(m1(array)[0]); System.out.println(m1(array)[1]); }

public static int[] m1(int[][] m) { int[] result = new int[2]; result[0] = m.length; result[1] = m[0].length; return result; } }

8.5 Case Study: Grading a Multiple-Choice Test The problem is to write a program that will grade multiple-choice tests.

Suppose you need to write a program that grades multiple-choice tests. Assume there are eight students and ten questions, and the answers are stored in a two-dimensional array. Each row records a student’s answers to the questions, as shown in the following array:

Point Check

Point Key

VideoNote

Grade multiple-choice test

M08_LIAN1878_11_GE_C08.indd 318 12/29/17 4:06 PM

8.5 Case Study: Grading a Multiple-Choice Test 319

Students’ Answers to the Questions:

0 1 2 3 4 5 6 7 8 9

Student 0 A B A C C D E E A D Student 1 D B A B C A E E A D Student 2 E D D A C B E E A D Student 3 C B A E D C E E A D Student 4 A B D C C D E E A D Student 5 B B E C C D E E A D Student 6 B B A C C D E E A D Student 7 E B E C C D E E A D

The key is stored in a one-dimensional array:

Key to the Questions:

0 1 2 3 4 5 6 7 8 9

Key D B D C C D A E A D

Your program grades the test and displays the result. It compares each student’s answers with the key, counts the number of correct answers, and displays it. Listing 8.2 gives the program.

Listing 8.2 GradeExam.java 1 public class GradeExam { 2 /** Main method */ 3 public static void main(String[] args) { 4 // Students' answers to the questions 5 char[][] answers = { 6 {'A', 'B', 'A', 'C', 'C', 'D', 'E', 'E', 'A', 'D'}, 7 {'D', 'B', 'A', 'B', 'C', 'A', 'E', 'E', 'A', 'D'}, 8 {'E', 'D', 'D', 'A', 'C', 'B', 'E', 'E', 'A', 'D'}, 9 {'C', 'B', 'A', 'E', 'D', 'C', 'E', 'E', 'A', 'D'}, 10 {'A', 'B', 'D', 'C', 'C', 'D', 'E', 'E', 'A', 'D'}, 11 {'B', 'B', 'E', 'C', 'C', 'D', 'E', 'E', 'A', 'D'}, 12 {'B', 'B', 'A', 'C', 'C', 'D', 'E', 'E', 'A', 'D'}, 13 {'E', 'B', 'E', 'C', 'C', 'D', 'E', 'E', 'A', 'D'}}; 14 15 // Key to the questions 16 char[] keys = {'D', 'B', 'D', 'C', 'C', 'D', 'A', 'E', 'A', 'D'}; 17 18 // Grade all answers 19 for (int i = 0; i < answers.length; i++) { 20 // Grade one student 21 int correctCount = 0; 22 for (int j = 0; j < answers[i].length; j++) { 23 if (answers[i][j] == keys[j]) 24 correctCount++; 25 } 26 27 System.out.println("Student " + i + "'s correct count is " + 28 correctCount); 29 } 30 } 31 }

2-D array

1-D array

compare with key

M08_LIAN1878_11_GE_C08.indd 319 12/29/17 4:06 PM

320 Chapter 8 Multidimensional Arrays

The statement in lines 5–13 declares, creates, and initializes a two-dimensional array of characters and assigns the reference to answers of the char[][] type.

The statement in line 16 declares, creates, and initializes an array of char values and assigns the reference to keys of the char[] type.

Each row in the array answers stores a student’s answer, which is graded by comparing it with the key in the array keys. The result is displayed immediately after a student's answer is graded.

8.5.1 How do you modify the code so it also displays the highest count and the student with the highest count?

8.6 Case Study: Finding the Closest Pair This section presents a geometric problem for finding the closest pair of points.

Given a set of points, the closest-pair problem is to find the two points that are nearest to each other. In Figure 8.3, for example, points (1, 1) and (2, 0.5) are closest to each other. There are several ways to solve this problem. An intuitive approach is to compute the distances between all pairs of points and find the one with the minimum distance, as implemented in Listing 8.3.

Point Check

Point Key

closest-pair animation on the Companion Website

Figure 8.3 Points can be represented in a two-dimensional array.

(1, 1)

(–1, –1)

(–1, 3)

(2, 0.5)

(3, 3)

–1 3

x y

–1 –1 1 1 2 0.5 2 –1 3 3 4 2 4

0 1 2 3 4 5 6 7 –0.5

(4, 2)

(2, –1) (4, –0.5)

Student 0's correct count is 7

Student 1's correct count is 6

Student 2's correct count is 5

Student 3's correct count is 4

Student 4's correct count is 8

Student 5's correct count is 7

Student 6's correct count is 7

Student 7's correct count is 7

Listing 8.3 FindNearestPoints.java 1 import java.util.Scanner; 2 3 public class FindNearestPoints { 4 public static void main(String[] args) { 5 Scanner input = new Scanner(System.in); 6 System.out.print("Enter the number of points: "); 7 int numberOfPoints = input.nextInt(); 8 9 // Create an array to store points

number of points

M08_LIAN1878_11_GE_C08.indd 320 12/29/17 4:06 PM

8.6 Case Study: Finding the Closest Pair 321

10 double[][] points = new double[numberOfPoints][2]; 11 System.out.print("Enter " + numberOfPoints + " points: "); 12 for (int i = 0; i < points.length; i++) { 13 points[i][0] = input.nextDouble(); 14 points[i][1] = input.nextDouble(); 15 } 16 17 // p1 and p2 are the indices in the points' array 18 int p1 = 0, p2 = 1; // Initial two points 19 double shortestDistance = distance(points[p1][0], points[p1][1], 20 points[p2][0], points[p2][1]); // Initialize shortestDistance 21 22 // Compute distance for every two points 23 for (int i = 0; i < points.length; i++) { 24 for (int j = i + 1; j < points.length; j++) { 25 double distance = distance(points[i][0], points[i][1], 26 points[j][0], points[j][1]); // Find distance 27 28 if (shortestDistance > distance) { 29 p1 = i; // Update p1 30 p2 = j; // Update p2 31 shortestDistance = distance; // Update shortestDistance 32 } 33 } 34 } 35 36 // Display result 37 System.out.println("The closest two points are " + 38 "(" + points[p1][0] + ", " + points[p1][1] + ") and (" + 39 points[p2][0] + ", " + points[p2][1] + ")"); 40 } 41 42 /** Compute the distance between two points (x1, y1) and (x2, y2)*/ 43 public static double distance( 44 double x1, double y1, double x2, double y2) { 45 return Math.sqrt((x2 − x1) * (x2 − x1) + (y2 − y1) * (y2 − y1)); 46 } 47 }

read points

2-D array

track shortestDistance

for each point i for each point j

track two points

distance between two points

update shortestDistance

distance between i and j

Enter the number of points: 8

Enter 8 points: −1 3 −1 −1 1 1 2 0.5 2 −1 3 3 4 2 4 −0.5

The closest two points are (1, 1) and (2, 0.5)

The program prompts the user to enter the number of points (lines 6 and 7). The points are read from the console and stored in a two-dimensional array named points (lines 12–15). The program uses the variable shortestDistance (line 19) to store the distance between the two nearest points, and the indices of these two points in the points array are stored in p1 and p2 (line 18).

For each point at index i, the program computes the distance between points[i] and points[j] for all j > i (lines 23–34). Whenever a shorter distance is found, the variable shortestDistance and p1 and p2 are updated (lines 28–32).

The distance between two points (x1, y1) and (x2, y2) can be computed using the formula 2(x2 - x1)2 + (y2 - y1)2 (lines 43–46).

The program assumes the plane has at least two points. You can easily modify the program to handle the case if the plane has zero or one point.

M08_LIAN1878_11_GE_C08.indd 321 12/29/17 4:06 PM

322 Chapter 8 Multidimensional Arrays

Note that there might be more than one closest pair of points with the same minimum dis- tance. The program finds one such pair. You may modify the program to find all closest pairs in Programming Exercise 8.8.

Tip It is cumbersome to enter all points from the keyboard. You may store the input in a file, say FindNearestPoints.txt, and run the program using the following command:

java FindNearestPoints < FindNearestPoints.txt

8.6.1 What happens if the input has only one point?

8.7 Case Study: Sudoku The problem is to check whether a given Sudoku solution is correct.

This section presents an interesting problem of a sort that appears in the newspaper every day. It is a number-placement puzzle, commonly known as Sudoku. This is a very challeng- ing problem. To make it accessible to the novice, this section presents a simplified version of the Sudoku problem, which is to verify whether a Sudoku solution is correct. The complete program for finding a Sudoku solution is presented in Supplement VI.C.

Sudoku is a 9 * 9 grid divided into smaller 3 * 3 boxes (also called regions or blocks), as shown in Figure 8.4a. Some cells, called fixed cells, are populated with numbers from 1 to 9. The objective is to fill the empty cells, also called free cells, with the numbers 1 to 9 so every row, every column, and every 3 * 3 box contains the numbers 1 to 9, as shown in Figure 8.4b.

multiple closest pairs

input file

Point Check

Point Key

VideoNote

Sudoku

fixed cells

free cells

Figure 8.4 The Sudoku puzzle in (a) is solved in (b).

5 3 7

6 1 9 5

9 8 6

8 6 3

4 8 3 1

7 2 6

4 1 9 5

8 7 9

(a) Puzzle

Solution

(b) Solution

5 3 4 6 7 8 9 1 2

6 7 2 1 9 5 3 4 8

1 9 8 3 4 2 5 6 7

8 5 9 7 6 1 4 2 3

4 2 6 8 5 3 7 9 1

7 1 3 9 2 4 8 5 6

9 6 1 5 3 7 2 8 4

2 8 7 4 1 9 6 3 5

3 4 5 2 8 6 1 7 9

Figure 8.5 A grid can be represented using a two-dimensional array.

5 3 0 0 0 0 0 0

0 0

00 0

0 0

0 00

0 080 00

0 00

0 06

6 0 0

0 0

0 0 0

1 9 5

9 8 6

8 6 3

4 8 3 1

7 2 6

4 1 9 5

7 9

(a) (b)

int[][] grid = {{5, 3, 0, 0, 7, 0, 0, 0, 0}, {6, 0, 0, 1, 9, 5, 0, 0, 0}, {0, 9, 8, 0, 0, 0, 0, 6, 0}, {8, 0, 0, 0, 6, 0, 0, 0, 3}, {4, 0, 0, 8, 0, 3, 0, 0, 1}, {7, 0, 0, 0, 2, 0, 0, 0, 6}, {0, 6, 0, 0, 0, 0, 2, 8, 0}, {0, 0, 0, 4, 1, 9, 0, 0, 5}, {0, 0, 0, 0, 8, 0, 0, 7, 9} };

For convenience, we use value 0 to indicate a free cell, as shown in Figure 8.5a. The grid can be naturally represented using a two-dimensional array, as shown in Figure 8.5b.representing a grid

M08_LIAN1878_11_GE_C08.indd 322 12/29/17 4:06 PM

8.7 Case Study: Sudoku 323

To find a solution for the puzzle, we must replace each 0 in the grid with an appropriate number from 1 to 9. For the solution to the puzzle in Figure 8.5, the grid should be as shown in Figure 8.6.

Once a solution to a Sudoku puzzle is found, how do you verify that it is correct? Here are two approaches:

1. Check if every row has numbers from 1 to 9, every column has numbers from 1 to 9, and every small box has numbers from 1 to 9.

2. Check each cell. Each cell must be a number from 1 to 9 and the cell must be unique on every row, every column, and every small box.

Figure 8.6 A solution is stored in grid.

A solution grid is {{5, 3, 4, 6, 7, 8, 9, 1, 2}, {6, 7, 2, 1, 9, 5, 3, 4, 8}, {1, 9, 8, 3, 4, 2, 5, 6, 7}, {8, 5, 9, 7, 6, 1, 4, 2, 3}, {4, 2, 6, 8, 5, 3, 7, 9, 1}, {7, 1, 3, 9, 2, 4, 8, 5, 6}, {9, 6, 1, 5, 3, 7, 2, 8, 4}, {2, 8, 7, 4, 1, 9, 6, 3, 5}, {3, 4, 5, 2, 8, 6, 1, 7, 9} };

The program in Listing 8.4 prompts the user to enter a solution and reports whether it is valid. We use the second approach in the program to check whether the solution is correct.

Listing 8.4 CheckSudokuSolution.java 1 import java.util.Scanner; 2 3 public class CheckSudokuSolution { 4 public static void main(String[] args) { 5 // Read a Sudoku solution 6 int[][] grid = readASolution(); 7 8 System.out.println(isValid(grid) ? "Valid solution" : 9 "Invalid solution"); 10 } 11 12 /** Read a Sudoku solution from the console */ 13 public static int[][] readASolution() { 14 // Create a Scanner 15 Scanner input = new Scanner(System.in); 16 17 System.out.println("Enter a Sudoku puzzle solution:"); 18 int[][] grid = new int[9][9]; 19 for (int i = 0; i < 9; i++) 20 for (int j = 0; j < 9; j++) 21 grid[i][j] = input.nextInt(); 22 23 return grid; 24 } 25 26 /** Check whether a solution is valid */ 27 public static boolean isValid(int[][] grid) {

read input

solution valid?

read solution

check solution

M08_LIAN1878_11_GE_C08.indd 323 12/29/17 4:06 PM

324 Chapter 8 Multidimensional Arrays

28 for (int i = 0; i < 9; i++) 29 for (int j = 0; j < 9; j++) 30 if (grid[i][j] < 1 || grid[i][j] > 9 31 || !isValid(i, j, grid)) 32 return false; 33 return true; // The solution is valid 34 } 35 36 /** Check whether grid[i][j] is valid in the grid */ 37 public static boolean isValid(int i, int j, int[][] grid) { 38 // Check whether grid[i][j] is unique in i's row 39 for (int column = 0; column < 9; column++) 40 if (column != j && grid[i][column] == grid[i][j]) 41 return false; 42 43 // Check whether grid[i][j] is unique in j's column 44 for (int row = 0; row < 9; row++) 45 if (row != i && grid[row][j] == grid[i][j]) 46 return false; 47 48 // Check whether grid[i][j] is unique in the 3−by−3 box 49 for (int row = (i / 3) * 3; row < (i / 3) * 3 + 3; row++) 50 for (int col = (j / 3) * 3; col < (j / 3) * 3 + 3; col++) 51 if (!(row == i && col == j) && grid[row][col] == grid[i][j]) 52 return false; 53 54 return true; // The current value at grid[i][j] is valid 55 } 56 }

check small boxes

check columns

check rows

Enter a Sudoku puzzle solution:

9 6 3 1 7 4 2 5 8 1 7 8 3 2 5 6 4 9 2 5 4 6 8 9 7 3 1

8 2 1 4 3 7 5 9 6

4 9 6 8 5 2 3 1 7

7 3 5 9 6 1 8 2 4

5 8 9 7 1 3 4 6 2 3 1 7 2 4 6 9 8 5 6 4 2 5 9 8 1 7 3 Valid solution

The program invokes the readASolution() method (line 6) to read a Sudoku solution and return a two-dimensional array representing a Sudoku grid.

The isValid(grid) method checks whether the values in the grid are valid by verifying that each value is between 1 and 9, and that each value is valid in the grid (lines 27–34).

The isValid(i, j, grid) method checks whether the value at grid[i][j] is valid. It checks whether grid[i][j] appears more than once in row i (lines 39–41), in column j (lines 44–46), and in the 3 * 3 box (lines 49–52).

How do you locate all the cells in the same box? For any grid[i][j], the starting cell of the 3 * 3 box that contains it is grid[(i / 3) * 3][(j / 3) * 3], as illustrated in Figure 8.7.

isValid method

overloaded isValid method

M08_LIAN1878_11_GE_C08.indd 324 12/29/17 4:06 PM

8.8 Multidimensional Arrays 325

With this observation, you can easily identify all the cells in the box. For instance, if grid[r][c] is the starting cell of a 3 * 3 box, the cells in the box can be traversed in a nested loop as follows:

// Get all cells in a 3−by−3 box starting at grid[r][c] for (int row = r; row < r + 3; row++) for (int col = c; col < c + 3; col++) // grid[row][col] is in the box

It is cumbersome to enter 81 numbers from the console. When you test the program, you may store the input in a file, say CheckSudokuSolution.txt (see liveexample.pearsoncmg. com/data/CheckSudokuSolution.txt) and run the program using the following command:

java CheckSudokuSolution < CheckSudokuSolution.txt

8.7.1 What happens if the code in line 51 in Listing 8.4 is changed to if (row != i && col != j && grid[row][col] == grid[i][j])

8.8 Multidimensional Arrays A two-dimensional array is an array of one-dimensional arrays, and a three-dimen- sional array is an array of two-dimensional arrays.

In the preceding section, you used a two-dimensional array to represent a matrix or a table. Occasionally, you will need to represent n-dimensional data structures. In Java, you can create n-dimensional arrays for any positive integer n.

The way to declare two-dimensional array variables and create two-dimensional arrays can be generalized to declare n-dimensional array variables and create n-dimensional arrays for n 7 = 3. For example, you may use a three-dimensional array to store exam scores for a class of six students with five exams, and each exam has two parts (multiple-choice and essay type questions). The following syntax declares a three-dimensional array variable scores, creates an array, and assigns its reference to scores.

double[][][] scores = new double[6][5][2];

You can also use the array initializer to create and initialize the array as follows:

double[][][] scores = { {{7.5, 20.5}, {9.0, 22.5}, {15, 33.5}, {13, 21.5}, {15, 2.5}}, {{4.5, 21.5}, {9.0, 22.5}, {15, 34.5}, {12, 20.5}, {14, 9.5}}, {{6.5, 30.5}, {9.4, 10.5}, {11, 33.5}, {11, 23.5}, {10, 2.5}}, {{6.5, 23.5}, {9.4, 32.5}, {13, 34.5}, {11, 20.5}, {16, 7.5}}, {{8.5, 26.5}, {9.4, 52.5}, {13, 36.5}, {13, 24.5}, {16, 2.5}}, {{9.5, 20.5}, {9.4, 42.5}, {13, 31.5}, {12, 20.5}, {16, 6.5}}};

input file

Point Check

Point Key

Figure 8.7 The location of the first cell in a 3 * 3 box determines the locations of other cells in the box.

grid[0][6]

grid[6][3]

For any grid[i][j] in this 3 by 3 box, its starting cell is grid[3×(i/3)] [3×(j/3)](i.e., grid[6][3]). For example, for grid[8][5], i=8 and j=5, 3×(i/3)=6 and 3×(j/3)=3.

For any grid[i][j] in this 3 by 3 box, its starting cell is grid[3×(i/3)][3×(j/3)] (i.e., grid[0][6]). For example, for grid[2][8], i=2 and j=8, 3×(i/3)=0 and 3×(j/3)=6.

grid[0][0]

M08_LIAN1878_11_GE_C08.indd 325 12/29/17 4:06 PM

326 Chapter 8 Multidimensional Arrays

scores[0][1][0] refers to the multiple-choice score for the first student’s second exam, which is 9.0. scores[0][1][1] refers to the essay score for the first student’s second exam, which is 22.5. This is depicted in the following figure:

scores [i] [j] [k]

Which student Which exam Multiple-choice or essay score

A multidimensional array is actually an array in which each element is another array. A three- dimensional array is an array of two-dimensional arrays. A two-dimensional array is an array of one-dimensional arrays. For example, suppose that x = new int[2][2][5] and x[0] and x[1] are two-dimensional arrays. x[0][0], x[0][1], x[1][0], and x[1][1] are one- dimensional arrays and each contains five elements. x.length is 2, x[0].length and x[1]. length are 2, and x[0][0].length, x[0][1].length, x[1][0].length, and x[1][1]. length are 5.

8.8.1 Case Study: Daily Temperature and Humidity Suppose a meteorology station records the temperature and humidity every hour of every day, and stores the data for the past 10 days in a text file named Weather.txt (see liveexample .pearsoncmg.com/data/Weather.txt). Each line of the file consists of four numbers that indi- cate the day, hour, temperature, and humidity. The contents of the file may look like those in (a).

Day Hour

Temperature Humidity

Day Hour

Temperature Humidity

1 1 76.4 0.92 10 24 98.7 0.74 1 2 77.7 0.93 1 2 77.7 0.93

... ... 10 23 97.7 0.71 10 23 97.7 0.71 10 24 98.7 0.74 1 1 76.4 0.92

(a) (b)

Note the lines in the file are not necessarily in increasing order of day and hour. For example, the file may appear as shown in (b).

Your task is to write a program that calculates the average daily temperature and humid- ity for the 10 days. You can use the input redirection to read the file and store the data in a three-dimensional array named data. The first index of data ranges from 0 to 9 and represents 10 days, the second index ranges from 0 to 23 and represents 24 hours, and the third index ranges from 0 to 1 and represents temperature and humidity, as depicted in the following figure:

Which day Which hour Temperature or humidity

data [ i ] [ j ] [ k ]

M08_LIAN1878_11_GE_C08.indd 326 12/29/17 4:06 PM

8.8 Multidimensional Arrays 327

Note the days are numbered from 1 to 10 and the hours from 1 to 24 in the file. Because the array index starts from 0, data[0][0][0] stores the temperature in day 1 at hour 1 and data[9][23][1] stores the humidity in day 10 at hour 24.

The program is given in Listing 8.5.

Listing 8.5 Weather.java 1 import java.util.Scanner; 2 3 public class Weather { 4 public static void main(String[] args) { 5 final int NUMBER_OF_DAYS = 10; 6 final int NUMBER_OF_HOURS = 24; 7 double[][][] data 8 = new double[NUMBER_OF_DAYS][NUMBER_OF_HOURS][2]; 9 10 Scanner input = new Scanner(System.in); 11 // Read input using input redirection from a file 12 for (int k = 0; k < NUMBER_OF_DAYS * NUMBER_OF_HOURS; k++) { 13 int day = input.nextInt(); 14 int hour = input.nextInt(); 15 double temperature = input.nextDouble(); 16 double humidity = input.nextDouble(); 17 data[day − 1][hour − 1][0] = temperature; 18 data[day − 1][hour − 1][1] = humidity; 19 } 20 21 // Find the average daily temperature and humidity 22 for (int i = 0; i < NUMBER_OF_DAYS; i++) { 23 double dailyTemperatureTotal = 0, dailyHumidityTotal = 0; 24 for (int j = 0; j < NUMBER_OF_HOURS; j++) { 25 dailyTemperatureTotal += data[i][j][0]; 26 dailyHumidityTotal += data[i][j][1]; 27 } 28 29 // Display result 30 System.out.println("Day " + i + "'s average temperature is " 31 + dailyTemperatureTotal / NUMBER_OF_HOURS); 32 System.out.println("Day " + i + "'s average humidity is " 33 + dailyHumidityTotal / NUMBER_OF_HOURS); 34 } 35 } 36 }

three-dimensional array

Day 0's average temperature is 77.7708 Day 0's average humidity is 0.929583 Day 1's average temperature is 77.3125 Day 1's average humidity is 0.929583 ... Day 9's average temperature is 79.3542 Day 9's average humidity is 0.9125

You can use the following command to run the program:

java Weather < Weather.txt

A three-dimensional array for storing temperature and humidity is created in line 8. The loop in lines 12–19 reads the input to the array. You can enter the input from the keyboard, but

M08_LIAN1878_11_GE_C08.indd 327 12/29/17 4:06 PM

328 Chapter 8 Multidimensional Arrays

doing so will be awkward. For convenience, we store the data in a file and use input redirection to read the data from the file. The loop in lines 24–27 adds all temperatures for each hour in a day to dailyTemperatureTotal, and all humidity for each hour to dailyHumidityTotal. The average daily temperature and humidity are displayed in lines 30–33.

8.8.2 Case Study: Guessing Birthdays Listing 4.3, GuessBirthday.java, gives a program that guesses a birthday. The program can be simplified by storing the numbers in five sets in a three-dimensional array and it prompts the user for the answers using a loop, as given in Listing 8.6. The sample run of the program can be the same as given in Listing 4.3.

Listing 8.6 GuessBirthdayUsingArray.java 1 import java.util.Scanner; 2 3 public class GuessBirthdayUsingArray { 4 public static void main(String[] args) { 5 int day = 0; // Day to be determined 6 int answer; 7 8 int[][][] dates = { 9 {{ 1, 3, 5, 7}, 10 { 9, 11, 13, 15}, 11 {17, 19, 21, 23}, 12 {25, 27, 29, 31}}, 13 {{ 2, 3, 6, 7}, 14 {10, 11, 14, 15}, 15 {18, 19, 22, 23}, 16 {26, 27, 30, 31}}, 17 {{ 4, 5, 6, 7}, 18 {12, 13, 14, 15}}, 19 {20, 21, 22, 23}, 20 {28, 29, 30, 31}}, 21 {{ 8, 9, 10, 11}, 22 {12, 13, 14, 15}, 23 {24, 25, 26, 27}, 24 {28, 29, 30, 31}}, 25 {{16, 17, 18, 19}, 26 {20, 21, 22, 23}, 27 {24, 25, 26, 27}, 28 {28, 29, 30, 31}}}; 29 30 // Create a Scanner 31 Scanner input = new Scanner(System.in); 32 33 for (int i = 0; i < 5; i++) { 34 System.out.println("Is your birthday in Set" + (i + 1) + "?"); 35 for (int j = 0; j < 4; j++) { 36 for (int k = 0; k < 4; k++) 37 System.out.printf("%4d", dates[i][j][k]); 38 System.out.println(); 39 } 40 41 System.out.print("\nEnter 0 for No and 1 for Yes: "); 42 answer = input.nextInt(); 43 44 if (answer == 1) 45 day += dates[i][0][0];

three-dimensional array

Set i

add to day

M08_LIAN1878_11_GE_C08.indd 328 12/29/17 4:06 PM

Programming Exercises 329

46 } 47 48 System.out.println("Your birthday is " + day); 49 } 50 }

A three-dimensional array dates is created in lines 8–28. This array stores five sets of numbers. Each set is a 4-by-4 two-dimensional array.

The loop starting from line 33 displays the numbers in each set and prompts the user to answer whether the birthday is in the set (lines 41 and 42). If the day is in the set, the first number (dates[i][0][0]) in the set is added to variable day (line 45).

8.8.1 Declare an array variable for a three-dimensional array, create a 4 * 6 * 5 int array, and assign its reference to the variable.

8.8.2 Assume char[][][] x = new char[12][5][2], how many elements are in the array? What are x.length, x[2].length, and x[0][0].length?

8.8.3 Show the output of the following code: int[][][] array = {{{1, 2}, {3, 4}}, {{5, 6},{7, 8}}}; System.out.println(array[0][0][0]); System.out.println(array[1][1][1]);

ChapTer summary 1. A two-dimensional array can be used to store a table.

2. A variable for two-dimensional arrays can be declared using the syntax: elementType[][] arrayVar.

3. A two-dimensional array can be created using the syntax: new elementType[ROW_SIZE] [COLUMN_SIZE].

4. Each element in a two-dimensional array is represented using the syntax: arrayVar[rowIndex][columnIndex].

5. You can create and initialize a two-dimensional array using an array initializer with the syntax: elementType[][] arrayVar = {{row values}, . . ., {row values}}.

6. You can use arrays of arrays to form multidimensional arrays. For example, a variable for three-dimensional arrays can be declared as elementType[][][] arrayVar and a three-dimensional array can be created using new elementType[size1][size2] [size3].

Quiz Answer the quiz for this chapter online at the book Companion Website.

programming exerCises *8.1 (Sum elements row by row) Write a method that returns the sum of all the ele-

ments in a specified row in a matrix using the following header:

public static double sumRow(double[][] m, int rowIndex)

Point Check

M08_LIAN1878_11_GE_C08.indd 329 12/29/17 4:06 PM

330 Chapter 8 Multidimensional Arrays

Write a test program that reads a 3-by-4 matrix and displays the sum of each row. Here is a sample run:

Enter a 3−by−4 matrix row by row:

1.5 2 3 4

5.5 6 7 8

9.5 1 3 1 Sum of the elements at row 0 is 10.5 Sum of the elements at row 1 is 26.5 Sum of the elements at row 2 is 14.5

Enter a 4−by−4 matrix row by row:

1 2 3 4.0 5 6.5 7 8 9 10 11 12 13 14 15 16 Average of the elements in the major diagonal is 8.625

*8.2 (Average the major diagonal in a matrix) Write a method that averages all the numbers in the major diagonal in an n * n matrix of double values using the following header:

public static double averageMajorDiagonal(double[][] m)

Write a test program that reads a 4-by-4 matrix and displays the average of all its elements on the major diagonal. Here is a sample run:

*8.3 (Sort students on grades) Rewrite Listing 8.2, GradeExam.java, to display stu- dents in decreasing order of the number of correct answers.

**8.4 (Compute the weekly hours for each employee) Suppose the weekly hours for all employees are stored in a two-dimensional array. Each row records an employee’s seven-day work hours with seven columns. For example, the fol- lowing array stores the work hours for eight employees. Write a program that displays employees and their total hours in increasing order of the total hours.

Su M T W Th F Sa

Employee 0 2 4 3 4 5 8 8

Employee 1 7 3 4 3 3 4 4

Employee 2 3 3 4 3 3 2 2

Employee 3 9 3 4 7 3 4 1

Employee 4 3 5 4 3 6 3 8

Employee 5 3 4 4 6 3 4 4

Employee 6 3 7 4 8 3 8 4

Employee 7 6 3 5 9 2 7 9

8.5 (Algebra: add two matrices) Write a method to add two matrices. The header of the method is as follows:

public static double[][] addMatrix(double[][] a, double[][] b)

M08_LIAN1878_11_GE_C08.indd 330 12/29/17 4:06 PM

Programming Exercises 331

In order to be added, the two matrices must have the same dimensions and the same or compatible types of elements. Let c be the resulting matrix. Each ele- ment cij is aij + bij. For example, for two 2 * 2 matrices a and b, c is

aa11 a12 a21 a22

b + ab11 b12 b21 b22

b = aa11 + b11 a12 + b12 a21 + b21 a22 + b22

Write a test program that prompts the user to enter two 2 * 2 matrices and displays their sum. Here is a sample run:

Enter matrix1: 1 2 3 4

Enter matrix2: 0 2 4 1

The matrices are added as follows

1.0 2.0 0.0 2.0 1.0 4.0

3.0 4.0 + 4.0 1.0 = 7.0 5.0

**8.6 (Algebra: multiply two matrices) Write a method to multiply two matrices. The header of the method is:

public static double[][] multiplyMatrix(double[][] a, double[][] b)

To multiply matrix a by matrix b, the number of columns in a must be the same as the number of rows in b, and the two matrices must have elements of the same or compatible types. Let c be the result of the multiplication. Assume the column size of matrix a is n. Each element cij is ai1 * b1j + ai2 * b2j + g + ain * bnj. For example, for two 3 * 3 matrices a and b, c is£a11 a12 a13a21 a22 a23

a31 a32 a33

≥ * £b11 b12 b13b21 b22 b23 b31 b32 b33

≥ = £c11 c12 c13c21 c22 c23 c31 c32 c33

≥ where cij = ai1 * b1j + ai2 * b2j + ai3 * b3j. Write a test program that prompts the user to enter two 3 * 3 matrices and

displays their product. Here is a sample run:

VideoNote

Multiply two matrices

Enter matrix1: 1 2 3 4 5 6 7 8 9

Enter matrix2: 0 2 4 1 4.5 2.2 1.1 4.3 5.2

The multiplication of the matrices is

1 2 3 0 2.0 4.0 5.3 23.9 24

4 5 6 * 1 4.5 2.2 = 11.6 56.3 58.2

7 8 9 1.1 4.3 5.2 17.9 88.7 92.4

*8.7 (Points nearest to each other) Listing 8.3 gives a program that finds two points in a two-dimensional space nearest to each other. Revise the program so it finds two points in a three-dimensional space nearest to each other. Use a two-dimensional array to represent the points. Test the program using the following points:

double[][] points = {{−1, 0, 3}, {−1, −1, −1}, {4, 1, 1}, {2, 0.5, 9}, {3.5, 2, −1}, {3, 1.5, 3}, {−1.5, 4, 2}, {5.5, 4, −0.5}};

M08_LIAN1878_11_GE_C08.indd 331 12/29/17 4:06 PM

332 Chapter 8 Multidimensional Arrays

The formula for computing the distance between two points (x1, y1, z1) and (x2, y2, z2) is 2(x2 - x1)2 + (y2 - y1)2 + (z2 - z1)2.

**8.8 (All closest pairs of points) Revise Listing 8.3, FindNearestPoints.java, to display all closest pairs of points with the same minimum distance. Here is a sample run:

Enter the number of points: 8 Enter 8 points: 0 0 1 1 −1 −1 2 2 −2 −2 −3 −3 −4 −4 5 5 The closest two points are (0.0, 0.0) and (1.0, 1.0) The closest two points are (0.0, 0.0) and (−1.0, −1.0) The closest two points are (1.0, 1.0) and (2.0, 2.0) The closest two points are (−1.0, −1.0) and (−2.0, −2.0) The closest two points are (−2.0, −2.0) and (−3.0, −3.0) The closest two points are (−3.0, −3.0) and (−4.0, −4.0) Their distance is 1.4142135623730951

Enter a column (0, 1, or 2) for player X: 1

| | | | | | X | | | | | | Enter a row (0, 1, or 2) for player O: 1

Enter a column (0, 1, or 2) for player O: 2

| | | | | | X | O | | | | |

***8.9 (Game: play a tic-tac-toe game) In a game of tic-tac-toe, two players take turns marking an available cell in a 3 * 3 grid with their respective tokens (either X or O). When one player has placed three tokens in a horizontal, vertical, or diago- nal row on the grid, the game is over and that player has won. A draw (no winner) occurs when all the cells on the grid have been filled with tokens and neither player has achieved a win. Create a program for playing a tic-tac-toe game.

The program prompts two players to alternately enter an X token and O token. Whenever a token is entered, the program redisplays the board on the console and determines the status of the game (win, draw, or continue). Here is a sample run:

M08_LIAN1878_11_GE_C08.indd 332 12/29/17 4:06 PM

Programming Exercises 333

*8.10 (Largest row and column) Write a program that randomly fills in 0s and 1s into a 5-by-5 matrix, prints the matrix, and finds the first row and column with the most 1s. Here is a sample run of the program:

Enter a row (0, 1, or 2) for player X:

. . .

| X | | | | O | X | O | | | | X | X player won

Enter a number between 0 and 511: 7

H H H

T T T

01101 01011 10010 11111 00101 The largest row index: 3 The largest column index: 4

The user entered 7, which corresponds to 000000111. Since 0 stands for H and 1 for T, the output is correct.

** 8.12 (Financial application: compute tax) Rewrite Listing 3.5, ComputeTax.java, using arrays. For each filing status, there are six tax rates. Each rate is applied to a certain amount of taxable income. For example, from the taxable income of $400,000 for a single filer, $8,350 is taxed at 10%, (33,950 - 8,350) at 15%,

**8.11 (Game: nine heads and tails) Nine coins are placed in a 3-by-3 matrix with some face up and some face down. You can represent the state of the coins using a 3-by-3 matrix with values 0 (heads) and 1 (tails). Here are some examples:

0 0 0 1 0 1 1 1 0 1 0 1 1 0 0 0 1 0 0 0 1 1 0 0 1 1 0 1 1 1 0 0 0 1 0 0 0 0 1 1 0 0 1 1 0

Each state can also be represented using a binary number. For example, the pre- ceding matrices correspond to the numbers

000010000 101001100 110100001 101110100 100111110

There are a total of 512 possibilities, so you can use decimal numbers 0, 1, 2, 3, . . . , and 511 to represent all states of the matrix. Write a program that prompts the user to enter a number between 0 and 511 and displays the corresponding matrix with the characters H and T. Here is a sample run:

M08_LIAN1878_11_GE_C08.indd 333 12/29/17 4:06 PM

334 Chapter 8 Multidimensional Arrays

(82,250 - 33,950) at 25%, (171,550 - 82,550) at 28%, (372,550 - 82,250) at 33%, and (400,000 - 372,950) at 36%. The six rates are the same for all filing statuses, which can be represented in the following array:

double[] rates = {0.10, 0.15, 0.25, 0.28, 0.33, 0.35};

The brackets for each rate for all the filing statuses can be represented in a two- dimensional array as follows:

int[][] brackets = { {8350, 33950, 82250, 171550, 372950}, // Single filer {16700, 67900, 137050, 20885, 372950}, // Married jointly // −or qualifying widow(er) {8350, 33950, 68525, 104425, 186475}, // Married separately {11950, 45500, 117450, 190200, 372950} // Head of household

};

Suppose the taxable income is $400,000 for single filers. The tax can be com- puted as follows:

tax = brackets[0][0] * rates[0] + (brackets[0][1] – brackets[0][0]) * rates[1] + (brackets[0][2] – brackets[0][1]) * rates[2] + (brackets[0][3] – brackets[0][2]) * rates[3] + (brackets[0][4] – brackets[0][3]) * rates[4] + (400000 – brackets[0][4]) * rates[5];

*8.13 (Locate the smallest element) Write the following method that returns the location of the smallest element in a two-dimensional array.

public static int[] locateSmallest(double[][] a)

The return value is a one-dimensional array that contains two elements. These two elements indicate the row and column indices of the smallest element in the two-dimensional array. Write a test program that prompts the user to enter a two- dimensional array and displays the location of the smallest element in the array. Here is a sample run:

Enter the number of rows and columns of the array: 3 4

Enter the array:

23.5 35 2 10 4.5 3 45 3.5

35 44 5.5 9.6

The location of the smallest element is at (0, 2)

** 8.14 (Explore matrix) Write a program that prompts the user to enter the length of a square matrix, randomly fills in 0s and 1s into the matrix, prints the matrix, and finds the rows, columns, and diagonals with all 0s or 1s. Here is a sample run of the program:

M08_LIAN1878_11_GE_C08.indd 334 12/29/17 4:06 PM

Programming Exercises 335

*8.15 (Geometry: same line?) Programming Exercise 6.39 gives a method for testing whether three points are on the same line.

Write the following method to test whether all the points in the array points are on the same line:

public static boolean sameLine(double[][] points)

Write a program that prompts the user to enter five points and displays whether they are on the same line. Here are sample runs:

Enter the size for the matrix: 4

0111

0000

0100

1111

All 0s on row 2

All 1s on row 4

No same numbers on a column

No same numbers on the major diagonal

No same numbers on the sub−diagonal

Enter five points: 3.4 2 6.5 9.5 2.3 2.3 5.5 5 −5 4

The five points are not on the same line

Enter five points: 1 1 2 2 3 3 4 4 5 5

The five points are on the same line

*8.16 (Sort two-dimensional array) Write a method to sort a two-dimensional array using the following header:

public static void sort(int m[][])

The method performs a primary sort on rows, and a secondary sort on columns. For example, the following array

{{4, 2},{1, 7},{4, 5},{1, 2},{1, 1},{4, 1}}

will be sorted to

{{4, 5},{4, 2},{4, 1},{1, 7},{1, 2},{1, 1}}.

***8.17 (Financial tsunami) Banks lend money to each other. In tough economic times, if a bank goes bankrupt, it may not be able to pay back the loan. A bank’s total assets are its current balance plus its loans to other banks. The diagram in Figure 8.8 shows five banks. The banks’ current balances are 25, 125, 175, 75, and 181 million dol- lars, respectively. The directed edge from node 1 to node 2 indicates that bank 1 lends 40 million dollars to bank 2.

M08_LIAN1878_11_GE_C08.indd 335 12/29/17 4:06 PM

336 Chapter 8 Multidimensional Arrays

If a bank’s total assets are under a certain limit, the bank is unsafe. The money it borrowed cannot be returned to the lender, and the lender cannot count the loan in its total assets. Consequently, the lender may also be unsafe, if its total assets are under the limit. Write a program to find all the unsafe banks. Your program reads the input as follows. It first reads two integers n and limit, where n indicates the number of banks and limit is the minimum total assets for keeping a bank safe. It then reads n lines that describe the information for n banks with IDs from 0 to n−1.

The first number in the line is the bank’s balance, the second number indicates the number of banks that borrowed money from the bank, and the rest are pairs of two numbers. Each pair describes a borrower. The first number in the pair is the borrower’s ID and the second is the amount borrowed. For example, the input for the five banks in Figure 8.8 is as follows (note the limit is 201):

5 201 25 2 1 100.5 4 320.5 125 2 2 40 3 85 175 2 0 125 3 75 75 1 0 125 181 1 2 125

The total assets of bank 3 are (75 + 125), which is under 201, so bank 3 is unsafe. After bank 3 becomes unsafe, the total assets of bank 1 fall below (125 + 40). Thus, bank 1 is also unsafe. The output of the program should be

Unsafe banks are 3 1

(Hint: Use a two-dimensional array borrowers to represent loans. borrowers[i][j] indicates the loan that bank i provides to bank j. Once bank j becomes unsafe, borrowers[i][j] should be set to 0.)

*8.18 (Shuffle rows) Write a method that shuffles the rows in a two-dimensional int array using the following header:

public static void shuffle(int[][] m)

Write a test program that shuffles the following matrix:

int[][] m = {{1, 2}, {3, 4}, {5, 6}, {7, 8}, {9, 10}};

**8.19 (Pattern recognition: four consecutive equal numbers) Write the following method that tests whether a two-dimensional array has four consecutive numbers of the same value, either horizontally, vertically, or diagonally:

public static boolean isConsecutiveFour(int[][] values)

Figure 8.8 Banks lend money to each other.

100.5

125

40 75

175125

125

320.5

181

25 75

M08_LIAN1878_11_GE_C08.indd 336 12/29/17 4:06 PM

Programming Exercises 337

Write a test program that prompts the user to enter the number of rows and columns of a two-dimensional array then the values in the array, and displays true if the array contains four consecutive numbers with the same value. Otherwise, the program displays false. Here are some examples of the true cases:

0 1 0 3 1 6 1

0 1 6 8 6 0 1

9 6 2 1 8 2 9

6 9 6 1 1 9 1

1 3 9 1 4 0 7

3 3 3 9 4 0 7

0 1 0 3 1 6 1

0 1 6 8 6 0 1

5 6 2 1 8 2 9

6 5 6 1 1 9 1

1 3 6 1 4 0 7

3 3 3 3 4 0 7

0 1 0 3 1 6 1

0 1 6 8 6 0 1

5 5 2 1 8 2 9

6 5 6 1 1 9 1

1 5 6 1 4 0 7

3 5 3 3 4 0 7

0 1 0 3 1 6 1

0 1 6 8 6 0 1

5 6 2 1 6 2 9

6 5 6 6 1 9 1

1 3 6 1 4 0 7

3 6 3 3 4 0 7

***8.20 (Game: connect four) Connect four is a two-player board game in which the players alternately drop colored disks into a seven-column, six-row vertically suspended grid, as shown below.

The objective of the game is to connect four same-colored disks in a row, a col- umn, or a diagonal before your opponent can do likewise. The program prompts two players to drop a red or yellow disk alternately. In the preceding figure, the red disk is shown in a dark color and the yellow in a light color. Whenever a disk is dropped, the program redisplays the board on the console and determines the status of the game (win, draw, or continue). Here is a sample run:

| | | | | | | |

| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Drop a red disk at column (0–6): 0 | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | |R| | | | | | |

M08_LIAN1878_11_GE_C08.indd 337 12/29/17 4:06 PM

338 Chapter 8 Multidimensional Arrays

*8.21 (Central city) Given a set of cities, the central city is the city that has the shortest total distance to all other cities. Write a program that prompts the user to enter the number of cities and the locations of the cities (coordinates), and finds the central city and its total distance to all other cities.

Enter the number of cities: 5

Enter the coordinates of the cities:

2.5 5 5.1 3 1 9 5.4 54 5.5 2.1

The central city is at (2.5, 5.0)

The total distance to all other cities is 60.81

Drop a yellow disk at column (0–6): 3 | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | |R| | |Y| | | |

. . . . . . . . .

Drop a yellow disk at column (0–6): 6

| | | | | | | | | | | | | | | | | | | |R| | | | | | | |Y|R|Y| | | | |R|Y|Y|Y|Y| |R|Y|R|Y|R|R|R| The yellow player won

*8.22 (Even number of 1s) Write a program that generates a 6-by-6 two-dimensional matrix filled with 0s and 1s, displays the matrix, and checks if every row and every column have an even number of 1s.

*8.23 (Game: find the flipped cell) Suppose you are given a 6-by-6 matrix filled with 0s and 1s. All rows and all columns have an even number of 1s. Let the user flip one cell (i.e., flip from 1 to 0 or from 0 to 1) and write a program to find which cell was flipped. Your program should prompt the user to enter a 6-by-6 array with 0s and 1s and find the first row r and first column c where the even number of the 1s property is violated (i.e., the number of 1s is not even). The flipped cell is at (r, c). Here is a sample run:

VideoNote

Even number of 1s

Enter a 6−by−6 matrix row by row:

1 1 1 0 1 1 1 1 1 1 0 0

0 1 0 1 1 1 1 1 1 1 1 1 0 1 1 1 1 0 1 0 0 0 0 1

The flipped cell is at (0, 1)

M08_LIAN1878_11_GE_C08.indd 338 12/29/17 4:06 PM

Programming Exercises 339

*8.24 (Check Sudoku solution) Listing 8.4 checks whether a solution is valid by check- ing whether every number is valid in the board. Rewrite the program by checking whether every row, every column, and every small box has the numbers 1 to 9.

*8.25 (Markov matrix) An n * n matrix is called a positive Markov matrix if each element is positive and the sum of the elements in each column is 1. Write the following method to check whether a matrix is a Markov matrix:

public static boolean isMarkovMatrix(double[][] m)

Write a test program that prompts the user to enter a 3 * 3 matrix of double values and tests whether it is a Markov matrix. Here are sample runs:

Enter a 3−by−3 matrix row by row:

0.15 0.875 0.375

0.55 0.005 0.225

0.30 0.12 0.4

It is a Markov matrix

Enter a 3−by−3 matrix row by row:

0.95 −0.875 0.375 0.65 0.005 0.225

0.30 0.22 −0.4

It is not a Markov matrix

Enter a 4−by−4 matrix row by row:

0.15 0.875 0.375 0.225

0.55 0.005 0.225 0.015

0.30 0.12 0.4 0.008

0.07 0.021 0.14 0.2 The row−sorted array is

0.15 0.225 0.375 0.875

0.005 0.015 0.225 0.55

0.008 0.12 0.30 0.4

0.021 0.07 0.14 0.2

*8.26 (Row sorting) Implement the following method to sort the rows in a two- dimensional array. A new array is returned and the original array is intact.

public static double[][] sortRows(double[][] m)

Write a test program that prompts the user to enter a 4 * 4 matrix of double values and displays a new row-sorted matrix. Here is a sample run:

*8.27 (Column sorting) Implement the following method to sort the columns in a two dimensional array. A new array is returned and the original array is intact.

public static double[][] sortColumns(double[][] m)

M08_LIAN1878_11_GE_C08.indd 339 12/29/17 4:06 PM

340 Chapter 8 Multidimensional Arrays

Write a test program that prompts the user to enter a 4 * 4 matrix of double values and displays a new column-sorted matrix. Here is a sample run:

Enter a 4−by−4 matrix row by row:

0.15 0.875 0.375 0.225

0.55 0.005 0.225 0.015

0.30 0.12 0.4 0.008 0.07 0.021 0.14 0.2 The column−sorted array is

0.07 0.005 0.14 0.008

0.15 0.021 0.225 0.015

0.30 0.12 0.375 0.2

0.55 0.875 0.4 0.225

Enter list1: 51 22 25 6 1 4 24 54 6

Enter list2: 51 22 25 6 1 4 24 54 6

The two arrays are strictly identical

Enter list1: 51 25 22 6 1 4 24 54 6

Enter list2: 51 22 25 6 1 4 24 54 6

The two arrays are not strictly identical

Enter list1: 51 25 22 6 1 4 24 54 6

Enter list2: 51 22 25 6 1 4 24 54 6

The two arrays are identical

Enter list1: 51 5 22 6 1 4 24 54 6

Enter list2: 51 22 25 6 1 4 24 54 6

The two arrays are not identical

8.28 (Strictly identical arrays) The two-dimensional arrays m1 and m2 are strictly identical if their corresponding elements are equal. Write a method that returns true if m1 and m2 are strictly identical, using the following header:

public static boolean equals(int[][] m1, int[][] m2)

Write a test program that prompts the user to enter two 3 * 3 arrays of integers and displays whether the two are strictly identical. Here are the sample runs:

8.29 (Identical arrays) The two-dimensional arrays m1 and m2 are identical if they have the same contents. Write a method that returns true if m1 and m2 are identi- cal, using the following header:

public static boolean equals(int[][] m1, int[][] m2)

Write a test program that prompts the user to enter two 3 * 3 arrays of integers and displays whether the two are identical. Here are the sample runs:

M08_LIAN1878_11_GE_C08.indd 340 12/29/17 4:06 PM

Programming Exercises 341

*8.30 (Algebra: solve linear equations) Write a method that solves the following 2 * 2 system of linear equations:

a00x + a01y = b0 a10x + a11y = b1

x = b0a11 - b1a01

a00a11 - a01a10 y =

b1a00 - b0a10 a00a11 - a01a10

The method header is:

public static double[] linearEquation(double[][] a, double[] b)

The method returns null if a00a11 - a01a10 is 0. Write a test program that prompts the user to enter a00, a01, a10, a11, b0, and b1 and displays the result. If a00a11 - a01a10 is 0, report that “The equation has no solution.” A sample run is similar to Programming Exercise 3.3.

*8.31 (Geometry: intersecting point) Write a method that returns the intersecting point of two lines. The intersecting point of the two lines can be found by using the formula given in Programming Exercise 3.25. Assume that (x1, y1) and (x2, y2) are the two points on line 1 and (x3, y3) and (x4, y4) are on line 2. The method header is:

public static double[] getIntersectingPoint(double[][] points)

The points are stored in a 4-by-2 two-dimensional array points with (points [0][0], points[0][1]) for (x1, y1). The method returns the intersecting point or null if the two lines are parallel. Write a program that prompts the user to enter four points and displays the intersecting point. See Programming Exercise 3.25 for a sample run.

*8.32 (Geometry: area of a triangle) Write a method that returns the area of a triangle using the following header:

public static double getTriangleArea(double[][] points)

The points are stored in a 3-by-2 two-dimensional array points with points [0][0] and points[0][1] for (x1, y1). The triangle area can be computed using the formula in Programming Exercise 2.19. The method returns 0 if the three points are on the same line. Write a program that prompts the user to enter three points of a triangle and displays the triangle’s area. Here are the sample runs:

Enter x1, y1, x2, y2, x3, y3: 2.5 2 5 −1.0 4.0 2.0

The area of the triangle is 2.25

Enter x1, y1, x2, y2, x3, y3: 2 2 4.5 4.5 6 6

The three points are on the same line

*8.33 (Geometry: polygon subareas) A convex four-vertex polygon is divided into four triangles, as shown in Figure 8.9.

Write a program that prompts the user to enter the coordinates of four vertices and displays the areas of the four triangles in increasing order. Here is a sample run:

Enter x1, y1, x2, y2, x3, y3, x4, y4:

−2.5 2 4 4 3 −2 −2 −3.5

The areas are 6.17 7.96 8.08 10.42

M08_LIAN1878_11_GE_C08.indd 341 12/29/17 4:06 PM

342 Chapter 8 Multidimensional Arrays

*8.34 (Geometry: rightmost lowest point) In computational geometry, often you need to find the rightmost lowest point in a set of points. Write the following method that returns the rightmost lowest point in a set of points:

public static double[] getRightmostLowestPoint(double[][] points)

Write a test program that prompts the user to enter the coordinates of six points and displays the rightmost lowest point. Here is a sample run:

Figure 8.9 A four-vertex polygon is defined by four vertices.

v2 (x2, y2)

v3 (x3, y3)

v4 (x4, y4)

v1 (x1, y1)

Enter 6 points: 1.5 2.5 −3 4.5 5.6 −7 6.5 −7 8 1 10 2.5

The rightmost lowest point is (6.5, −7.0)

Enter the number of rows in the matrix: 5 Enter the matrix row by row:

1 0 1 0 1

1 1 1 0 1

1 0 1 1 1

The maximum square submatrix is at (2, 2) with size 3

Your program should implement and use the following method to find the maxi- mum square submatrix:

public static int[] findLargestBlock(int[][] m)

The return value is an array that consists of three values. The first two values are the row and column indices for the first element in the submatrix, and the third value is the number of the rows in the submatrix.

**8.36 (Latin square) A Latin square is an n-by-n array filled with n different Latin let- ters, each occurring exactly once in each row and once in each column. Write a

**8.35 (Largest block) Given a square matrix with the elements 0 or 1, write a program to find a maximum square submatrix whose elements are all 1s. Your program should prompt the user to enter the number of rows in the matrix. The program then dis- plays the location of the first element in the maximum square submatrix and the number of rows in the submatrix. Here is a sample run:

M08_LIAN1878_11_GE_C08.indd 342 12/29/17 4:06 PM

Programming Exercises 343

program that prompts the user to enter the number n and the array of characters, as shown in the sample output, and checks if the input array is a Latin square. The characters are the first n characters starting from A.

Enter number n: 4

Enter 4 rows of letters separated by spaces:

A B C D

B A D C

C D B A

D C A B

The input array is a Latin square

Enter number n: 3

Enter 3 rows of letters separated by spaces:

A F D

Wrong input: the letters must be from A to C

What is the capital of Alabama? Montogomery

The correct answer should be Montgomery

What is the capital of Alaska? Juneau

Your answer is correct

What is the capital of Arizona? ...

...

The correct count is 35

**8.37 (Guess the capitals) Write a program that repeatedly prompts the user to enter a capital for a state. Upon receiving the user input, the program reports whether the answer is correct. Assume that 50 states and their capitals are stored in a two-dimensional array, as shown in Figure 8.10. The program prompts the user to answer all states’ capitals and displays the total correct count. The user’s answer is not case-sensitive.

Alabama Montgomery

Alaska Juneau

Arizona Phoenix

... ...

Figure 8.10 A two-dimensional array stores states and their capitals.

Here is a sample run:

M08_LIAN1878_11_GE_C08.indd 343 12/29/17 4:06 PM

A01_GIPE3898_09_SE_FM.indd 3 12/20/17 5:56 PM

This page intentionally left blank

Objects and Classes

Objectives ■■ To describe objects and classes, and use classes to model objects (§9.2).

■■ To use UML graphical notation to describe classes and objects (§9.2).

■■ To demonstrate how to define classes and create objects (§9.3).

■■ To create objects using constructors (§9.4).

■■ To access objects via object reference variables (§9.5).

■■ To define a reference variable using a reference type (§9.5.1).

■■ To access an object’s data and methods using the object member access operator (.) (§9.5.2).

■■ To define data fields of reference types and assign default values for an object’s data fields (§9.5.3).

■■ To distinguish between object reference variables and primitive-data- type variables (§9.5.4).

■■ To use the Java library classes Date, Random, and Point2D (§9.6).

■■ To distinguish between instance and static variables and methods (§9.7).

■■ To define private data fields with appropriate getter and setter methods (§9.8).

■■ To encapsulate data fields to make classes easy to maintain (§9.9).

■■ To develop methods with object arguments and differentiate between primitive-type arguments and object-type arguments (§9.10).

■■ To store and process objects in arrays (§9.11).

■■ To create immutable objects from immutable classes to protect the contents of objects (§9.12).

■■ To determine the scope of variables in the context of a class (§9.13).

■■ To use the keyword this to refer to the calling object itself (§9.14).

Chapter

M09_LIAN1878_11_GE_C09.indd 345 12/29/17 4:11 PM

346 Chapter 9 Objects and Classes

9.1 Introduction Object-oriented programming enables you to develop large-scale software and GUIs effectively.

Object-oriented programming is essentially a technology for developing reusable soft- ware. Having learned the material in the preceding chapters, you are able to solve many pro- gramming problems using selections, loops, methods, and arrays. However, these Java features are not sufficient for developing graphical user interfaces and large-scale software systems. Suppose you want to develop a graphical user interface (GUI, pronounced goo-ee) as shown in Figure 9.1. How would you program it?

Point Key

why OOP?

Figure 9.1 The GUI objects are created from classes.

Button Text Field Radio Button Combo BoxLabel Check Box

This chapter introduces object-oriented programming, which you can use to develop GUI and large-scale software systems.

9.2 Defining Classes for Objects A class defines the properties and behaviors for objects.

Object-oriented programming (OOP) involves programming using objects. An object repre- sents an entity in the real world that can be distinctly identified. For example, a student, a desk, a circle, a button, and even a loan can all be viewed as objects. An object has a unique identity, state, and behavior.

■■ The state of an object (also known as its properties or attributes) is represented by data fields with their current values. A circle object, for example, has a data field radius, which is the property that characterizes a circle. A rectangle object, for example, has the data fields width and height, which are the properties that characterize a rectangle.

■■ The behavior of an object (also known as its actions) is defined by methods. To invoke a method on an object is to ask the object to perform an action. For example, you may define methods named getArea() and getPerimeter() for circle objects. A circle object may invoke getArea() to return its area and getPerimeter() to return its perimeter. You may also define the setRadius(radius) method. A circle object can invoke this method to change its radius.

Objects of the same type are defined using a common class. A class is a template, blueprint, or contract that defines what an object’s data fields and methods will be. An object is an instance of a class. You can create many instances of a class. Creating an instance is referred to as instantiation. The terms object and instance are often interchangeable. The relationship between classes and objects is analogous to that between an apple-pie recipe and apple pies: You can make as many apple pies as you want from a single recipe. Figure 9.2 shows a class named Circle and its three objects.

A Java class uses variables to define data fields and methods to define actions. In addition, a class provides methods of a special type, known as constructors, which are invoked to create a new object. A constructor can perform any action, but constructors are designed to perform initializing actions, such as initializing the data fields of objects. Figure 9.3 shows an example of defining the class for circle objects.

Point Key

VideoNote Define classes and objects

object state of an object

properties attributes

data fields behavior actions

class

contract

instantiation

instance

data field method

constructors

M09_LIAN1878_11_GE_C09.indd 346 12/29/17 4:11 PM

9.2 Defining Classes for Objects 347

The Circle class is different from all of the other classes you have seen thus far. It does not have a main method, and therefore cannot be run; it is merely a definition for circle objects. The class that contains the main method will be referred to in this book, for convenience, as the main class.

The illustration of class templates and objects in Figure 9.2 can be standardized using Unified Modeling Language (UML) notation. This notation, as shown in Figure 9.4, is called a UML class diagram, or simply a class diagram. In the class diagram, the data field is denoted as

dataFieldName: dataFieldType

The constructor is denoted as

ClassName(parameterName: parameterType)

main class Unified Modeling Language

(UML)

class diagram

Figure 9.2 A class is a template for creating objects.

Class Name: Circle

Data Fields: radius is _____

Methods: getArea getPerimeter setRadius

Circle Object 1

Data Fields: radius is 1

Circle Object 2

Data Fields: radius is 25

Circle Object 3

Data Fields: radius is 125

A class template

Three objects of the Circle class

Figure 9.3 A class is a construct that defines objects of the same type.

Data �eld

Constructors

Method

class Circle { /** The radius of this circle */ double radius = 1;

/** Construct a circle object */ Circle() { }

/** Construct a circle object */ Circle(double newRadius) { radius = newRadius; }

/** Return the area of this circle */ double getArea() { return radius * radius * Math.PI; }

/** Return the perimeter of this circle */ double getPerimeter() { return 2 * radius * Math.PI; }

/** Set a new radius for this circle */ void setRadius(double newRadius) { radius = newRadius; } }

M09_LIAN1878_11_GE_C09.indd 347 12/29/17 4:11 PM

348 Chapter 9 Objects and Classes

The method is denoted as

methodName(parameterName: parameterType): returnType

9.3 Example: Defining Classes and Creating Objects Classes are definitions for objects and objects are created from classes.

This section gives two examples of defining classes and uses the classes to create objects. Listing 9.1 is a program that defines the Circle class and uses it to create objects. The pro- gram constructs three circle objects with radius 1, 25, and 125 and displays the radius and area of each of the three circles. It then changes the radius of the second object to 100 and displays its new radius and area.

Listing 9.1 TestCircle.java 1 public class TestCircle { 2 /** Main method */ 3 public static void main(String[] args) { 4 // Create a circle with radius 1 5 Circle circle1 = new Circle(); 6 System.out.println("The area of the circle of radius " 7 + circle1.radius + " is " + circle1.getArea()); 8 9 // Create a circle with radius 25 10 Circle circle2 = new Circle(25); 11 System.out.println("The area of the circle of radius " 12 + circle2.radius + " is " + circle2.getArea()); 13 14 // Create a circle with radius 125 15 Circle circle3 = new Circle(125); 16 System.out.println("The area of the circle of radius " 17 + circle3.radius + " is " + circle3.getArea()); 18 19 // Modify circle radius 20 circle2.radius = 100; // or circle2.setRadius(100) 21 System.out.println("The area of the circle of radius " 22 + circle2.radius + " is " + circle2.getArea()); 23 } 24 } 25

Point Key

main class

main method

create object

Figure 9.4 Classes and objects can be represented using UML notation.

Class nameCircle

radius: double

Circle()

Circle(newRadius: double)

getArea(): double

getPerimeter(): double

setRadius(newRadius: double): void

Data �elds

Constructors and methods

UML Class Diagram

UML notation for objects

circle2: Circle

radius = 25

circle3: Circle

radius = 125

circle1: Circle

radius = 1

M09_LIAN1878_11_GE_C09.indd 348 12/29/17 4:11 PM

9.3 Example: Defining Classes and Creating Objects 349

26 // Define the circle class with two constructors 27 class Circle { 28 double radius; 29 30 /** Construct a circle with radius 1 */ 31 Circle() { 32 radius = 1; 33 } 34 35 /** Construct a circle with a specified radius */ 36 Circle(double newRadius) { 37 radius = newRadius; 38 } 39 40 /** Return the area of this circle */ 41 double getArea() { 42 return radius * radius * Math.PI; 43 } 44 45 /** Return the perimeter of this circle */ 46 double getPerimeter() { 47 return 2 * radius * Math.PI; 48 } 49 50 /** Set a new radius for this circle */ 51 void setRadius(double newRadius) { 52 radius = newRadius; 53 } 54 }

class Circle data field

no-arg constructor

second constructor

getArea

getPerimeter

setRadius

The area of the circle of radius 1.0 is 3.141592653589793

The area of the circle of radius 25.0 is 1963.4954084936207

The area of the circle of radius 125.0 is 49087.385212340516

The area of the circle of radius 100.0 is 31415.926535897932

Figure 9.5 Each class in the source code file is compiled into a .class file.

Java Compilercompiled

generates

// File TestCircle.java

public class TestCircle { … }

class Circle { … }

TestCircle.class

Circle.class

The program contains two classes. The first of these, TestCircle, is the main class. Its sole purpose is to test the second class, Circle. Such a program that uses the class is often referred to as a client of the class. When you run the program, the Java runtime system invokes the main method in the main class.

You can put the two classes into one file, but only one class in the file can be a public class. Furthermore, the public class must have the same name as the file name. Therefore, the file name is TestCircle.java, since TestCircle is public. Each class in the source code is com- piled into a .class file. When you compile TestCircle.java, two class files TestCircle.class and Circle.class are generated, as shown in Figure 9.5.

client

public class

M09_LIAN1878_11_GE_C09.indd 349 12/29/17 4:11 PM

350 Chapter 9 Objects and Classes

The main class contains the main method (line 3) that creates three objects. As in creating an array, the new operator is used to create an object from the constructor: new Circle() creates an object with radius 1 (line 5), new Circle(25) creates an object with radius 25 (line 10), and new Circle(125) creates an object with radius 125 (line 15).

These three objects (referenced by circle1, circle2, and circle3) have different data but the same methods. Therefore, you can compute their respective areas by using the getArea() method. The data fields can be accessed via the reference of the object using circle1.radius, circle2.radius, and circle3.radius, respectively. The object can invoke its method via the reference of the object using circle1.getArea(), circle2. getArea(), and circle3.getArea(), respectively.

These three objects are independent. The radius of circle2 is changed to 100 in line 20. The object’s new radius and area are displayed in lines 21 and 22.

There are many ways to write Java programs. For instance, you can combine the two classes in the preceding example into one, as given in Listing 9.2.

Listing 9.2 Circle.java (AlternativeCircle.java) 1 public class Circle { 2 /** Main method */ 3 public static void main(String[] args) { 4 // Create a circle with radius 1 5 Circle circle1 = new Circle(); 6 System.out.println("The area of the circle of radius " 7 + circle1.radius + " is " + circle1.getArea()); 8 9 // Create a circle with radius 25 10 Circle circle2 = new Circle(25); 11 System.out.println("The area of the circle of radius " 12 + circle2.radius + " is " + circle2.getArea()); 13 14 // Create a circle with radius 125 15 Circle circle3 = new Circle(125); 16 System.out.println("The area of the circle of radius " 17 + circle3.radius + " is " + circle3.getArea()); 18 19 // Modify circle radius 20 circle2.radius = 100; 21 System.out.println("The area of the circle of radius " 22 + circle2.radius + " is " + circle2.getArea()); 23 } 24 25 double radius; 26 27 /** Construct a circle with radius 1 */ 28 Circle() { 29 radius = 1; 30 } 31 32 /** Construct a circle with a specified radius */ 33 Circle(double newRadius) { 34 radius = newRadius; 35 } 36 37 /** Return the area of this circle */ 38 double getArea() { 39 return radius * radius * Math.PI; 40 } 41

main method

data field

no-arg constructor

second constructor

method

M09_LIAN1878_11_GE_C09.indd 350 12/29/17 4:11 PM

9.3 Example: Defining Classes and Creating Objects 351

42 /** Return the perimeter of this circle */ 43 double getPerimeter() { 44 return 2 * radius * Math.PI; 45 } 46 47 /** Set a new radius for this circle */ 48 void setRadius(double newRadius) { 49 radius = newRadius; 50 } 51 }

Since the combined class has a main method, it can be executed by the Java interpreter. The main method is the same as that in Listing 9.1. This demonstrates that you can test a class by simply adding a main method in the same class.

As another example, consider television sets. Each TV is an object with states (current channel, current volume level, and power on or off) and behaviors (change channels, adjust volume, and turn on/off). You can use a class to model TV sets. The UML diagram for the class is shown in Figure 9.6.

Figure 9.6 The TV class models TV sets.

The current channel (1–120) of this TV.

The current volume level (1–7) of this TV.

Indicates whether this TV is on/off.

channel: int

volumeLevel: int

on: boolean

+TV()

+turnOn(): void

+turnOf f(): void

+setChannel(newChannel: int): void

+setVolume(newVolumeLevel: int): void

+channelUp(): void

+channelDown(): void

+volumeUp(): void

+volumeDown(): void

Constructs a default TV object.

Turns on this TV.

Turns off this TV.

Sets a new channel for this TV.

Sets a new volume level for this TV.

Increases the channel number by 1.

Decreases the channel number by 1.

Increases the volume level by 1.

Decreases the volume level by 1.

The + sign indicates public modi�er

Listing 9.3 gives a program that defines the TV class.

Listing 9.3 TV.java 1 public class TV { 2 int channel = 1; // Default channel is 1 3 int volumeLevel = 1; // Default volume level is 1 4 boolean on = false; // TV is off 5 6 public TV() { 7 } 8 9 public void turnOn() { 10 on = true; 11 } 12 13 public void turnOff() {

data fields

constructor

turn on TV

turn off TV

M09_LIAN1878_11_GE_C09.indd 351 12/29/17 4:11 PM

352 Chapter 9 Objects and Classes

14 on = false; 15 } 16 17 public void setChannel(int newChannel) { 18 if (on && newChannel >= 1 && newChannel <= 120) 19 channel = newChannel; 20 } 21 22 public void setVolume(int newVolumeLevel) { 23 if (on && newVolumeLevel >= 1 && newVolumeLevel <= 7) 24 volumeLevel = newVolumeLevel; 25 } 26 27 public void channelUp() { 28 if (on && channel < 120) 29 channel++; 30 } 31 32 public void channelDown() { 33 if (on && channel > 1) 34 channel—–; 35 } 36 37 public void volumeUp() { 38 if (on && volumeLevel < 7) 39 volumeLevel++; 40 } 41 42 public void volumeDown() { 43 if (on && volumeLevel > 1) 44 volumeLevel—–; 45 } 46 }

The constructor and methods in the TV class are defined public so they can be accessed from other classes. Note the channel and volume level are not changed if the TV is not on. Before either of these is changed, its current value is checked to ensure it is within the correct range.

Listing 9.4 gives a program that uses the TV class to create two objects.

Listing 9.4 TestTV.java 1 public class TestTV { 2 public static void main(String[] args) { 3 TV tv1 = new TV(); 4 tv1.turnOn(); 5 tv1.setChannel(30); 6 tv1.setVolume(3); 7 8 TV tv2 = new TV(); 9 tv2.turnOn(); 10 tv2.channelUp(); 11 tv2.channelUp(); 12 tv2.volumeUp(); 13 14 System.out.println("tv1's channel is " + tv1.channel 15 + " and volume level is " + tv1.volumeLevel); 16 System.out.println("tv2's channel is " + tv2.channel 17 + " and volume level is " + tv2.volumeLevel); 18 } 19 }

set a new channel

set a new volume

increase channel

decrease channel

increase volume

decrease volume

main method create a TV turn on set a new channel set a new volume

create a TV turn on increase channel

increase volume

display state

M09_LIAN1878_11_GE_C09.indd 352 12/29/17 4:11 PM

9.4 Constructing Objects Using Constructors 353

The program creates two objects in lines 3 and 8 and invokes the methods on the objects to perform actions for setting channels and volume levels and for increasing channels and vol- umes. The program displays the state of the objects in lines 14–17. The methods are invoked using syntax such as tv1.turnOn() (line 4). The data fields are accessed using syntax such as tv1.channel (line 14).

These examples have given you a glimpse of classes and objects. You may have many ques- tions regarding constructors, objects, reference variables, accessing data fields, and invoking object’s methods. The sections that will follow discuss these issues in detail.

9.3.1 Describe the relationship between an object and its defining class. 9.3.2 How do you define a class? 9.3.3 How do you declare an object’s reference variable? 9.3.4 How do you create an object?

9.4 Constructing Objects Using Constructors A constructor is invoked to create an object using the new operator.

Constructors are a special kind of method. They have three peculiarities:

■■ A constructor must have the same name as the class itself.

■■ Constructors do not have a return type—not even void.

■■ Constructors are invoked using the new operator when an object is created. Constructors play the role of initializing objects.

The constructor has exactly the same name as its defining class. Like regular methods, con- structors can be overloaded (i.e., multiple constructors can have the same name but different signatures), making it easy to construct objects with different initial data values.

It is a common mistake to put the void keyword in front of a constructor. For example,

public void Circle() { }

In this case, Circle() is a method, not a constructor. Constructors are used to construct objects. To construct an object from a class, invoke a

constructor of the class using the new operator, as follows:

new ClassName(arguments);

For example, new Circle() creates an object of the Circle class using the first construc- tor defined in the Circle class, and new Circle(25) creates an object using the second constructor defined in the Circle class.

A class normally provides a constructor without arguments (e.g., Circle()). Such a con- structor is referred to as a no-arg or no-argument constructor.

A class may be defined without constructors. In this case, a public no-arg constructor with an empty body is implicitly defined in the class. This constructor, called a default constructor, is provided automatically only if no constructors are explicitly defined in the class.

9.4.5 What are the differences between constructors and methods? 9.4.6 When will a class have a default constructor?

Point Check

Point Key

constructor’s name

no return type

new operator

overloaded constructors

no void

constructing objects

no-arg constructor

default constructor

Point Check

tv1's channel is 30 and volume level is 3

tv2's channel is 3 and volume level is 2

M09_LIAN1878_11_GE_C09.indd 353 12/29/17 4:11 PM

354 Chapter 9 Objects and Classes

9.5 Accessing Objects via Reference Variables An object’s data and methods can be accessed through the dot (.) operator via the object’s reference variable.

Newly created objects are allocated in the memory. They can be accessed via reference variables.

9.5.1 Reference Variables and Reference Types Objects are accessed via the object’s reference variables, which contain references to the objects. Such variables are declared using the following syntax:

ClassName objectRefVar;

A class is essentially a programmer-defined type. A class is a reference type, which means that a variable of the class type can reference an instance of the class. The following statement declares the variable myCircle to be of the Circle type:

Circle myCircle;

The variable myCircle can reference a Circle object. The next statement creates an object and assigns its reference to myCircle:

myCircle = new Circle();

You can write a single statement that combines the declaration of an object reference variable, the creation of an object, and the assigning of an object reference to the variable with the fol- lowing syntax:

ClassName objectRefVar = new ClassName();

Here is an example:

Circle myCircle = new Circle();

The variable myCircle holds a reference to a Circle object.

Note An object reference variable that appears to hold an object actually contains a reference to that object. Strictly speaking, an object reference variable and an object are different, but most of the time the distinction can be ignored. Therefore, it is fine, for simplicity, to say that myCircle is a Circle object rather than use the long-winded description that myCircle is a variable that contains a reference to a Circle object.

Note Arrays are treated as objects in Java. Arrays are created using the new operator. An array variable is actually a variable that contains a reference to an array.

9.5.2 Accessing an Object’s Data and Methods In OOP terminology, an object’s member refers to its data fields and methods. After an object is created, its data can be accessed and its methods can be invoked using the dot operator (.), also known as the object member access operator:

■■ objectRefVar.dataField references a data field in the object.

■■ objectRefVar.method(arguments) invokes a method on the object.

Point Key

reference variable

reference type

object vs. object reference variable

array object

dot operator (.)

M09_LIAN1878_11_GE_C09.indd 354 12/29/17 4:11 PM

9.5 Accessing Objects via Reference Variables 355

For example, myCircle.radius references the radius in myCircle and myCircle .getArea() invokes the getArea method on myCircle. Methods are invoked as operations on objects.

The data field radius is referred to as an instance variable because it is dependent on a specific instance. For the same reason, the method getArea is referred to as an instance method because you can invoke it only on a specific instance. The object on which an instance method is invoked is called a calling object.

Caution Recall that you use Math.methodName(arguments) (e.g., Math.pow(3, 2.5)) to invoke a method in the Math class. Can you invoke getArea() using Circle. getArea()? The answer is no. All the methods in the Math class are static methods, which are defined using the static keyword. However, getArea() is an instance method, and thus nonstatic. It must be invoked from an object using objectRefVar. methodName(arguments) (e.g., myCircle.getArea()). Further explanation will be given in Section 9.7, Static Variables, Constants, and Methods.

Note Usually you create an object and assign it to a variable, then later you can use the variable to reference the object. Occasionally, an object does not need to be referenced later. In this case, you can create an object without explicitly assigning it to a variable using the syntax:

new Circle();

System.out.println("Area is " + new Circle(5).getArea());

The former statement creates a Circle object. The latter creates a Circle object and invokes its getArea method to return its area. An object created in this way is known as an anonymous object.

9.5.3 Reference Data Fields and the null Value The data fields can be of reference types. For example, the following Student class contains a data field name of the String type. String is a predefined Java class.

class Student { String name; // name has the default value null int age; // age has the default value 0 boolean isScienceMajor; // isScienceMajor has default value false char gender; // gender has default value '\u0000' }

If a data field of a reference type does not reference any object, the data field holds a special Java value, null. null is a literal just like true and false. While true and false are Boolean literals, null is a literal for a reference type.

The default value of a data field is null for a reference type, 0 for a numeric type, false for a boolean type, and \u0000 for a char type. However, Java assigns no default value to a local variable inside a method. The following code displays the default values of the data fields name, age, isScienceMajor, and gender for a Student object:

class TestStudent { public static void main(String[] args) { Student student = new Student(); System.out.println("name? " + student.name);

instance variable

instance method

calling object

invoking methods

anonymous object

reference data fields

null value

default field values

M09_LIAN1878_11_GE_C09.indd 355 12/29/17 4:11 PM

356 Chapter 9 Objects and Classes

System.out.println("age? " + student.age); System.out.println("isScienceMajor? " + student.isScienceMajor); System.out.println("gender? " + student.gender); } }

The following code has a compile error, because the local variables x and y are not initialized:

class TestLocalVariables { public static void main(String[] args) { int x; // x has no default value String y; // y has no default value System.out.println("x is " + x); System.out.println("y is " + y); } }

Caution NullPointerException is a common runtime error. It occurs when you invoke a method on a reference variable with a null value. Make sure you assign an object refer- ence to the variable before invoking the method through the reference variable (see CheckPoint Question 9.5.5c).

9.5.4 Differences between Variables of Primitive Types and Reference Types

Every variable represents a memory location that holds a value. When you declare a variable, you are telling the compiler what type of value the variable can hold. For a variable of a primi- tive type, the value is of the primitive type. For a variable of a reference type, the value is a reference to where an object is located. For example, as shown in Figure 9.7, the value of int variable i is int value 1, and the value of Circle object c holds a reference to where the contents of the Circle object are stored in memory.

NullPointerException

Figure 9.7 A variable of a primitive type holds a value of the primitive type, and a variable of a reference type holds a reference to where an object is stored in memory.

Primitive type

Object type

int i = 1

Circle c

c reference

Created using new Circle()

radius = 1

c: Circle

Figure 9.8 Primitive variable j is copied to variable i.

Primitive type assignment i = j

Before i = j After i = j

When you assign one variable to another, the other variable is set to the same value. For a variable of a primitive type, the real value of one variable is assigned to the other variable. For a variable of a reference type, the reference of one variable is assigned to the other vari- able. As shown in Figure 9.8, the assignment statement i = j copies the contents of j into i

M09_LIAN1878_11_GE_C09.indd 356 12/29/17 4:11 PM

9.5 Accessing Objects via Reference Variables 357

Note As illustrated in Figure 9.9, after the assignment statement c1 = c2, c1 points to the same object referenced by c2. The object previously referenced by c1 is no longer useful and therefore is now known as garbage. Garbage occupies memory space, so the Java runtime system detects garbage and automatically reclaims the space it occupies. This process is called garbage collection.

Tip If you know that an object is no longer needed, you can explicitly assign null to a refer- ence variable for the object. The JVM will automatically collect the space if the object is not referenced by any reference variable.

9.5.1 Which operator is used to access a data field or invoke a method from an object? 9.5.2 What is an anonymous object? 9.5.3 What is NullPointerException? 9.5.4 Is an array an object or a primitive-type value? Can an array contain elements of an

object type? Describe the default value for the elements of an array.

9.5.5 What is wrong with each of the following programs?

1 public class ShowErrors { 2 public static void main(String[] args) { 3 ShowErrors t = new ShowErrors(5); 4 } 5 }

1 public class ShowErrors { 2 public static void main(String[] args) { 3 ShowErrors t = new ShowErrors(); 4 t.x(); 5 } 6 }

(a) (b)

1 public class ShowErrors { 2 public void method1() { 3 Circle c; 4 System.out.println("What is radius " 5 + c.getRadius()); 6 c = new Circle(); 7 } 8 }

1 public class ShowErrors { 2 public static void main(String[] args) { 3 C c = new C(5.0); 4 System.out.println(c.value); 5 } 6 } 7 8 class C { 9 int value = 2; 10 }

garbage

garbage collection

Point Check

for primitive variables. As shown in Figure 9.9, the assignment statement c1 = c2 copies the reference of c2 into c1 for reference variables. After the assignment, variables c1 and c2 refer to the same object.

Figure 9.9 Reference variable c2 is copied to variable c1.

Object type assignment c1 = c2

After c1 = c2

Before c1 = c2

radius = 9

c2: Circle

radius = 5

c1: Circle

radius = 5

c1: Circle

radius = 9

c2: Circle

M09_LIAN1878_11_GE_C09.indd 357 12/29/17 4:11 PM

358 Chapter 9 Objects and Classes

9.5.6 What is wrong in the following code? 1 class Test { 2 public static void main(String[] args) { 3 A a = new A(); 4 a.print(); 5 } 6 } 7 8 class A { 9 String s; 10 11 A(String newS) { 12 s = newS; 13 } 14 15 public void print() { 16 System.out.print(s); 17 } 18 }

9.5.7 What is the output of the following code? public class A { boolean x;

public static void main(String[] args) { A a = new A(); System.out.println(a.x); } }

9.6 Using Classes from the Java Library The Java API contains a rich set of classes for developing Java programs.

Listing 9.1 defined the Circle class and created objects from the class. You will frequently use the classes in the Java library to develop programs. This section gives some examples of the classes in the Java library.

9.6.1 The Date Class In Listing 2.7, ShowCurrentTime.java, you learned how to obtain the current time using System.currentTimeMillis(). You used the division and remainder operators to extract the current second, minute, and hour. Java provides a system-independent encapsulation of date and time in the java.util.Date class, as shown in Figure 9.10.

Point Key

VideoNote

Use classes

java.util.Date class

Figure 9.10 A Date object represents a specific date and time.

java.util.Date

+Date()

+Date(elapseTime: long)

+toString(): String

+getTime(): long

+setTime(elapseTime: long): void

Returns a string representing the date and time.

Returns the number of milliseconds since January 1,

Sets a new elapse time in the object.

Constructs a Date object for the current time.

Constructs a Date object for a given time in milliseconds elapsed since January 1, 1970, GMT.

1970, GMT.

M09_LIAN1878_11_GE_C09.indd 358 12/29/17 4:11 PM

9.6 Using Classes from the Java Library 359

You can use the no-arg constructor in the Date class to create an instance for the current date and time, the getTime() method to return the elapsed time in milliseconds since January 1, 1970, GMT, and the toString() method to return the date and time as a string. For example, the following code

java.util.Date date = new java.util.Date(); System.out.println("The elapsed time since Jan 1, 1970 is " + date.getTime() + " milliseconds"); System.out.println(date.toString());

displays the output as follows:

The elapsed time since Jan 1, 1970 is 1324903419651 milliseconds Mon Dec 26 07:43:39 EST 2011

The Date class has another constructor, Date(long elapseTime), which can be used to construct a Date object for a given time in milliseconds elapsed since January 1, 1970, GMT.

9.6.2 The Random Class You have used Math.random() to obtain a random double value between 0.0 and 1.0 (excluding 1.0). Another way to generate random numbers is to use the java.util.Random class, as shown in Figure 9.11, which can generate a random int, long, double, float, and boolean value.

create object

get elapsed time invoke toString

Figure 9.11 A Random object can be used to generate random values.

+Random()

+Random(seed: long)

+nextInt(): int

+nextInt(n: int): int

+nextLong(): long

+nextDouble(): double

+nextFloat(): float

+nextBoolean(): boolean

Constructs a Random object with the current time as its seed.

Constructs a Random object with a speci�ed seed.

Returns a random int value.

Returns a random int value between 0 and n (excluding n).

Returns a random long value.

Returns a random double value between 0.0 and 1.0 (excluding 1.0).

Returns a random float value between 0.0F and 1.0F (excluding 1.0F).

Returns a random boolean value.

java.util.Random

When you create a Random object, you have to specify a seed or use the default seed. A seed is a number used to initialize a random number generator. The no-arg constructor creates a Random object using the current elapsed time as its seed. If two Random objects have the same seed, they will generate identical sequences of numbers. For example, the following code creates two Random objects with the same seed, 3:

Random generator1 = new Random(3); System.out.print("From generator1: "); for (int i = 0; i < 10; i++) System.out.print(generator1.nextInt(1000) + " ");

Random generator2 = new Random(3); System.out.print("\nFrom generator2: "); for (int i = 0; i < 10; i++) System.out.print(generator2.nextInt(1000) + " ");

The code generates the same sequence of random int values:

From generator1: 734 660 210 581 128 202 549 564 459 961 From generator2: 734 660 210 581 128 202 549 564 459 961

M09_LIAN1878_11_GE_C09.indd 359 12/29/17 4:11 PM

360 Chapter 9 Objects and Classes

Note The ability to generate the same sequence of random values is useful in software testing and many other applications. In software testing, often you need to reproduce the test cases from a fixed sequence of random numbers.

Note You can generate random numbers using the java.security.SecureRandom class rather than the Random class. The random numbers generated from the Random are deterministic and they can be predicated by hackers. The random numbers generated from the SecureRandom class are nondeterministic and are secure.

9.6.3 The Point2D Class Java API has a convenient Point2D class in the javafx.geometry package for representing a point in a two-dimensional plane. The UML diagram for the class is shown in Figure 9.12.

same sequence

SecureRandom

Figure 9.12 A Point2D object represents a point with x- and y-coordinates.

Constructs a Point2D object with the speci�ed x- and y-coordinates.

Returns the distance between this point and the speci�ed point (x, y).

Returns the distance between this point and the speci�ed point p.

Returns the x-coordinate from this point.

Returns the y-coordinate from this point.

Returns the midpoint between this point and point p.

Returns a string representation for the point.

+Point2D(x: double, y: double)

+distance(x: double, y: double): double

+distance(p: Point2D): double

+getX(): double

+getY(): double

+midpoint(p: Point2D): Point2D

+toString(): String

javafx.geometry.Point2D

You can create a Point2D object for a point with the specified x- and y-coordinates, use the distance method to compute the distance from this point to another point, and use the toString() method to return a string representation of the point. Listing 9.5 gives an example of using this class.

Listing 9.5 TestPoint2D.java 1 import java.util.Scanner; 2 import javafx.geometry.Point2D; 3 4 public class TestPoint2D { 5 public static void main(String[] args) { 6 Scanner input = new Scanner(System.in); 7 8 System.out.print("Enter point1's x-, y–coordinates: "); 9 double x1 = input.nextDouble(); 10 double y1 = input.nextDouble(); 11 System.out.print("Enter point2's x-, y–coordinates: "); 12 double x2 = input.nextDouble(); 13 double y2 = input.nextDouble(); 14 15 Point2D p1 = new Point2D(x1, y1); 16 Point2D p2 = new Point2D(x2, y2); 17 System.out.println("p1 is " + p1.toString()); 18 System.out.println("p2 is " + p2.toString()); 19 System.out.println("The distance between p1 and p2 is " + 20 p1.distance(p2));

create an object

invoke toString()

get distance

M09_LIAN1878_11_GE_C09.indd 360 12/29/17 4:11 PM

9.7 Static Variables, Constants, and Methods 361

This program creates two objects of the Point2D class (lines 15 and 16). The toString() method returns a string that describes the object (lines 17 and 18). Invoking p1.distance(p2) returns the distance between the two points (line 20). Invoking p1.midpoint(p2) returns the midpoint between the two points (line 22).

9.6.1 How do you create a Date for the current time? How do you display the current time? 9.6.2 How do you create a Point2D? Suppose p1 and p2 are two instances of Point2D,

how do you obtain the distance between the two points? How do you obtain the midpoint between the two points?

9.6.3 Which packages contain the classes Date, Random, Point2D, System, and Math?

9.7 Static Variables, Constants, and Methods A static variable is shared by all objects of the class. A static method cannot access instance members (i.e., instance data fields and methods) of the class.

The data field radius in the circle class is known as an instance variable. An instance variable is tied to a specific instance of the class; it is not shared among objects of the same class. For example, suppose that you create the following objects:

Circle circle1 = new Circle(); Circle circle2 = new Circle(5);

The radius in circle1 is independent of the radius in circle2 and is stored in a different memory location. Changes made to circle1’s radius do not affect circle2’s radius, and vice versa.

If you want all the instances of a class to share data, use static variables, also known as class variables. Static variables store values for the variables in a common memory location. Because of this common location, if one object changes the value of a static variable, all objects of the same class are affected. Java supports static methods as well as static variables. Static methods can be called without creating an instance of the class.

Let’s modify the Circle class by adding a static variable numberOfObjects to count the number of circle objects created. When the first object of this class is created, numberOfObjects is 1. When the second object is created, numberOfObjects becomes 2. The UML of the new circle class is shown in Figure 9.13. The Circle class defines the instance variable radius and the static variable numberOfObjects, the instance methods getRadius, setRadius, and getArea, and the static method getNumberOfObjects. (Note static variables and methods are underlined in the UML class diagram.)

Point Check

Point Key

VideoNote

Static vs. instance

instance variable

static variable

static method

Enter point1's x-, y-coordinates: 1.5 5.5

Enter point2's x-, y-coordinates: −5.3 −4.4 p1 is Point2D [x = 1.5, y = 5.5]

p2 is Point2D [x = −5.3, y = −4.4]

The distance between p1 and p2 is 12.010412149464313

The midpoint between p1 and p2 is

Point2D [x = −1.9, y = 0.5499999999999998]

21 System.out.println("The midpoint between p1 and p2 is " + 22 p1.midpoint(p2).toString()); 23 } 24 }

get midpoint

M09_LIAN1878_11_GE_C09.indd 361 12/29/17 4:11 PM

362 Chapter 9 Objects and Classes

To declare a static variable or define a static method, put the modifier static in the vari- able or method declaration. The static variable numberOfObjects and the static method getNumberOfObjects() can be declared as follows:

static int numberOfObjects;

static int getNumberObjects() { return numberOfObjects; }

Constants in a class are shared by all objects of the class. Thus, constants should be declared as final static. For example, the constant PI in the Math class is defined as follows:

final static double PI = 3.14159265358979323846;

The new circle class is defined in Listing 9.6.

Listing 9.6 Circle.java (for CircleWithStaticMembers) 1 public class Circle { 2 /** The radius of the circle */ 3 double radius; 4 5 /** The number of objects created */ 6 static int numberOfObjects = 0; 7 8 /** Construct a circle with radius 1 */ 9 Circle() { 10 radius = 1; 11 numberOfObjects++; 12 } 13 14 /** Construct a circle with a specified radius */ 15 Circle(double newRadius) { 16 radius = newRadius; 17 numberOfObjects++; 18 } 19 20 /** Return numberOfObjects */ 21 static int getNumberOfObjects() { 22 return numberOfObjects; 23 } 24 25 /** Return the area of this circle */

declare static variable

define static method

declare constant

static variable

increase by 1

static method

Figure 9.13 Instance variables belong to the instances and have memory storage independent of one another. Static variables are shared by all the instances of the same class.

radius

numberOfObjects

radius

Memory instantiate

instantiate

Circle

circle2: Circle

radius: double numberOfObjects: int

getNumberOfObjects(): int getArea(): double

radius = 1 numberOfObjects = 2

radius = 5 numberOfObjects = 2

circle1: Circle

UML Notation: underline: static variables or methods

After two Circle Objects were created, numberOfObjects is 2.

M09_LIAN1878_11_GE_C09.indd 362 12/29/17 4:11 PM

9.7 Static Variables, Constants, and Methods 363

26 double getArea() { 27 return radius * radius * Math.PI; 28 } 29 }

Method getNumberOfObjects() in Circle is a static method. All the methods in the Math class are static. The main method is static, too.

Instance methods (e.g., getArea()) and instance data (e.g., radius) belong to instances and can be used only after the instances are created. They are accessed via a reference variable. Static methods (e.g., getNumberOfObjects()) and static data (e.g., numberOfObjects) can be accessed from a reference variable or from their class name.

The program in Listing 9.7 demonstrates how to use instance and static variables and meth- ods and illustrates the effects of using them.

Listing 9.7 TestCircleWithStaticMembers.java 1 public class TestCircleWithStaticMembers { 2 /** Main method */ 3 public static void main(String[] args) { 4 System.out.println("Before creating objects"); 5 System.out.println("The number of Circle objects is " + 6 Circle.numberOfObjects); 7 8 // Create c1 9 Circle c1 = new Circle(); // Use the Circle class in Listing 9.6 10 11 // Display c1 BEFORE c2 is created 12 System.out.println("\nAfter creating c1"); 13 System.out.println("c1: radius (" + c1.radius + 14 ") and number of Circle objects (" + 15 c1.numberOfObjects + ")"); 16 17 // Create c2 18 Circle c2 = new Circle(5); 19 20 // Modify c1 21 c1.radius = 9; 22 23 // Display c1 and c2 AFTER c2 was created 24 System.out.println("\nAfter creating c2 and modifying c1"); 25 System.out.println("c1: radius (" + c1.radius + 26 ") and number of Circle objects (" + 27 c1.numberOfObjects + ")"); 28 System.out.println("c2: radius (" + c2.radius + 29 ") and number of Circle objects (" + 30 c2.numberOfObjects + ")"); 31 } 32 }

static variable

instance variable

static variable

instance variable

static variable

Before creating objects

The number of Circle objects is 0

After creating c1

c1: radius (1.0) and number of Circle objects (1)

After creating c2 and modifying c1

c1: radius (9.0) and number of Circle objects (2)

c2: radius (5.0) and number of Circle objects (2)

M09_LIAN1878_11_GE_C09.indd 363 12/29/17 4:11 PM

364 Chapter 9 Objects and Classes

When you compile TestCircleWithStaticMembers.java, the Java compiler automati- cally compiles Circle.java if it has not been compiled since the last change.

Static variables and methods can be accessed without creating objects. Line 6 displays the number of objects, which is 0, since no objects have been created.

The main method creates two circles, c1 and c2 (lines 9 and18). The instance variable radius in c1 is modified to become 9 (line 21). This change does not affect the instance variable radius in c2, since these two instance variables are independent. The static vari- able numberOfObjects becomes 1 after c1 is created (line 9), and it becomes 2 after c2 is created (line 18).

Note PI is a constant defined in Math and Math.PI references the constant. c1.numberO- fObjects (line 27) and c2.numberOfObjects (line 30) are better replaced by Circle. numberOfObjects. This improves readability because other programmers can easily rec- ognize the static variable. You can also replace Circle.numberOfObjects with Circle. getNumberOfObjects().

Tip Use ClassName.methodName(arguments) to invoke a static method and Class- Name.staticVariable to access a static variable. This improves readability because this makes static methods and data easy to spot.

An instance method can invoke an instance or static method, and access an instance or static data field. A static method can invoke a static method and access a static data field. However, a static method cannot invoke an instance method or access an instance data field, since static methods and static data fields don’t belong to a particular object. The relationship between static and instance members is summarized in the following diagram:

use class name

An instance method

invoke

access

invoke

access

An instance method

An instance data field

A static method

A static data field

A static method

invoke

access

invoke

access

An instance method

An instance data field

A static method

A static data field

For example, the following code is wrong.

1 public class A { 2 int i = 5; 3 static int k = 2; 4 5 public static void main(String[] args) { 6 int j = i; // Wrong because i is an instance variable 7 m1(); // Wrong because m1() is an instance method 8 } 9 10 public void m1() { 11 // Correct since instance and static variables and methods 12 // can be used in an instance method 13 i = i + k + m2(i, k); 14 } 15 16 public static int m2(int i, int j) { 17 return (int)(Math.pow(i, j)); 18 } 19 }

M09_LIAN1878_11_GE_C09.indd 364 12/29/17 4:11 PM

9.7 Static Variables, Constants, and Methods 365

Note if you replace the preceding code with the following new code, the program would be fine, because the instance data field i and method m1 are now accessed from an object a (lines 7 and 8):

1 public class A { 2 int i = 5; 3 static int k = 2; 4 5 public static void main(String[] args) { 6 A a = new A(); 7 int j = a.i; // OK, a.i accesses the object's instance variable 8 a.m1(); // OK, a.m1() invokes the object's instance method 9 } 10 11 public void m1() { 12 i = i + k + m2(i, k); 13 } 14 15 public static int m2(int i, int j) { 16 return (int)(Math.pow(i, j)); 17 } 18 }

Design Guide How do you decide whether a variable or a method should be instance or static? A variable or a method that is dependent on a specific instance of the class should be an instance variable or method. A variable or a method that is not dependent on a specific instance of the class should be a static variable or method. For example, every circle has its own radius, so the radius is dependent on a specific circle. Therefore, radius is an instance variable of the Circle class. Since the getArea method is dependent on a specific circle, it is an instance method. None of the methods in the Math class, such as random, pow, sin, and cos, is dependent on a specific instance. Therefore, these methods are static methods. The main method is static and can be invoked directly from a class.

Caution It is a common design error to define an instance method that should have been defined as static. For example, the method factorial(int n) should be defined as static, as shown next, because it is independent of any specific instance.

instance or static?

9.7.1 Suppose the class F is defined in (a). Let f be an instance of F. Which of the statements in (b) are correct? Point

Check

public class Test { public int factorial(int n) { int result = 1; for (int i = 1; i <= n; i++) result *= i;

return result; } }

(a) Wrong design

public class Test { public static int factorial(int n) { int result = 1; for (int i = 1; i <= n; i++) result *= i;

return result; } }

(b) Correct design

common design error

M09_LIAN1878_11_GE_C09.indd 365 12/29/17 4:11 PM

366 Chapter 9 Objects and Classes

9.7.2 Add the static keyword in the place of ? if appropriate. public class Test { int count;

public ? void main(String[] args) { ... }

public ? int getCount() { return count; }

public ? int factorial(int n) { int result = 1; for (int i = 1; i <= n; i++) result *= i;

return result; } }

9.7.3 Can you invoke an instance method or reference an instance variable from a static method? Can you invoke a static method or reference a static variable from an instance method? What is wrong in the following code?

1 public class C { 2 public static void main(String[] args) { 3 method1(); 4 } 5 6 public void method1() { 7 method2(); 8 } 9 10 public static void method2() { 11 System.out.println("What is radius " + c.getRadius()); 12 } 13 14 Circle c = new Circle(); 15 }

9.8 Visibility Modifiers Visibility modifiers can be used to specify the visibility of a class and its members.

You can use the public visibility modifier for classes, methods, and data fields to denote they can be accessed from any other classes. If no visibility modifier is used, then by default the classes, methods, and data fields are accessible by any class in the same package. This is known as package-private or package-access.

Point Key

package-private (or package-access)

public class F { int i; static String s;

void imethod() { }

static void smethod() { } }

(a)

System.out.println(f.i); System.out.println(f.s); f.imethod(); f.smethod(); System.out.println(F.i); System.out.println(F.s); F.imethod(); F.smethod();

(b)

M09_LIAN1878_11_GE_C09.indd 366 12/29/17 4:11 PM

9.8 Visibility Modifiers 367

Note Packages can be used to organize classes. To do so, you need to add the following line as the first noncomment and nonblank statement in the program:

package packageName;

If a class is defined without the package statement, it is said to be placed in the default package.

Java recommends that you place classes into packages rather than using a default package. For simplicity, however, this book uses default packages. For more information on packages, see Supplement III.E, Packages.

In addition to the public and default visibility modifiers, Java provides the private and protected modifiers for class members. This section introduces the private modifier. The protected modifier will be introduced in Section 11.14, The protected Data and Methods.

The private modifier makes methods and data fields accessible only from within its own class. Figure 9.14 illustrates how a public, default, and private data field or method in class C1 can be accessed from a class C2 in the same package, and from a class C3 in a different package.

using packages

Figure 9.14 The private modifier restricts access to its defining class, the default modifier restricts access to a package, and the public modifier enables unrestricted access.

package p1;

public class C1 { public int x; int y; private int z;

public void m1() { } void m2() { } private void m3() { } }

package p1;

public class C2 { void aMethod() { C1 c1 = new C1(); can access c1.x; can access c1.y; cannot access o.z;

can invoke c1.m1(); can invoke c1.m2(); cannot invoke c1.m3(); } }

package p2;

public class C3 { void aMethod() { C1 c1 = new C1(); can access c1.x; cannot access c1.y; cannot access c1.z;

can invoke c1.m1(); cannot invoke c1.m2(); cannot invoke c1.m3(); } }

If a class is not defined as public, it can be accessed only within the same package. As shown in Figure 9.15, C1 can be accessed from C2, but not from C3.

package p1;

class C1 { ... }

package p1;

public class C2 { can access C1 }

package p2;

public class C3 { cannot access p1.C1; can access p1.C2; }

Figure 9.15 A nonpublic class has package access.

A visibility modifier specifies how data fields and methods in a class can be accessed from outside the class. There is no restriction on accessing data fields and methods from inside the class. As shown in Figure 9.16b, an object c of class C cannot access its private members, because c is in the Test class. As shown in Figure 9.16a, an object c of class C can access its private members, because c is defined inside its own class.

inside access

M09_LIAN1878_11_GE_C09.indd 367 12/29/17 4:11 PM

368 Chapter 9 Objects and Classes

Caution The private modifier applies only to the members of a class. The public modifier can apply to a class or members of a class. Using the modifiers public and private on local variables would cause a compile error.

Note In most cases, the constructor should be public. However, if you want to prohibit the user from creating an instance of a class, use a private constructor. For example, there is no reason to create an instance from the Math class, because all of its data fields and methods are static. To prevent the user from creating objects from the Math class, the constructor in java.lang.Math is defined as follows:

private Math() { }

9.9 Data Field Encapsulation Making data fields private protects data and makes the class easy to maintain.

The data fields radius and numberOfObjects in the Circle class in Listing 9.6 can be modified directly (e.g., c1.radius = 5 or Circle.numberOfObjects = 10). This is not a good practice—for two reasons:

1. Data may be tampered with. For example, numberOfObjects is to count the number of objects created, but it may be mistakenly set to an arbitrary value (e.g., Circle. numberOfObjects = 10).

2. The class becomes difficult to maintain and vulnerable to bugs. Suppose that you want to modify the Circle class to ensure that the radius is nonnegative after other programs have already used the class. You have to change not only the Circle class but also the programs that use it because the clients may have modified the radius directly (e.g., c1.radius = –5).

To prevent direct modifications of data fields, you should declare the data fields private, using the private modifier. This is known as data field encapsulation.

private constructor

Point Key

Data field encapsulation

VideoNote

Data field encapsulation

data field encapsulation

Figure 9.16 An object can access its private members if it is defined in its own class.

public class C { private boolean x;

public static void main(string[] args) { C c = new C(); system.out.println(c.x); system.out.println(c.convert()); }

private int convert() { return x ? 1 : –1; } }

(a) This is okay because object c is used inside the class C.

public class Test { public static void main(string[] args) { C c = new C(); system.out.println(c.x); system.out.println(c.convert()); } }

(b) This is wrong because x and convert are private in class C.

M09_LIAN1878_11_GE_C09.indd 368 12/29/17 4:11 PM

9.9 Data Field Encapsulation 369

A private data field cannot be accessed by an object from outside the class that defines the private field. However, a client often needs to retrieve and modify a data field. To make a private data field accessible, provide a getter method to return its value. To enable a private data field to be updated, provide a setter method to set a new value. A getter method is also referred to as an accessor and a setter to a mutator. A getter method has the following signature:

public returnType getPropertyName()

If the returnType is boolean, the getter method should be defined as follows by convention:

public boolean isPropertyName()

A setter method has the following signature:

public void setPropertyName(dataType propertyValue)

Let’s create a new circle class with a private data-field radius and its associated accessor and mutator methods. The class diagram is shown in Figure 9.17. The new circle class is defined in Listing 9.8:

getter (or accessor) setter (or mutator)

boolean accessor

Figure 9.17 The Circle class encapsulates circle properties and provides getter/setter and other methods.

The - sign indicates a private modi�er

Circle

-radius: double

-numberOfObjects: int

+Circle()

+getArea(): double

+getNumberOfObjects(): int

+setRadius(radius: double): void

+getRadius(): double

+Circle(radius: double)

The radius of this circle (default: 1.0).

Constructs a default circle object.

The number of circle objects created.

Returns the area of this circle.

Returns the number of circle objects created.

Sets a new radius for this circle. Returns the radius of this circle.

Constructs a circle object with the speci�ed radius.

Listing 9.8 Circle.java(for CircleWithPrivateDataFields) 1 public class Circle { 2 /** The radius of the circle */ 3 private double radius = 1; 4 5 /** The number of objects created */ 6 private static int numberOfObjects = 0; 7 8 /** Construct a circle with radius 1 */ 9 public Circle() { 10 numberOfObjects++; 11 } 12 13 /** Construct a circle with a specified radius */ 14 public Circle(double newRadius) { 15 radius = newRadius; 16 numberOfObjects++;

encapsulate radius

encapsulate numberOfObjects

M09_LIAN1878_11_GE_C09.indd 369 12/29/17 4:11 PM

370 Chapter 9 Objects and Classes

17 } 18 19 /** Return radius */ 20 public double getRadius() { 21 return radius; 22 } 23 24 /** Set a new radius */ 25 public void setRadius(double newRadius) { 26 radius = (newRadius >= 0) ? newRadius : 0; 27 } 28 29 /** Return numberOfObjects */ 30 public static int getNumberOfObjects() { 31 return numberOfObjects; 32 } 33 34 /** Return the area of this circle */ 35 public double getArea() { 36 return radius * radius * Math.PI; 37 } 38 }

The getRadius() method (lines 20–22) returns the radius and the setRadius(newRadius) method (lines 25–27) sets a new radius for the object. If the new radius is negative, 0 is set as the radius for the object. Since these methods are the only ways to read and modify the radius, you have total control over how the radius property is accessed. If you have to change the implementation of these methods, you don’t need to change the client programs. This makes the class easy to maintain.

Listing 9.9 gives a client program that uses the Circle class to create a Circle object, and modifies the radius using the setRadius method.

Listing 9.9 TestCircleWithPrivateDataFields.java 1 public class TestCircleWithPrivateDataFields { 2 /** Main method */ 3 public static void main(String[] args) { 4 // Create a circle with radius 5.0 5 Circle myCircle = new Circle(5.0); 6 System.out.println("The area of the circle of radius " 7 + myCircle.getRadius() + " is " + myCircle.getArea()); 8 9 // Increase myCircle's radius by 10% 10 myCircle.setRadius(myCircle.getRadius() * 1.1); 11 System.out.println("The area of the circle of radius " 12 + myCircle.getRadius() + " is " + myCircle.getArea()); 13 14 System.out.println("The number of objects created is " 15 + Circle.getNumberOfObjects()); 16 } 17 }

The data field radius is declared private. Private data can be accessed only within their defin- ing class, so you cannot use myCircle.radius in the client program. A compile error would occur if you attempted to access private data from a client.

Since numberOfObjects is private, it cannot be modified. This prevents tampering. For example, the user cannot set numberOfObjects to 100. The only way to make it 100 is to create 100 objects of the Circle class.

accessor method

mutator method

accessor method

invoke public method

M09_LIAN1878_11_GE_C09.indd 370 12/29/17 4:11 PM

9.10 Passing Objects to Methods 371

Suppose you combined TestCircleWithPrivateDataFields and Circle into one class by moving the main method in TestCircleWithPrivateDataFields into Circle. Could you use myCircle.radius in the main method? See CheckPoint Question 9.9.3 for the answer.

Design Guide To prevent data from being tampered with and to make the class easy to maintain, declare data fields private.

Note From now on, all data fields should be declared private, and all constructors and methods should be defined public, unless specified otherwise.

9.9.1 What is an accessor method? What is a mutator method? What are the naming con- ventions for accessor methods and mutator methods?

9.9.2 What are the benefits of data field encapsulation? 9.9.3 In the following code, radius is private in the Circle class, and myCircle is an object

of the Circle class. Does the highlighted code cause any problems? If so, explain why.

public class Circle { private double radius = 1;

/** Find the area of this circle */ public double getArea() { return radius * radius * Math.PI; }

public static void main(String[] args) { Circle myCircle = new Circle(); System.out.println("Radius is " + myCircle.radius); } }

9.10 Passing Objects to Methods Passing an object to a method is to pass the reference of the object.

You can pass objects to methods. Like passing an array, passing an object is actually passing the reference of the object. The following code passes the myCircle object as an argument to the printCircle method:

1 public class Test { 2 public static void main(String[] args) { 3 // Circle is defined in Listing 9.8 4 Circle myCircle = new Circle(5.0); 5 printCircle(myCircle); 6 } 7 8 public static void printCircle(Circle c) { 9 System.out.println("The area of the circle of radius " 10 + c.getRadius() + " is " + c.getArea()); 11 } 12 }

Java uses exactly one mode of passing arguments: pass-by-value. In the preceding code, the value of myCircle is passed to the printCircle method. This value is a reference to a Circle object.

The program in Listing 9.10 demonstrates the difference between passing a primitive-type value and passing a reference value.

Point Check

Point Key

pass an object

pass-by-value

M09_LIAN1878_11_GE_C09.indd 371 12/29/17 4:11 PM

372 Chapter 9 Objects and Classes

Listing 9.10 TestPassObject.java 1 public class TestPassObject { 2 /** Main method */ 3 public static void main(String[] args) { 4 // Create a Circle object with radius 1 5 Circle myCircle = 6 new Circle(1); // Use the Circle class in Listing 9.8 7 8 // Print areas for radius 1, 2, 3, 4, and 5. 9 int n = 5; 10 printAreas(myCircle, n); 11 12 // See myCircle.radius and times 13 System.out.println("\n" + "Radius is " + myCircle.getRadius()); 14 System.out.println("n is " + n); 15 } 16 17 /** Print a table of areas for radius */ 18 public static void printAreas(Circle c, int times) { 19 System.out.println("Radius \t\tArea"); 20 while (times >= 1) { 21 System.out.println(c.getRadius() + "\t\t" + c.getArea()); 22 c.setRadius(c.getRadius() + 1); 23 times——; 24 } 25 } 26 }

pass object

object parameter

Radius Area

1.0 3.141592653589793

2.0 12.566370614359172

3.0 28.274333882308138

4.0 50.26548245743669

5.0 78.53981633974483

Radius is 6.0

n is 5

The Circle class is defined in Listing 9.8. The program passes a Circle object myCircle and an integer value from n to invoke printAreas(myCircle, n) (line 10), which prints a table of areas for radii 1, 2, 3, 4, and 5, as presented in the sample output.

Figure 9.18 shows the call stack for executing the methods in the program. Note the objects are stored in a heap (see Section 7.6).

When passing an argument of a primitive data type, the value of the argument is passed. In this case, the value of n (5) is passed to times. Inside the printAreas method, the content of times is changed; this does not affect the content of n.

When passing an argument of a reference type, the reference of the object is passed. In this case, c contains a reference for the object that is also referenced via myCircle. Therefore, changing the properties of the object through c inside the printAreas method has the same effect as doing so outside the method through the variable myCircle. Pass-by-value on refer- ences can be best described semantically as pass-by-sharing; that is, the object referenced in the method is the same as the object being passed.

pass-by-sharing

M09_LIAN1878_11_GE_C09.indd 372 12/29/17 4:11 PM

9.10 Passing Objects to Methods 373

9.10.1 Describe the difference between passing a parameter of a primitive type and passing a parameter of a reference type. Show the output of the following programs: Point

Check

Figure 9.18 The value of n is passed to times, and the reference to myCircle is passed to c in the printAreas method.

Activation record for the main method int n: 5 myCircle:

Stack

Activation record for the printAreas method int times: 5 Circle c:

A Circle object

Heap

reference

Pass-by-value (here the value is the reference for the object)

Pass-by-value (here the value is 5)

public class Test { public static void main(String[] args) { Count myCount = new Count(); int times = 0;

for (int i = 0; i < 100; i++) increment(myCount, times);

System.out.println("count is " + myCount.count); System.out.println("times is " + times); }

public static void increment(Count c, int times) { c.count++; times++; } }

public class Count {

public int count;

public Count (int c) { count = c; }

public Count () { count = 1; } }

9.10.2 Show the output of the following program: public class Test { public static void main(String[] args) { Circle circle1 = new Circle(1); Circle circle2 = new Circle(2);

swap1(circle1, circle2); System.out.println("After swap1: circle1 = " + circle1.radius + " circle2 = " + circle2.radius);

swap2(circle1, circle2); System.out.println("After swap2: circle1 = " + circle1.radius + " circle2 = " + circle2.radius); }

public static void swap1(Circle x, Circle y) { Circle temp = x; x = y; y = temp; }

M09_LIAN1878_11_GE_C09.indd 373 12/29/17 4:11 PM

374 Chapter 9 Objects and Classes

public static void swap2(Circle x, Circle y) { double temp = x.radius; x.radius = y.radius; y.radius = temp; } }

class Circle { double radius;

Circle(double newRadius) { radius = newRadius; } }

9.10.3 Show the output of the following code:

public class Test { public static void main(String[] args) { int[] a = {1, 2}; swap(a[0], a[1]); System.out.println("a[0] = " + a[0] + " a[1] = " + a[1]); }

public static void swap(int n1, int n2) { int temp = n1; n1 = n2; n2 = temp; } }

(a)

public class Test { public static void main(String[] args) { int[] a = {1, 2}; swap(a); System.out.println("a[0] = " + a[0] + " a[1] = " + a[1]); }

public static void swap(int[] a) { int temp = a[0]; a[0] = a[1]; a[1] = temp; } }

(b)

public class Test { public static void main(String[] args) { T t = new T(); swap(t); System.out.println("e1 = " + t.e1 + " e2 = " + t.e2); }

public static void swap(T t) { int temp = t.e1; t.e1 = t.e2; t.e2 = temp; } }

class T { int e1 = 1; int e2 = 2; }

(c)

public class Test { public static void main(String[] args) { T t1 = new T(); T t2 = new T(); System.out.println("t1's i = " + t1.i + " and j = " + t1.j); System.out.println("t2's i = " + t2.i + " and j = " + t2.j); } }

class T { static int i = 0; int j = 0;

T() { i++; j = 1; } }

(d)

M09_LIAN1878_11_GE_C09.indd 374 12/29/17 4:11 PM

9.11 Array of Objects 375

9.10.4 What is the output of the following programs?

import java.util.Date;

public class Test { public static void main(String[] args) { Date date = null; m1(date); System.out.println(date); }

public static void m1(Date date) { date = new Date(); } }

(a)

import java.util.Date;

public class Test { public static void main(String[] args) { Date date = new Date(1234567); m1(date); System.out.println(date.getTime()); }

public static void m1(Date date) { date = new Date(7654321); } }

(b)

9.11 Array of Objects An array can hold objects as well as primitive-type values.

Chapter 7, Single-Dimensional Arrays, described how to create arrays of primitive-type ele- ments. You can also create arrays of objects. For example, the following statement declares and creates an array of 10 Circle objects:

Circle[] circleArray = new Circle[10];

To initialize circleArray, you can use a for loop as follows:

for (int i = 0; i < circleArray.length; i++) { circleArray[i] = new Circle(); }

An array of objects is actually an array of reference variables. Thus, invoking circleArray[1] .getArea() involves two levels of referencing, as shown in Figure 9.19. circleArray references the entire array, and circleArray[1] references a Circle object.

Note When an array of objects is created using the new operator, each element in the array is a reference variable with a default value of null.

Point Key

import java.util.Date;

public class Test { public static void main(String[] args) { Date date = new Date(1234567); m1(date); System.out.println(date.getTime()); }

public static void m1(Date date) { date.setTime(7654321); } }

(c)

import java.util.Date;

public class Test { public static void main(String[] args) { Date date = new Date(1234567); m1(date); System.out.println(date.getTime()); }

public static void m1(Date date) { date = null; } }

(d)

M09_LIAN1878_11_GE_C09.indd 375 12/29/17 4:11 PM

376 Chapter 9 Objects and Classes

Listing 9.11 gives an example that demonstrates how to use an array of objects. The pro- gram summarizes the areas of an array of circles. The program creates circleArray, an array composed of five Circle objects; it then initializes circle radii with random values and displays the total area of the circles in the array.

Listing 9.11 TotalArea.java 1 public class TotalArea { 2 /** Main method */ 3 public static void main(String[] args) { 4 // Declare circleArray 5 Circle[] circleArray; 6 7 // Create circleArray 8 circleArray = createCircleArray(); 9 10 // Print circleArray and total areas of the circles 11 printCircleArray(circleArray); 12 } 13 14 /** Create an array of Circle objects */ 15 public static Circle[] createCircleArray() { 16 Circle[] circleArray = new Circle[5]; 17 18 for (int i = 0; i < circleArray.length; i++) { 19 circleArray[i] = new Circle(Math.random() * 100); 20 } 21 22 // Return Circle array 23 return circleArray; 24 } 25 26 /** Print an array of circles and their total area */ 27 public static void printCircleArray(Circle[] circleArray) { 28 System.out.printf("%–30s%–15s\n", "Radius", "Area"); 29 for (int i = 0; i < circleArray.length; i++) { 30 System.out.printf("%–30f%–15f\n", circleArray[i].getRadius(), 31 circleArray[i].getArea()); 32 } 33 34 System.out.println("— — — — — — — — — — — — — — — — — — — — — — — — — — — —"); 35 36 // Compute and display the result 37 System.out.printf("%–30s%–15f\n", "The total area of circles is", 38 sum(circleArray)); 39 } 40

array of objects

return array of objects

pass array of objects

Figure 9.19 In an array of objects, an element of the array contains a reference to an object.

…

circleArray[9]

circleArray[1]

circleArray[0]reference Circle object 0

Circle object 1

Circle object 9

circleArray

M09_LIAN1878_11_GE_C09.indd 376 12/29/17 4:11 PM

9.12 Immutable Objects and Classes 377

41 /** Add circle areas */ 42 public static double sum(Circle[] circleArray) { 43 // Initialize sum 44 double sum = 0; 45 46 // Add areas to sum 47 for (int i = 0; i < circleArray.length; i++) 48 sum += circleArray[i].getArea(); 49 50 return sum; 51 } 52 }

pass array of objects

Radius Area

70.577708 15649.941866

44.152266 6124.291736

24.867853 1942.792644

5.680718 101.380949

36.734246 4239.280350

—————————————————————————————————————————————

The total area of circles is 28056.687544

The program invokes createCircleArray() (line 8) to create an array of five circle objects. Several circle classes were introduced in this chapter. This example uses the Circle class introduced in Section 9.9, Data Field Encapsulation.

The circle radii are randomly generated using the Math.random() method (line 19). The createCircleArray method returns an array of Circle objects (line 23). The array is passed to the printCircleArray method, which displays the radius and area of each circle and the total area of the circles.

The sum of the circle areas is computed by invoking the sum method (line 38), which takes the array of Circle objects as the argument and returns a double value for the total area.

9.11.1 What is wrong in the following code? 1 public class Test { 2 public static void main(String[] args) { 3 java.util.Date[] dates = new java.util.Date[10]; 4 System.out.println(dates[0]); 5 System.out.println(dates[0].toString()); 6 } 7 }

9.12 Immutable Objects and Classes You can define immutable classes to create immutable objects. The contents of immutable objects cannot be changed.

Normally, you create an object and allow its contents to be changed later. However, occasion- ally it is desirable to create an object whose contents cannot be changed once the object has been created. We call such an object as immutable object and its class as immutable class. The String class, for example, is immutable. If you deleted the setter method in the Circle class in Listing 9.8, the class would be immutable because radius is private and cannot be changed without a setter method.

If a class is immutable, then all its data fields must be private and it cannot contain public setter methods for any data fields. A class with all private data fields and no mutators is not

Point Key

immutable class

VideoNote Immutable objects and this keyword

immutable object

M09_LIAN1878_11_GE_C09.indd 377 12/29/17 4:11 PM

378 Chapter 9 Objects and Classes

necessarily immutable. For example, the following Student class has all private data fields and no setter methods, but it is not an immutable class:

1 public class Student { 2 private int id; 3 private String name; 4 private java.util.Date dateCreated; 5 6 public Student(int ssn, String newName) { 7 id = ssn; 8 name = newName; 9 dateCreated = new java.util.Date(); 10 } 11 12 public int getId() { 13 return id; 14 } 15 16 public String getName() { 17 return name; 18 } 19 20 public java.util.Date getDateCreated() { 21 return dateCreated; 22 } 23 }

As shown in the following code, the data field dateCreated is returned using the get- DateCreated() method. This is a reference to a Date object. Through this reference, the content for dateCreated can be changed.

public class Test { public static void main(String[] args) { Student student = new Student(111223333, "John"); java.util.Date dateCreated = student.getDateCreated(); dateCreated.setTime(200000); // Now dateCreated field is changed! } }

For a class to be immutable, it must meet the following requirements:

■■ All data fields must be private.

■■ There can’t be any mutator methods for data fields.

■■ No accessor methods can return a reference to a data field that is mutable.

Interested readers may refer to Supplement III.U for an extended discussion on immutable objects.

9.12.1 If a class contains only private data fields and no setter methods, is the class immutable?

9.12.2 If all the data fields in a class are private and of primitive types, and the class doesn’t contain any setter methods, is the class immutable?

9.12.3 Is the following class immutable? public class A { private int[] values;

public int[] getValues() {

Student class

Point Check

M09_LIAN1878_11_GE_C09.indd 378 12/29/17 4:11 PM

9.13 The Scope of Variables 379

return values; } }

9.13 The Scope of Variables The scope of instance and static variables is the entire class, regardless of where the variables are declared.

Section 6.9 discussed local variables and their scope rules. Local variables are declared and used inside a method locally. This section discusses the scope rules of all the variables in the context of a class.

Instance and static variables in a class are referred to as the class’s variables or data fields. A variable defined inside a method is referred to as a local variable. The scope of a class’s variables is the entire class, regardless of where the variables are declared. A class’s variables and methods can appear in any order in the class, as shown in Figure 9.20a. The exception is when a data field is initialized based on a reference to another data field. In such cases, the other data field must be declared first, as shown in Figure 9.20b. For consistency, this book declares data fields at the beginning of the class.

Point Key

class’s variables

Figure 9.20 Members of a class can be declared in any order, with one exception.

public class Circle { public double getArea() { return radius * radius * Math.PI; }

private double radius = 1; }

(a) The variable radius and method getArea() can be declared in any order.

public class F { private int i; private int j = i + 1; }

(b) i has to be declared before j because j’s initial value is dependent on i.

You can declare a class’s variable only once, but you can declare the same variable name in a method many times in different nonnesting blocks.

If a local variable has the same name as a class’s variable, the local variable takes prece- dence and the class’s variable with the same name is hidden. For example, in the following program, x is defined both as an instance variable and as a local variable in the method:

public class F { private int x = 0; // Instance variable private int y = 0;

public F() { }

public void p() { int x = 1; // Local variable System.out.println("x = " + x); System.out.println("y = " + y); } }

What is the output for f.p(), where f is an instance of F? The output for f.p() is 1 for x and 0 for y. Here is why:

■■ x is declared as a data field with the initial value of 0 in the class, but it is also declared in the method p() with an initial value of 1. The latter x is referenced in the System.out.println statement.

■■ y is declared outside the method p(), but y is accessible inside the method.

hidden variables

M09_LIAN1878_11_GE_C09.indd 379 12/29/17 4:11 PM

380 Chapter 9 Objects and Classes

Tip To avoid confusion and mistakes, do not use the names of instance or static variables as local variable names, except for method parameters. We will discuss hidden data fields by method parameters in the next section.

9.13.1 What is the output of the following program? public class Test { private static int i = 0; private static int j = 0;

public static void main(String[] args) { int i = 2; int k = 3;

{ int j = 3; System.out.println("i + j is " + i + j); }

k = i + j; System.out.println("k is " + k); System.out.println("j is " + j); } }

9.14 The this Reference The keyword this refers to the object itself. It can also be used inside a constructor to invoke another constructor of the same class.

The this keyword is the name of a reference that an object can use to refer to itself. You can use the this keyword to reference the object’s instance members. For example, the following code in (a) uses this to reference the object’s radius and invokes its getArea() method explicitly. The this reference is normally omitted for brevity as shown in (b). However, the this reference is needed to reference a data field hidden by a method or constructor parameter, or to invoke an overloaded constructor.

Point Check

Point Key

this keyword

VideoNote

The this keyword

public class Circle { private double radius;

...

public double getArea() { return this.radius * this.radius * Math.PI; }

public String toString() { return "radius: " + this.radius + "area: " + this.getArea(); } }

(a)

public class Circle { private double radius;

...

public double getArea() { return radius * radius * Math.PI; }

public String toString() { return "radius: " + radius + "area: " + getArea(); } }

(b)

Equivalent

M09_LIAN1878_11_GE_C09.indd 380 12/29/17 4:11 PM

9.14 The this Reference 381

9.14.1 Using this to Reference Data Fields It is a good practice to use the data field as the parameter name in a setter method or a con- structor to make the code easy to read and to avoid creating unnecessary names. In this case, you need to use the this keyword to reference the data field in the setter method. For exam- ple, the setRadius method can be implemented as shown in (a). It would be wrong if it is implemented as shown in (b).

The data field radius is hidden by the parameter radius in the setter method. You need to reference the data field name in the method using the syntax this.radius. A hidden static variable can be accessed simply by using the ClassName.staticVariable reference. A hidden instance variable can be accessed by using the keyword this, as shown in Figure 9.21a.

reference data fields

public class F { private int i = 5; private static double k = 0;

public void setI(int i) { this.i = i; }

public static void setK(double k) { F.k = k; }

// other methods omitted }

(a)

Suppose that f1 and f2 are two objects of F.

Invoking f1.setI(10) is to execute

this.i = 10, where this refers f1

Invoking f2.setI(45) is to execute

this.i = 45, where this refers f2

Invoking F.setK(33) is to execute

F.k = 33. setK is a static method

(b)

Figure 9.21 The keyword this refers to the calling object that invokes the method.

The this keyword gives us a way to reference the object that invokes an instance method. To invoke f1.setI(10), this.i = i is executed, which assigns the value of parameter i to the data field i of this calling object f1. The keyword this refers to the object that invokes the instance method setI, as shown in Figure 9.21b. The line F.k = k means the value in parameter k is assigned to the static data field k of the class, which is shared by all the objects of the class.

private double radius;

public void setRadius(double radius) { this.radius = radius; }

private double radius = 1;

public void setRadius(double radius) { radius = radius; }

Refers to data field radius in this object.

Here, radius is the parameter in the method.

(a) this.radius refers the radius data field in this object.

(b) radius is the parameter defined in the method header.

M09_LIAN1878_11_GE_C09.indd 381 12/29/17 4:11 PM

382 Chapter 9 Objects and Classes

The line this(1.0) in the second constructor invokes the first constructor with a double value argument.

Note Java requires that the this(arg-list) statement appear first in the constructor before any other executable statements.

Tip If a class has multiple constructors, it is better to implement them using this(arg-list) as much as possible. In general, a constructor with no or fewer arguments can invoke a constructor with more arguments using this(arg-list). This syntax often simplifies coding and makes the class easier to read and to maintain.

9.14.1 Describe the role of the this keyword. 9.14.2 What is wrong in the following code?

1 public class C { 2 private int p; 3 4 public C() { 5 System.out.println("C's no-arg constructor invoked"); 6 this(0); 7 } 8 9 public C(int p) { 10 p = p; 11 } 12 13 public void setP(int p) { 14 p = p; 15 } 16 }

9.14.3 What is wrong in the following code? public class Test { private int id;

public void m1() { this.id = 45; }

Point Check

public class Circle { private double radius; public Circle(double radius) { this.radius = radius; } The this keyword is used to reference the

data field radius of the object being constructed. public Circle() { this(1.0); } The this keyword is used to invoke another

constructor. ... }

9.14.2 Using this to Invoke a Constructor The this keyword can be used to invoke another constructor of the same class. For example, you can rewrite the Circle class as follows:

M09_LIAN1878_11_GE_C09.indd 382 12/29/17 4:11 PM

Chapter Summary 383

public void m2() { Test.id = 45; } }

Key Terms action 346 anonymous object 355 attribute 346 behavior 346 class 346 class’s variable 379 client 349 constructor 346 data field 346 data field encapsulation 368 default constructor 353 dot operator (.) 354 getter (or accessor) 369 instance 346 instance method 355 instance variable 355 instantiation 346 immutable class 377

immutable object 377 no-arg constructor 349 null value 355 object 346 object-oriented programming (OOP) package-private (or package-access) 366 private constructor 368 property 346 public class 349 reference type 354 reference variable 354 setter (or mutator) 369 state 346 static method 361 static variable 361 this keyword 380 Unified Modeling Language (UML) 347

ChapTer summary 1. A class is a template for objects. It defines the properties of objects and provides con-

structors for creating objects and methods for manipulating them.

2. A class is also a data type. You can use it to declare object reference variables. An object reference variable that appears to hold an object actually contains a reference to that object. Strictly speaking, an object reference variable and an object are different, but most of the time the distinction can be ignored.

3. An object is an instance of a class. You use the new operator to create an object and the dot operator (.) to access members of that object through its reference variable.

4. An instance variable or method belongs to an instance of a class. Its use is associated with individual instances. A static variable is a variable shared by all instances of the same class. A static method is a method that can be invoked without using instances.

5. Every instance of a class can access the class’s static variables and methods. For clar- ity, however, it is better to invoke static variables and methods using ClassName . variable and ClassName.method.

6. Visibility modifiers specify how the class, method, and data are accessed. A public class, method, or data is accessible to all clients. A private method or data is acces- sible only inside the class.

7. You can provide a getter (accessor) method or a setter (mutator) method to enable clients to see or modify the data.

M09_LIAN1878_11_GE_C09.indd 383 12/29/17 4:11 PM

384 Chapter 9 Objects and Classes

8. A getter method has the signature public returnType getPropertyName(). If the returnType is boolean, the getter method should be defined as public boolean isPropertyName(). A setter method has the signature public void setPropertyName(dataType propertyValue).

9. All parameters are passed to methods using pass-by-value. For a parameter of a primi- tive type, the actual value is passed; for a parameter of a reference type, the reference for the object is passed.

10. A Java array is an object that can contain primitive-type values or object-type values. When an array of objects is created, its elements are assigned the default value of null.

11. Once it is created, an immutable object cannot be modified. To prevent users from modi- fying an object, you can define immutable classes.

12. The scope of instance and static variables is the entire class, regardless of where the vari- ables are declared. Instance and static variables can be declared anywhere in the class. For consistency, they are declared at the beginning of the class in this book.

13. The keyword this can be used to refer to the calling object. It can also be used inside a constructor to invoke another constructor of the same class.

Quiz Answer the quiz for this chapter online at the book Companion Website.

programming exerCises

Pedagogical Note The exercises in Chapters 9–13 help you to achieve three objectives:

1. Design classes and draw UML class diagrams. 2. Implement classes from the UML. 3. Use classes to develop applications.

Students can download solutions for the UML diagrams for the even-numbered exercises from the Companion Website and instructors can download all solutions from the same site.

Starting from Section 9.7, all data fields should be declared private and all constructors and methods should be defined public unless specified otherwise.

Sections 9.2–9.5 9.1 (The Rectangle class) Following the example of the Circle class in Section 9.2,

design a class named Rectangle to represent a rectangle. The class contains:

■■ Two double data fields named width and height that specify the width and height of the rectangle. The default values are 1 for both width and height.

■■ A no-arg constructor that creates a default rectangle. ■■ A constructor that creates a rectangle with the specified width and height. ■■ A method named getArea() that returns the area of this rectangle. ■■ A method named getPerimeter() that returns the perimeter.

Draw the UML diagram for the class then implement the class. Write a test pro- gram that creates two Rectangle objects—one with width 4 and height 40, and

three objectives

M09_LIAN1878_11_GE_C09.indd 384 12/29/17 4:11 PM

Programming Exercises 385

the other with width 3.5 and height 35.9. Display the width, height, area, and perimeter of each rectangle in this order.

9.2 (The Stock class) Following the example of the Circle class in Section 9.2, design a class named Stock that contains:

■■ A string data field named symbol for the stock’s symbol. ■■ A string data field named name for the stock’s name. ■■ A double data field named previousClosingPrice that stores the stock

price for the previous day. ■■ A double data field named currentPrice that stores the stock price for the

current time. ■■ A constructor that creates a stock with the specified symbol and name. ■■ A method named getChangePercent() that returns the percentage changed

from previousClosingPrice to currentPrice.

Draw the UML diagram for the class then implement the class. Write a test pro- gram that creates a Stock object with the stock symbol ORCL, the name Oracle Corporation, and the previous closing price of 34.5. Set a new current price to 34.35 and display the price-change percentage.

Section 9.6 *9.3 (Use the Date class) Write a program that creates a Date object, sets its elapsed

time to 10000, 100000, 1000000, 10000000, 100000000, 1000000000, 10000000000, and 100000000000, and displays the date and time using the toString() method, respectively.

*9.4 (Use the Random class) Write a program that creates a Random object with seed 1000 and displays the first 50 random integers between 0 and 100 using the nextInt(100) method.

*9.5 (Use the GregorianCalendar class) Java API has the GregorianCalendar class in the java.util package, which you can use to obtain the year, month, and day of a date. The no-arg constructor constructs an instance for the current date, and the methods get(GregorianCalendar.YEAR), get(GregorianCalendar. MONTH), and get(GregorianCalendar.DAY_OF_MONTH) return the year, month, and day. Write a program to perform two tasks:

1. Display the current year, month, and day. 2. The GregorianCalendar class has the setTimeInMillis(long), which

can be used to set a specified elapsed time since January 1, 1970. Set the value to 1234567898765L and display the year, month, and day.

Sections 9.7–9.9 *9.6 (Stopwatch) Design a class named StopWatch. The class contains:

■■ Private data fields startTime and endTime with getter methods. ■■ A no-arg constructor that initializes startTime with the current time. ■■ A method named start() that resets the startTime to the current time. ■■ A method named stop() that sets the endTime to the current time. ■■ A method named getElapsedTime() that returns the elapsed time for the

stopwatch in milliseconds.

Draw the UML diagram for the class then implement the class. Write a test program that measures the execution time of sorting 100,000 numbers using selection sort.

9.7 (The Account class) Design a class named Account that contains: ■■ A private int data field named id for the account (default 0). ■■ A private double data field named balance for the account (default 0).

M09_LIAN1878_11_GE_C09.indd 385 12/29/17 4:11 PM

386 Chapter 9 Objects and Classes

■■ A private double data field named annualInterestRate that stores the cur- rent interest rate (default 0). Assume that all accounts have the same interest rate.

■■ A private Date data field named dateCreated that stores the date when the account was created.

■■ A no-arg constructor that creates a default account. ■■ A constructor that creates an account with the specified id and initial balance. ■■ The accessor and mutator methods for id, balance, and annualInterestRate. ■■ The accessor method for dateCreated. ■■ A method named getMonthlyInterestRate() that returns the monthly

interest rate. ■■ A method named getMonthlyInterest() that returns the monthly interest. ■■ A method named withdraw that withdraws a specified amount from the

account. ■■ A method named deposit that deposits a specified amount to the account.

Draw the UML diagram for the class then implement the class. (Hint: The method getMonthlyInterest() is to return monthly interest, not the interest rate. Monthly interest is balance * monthlyInterestRate. monthlyInteres- tRate is annualInterestRate / 12. Note annualInterestRate is a per- centage, for example 4.5%. You need to divide it by 100.)

Write a test program that creates an Account object with an account ID of 1122, a balance of $20,000, and an annual interest rate of 4.5%. Use the withdraw method to withdraw $2,500, use the deposit method to deposit $3,000, and print the bal- ance, the monthly interest, and the date when this account was created.

9.8 (The Fan class) Design a class named Fan to represent a fan. The class contains: ■■ Three constants named SLOW, MEDIUM, and FAST with the values 1, 2, and 3

to denote the fan speed. ■■ A private int data field named speed that specifies the speed of the fan (the

default is SLOW). ■■ A private boolean data field named on that specifies whether the fan is on (the

default is false). ■■ A private double data field named radius that specifies the radius of the fan

(the default is 5). ■■ A string data field named color that specifies the color of the fan (the default

is blue). ■■ The accessor and mutator methods for all four data fields. ■■ A no-arg constructor that creates a default fan. ■■ A method named toString() that returns a string description for the fan. If

the fan is on, the method returns the fan speed, color, and radius in one com- bined string. If the fan is not on, the method returns the fan color and radius along with the string “fan is off” in one combined string.

Draw the UML diagram for the class then implement the class. Write a test program that creates two Fan objects. Assign maximum speed, radius 10, color yellow, and turn it on to the first object. Assign medium speed, radius 5, color blue, and turn it off to the second object. Display the objects by invoking their toString method.

**9.9 (Geometry: n-sided regular polygon) In an n-sided regular polygon, all sides have the same length and all angles have the same degree (i.e., the polygon is both equi- lateral and equiangular). Design a class named RegularPolygon that contains:

■■ A private int data field named n that defines the number of sides in the polygon with default value 3.

VideoNote

The Fan class

M09_LIAN1878_11_GE_C09.indd 386 12/29/17 4:11 PM

Programming Exercises 387

■■ A private double data field named side that stores the length of the side with default value 1.

■■ A private double data field named x that defines the x-coordinate of the poly- gon’s center with default value 0.

■■ A private double data field named y that defines the y-coordinate of the poly- gon’s center with default value 0.

■■ A no-arg constructor that creates a regular polygon with default values. ■■ A constructor that creates a regular polygon with the specified number of sides

and length of side, centered at (0, 0). ■■ A constructor that creates a regular polygon with the specified number of sides,

length of side, and x- and y-coordinates. ■■ The accessor and mutator methods for all data fields. ■■ The method getPerimeter() that returns the perimeter of the polygon. ■■ The method getArea() that returns the area of the polygon. The formula for

computing the area of a regular polygon is

Area = n * s2

4 * tanap n b

Draw the UML diagram for the class then implement the class. Write a test pro- gram that creates three RegularPolygon objects, created using the no-arg con- structor, using RegularPolygon(6, 4), and using RegularPolygon(10, 4, 5.6, 7.8). For each object, display its perimeter and area.

*9.10 (Algebra: quadratic equations) Design a class named QuadraticEquation for a quadratic equation ax2 + bx + c = 0. The class contains:

■■ Private data fields a, b, and c that represent three coefficients. ■■ A constructor with the arguments for a, b, and c. ■■ Three getter methods for a, b, and c. ■■ A method named getDiscriminant() that returns the discriminant, which

is b2 - 4ac. ■■ The methods named getRoot1() and getRoot2() for returning two roots

of the equation

r1 = -b + 2b2 - 4ac

2a and r2 =

-b - 2b2 - 4ac 2a

These methods are useful only if the discriminant is nonnegative. Let these methods return 0 if the discriminant is negative.

Draw the UML diagram for the class then implement the class. Write a test pro- gram that prompts the user to enter values for a, b, and c and displays the result based on the discriminant. If the discriminant is positive, display the two roots. If the discriminant is 0, display the one root. Otherwise, display “The equation has no roots.” See Programming Exercise 3.1 for sample runs.

*9.11 (Algebra: 2 * 2 linear equations) Design a class named LinearEquation for a 2 * 2 system of linear equations:

ax + by = e cx + dy = f

x = ed - bf ad - bc

y = af - ec ad - bc

The class contains:

M09_LIAN1878_11_GE_C09.indd 387 12/29/17 4:11 PM

388 Chapter 9 Objects and Classes

■■ Private data fields a, b, c, d, e, and f. ■■ A constructor with the arguments for a, b, c, d, e, and f. ■■ Six getter methods for a, b, c, d, e, and f. ■■ A method named isSolvable() that returns true if ad - bc is not 0. ■■ Methods getX() and getY() that return the solution for the equation.

Draw the UML diagram for the class then implement the class. Write a test pro- gram that prompts the user to enter a, b, c, d, e, and f and displays the result. If ad - bc is 0, report that “The equation has no solution.” See Programming Exer- cise 3.3 for sample runs.

**9.12 (Geometry: intersecting point) Suppose two line segments intersect. The two end- points for the first line segment are (x1, y1) and (x2, y2) and for the second line segment are (x3, y3) and (x4, y4). Write a program that prompts the user to enter these four endpoints and displays the intersecting point. As discussed in Program- ming Exercise 3.25, the intersecting point can be found by solving a linear equa- tion. Use the LinearEquation class in Programming Exercise 9.11 to solve this equation. See Programming Exercise 3.25 for sample runs.

**9.13 (The Location class) Design a class named Location for locating a maximal value and its location in a two-dimensional array. The class contains public data fields row, column, and maxValue that store the maximal value and its indices in a two-dimensional array with row and column as int types and maxValue as a double type.

Write the following method that returns the location of the largest element in a two-dimensional array:

public static Location locateLargest(double[][] a)

The return value is an instance of Location. Write a test program that prompts the user to enter a two-dimensional array and displays the location of the largest element in the array. Here is a sample run:

Enter the number of rows and columns in the array: 3 4 Enter the array:

23.5 35 2 10

4.5 3 45 3.5

35 44 5.5 9.6 The location of the largest element is 45 at (1, 2)

M09_LIAN1878_11_GE_C09.indd 388 12/29/17 4:11 PM

Objectives ■■ To apply class abstraction to develop software (§10.2).

■■ To explore the differences between the procedural paradigm and object-oriented paradigm (§10.3).

■■ To discover the relationships between classes (§10.4).

■■ To design programs using the object-oriented paradigm (§§10.5 and 10.6).

■■ To create objects for primitive values using the wrapper classes (Byte, Short, Integer, Long, Float, Double, Character, and Boolean) (§10.7).

■■ To simplify programming using automatic conversion between primitive types and wrapper class types (§10.8).

■■ To use the BigInteger and BigDecimal classes for computing very large numbers with arbitrary precisions (§10.9).

■■ To use the String class to process immutable strings (§10.10).

■■ To use the StringBuilder and StringBuffer classes to process mutable strings (§10.11).

Object-Oriented Thinking

Chapter

M10_LIAN1878_11_GE_C10.indd 389 12/29/17 4:15 PM

390 Chapter 10 Object-Oriented Thinking

10.1 Introduction The focus of this chapter is on class design and to explore the differences between procedural programming and object-oriented programming.

The preceding chapter introduced objects and classes. You learned how to define classes, create objects, and use objects. This book’s approach is to teach problem solving and funda- mental programming techniques before object-oriented programming. This chapter shows how procedural and object-oriented programming differ. You will see the benefits of object-oriented programming and learn to use it effectively.

Our focus here is on class design. We will use several examples to illustrate the advantages of the object-oriented approach. The examples involve designing new classes and using them in applications and introducing new classes in the Java API.

10.2 Class Abstraction and Encapsulation Class abstraction is separation of class implementation from the use of a class. The details of implementation are encapsulated and hidden from the user. This is known as class encapsulation.

In Chapter 6, you learned about method abstraction and used it in stepwise refinement. Java provides many levels of abstraction, and class abstraction separates class implementation from how the class is used. The creator of a class describes the functions of the class and lets the user know how the class can be used. The collection of public constructors, meth- ods, and fields that are accessible from outside the class, together with the description of how these members are expected to behave, serves as the class’s contract. As shown in Figure 10.1, the user of the class does not need to know how the class is implemented. The details of implementation are encapsulated and hidden from the user. This is called class encapsulation. For example, you can create a Circle object and find the area of the circle without knowing how the area is computed. For this reason, a class is also known as an abstract data type (ADT).

Point Key

class abstraction

class’s contract

class encapsulation

abstract data type

Figure 10.1 Class abstraction separates class implementation from the use of the class.

Class Contract (signatures of public

constructors, methods, and data �elds)

Class

Class implementation is like a black box hidden from the clients

Clients use the class through the

contract of the class

Class abstraction and encapsulation are two sides of the same coin. Many real-life examples illustrate the concept of class abstraction. Consider, for instance, building a computer system. Your personal computer has many components—a CPU, memory, disk, motherboard, fan, and so on. Each component can be viewed as an object that has properties and methods. To get the components to work together, you need to know only how each component is used and how it interacts with the others. You don’t need to know how the components work internally. The internal implementation is encapsulated and hidden from you. You can build a computer without knowing how a component is implemented.

The computer-system analogy precisely mirrors the object-oriented approach. Each com- ponent can be viewed as an object of the class for the component. For example, you might have a class that models all kinds of fans for use in a computer, with properties such as fan size and speed and methods such as start and stop. A specific fan is an instance of this class with specific property values.

As another example, consider getting a loan. A specific loan can be viewed as an object of a Loan class. The interest rate, loan amount, and loan period are its data properties and

M10_LIAN1878_11_GE_C10.indd 390 12/29/17 4:15 PM

10.2 Class Abstraction and Encapsulation 391

computing the monthly and total payments are its methods. When you buy a car, a loan object is created by instantiating the class with your loan interest rate, loan amount, and loan period. You can then use the methods to find the monthly payment and total payment of your loan. As a user of the Loan class, you don’t need to know how these methods are implemented.

Listing 2.9, ComputeLoan.java, presented a program for computing loan payments. That program cannot be reused in other programs because the code for computing the payments is in the main method. One way to fix this problem is to define static methods for computing the monthly payment and the total payment. However, this solution has limitations. Suppose that you wish to associate a date with the loan. There is no good way to tie a date with a loan without using objects. The traditional procedural programming paradigm is action-driven, and data are separated from actions. The object-oriented programming paradigm focuses on objects, and actions are defined along with the data in objects. To tie a date with a loan, you can define a loan class with a date along with the loan’s other properties as data fields. A loan object now contains data and actions for manipulating and processing data, and the loan data and actions are integrated in one object. Figure 10.2 shows the UML class diagram for the Loan class.

VideoNote

The Loan class

Figure 10.2 The Loan class models the properties and behaviors of loans.

Loan

The annual interest rate of the loan (default: 2.5).

Returns the annual interest rate of this loan.

Sets a new annual interest rate for this loan.

Sets a new number of years for this loan.

Sets a new amount for this loan.

Returns the monthly payment for this loan.

The number of years for the loan (default: 1).

The loan amount (default: 1000).

The date this loan was created.

Constructs a loan with speci�ed interest rate, years, and loan amount.

Returns the number of years of this loan.

Returns the amount of this loan.

Returns the date of the creation of this loan.

Returns the total payment for this loan.

+Loan(annualInterestRate: double, numberOfYears: int, loanAmount: double)

+getAnnualInterestRate(): double

+setNumberOfYears( numberOfYears: int): void

+setLoanAmount( loanAmount: double): void

+getMonthlyPayment(): double

+setAnnualInterestRate( annualInterestRate: double): void

+Loan()

+getNumberOfYears(): int

+getLoanAmount(): double

+getLoanDate(): java.util.Date

+getTotalPayment(): double

Constructs a default Loan object.

–annualInterestRate: double

–numberOfYears: int

–loanAmount: double

–loanDate: java.util.Date

The UML diagram in Figure 10.2 serves as the contract for the Loan class. Throughout this book, you will play the roles of both class user and class developer. Remember that a class user can use the class without knowing how the class is implemented.

Assume the Loan class is available. The program in Listing 10.1 uses that class.

Listing 10.1 TestLoanClass.java 1 import java.util.Scanner; 2 3 public class TestLoanClass { 4 /** Main method */ 5 public static void main(String[] args) {

M10_LIAN1878_11_GE_C10.indd 391 12/29/17 4:15 PM

392 Chapter 10 Object-Oriented Thinking

6 // Create a Scanner 7 Scanner input = new Scanner(System.in); 8 9 // Enter annual interest rate 10 System.out.print( 11 "Enter annual interest rate, for example, 8.25: "); 12 double annualInterestRate = input.nextDouble(); 13 14 // Enter number of years 15 System.out.print("Enter number of years as an integer: "); 16 int numberOfYears = input.nextInt(); 17 18 // Enter loan amount 19 System.out.print("Enter loan amount, for example, 120000.95: "); 20 double loanAmount = input.nextDouble(); 21 22 // Create a Loan object 23 Loan loan = 24 new Loan(annualInterestRate, numberOfYears, loanAmount); 25 26 // Display loan date, monthly payment, and total payment 27 System.out.printf("The loan was created on %s\n" + 28 "The monthly payment is %.2f\nThe total payment is %.2f\n", 29 loan.getLoanDate().toString(), loan.getMonthlyPayment(), 30 loan.getTotalPayment()); 31 } 32 }

create loan object

invoke instance method invoke instance method

Enter annual interest rate, for example, 8.25: 2.5

Enter number of years as an integer: 5

Enter loan amount, for example, 120000.95: 1000

The loan was created on Sat Jun 16 21:12:50 EDT 2012

The monthly payment is 17.75

The total payment is 1064.84

The main method reads the interest rate, the payment period (in years), and the loan amount; creates a Loan object; then obtains the monthly payment (line 29) and the total payment (line 30) using the instance methods in the Loan class.

The Loan class can be implemented as in Listing 10.2.

Listing 10.2 Loan.java 1 public class Loan { 2 private double annualInterestRate; 3 private int numberOfYears; 4 private double loanAmount; 5 private java.util.Date loanDate; 6 7 /** Default constructor */ 8 public Loan() { 9 this(2.5, 1, 1000); 10 } 11

no-arg constructor

M10_LIAN1878_11_GE_C10.indd 392 12/29/17 4:15 PM

10.2 Class Abstraction and Encapsulation 393

12 /** Construct a loan with specified annual interest rate, 13 number of years, and loan amount 14 */ 15 public Loan(double annualInterestRate, int numberOfYears, 16 double loanAmount) { 17 this.annualInterestRate = annualInterestRate; 18 this.numberOfYears = numberOfYears; 19 this.loanAmount = loanAmount; 20 loanDate = new java.util.Date(); 21 } 22 23 /** Return annualInterestRate */ 24 public double getAnnualInterestRate() { 25 return annualInterestRate; 26 } 27 28 /** Set a new annualInterestRate */ 29 public void setAnnualInterestRate(double annualInterestRate) { 30 this.annualInterestRate = annualInterestRate; 31 } 32 33 /** Return numberOfYears */ 34 public int getNumberOfYears() { 35 return numberOfYears; 36 } 37 38 /** Set a new numberOfYears */ 39 public void setNumberOfYears(int numberOfYears) { 40 this.numberOfYears = numberOfYears; 41 } 42 43 /** Return loanAmount */ 44 public double getLoanAmount() { 45 return loanAmount; 46 } 47 48 /** Set a new loanAmount */ 49 public void setLoanAmount(double loanAmount) { 50 this.loanAmount = loanAmount; 51 } 52 53 /** Find monthly payment */ 54 public double getMonthlyPayment() { 55 double monthlyInterestRate = annualInterestRate / 1200; 56 double monthlyPayment = loanAmount * monthlyInterestRate / (1 – 57 (1 / Math.pow(1 + monthlyInterestRate, numberOfYears * 12))); 58 return monthlyPayment; 59 } 60 61 /** Find total payment */ 62 public double getTotalPayment() { 63 double totalPayment = getMonthlyPayment() * numberOfYears * 12; 64 return totalPayment; 65 } 66 67 /** Return loan date */ 68 public java.util.Date getLoanDate() { 69 return loanDate; 70 } 71 }

constructor

M10_LIAN1878_11_GE_C10.indd 393 12/29/17 4:15 PM

394 Chapter 10 Object-Oriented Thinking

From a class developer’s perspective, a class is designed for use by many different customers. In order to be useful in a wide range of applications, a class should provide a variety of ways for customization through constructors, properties, and methods.

The Loan class contains two constructors, four getter methods, three setter methods, and the methods for finding the monthly payment and the total payment. You can con- struct a Loan object by using the no-arg constructor or the constructor with three param- eters: annual interest rate, number of years, and loan amount. When a loan object is created, its date is stored in the loanDate field. The getLoanDate method returns the date. The methods—getAnnualInterest, getNumberOfYears, and getLoanAmount—return the annual interest rate, payment years, and loan amount, respectively. All the data properties and methods in this class are tied to a specific instance of the Loan class. Therefore, they are instance variables and methods.

Important Pedagogical Tip Use the UML diagram for the Loan class shown in Figure 10.2 to write a test program that uses the Loan class even though you don’t know how the Loan class is imple- mented. This has three benefits:

■ ■■ It demonstrates that developing a class and using a class are two separate tasks.

■ ■■ It enables you to skip the complex implementation of certain classes without inter- rupting the sequence of this book.

■ ■■ It is easier to learn how to implement a class if you are familiar with it by using the class.

For all the class examples from now on, create an object from the class and try using its methods before turning your attention to its implementation.

10.2.1 If you redefine the Loan class in Listing 10.2 without setter methods, is the class immutable?

10.3 Thinking in Objects The procedural paradigm focuses on designing methods. The object-oriented paradigm couples data and methods together into objects. Software design using the object-oriented paradigm focuses on objects and operations on objects.

Chapters 1 through 8 introduced fundamental programming techniques for problem solving using loops, methods, and arrays. Knowing these techniques lays a solid foundation for object- oriented programming. Classes provide more flexibility and modularity for building reusable software. This section improves the solution for a problem introduced in Chapter 3 using the object-oriented approach. From these improvements, you will gain insight into the differences between procedural and object-oriented programming, and see the benefits of developing reus- able code using objects and classes.

Listing 3.4, ComputeAndInterpretBMI.java, presented a program for computing the body mass index (BMI). The code cannot be reused in other programs, because the code is in the main method. To make it reusable, define a static method to compute body mass index as follows:

public static double getBMI(double weight, double height)

This method is useful for computing body mass index for a specified weight and height. How- ever, it has limitations. Suppose you need to associate the weight and height with a person’s name and birth date. You could declare separate variables to store these values, but these values would not be tightly coupled. The ideal way to couple them is to create an object that contains them all. Since these values are tied to individual objects, they should be stored in instance data fields. You can define a class named BMI as shown in Figure 10.3.

Point Check

Point Key

VideoNote

The BMI class

M10_LIAN1878_11_GE_C10.indd 394 12/29/17 4:15 PM

10.3 Thinking in Objects 395

Figure 10.3 The BMI class encapsulates BMI information.

BMI

–name: String

–age: int –weight: double

–height: double

+BMI(name: String, age: int, weight: double, height: double)

+BMI(name: String, weight: double, height: double)

+getBMI(): double

+getStatus(): String

The name of the person.

The age of the person. The weight of the person in pounds. The height of the person in inches.

Creates a BMI object with the speci�ed name, age, weight, and height.

Creates a BMI object with the speci�ed name, weight, height, and a default age 20.

Returns the BMI.

Returns the BMI status (e.g., normal, overweight, etc.).

The getter methods for these data �elds are provided in the class, but omitted in the UML diagram for brevity.

The BMI for Kim Yang is 20.81 Normal

The BMI for Susan King is 30.85 Obese

Assume the BMI class is available. Listing 10.3 gives a test program that uses this class.

Listing 10.3 UseBMIClass.java 1 public class UseBMIClass { 2 public static void main(String[] args) { 3 BMI bmi1 = new BMI("Kim Yang", 18, 145, 70); 4 System.out.println("The BMI for " + bmi1.getName() + " is " 5 + bmi1.getBMI() + " " + bmi1.getStatus()); 6 7 BMI bmi2 = new BMI("Susan King", 215, 70); 8 System.out.println("The BMI for " + bmi2.getName() + " is " 9 + bmi2.getBMI() + " " + bmi2.getStatus()); 10 } 11 }

create an object

invoke instance method create an object

invoke instance method

Line 3 creates the object bmi1 for Kim Yang, and line 7 creates the object bmi2 for Susan King. You can use the instance methods getName(), getBMI(), and getStatus() to return the BMI information in a BMI object.

The BMI class can be implemented as in Listing 10.4.

Listing 10.4 BMI.java 1 public class BMI { 2 private String name; 3 private int age; 4 private double weight; // in pounds 5 private double height; // in inches 6 public static final double KILOGRAMS_PER_POUND = 0.45359237; 7 public static final double METERS_PER_INCH = 0.0254; 8 9 public BMI(String name, int age, double weight, double height) { 10 this.name = name;

constructor

M10_LIAN1878_11_GE_C10.indd 395 12/29/17 4:15 PM

396 Chapter 10 Object-Oriented Thinking

11 this.age = age; 12 this.weight = weight; 13 this.height = height; 14 } 15 16 public BMI(String name, double weight, double height) { 17 this(name, 20, weight, height); 18 } 19 20 public double getBMI() { 21 double bmi = weight * KILOGRAMS_PER_POUND / 22 ((height * METERS_PER_INCH) * (height * METERS_PER_INCH)); 23 return Math.round(bmi * 100) / 100.0; 24 } 25 26 public String getStatus() { 27 double bmi = getBMI(); 28 if (bmi < 18.5) 29 return "Underweight"; 30 else if (bmi < 25) 31 return "Normal"; 32 else if (bmi < 30) 33 return "Overweight"; 34 else 35 return "Obese"; 36 } 37 38 public String getName() { 39 return name; 40 } 41 42 public int getAge() { 43 return age; 44 } 45 46 public double getWeight() { 47 return weight; 48 } 49 50 public double getHeight() { 51 return height; 52 } 53 }

The mathematical formula for computing the BMI using weight and height is given in Section 3.8. The instance method getBMI() returns the BMI. Since the weight and height are instance data fields in the object, the getBMI() method can use these properties to compute the BMI for the object.

The instance method getStatus() returns a string that interprets the BMI. The interpreta- tion is also given in Section 3.8.

This example demonstrates the advantages of the object-oriented paradigm over the proce- dural paradigm. The procedural paradigm focuses on designing methods. The object-oriented paradigm couples data and methods together into objects. Software design using the object- oriented paradigm focuses on objects and operations on objects. The object-oriented approach combines the power of the procedural paradigm with an added dimension that integrates data with operations into objects.

In procedural programming, data and operations on the data are separate, and this meth- odology requires passing data to methods. Object-oriented programming places data and

constructor

getBMI

getStatus

procedural vs. object-oriented paradigms

M10_LIAN1878_11_GE_C10.indd 396 12/29/17 4:15 PM

10.4 Class Relationships 397

the operations that pertain to them in an object. This approach solves many of the problems inherent in procedural programming. The object-oriented programming approach organizes programs in a way that mirrors the real world, in which all objects are associated with both attributes and activities. Using objects improves software reusability and makes programs easier to develop and easier to maintain. Programming in Java involves thinking in terms of objects; a Java program can be viewed as a collection of cooperating objects.

10.3.1 Is the BMI class defined in Listing 10.4 immutable?

10.4 Class Relationships To design classes, you need to explore the relationships among classes. The common relationships among classes are association, aggregation, composition, and inheritance.

This section explores association, aggregation, and composition. The inheritance relationship will be introduced in Chapter 11.

10.4.1 Association Association is a general binary relationship that describes an activity between two classes. For example, a student taking a course is an association between the Student class and the Course class, and a faculty member teaching a course is an association between the Faculty class and the Course class. These associations can be represented in UML graphical notation, as shown in Figure 10.4.

Point Check

Point Key

association

Figure 10.4 This UML diagram shows that a student may take any number of courses, a faculty member may teach at most three courses, a course may have from 5 to 60 students, and a course is taught by only one faculty member.

Teach

Teacher

Take

FacultyStudent 5..60 0..3 1

Course

An association is illustrated by a solid line between two classes with an optional label that describes the relationship. In Figure 10.4, the labels are Take and Teach. Each relationship may have an optional small black triangle that indicates the direction of the relationship. In this figure, the N indicates that a student takes a course (as opposed to a course taking a student).

Each class involved in the relationship may have a role name that describes the role it plays in the relationship. In Figure 10.4, teacher is the role name for Faculty.

Each class involved in an association may specify a multiplicity, which is placed at the side of the class to specify how many of the class’s objects are involved in the relationship in UML. A multiplicity could be a number or an interval that specifies how many of the class’s objects are involved in the relationship. The character * means an unlimited number of objects, and the interval m..n indicates that the number of objects is between m and n, inclusively. In Figure 10.4, each student may take any number of courses, and each course must have at least 5 and at most 60 students. Each course is taught by only one faculty member, and a faculty member may teach from 0 to 3 courses per semester.

In Java code, you can implement associations by using data fields and methods. For exam- ple, the relationships in Figure 10.4 may be implemented using the classes in Figure 10.5.

multiplicity

M10_LIAN1878_11_GE_C10.indd 397 12/29/17 4:15 PM

398 Chapter 10 Object-Oriented Thinking

The relation “a student takes a course” is implemented using the addCourse method in the Student class and the addStudent method in the Course class. The relation “a faculty teaches a course” is implemented using the addCourse method in the Faculty class and the setFaculty method in the Course class. The Student class may use a list to store the courses that the student is taking, the Faculty class may use a list to store the courses that the faculty is teaching, and the Course class may use a list to store students enrolled in the course and a data field to store the instructor who teaches the course.

Figure 10.6 Each student has a name and an address.

Aggregation

AddressName

Composition

1 1 1..3 1 Student

Figure 10.5 The association relations are implemented using data fields and methods in classes.

public class Student { private Course[] courseList;

public void addCourse( Course c) { ... } }

public class Course { private Student[] classList; private Faculty faculty;

public void addStudent( Student s) { ... }

public void setFaculty( Faculty faculty) { ... } }

public class Faculty { private Course[] courseList;

public void addCourse( Course c) { ... } }

Note There are many possible ways to implement relationships. For example, the student and faculty information in the Course class can be omitted, since they are already in the Student and Faculty class. Likewise, if you don’t need to know the courses a student takes or a faculty member teaches, the data field courseList and the addCourse method in Student or Faculty can be omitted.

10.4.2 Aggregation and Composition Aggregation is a special form of association that represents an ownership relationship between two objects. Aggregation models has-a relationships. The owner object is called an aggregating object, and its class is called an aggregating class. The subject object is called an aggregated object, and its class is called an aggregated class.

We refer aggregation between two objects as composition if the existence of the aggregated object is dependent on the aggregating object. In other words, if a relationship is composition, the aggregated object cannot exist on its own. For example, “a student has a name” is a composition relationship between the Student class and the Name class because Name is dependent on Student, whereas “a student has an address” is an aggregation relationship between the Student class and the Address class because an address can exist by itself. Composition implies exclusive ownership. One object owns another object. When the owner object is destroyed, the dependent object is destroyed as well. In UML, a filled diamond is attached to an aggregating class (in this case, Student) to denote the composition relationship with an aggregated class (Name), and an empty diamond is attached to an aggregating class (Student) to denote the aggregation relationship with an aggregated class (Address), as shown in Figure 10.6.

many possible implementations

aggregation

aggregating object aggregating class aggregated object aggregated class

composition

In Figure 10.6, each student has only one multiplicity—address—and each address can be shared by up to 3 students. Each student has one name, and the name is unique for each student.

M10_LIAN1878_11_GE_C10.indd 398 12/29/17 4:15 PM

10.4 Class Relationships 399

An aggregation relationship is usually represented as a data field in the aggregating class. For example, the relationships in Figure 10.6 may be implemented using the classes in Figure 10.7. The relation “a student has a name” and “a student has an address” are implemented in the data field name and address in the Student class.

Figure 10.8 A person may have a supervisor.

1 Supervisor

Person

Figure 10.9 A person can have several supervisors.

Person

Supervisor

(a) (b)

public class Person { ...

private Person[] supervisors; }

Figure 10.7 The composition relations are implemented using data fields in classes.

public class Name { ... }

public class Address { ... }

public class Student { private Name name; private Address address; ... }

Aggregated class Aggregating class Aggregated class

Aggregation may exist between objects of the same class. For example, a person may have a supervisor. This is illustrated in Figure 10.8.

In the relationship “a person has a supervisor,” a supervisor can be represented as a data field in the Person class, as follows:

public class Person { // The type for the data is the class itself private Person supervisor;

... }

If a person can have several supervisors, as shown in Figure 10.9a, you may use an array to store supervisors, as shown in Figure 10.9b.

Important Note Since aggregation and composition relationships are represented using classes in the same way, we will not differentiate them and call both compositions for simplicity.

10.4.1 What are common relationships among classes? 10.4.2 What is association? What is aggregation? What is composition? 10.4.3 What is UML notation of aggregation and composition? 10.4.4 Why both aggregation and composition are together referred to as composition?

aggregation or composition

Point Check

M10_LIAN1878_11_GE_C10.indd 399 12/29/17 4:15 PM

400 Chapter 10 Object-Oriented Thinking

10.5 Case Study: Designing the Course Class This section designs a class for modeling courses.

This book’s philosophy is teaching by example and learning by doing. The book provides a wide variety of examples to demonstrate object-oriented programming. This section and the next offer additional examples on designing classes.

Suppose you need to process course information. Each course has a name and has students enrolled. You should be able to add/drop a student to/from the course. You can use a class to model the courses, as shown in Figure 10.10.

Point Key

Figure 10.10 The Course class models the courses.

Course

–courseName: String –students: String[]

–numberOfStudents: int

+Course(courseName: String)

+getCourseName(): String

+addStudent(student: String): void

+dropStudent(student: String): void

+getStudents(): String[] +getNumberOfStudents(): int

The name of the course. An array to store the students for the course.

The number of students (default: 0).

Creates a course with the speci�ed name.

Returns the course name.

Adds a new student to the course.

Drops a student from the course. Returns the students for the course. Returns the number of students for the course.

A Course object can be created using the constructor Course(String name) by passing a course name. You can add students to the course using the addStudent(String student) method, drop a student from the course using the dropStudent(String student) method, and return all the students in the course using the getStudents() method. Suppose that the Course class is available; Listing 10.5 gives a test class that creates two courses and adds students to them.

Listing 10.5 TestCourse.java 1 public class TestCourse { 2 public static void main(String[] args) { 3 Course course1 = new Course("Data Structures"); 4 Course course2 = new Course("Database Systems"); 5 6 course1.addStudent("Peter Jones"); 7 course1.addStudent("Kim Smith"); 8 course1.addStudent("Anne Kennedy"); 9 10 course2.addStudent("Peter Jones"); 11 course2.addStudent("Steve Smith"); 12 13 System.out.println("Number of students in course1: " 14 + course1.getNumberOfStudents()); 15 String[] students = course1.getStudents(); 16 for (int i = 0; i < course1.getNumberOfStudents(); i++) 17 System.out.print(students[i] + ", "); 18 19 System.out.println(); 20 System.out.print("Number of students in course2: " 21 + course2.getNumberOfStudents()); 22 } 23 }

create a Course

add a Student

number of students return students

M10_LIAN1878_11_GE_C10.indd 400 12/29/17 4:15 PM

10.5 Case Study: Designing the Course Class 401

The Course class is implemented in Listing 10.6. It uses an array to store the students in the course. For simplicity, assume the maximum course enrollment is 100. The array is created using new String[100] in line 3. The addStudent method (line 10) adds a student to the array. Whenever a new student is added to the course, numberOfStudents is increased (line 12). The getStudents method returns the array. The dropStudent method (line 27) is left as an exercise.

Listing 10.6 Course.java 1 public class Course { 2 private String courseName; 3 private String[] students = new String[100]; 4 private int numberOfStudents; 5 6 public Course(String courseName) { 7 this.courseName = courseName; 8 } 9 10 public void addStudent(String student) { 11 students[numberOfStudents] = student; 12 numberOfStudents++; 13 } 14 15 public String[] getStudents() { 16 return students; 17 } 18 19 public int getNumberOfStudents() { 20 return numberOfStudents; 21 } 22 23 public String getCourseName() { 24 return courseName; 25 } 26 27 public void dropStudent(String student) { 28 // Left as an exercise in Programming Exercise 10.9 29 } 30 }

The array size is fixed to be 100 (line 3), so you cannot have more than 100 students in the course. You can improve the class by automatically increasing the array size in Programming Exercise 10.9.

When you create a Course object, an array object is created. A Course object contains a reference to the array. For simplicity, you can say the Course object contains the array.

The user can create a Course object and manipulate it through the public methods addStudent, dropStudent, getNumberOfStudents, and getStudents. However, the user doesn’t need to know how these methods are implemented. The Course class encapsu- lates the internal implementation. This example uses an array to store students, but you could use a different data structure to store students. The program that uses Course does not need to change as long as the contract of the public methods remains unchanged.

create students

add a course

return students

number of students

Number of students in course1: 3

Peter Jones, Kim Smith, Anne Kennedy,

Number of students in course2: 2

M10_LIAN1878_11_GE_C10.indd 401 12/29/17 4:15 PM

402 Chapter 10 Object-Oriented Thinking

10.5.1 Replace the statement in line 17 in Listing 10.5, TestCourse.java, so the loop dis- plays each student name followed by a comma except the last student name.

10.6 Case Study: Designing a Class for Stacks This section designs a class for modeling stacks.

Recall that a stack is a data structure that holds data in a last-in, first-out fashion, as shown in Figure 10.11.

Point Check

Point Keystack

Figure 10.11 A stack holds data in a last-in, first-out fashion.

Data1

Data2 Data1

Data3

Data3 Data2 Data1

Data1

Data2

Data1

Data2

Data2 Data1

Data3

Stacks have many applications. For example, the compiler uses a stack to process method invocations. When a method is invoked, its parameters and local variables are pushed into a stack. When a method calls another method, the new method’s parameters and local variables are pushed into the stack. When a method finishes its work and returns to its caller, its associ- ated space is released from the stack.

You can define a class to model stacks. For simplicity, assume the stack holds the int values. Thus, name the stack class StackOfIntegers. The UML diagram for the class is shown in Figure 10.12.

VideoNote

The StackOfIntegers class

Figure 10.12 The StackOfIntegers class encapsulates the stack storage and provides the operations for manipulating the stack.

StackOfIntegers

–elements: int[]

–size: int

+StackOfIntegers()

+StackOfIntegers(capacity: int)

+empty(): boolean +peek(): int

+push(value: int): void

+pop(): int

+getSize(): int

An array to store integers in the stack. The number of integers in the stack.

Constructs an empty stack with a default capacity of 16. Constructs an empty stack with a speci�ed capacity. Returns true if the stack is empty. Returns the integer at the top of the stack without removing it from the stack. Stores an integer into the top of the stack.

Removes the integer at the top of the stack and returns it.

Returns the number of elements in the stack.

Suppose the class is available. The test program in Listing 10.7 uses the class to create a stack (line 3), store 10 integers 0, 1, 2, . . . , and 9 (line 6), and displays them in reverse order (line 9).

Listing 10.7 TestStackOfIntegers.java 1 public class TestStackOfIntegers { 2 public static void main(String[] args) { 3 StackOfIntegers stack = new StackOfIntegers(); create a stack

M10_LIAN1878_11_GE_C10.indd 402 12/29/17 4:15 PM

10.6 Case Study: Designing a Class for Stacks 403

4 5 for (int i = 0; i < 10; i++) 6 stack.push(i); 7 8 while (!stack.empty()) 9 System.out.print(stack.pop() + " "); 10 } 11 }

push to stack

pop from stack

9 8 7 6 5 4 3 2 1 0

Figure 10.13 The StackOfIntegers uses an array to store the elements in a stack.

elements[0] elements[1]

elements[size 2 1] capacity

top

bottom

size

elements[capacity 2 1]

How do you implement the StackOfIntegers class? The elements in the stack are stored in an array named elements. When you create a stack, the array is also created. The no-arg constructor creates an array with the default capacity of 16. The variable size counts the number of elements in the stack, and size – 1 is the index of the element at the top of the stack, as shown in Figure 10.13. For an empty stack, size is 0.

The StackOfIntegers class is implemented in Listing 10.8. The methods empty(), peek(), pop(), and getSize() are easy to implement. To implement push(int value), assign value to elements[size] if size < capacity (line 24). If the stack is full (i.e., size >= capacity), create a new array of twice the current capacity (line 19), copy the con- tents of the current array to the new array (line 20), and assign the reference of the new array to the current array in the stack (line 21). Now you can add the new value to the array (line 24).

Listing 10.8 StackOfIntegers.java 1 public class StackOfIntegers { 2 private int[] elements; 3 private int size; 4 public static final int DEFAULT_CAPACITY = 16; 5 6 /** Construct a stack with the default capacity 16 */ 7 public StackOfIntegers() { 8 this(DEFAULT_CAPACITY); 9 } 10 11 /** Construct a stack with the specified maximum capacity */ 12 public StackOfIntegers(int capacity) { 13 elements = new int[capacity]; 14 } 15

max capacity 16

M10_LIAN1878_11_GE_C10.indd 403 12/29/17 4:15 PM

404 Chapter 10 Object-Oriented Thinking

16 /** Push a new integer to the top of the stack */ 17 public void push(int value) { 18 if (size >= elements.length) { 19 int[] temp = new int[elements.length * 2]; 20 System.arraycopy(elements, 0, temp, 0, elements.length); 21 elements = temp; 22 } 23 24 elements[size++] = value; 25 } 26 27 /** Return and remove the top element from the stack */ 28 public int pop() { 29 return elements[——size]; 30 } 31 32 /** Return the top element from the stack */ 33 public int peek() { 34 return elements[size – 1]; 35 } 36 37 /** Test whether the stack is empty */ 38 public boolean empty() { 39 return size == 0; 40 } 41 42 /** Return the number of elements in the stack */ 43 public int getSize() { 44 return size; 45 } 46 }

10.7 Processing Primitive Data Type Values as Objects A primitive-type value is not an object, but it can be wrapped in an object using a wrapper class in the Java API.

Owing to performance considerations, primitive data type values are not objects in Java. Because of the overhead of processing objects, the language’s performance would be adversely affected if primitive data type values were treated as objects. However, many Java methods require the use of objects as arguments. Java offers a convenient way to incorporate, or wrap, a primitive data type value into an object (e.g., wrapping an int into an Integer object, wrapping a double into a Double object, and wrapping a char into a Character object). By using a wrapper class, you can process primitive data type values as objects. Java provides Boolean, Character, Double, Float, Byte, Short, Integer, and Long wrapper classes in the java.lang package for primitive data types. The Boolean class wraps a Boolean value true or false. This section uses Integer and Double as examples to introduce the numeric wrapper classes.

Note Most wrapper class names for a primitive type are the same as the primitive data type name with the first letter capitalized. The exceptions are Integer for int and Character for char.

Numeric wrapper classes are very similar to each other. Each contains the methods doubleValue(), floatValue(), intValue(), longValue(), shortValue(), and byteValue(). These methods “convert” objects into primitive-type values. The key features of Integer and Double are shown in Figure 10.14.

double the capacity

add to stack

Point Key

why wrapper class?

naming convention

M10_LIAN1878_11_GE_C10.indd 404 12/29/17 4:15 PM

10.7 Processing Primitive Data Type Values as Objects 405

Figure 10.14 The wrapper classes provide constructors, constants, and conversion methods for manipulating various data types.

–value: int

+MAX_VALUE: int

+MIN_VALUE: int

+Integer(value: int)

+Integer(s: String)

+byteValue(): byte

+shortValue(): short

+intValue(): int

+longValue(): long

+floatValue(): float

+doubleValue(): double

+compareTo(o: Integer): int

+toString(): String

+valueOf(s: String): Integer

+valueOf(s: String, radix: int): Integer

+parseInt(s: String): int

+parseInt(s: String, radix: int): int

java.lang.Integer java.lang.Double

–value: double

+MAX_VALUE: double

+MIN_VALUE: double

+Double(value: double)

+Double(s: String)

+byteValue(): byte

+shortValue(): short

+intValue(): int

+longValue(): long

+floatValue(): float

+doubleValue(): double

+compareTo(o: Double): int

+toString(): String

+valueOf(s: String): Double

+valueOf(s: String, radix: int): Double

+parseDouble(s: String): double

+parseDouble(s: String, radix: int): double

You can construct a wrapper object either from a primitive data type value or from a string representing the numeric value—for example, new Double(5.0), new Double("5.0"), new Integer(5), and new Integer("5").

The wrapper classes do not have no-arg constructors. The instances of all wrapper classes are immutable; this means that, once the objects are created, their internal values cannot be changed.

Each numeric wrapper class has the constants MAX_VALUE and MIN_VALUE. MAX_VALUE represents the maximum value of the corresponding primitive data type. For Byte, Short, Integer, and Long, MIN_VALUE represents the minimum byte, short, int, and long values. Float and Double, MIN_VALUE represents the minimum positive float and double values. The following statements display the maximum integer (2,147,483,647), the minimum positive float (1.4E–45), and the maximum double floating-point number (1.79769313486231570e + 308d):

System.out.println("The maximum integer is " + Integer.MAX_VALUE); System.out.println("The minimum positive float is " + Float.MIN_VALUE); System.out.println( "The maximum double-precision floating-point number is " + Double.MAX_VALUE);

Each numeric wrapper class contains the methods doubleValue(), floatValue(), intValue(), longValue(), and shortValue() for returning a double, float, int, long, or short value for the wrapper object. For example,

new Double(12.4).intValue() returns 12; new Integer(12).doubleValue() returns 12.0;

Recall the String class contains the compareTo method for comparing two strings. The numeric wrapper classes contain the compareTo method for comparing two numbers and

constructors

no no-arg constructor immutable

constants

conversion methods

compareTo method

M10_LIAN1878_11_GE_C10.indd 405 12/29/17 4:15 PM

406 Chapter 10 Object-Oriented Thinking

returns 1, 0, or –1, if this number is greater than, equal to, or less than the other number. For example,

new Double(12.4).compareTo(new Double(12.3)) returns 1; new Double(12.3).compareTo(new Double(12.3)) returns 0; new Double(12.3).compareTo(new Double(12.51)) returns –1;

The numeric wrapper classes have a useful static method, valueOf(String s). This method creates a new object initialized to the value represented by the specified string. For example,

Double doubleObject = Double.valueOf("12.4"); Integer integerObject = Integer.valueOf("12");

You have used the parseInt method in the Integer class to parse a numeric string into an int value and the parseDouble method in the Double class to parse a numeric string into a double value. Each numeric wrapper class has two overloaded parsing methods to parse a numeric string into an appropriate numeric value based on 10 (decimal) or any specified radix (e.g., 2 for binary, 8 for octal, and 16 for hexadecimal).

// These two methods are in the Byte class public static byte parseByte(String s) public static byte parseByte(String s, int radix)

// These two methods are in the Short class public static short parseShort(String s) public static short parseShort(String s, int radix)

// These two methods are in the Integer class public static int parseInt(String s) public static int parseInt(String s, int radix)

// These two methods are in the Long class public static long parseLong(String s) public static long parseLong(String s, int radix)

// These two methods are in the Float class public static float parseFloat(String s) public static float parseFloat(String s, int radix)

// These two methods are in the Double class public static double parseDouble(String s) public static double parseDouble(String s, int radix)

For example,

Integer.parseInt("11", 2) returns 3; Integer.parseInt("12", 8) returns 10; Integer.parseInt("13", 10) returns 13; Integer.parseInt("1A", 16) returns 26;

Integer.parseInt("12", 2) would raise a runtime exception because 12 is not a binary number.

Note you can convert a decimal number into a hex number using the format method. For example,

String.format("%x", 26) returns 1A;

static valueOf methods

static parsing methods

converting decimal to hex

M10_LIAN1878_11_GE_C10.indd 406 12/29/17 4:15 PM

10.8 Automatic Conversion between Primitive Types and Wrapper Class Types 407

10.7.1 Describe primitive-type wrapper classes. 10.7.2 Can each of the following statements be compiled?

a. Integer i = new Integer("23");

b. Integer i = new Integer(23);

c. Integer i = Integer.valueOf("23");

d. Integer i = Integer.parseInt("23", 8);

e. Double d = new Double();

f. Double d = Double.valueOf("23.45");

g. int i = (Integer.valueOf("23")).intValue();

h. double d = (Double.valueOf("23.4")).doubleValue();

i. int i = (Double.valueOf("23.4")).intValue();

j. String s = (Double.valueOf("23.4")).toString();

10.7.3 How do you convert an integer into a string? How do you convert a numeric string into an integer? How do you convert a double number into a string? How do you convert a numeric string into a double value?

10.7.4 Show the output of the following code:

public class Test { public static void main(String[] args) { Integer x = new Integer(3); System.out.println(x.intValue()); System.out.println(x.compareTo(new Integer(4))); } }

10.7.5 What is the output of the following code?

public class Test { public static void main(String[] args) { System.out.println(Integer.parseInt("10")); System.out.println(Integer.parseInt("10", 10)); System.out.println(Integer.parseInt("10", 16)); System.out.println(Integer.parseInt("11")); System.out.println(Integer.parseInt("11", 10)); System.out.println(Integer.parseInt("11", 16)); } }

10.8 Automatic Conversion between Primitive Types and Wrapper Class Types

A primitive-type value can be automatically converted to an object using a wrapper class, and vice versa, depending on the context.

Converting a primitive value to a wrapper object is called boxing. The reverse conversion is called unboxing. Java allows primitive types and wrapper classes to be converted automati- cally. The compiler will automatically box a primitive value that appears in a context requiring an object, and unbox an object that appears in a context requiring a primitive value. This is called autoboxing and autounboxing.

Point Check

Point Key

boxing unboxing autoboxing

autounboxing

M10_LIAN1878_11_GE_C10.indd 407 12/29/17 4:15 PM

408 Chapter 10 Object-Oriented Thinking

For instance, the following statement in (a) can be simplified as in (b) using autoboxing.

Consider the following example:

1 Integer[] intArray = {1, 2, 3}; 2 System.out.println(intArray[0] + intArray1] + intArray[2]);

In line 1, the primitive values 1, 2, and 3 are automatically boxed into objects new Integer(1), new Integer(2), and new Integer(3). In line 2, the objects intArray[0], intArray[1], and intArray[2] are automatically unboxed into int values that are added together.

10.8.1 What are autoboxing and autounboxing? Are the following statements correct?

a. Integer x = 3 + new Integer(5);

b. Integer x = 3;

c. Double x = 3;

d. Double x = 3.0;

e. int x = new Integer(3);

f. int x = new Integer(3) + new Integer(4);

10.8.2 Show the output of the following code.

public class Test { public static void main(String[] args) { Double x = 3.5; System.out.println(x.intValue()); System.out.println(x.compareTo(4.5)); } }

10.9 The BigInteger and BigDecimal Classes The BigInteger and BigDecimal classes can be used to represent integers or decimal numbers of any size and precision.

If you need to compute with very large integers or high-precision floating-point values, you can use the BigInteger and BigDecimal classes in the java.math package. Both are immutable. The largest integer of the long type is Long.MAX_VALUE (i.e., 9223372036854775807). An instance of BigInteger can represent an integer of any size. You can use new BigInteger(String) and new BigDecimal(String) to create an instance of BigInteger and BigDecimal, use the add, subtract, multiply, divide, and remainder methods to perform arithmetic operations, and use the compareTo method to compare two big numbers. For example, the following code creates two BigInteger objects and multiplies them:

BigInteger a = new BigInteger("9223372036854775807"); BigInteger b = new BigInteger("2"); BigInteger c = a.multiply(b); // 9223372036854775807 * 2 System.out.println(c);

Point Check

Point Key

VideoNote

Process large numbers

immutable

Integer intObject = new Integer(2); Equivalent Integer intObject = 2;

autoboxing (a) (b)

int i = 1; (a) and (b) are

Equivalent

int i = new Integer(1);

(a) (b)

The following statement in (a) is the same as in (b) due to autounboxing.

M10_LIAN1878_11_GE_C10.indd 408 12/29/17 4:15 PM

10.9 The BigInteger and BigDecimal Classes 409

The output is 18446744073709551614. There is no limit to the precision of a BigDecimal object. The divide method may throw

an ArithmeticException if the result cannot be terminated. However, you can use the over- loaded divide(BigDecimal d, int scale, int roundingMode) method to specify a scale and a rounding mode to avoid this exception, where scale is the maximum number of digits after the decimal point. For example, the following code creates two BigDecimal objects and performs division with scale 20 and rounding mode BigDecimal.ROUND_UP:

BigDecimal a = new BigDecimal(1.0); BigDecimal b = new BigDecimal(3); BigDecimal c = a.divide(b, 20, BigDecimal.ROUND_UP); System.out.println(c);

The output is 0.33333333333333333334. Note the factorial of an integer can be very large. Listing 10.9 gives a method that can return

the factorial of any integer.

Listing 10.9 LargeFactorial.java 1 import java.util.Scanner; 2 import java.math.*; 3 4 public class LargeFactorial { 5 public static void main(String[] args) { 6 Scanner input = new Scanner(System.in); 7 System.out.print("Enter an integer: "); 8 int n = input.nextInt(); 9 System.out.println(n +"! is \n" + factorial(n)); 10 } 11 12 public static BigInteger factorial(long n) { 13 BigInteger result = BigInteger.ONE; 14 for (int i = 1; i <= n; i++) 15 result = result.multiply(new BigInteger(i + "")); 16 17 return result; 18 } 19 }

constant

multiply

Enter an integer: 50 50! is 30414093201713378043612608166064768844377641568960512000000000000

BigInteger.ONE (line 13) is a constant defined in the BigInteger class. BigInteger. ONE is the same as new BigInteger("1").

A new result is obtained by invoking the multiply method (line 15).

10.9.1 What is the output of the following code? public class Test { public static void main(String[] args) { java.math.BigInteger x = new java.math.BigInteger("3"); java.math.BigInteger y = new java.math.BigInteger("7"); java.math.BigInteger z = x.add(y); System.out.println("x is " + x); System.out.println("y is " + y); System.out.println("z is " + z); } }

Point Check

M10_LIAN1878_11_GE_C10.indd 409 12/29/17 4:15 PM

410 Chapter 10 Object-Oriented Thinking

10.10 The String Class A String object is immutable; its contents cannot be changed once the string is created.

Strings were introduced in Section 4.4. You know strings are objects. You can invoke the charAt(index) method to obtain a character at the specified index from a string, the length() method to return the size of a string, the substring method to return a substring in a string, the indexOf and lastIndexOf methods to return the first or last index of a matching character or a substring, the equals and compareTo methods to compare two strings, and the trim() method to trim whitespace characters from the two ends of a string, and the toLowerCase() and toUpperCase() methods to return the lowercase and uppercase from a string. We will take a closer look at strings in this section.

The String class has 13 constructors and more than 40 methods for manipulating strings. Not only is it very useful in programming, but it is also a good example for learning classes and objects.

10.10.1 Constructing a String You can create a string object from a string literal or from an array of characters. To create a string from a string literal, use the syntax:

String newString = new String(stringLiteral);

The argument stringLiteral is a sequence of characters enclosed in double quotes. The fol- lowing statement creates a String object message for the string literal "Welcome to Java":

String message = new String("Welcome to Java");

Java treats a string literal as a String object. Thus, the following statement is valid:

String message = "Welcome to Java";

You can also create a string from an array of characters. For example, the following statements create the string "Good Day":

char[] charArray = {'G', 'o', 'o', 'd', ' ', 'D', 'a', 'y'}; String message = new String(charArray);

Note A String variable holds a reference to a String object that stores a string value. Strictly speaking, the terms String variable, String object, and string value are different, but most of the time the distinctions between them can be ignored. For simplicity, the term string will often be used to refer to String variable, String object, and string value.

10.10.2 Immutable Strings and Interned Strings A String object is immutable; its contents cannot be changed. Does the following code change the contents of the string?

String s = "Java"; s = "HTML";

The answer is no. The first statement creates a String object with the content "Java" and assigns its reference to s. The second statement creates a new String object with the content "HTML" and assigns its reference to s. The first String object still exists after the assignment, but it can no longer be accessed, because variable s now points to the new object, as shown in Figure 10.15.

Point Key

VideoNote

The String class

string literal object

String variable, string object, string value

immutable

M10_LIAN1878_11_GE_C10.indd 410 12/29/17 4:15 PM

10.10 The String Class 411

Because strings are immutable and are ubiquitous in programming, the JVM uses a unique instance for string literals with the same character sequence in order to improve efficiency and save memory. Such an instance is called an interned string. For example, the following statements: interned string

String s1 = "Welcome to Java";

String s2 = new String("Welcome to Java");

String s3 = "Welcome to Java";

System.out.println("s1 == s2 is " + (s1 == s2)); System.out.println("s1 == s3 is " + (s1 == s3));

s3 Interned string object for "Welcome to Java"

: String

A string object for "Welcome to Java"

: String

Figure 10.16 The String class contains the methods for replacing and splitting strings.

java.lang.String

Returns a new string that replaces all matching characters in this string with the new character.

Returns a new string that replaces the �rst matching substring in this string with the new substring.

Returns a new string that replaces all matching substrings in this string with the new substring.

Returns an array of strings consisting of the substrings split by the delimiter.

+replace(oldChar: char, newChar: char): String

+replaceFirst(oldString: String, newString: String): String

+replaceAll(oldString: String, newString: String): String

+split(delimiter: String): String[]

display

s1 == s2 is false s1 == s3 is true

In the preceding statements, s1 and s3 refer to the same interned string—"Welcome to Java"—so s1 == s3 is true. However, s1 == s2 is false, because s1 and s2 are two different string objects, even though they have the same contents.

10.10.3 Replacing and Splitting Strings The String class provides the methods for replacing and splitting strings, as shown in Figure 10.16.

Figure 10.15 Strings are immutable; once created, their contents cannot be changed.

After executing String s = "Java"; After executing s = "HTML";

Contents cannot be changed

String object for "Java"

String object for "HTML"

This string object is now unreferenced

String object for "Java"

s : String : String

: String

M10_LIAN1878_11_GE_C10.indd 411 12/29/17 4:15 PM

412 Chapter 10 Object-Oriented Thinking

Once a string is created, its contents cannot be changed. The methods replace, replaceFirst, and replaceAll return a new string derived from the original string ( without changing the original string!). Several versions of the replace methods are provided to replace a character or a substring in the string with a new character or a new substring.

For example,

"Welcome".replace('e', 'A') returns a new string, WAlcomA. "Welcome".replaceFirst("e", "AB") returns a new string, WABlcome. "Welcome".replace("e", "AB") returns a new string, WABlcomAB. "Welcome".replace("el", "AB") returns a new string, WABcome.

The split method can be used to extract tokens from a string with the specified delimiters. For example, the following code

String[] tokens = "Java#HTML#Perl".split("#"); for (int i = 0; i < tokens.length; i++) System.out.print(tokens[i] + " ");

displays

Java HTML Perl

10.10.4 Matching, Replacing, and Splitting by Patterns Often you will need to write code that validates user input, such as to check whether the input is a number, a string with all lowercase letters, or a Social Security number. How do you write this type of code? A simple and effective way to accomplish this task is to use the regular expression.

A regular expression (abbreviated regex) is a string that describes a pattern for matching a set of strings. You can match, replace, or split a string by specifying a pattern. This is an extremely useful and powerful feature.

Let us begin with the matches method in the String class. At first glance, the matches method is very similar to the equals method. For example, the following two statements both evaluate to true:

"Java".matches("Java"); "Java".equals("Java");

However, the matches method is more powerful. It can match not only a fixed string, but also a set of strings that follow a pattern. For example, the following statements all evaluate to true:

"Java is fun".matches("Java.*") "Java is cool".matches("Java.*") "Java is powerful".matches("Java.*")

Java.* in the preceding statements is a regular expression. It describes a string pattern that begins with Java followed by any zero or more characters. Here, the substring matches any zero or more characters.

The following statement evaluates to true:

"440–02–4534".matches("\\d{3}–\\d{2}–\\d{4}")

Here, \\d represents a single digit, and \\d{3} represents three digits.

replace

replaceFirst

replace

split

why regular expression?

regular expression regex

matches(regex)

M10_LIAN1878_11_GE_C10.indd 412 12/29/17 4:15 PM

10.10 The String Class 413

The replaceAll, replaceFirst, and split methods can be used with a regular expression. For example, the following statement returns a new string that replaces $, +, or # in a+b$#c with the string NNN.

String s = "a+b$#c".replaceAll("[$+#]", "NNN"); System.out.println(s);

Here, the regular expression [$+#] specifies a pattern that matches $, +, or #. Thus, the output is aNNNbNNNNNNc.

The following statement splits the string into an array of strings delimited by punctuation marks.

String[] tokens = "Java,C?C#,C++".split("[.,:;?]");

for (int i = 0; i < tokens.length; i++) System.out.println(tokens[i]);

In this example, the regular expression [.,:;?] specifies a pattern that matches ., ,, :, ;, or ?. Each of these characters is a delimiter for splitting the string. Thus, the string is split into Java, C, C#, and C++, which are stored in array tokens.

Regular expression patterns are complex for beginning students to understand. For this reason, simple patterns are introduced in this section. Please refer to Appendix H, Regular Expressions, to learn more about these patterns.

10.10.5 Conversion between Strings and Arrays Strings are not arrays, but a string can be converted into an array and vice versa. To convert a string into an array of characters, use the toCharArray method. For example, the following statement converts the string Java to an array:

char[] chars = "Java".toCharArray();

Thus, chars[0] is J, chars[1] is a, chars[2] is v, and chars[3] is a. You can also use the getChars(int srcBegin, int srcEnd, char[] dst, int

dstBegin) method to copy a substring of the string from index srcBegin to index srcEnd–1 into a character array dst starting from index dstBegin. For example, the following code copies a substring "3720" in "CS3720" from index 2 to index 6–1 into the character array dst starting from index 4:

char[] dst = {'J', 'A', 'V', 'A', '1', '3', '0', '1'}; "CS3720".getChars(2, 6, dst, 4);

Thus, dst becomes {'J', 'A', 'V', 'A', '3', '7', '2', '0'}. To convert an array of characters into a string, use the String(char[]) constructor or the

valueOf(char[]) method. For example, the following statement constructs a string from an array using the String constructor:

String str = new String(new char[]{'J', 'a', 'v', 'a'});

The next statement constructs a string from an array using the valueOf method.

String str = String.valueOf(new char[]{'J', 'a', 'v', 'a'});

10.10.6 Converting Characters and Numeric Values to Strings Recall that you can use Double.parseDouble(str) or Integer.parseInt(str) to convert a string to a double value or an int value", and you can convert a character or a number into a string by using the string concatenating operator. Another way of converting a

replaceAll(regex)

split(regex)

further studies

toCharArray

getChars

valueOf

M10_LIAN1878_11_GE_C10.indd 413 12/29/17 4:15 PM

414 Chapter 10 Object-Oriented Thinking

number into a string is to use the overloaded static valueOf method. This method can also be used to convert a character or an array of characters into a string, as shown in Figure 10.17.

overloaded valueOf

Figure 10.17 The String class contains the static methods for creating strings from prim- itive-type values.

java.lang.String

+valueOf(c: char): String

+valueOf(data: char[]): String

+valueOf(d: double): String

+valueOf(f: f loat): String

+valueOf(i: int): String

+valueOf(l: long): String

Returns a string consisting of the character c.

Returns a string consisting of the characters in the array.

Returns a string representing the double value.

Returns a string representing the �oat value.

Returns a string representing the int value.

Returns a string representing the long value.

+valueOf(b: boolean): String Returns a string representing the boolean value.

For example, to convert a double value 5.44 to a string, use String.valueOf(5.44). The return value is a string consisting of the characters '5', '.', '4', and '4'.

10.10.7 Formatting Strings The String class contains the static format method to return a formatted string. The syntax to invoke this method is

String.format(format, item1, item2, ..., itemk);

This method is similar to the printf method except that the format method returns a format- ted string, whereas the printf method displays a formatted string. For example,

String s = String.format("%7.2f%6d%-4s", 45.556, 14, "AB"); System.out.println(s);

displays

45.56 14AB

where the square box ( ) denotes a blank space.

Note

System.out.printf(format, item1, item2, ..., itemk);

is equivalent to

System.out.print( String.format(format, item1, item2, ..., itemk));

10.10.1 Suppose s1, s2, s3, and s4 are four strings, given as follows:

String s1 = "Welcome to Java"; String s2 = s1; String s3 = new String("Welcome to Java"); String s4 = "Welcome to Java";

What are the results of the following expressions?

a. s1 == s2

b. s1 == s3

Point Check

M10_LIAN1878_11_GE_C10.indd 414 12/29/17 4:15 PM

10.10 The String Class 415

c. s1 == s4

d. s1.equals(s3)

e. s1.equals(s4)

f. "Welcome to Java".replace("Java", "HTML")

g. s1.replace('o', 'T')

h. s1.replaceAll("o", "T")

i. s1.replaceFirst("o", "T")

j. s1.toCharArray()

10.10.2 To create the string Welcome to Java, you may use a statement like this:

String s = "Welcome to Java";

String s = new String("Welcome to Java");

Which one is better? Why?

10.10.3 What is the output of the following code?

String s1 = "Welcome to Java"; String s2 = s1.replace("o", "abc"); System.out.println(s1); System.out.println(s2);

10.10.4 Let s1 be " Welcome " and s2 be " welcome ". Write the code for the fol- lowing statements:

a. Replace all occurrences of the character e with E in s1 and assign the new string to s3.

b. Split Welcome to Java and HTML into an array tokens delimited by a space and assign the first two tokens into s1 and s2.

10.10.5 Does any method in the String class change the contents of the string? 10.10.6 Suppose string s is created using new String(); what is s.length()? 10.10.7 How do you convert a char, an array of characters, or a number to a string? 10.10.8 Why does the following code cause a NullPointerException?

1 public class Test { 2 private String text; 3 4 public Test(String s) { 5 String text = s; 6 } 7 8 public static void main(String[] args) { 9 Test test = new Test("ABC"); 10 System.out.println(test.text.toLowerCase()); 11 } 12 }

10.10.9 What is wrong in the following program?

1 public class Test { 2 String text; 3

M10_LIAN1878_11_GE_C10.indd 415 12/29/17 4:15 PM

416 Chapter 10 Object-Oriented Thinking

4 public void Test(String s) { 5 text = s; 6 } 7 8 public static void main(String[] args) { 9 Test test = new Test("ABC"); 10 System.out.println(test); 11 } 12 }

10.10.10 Show the output of the following code:

public class Test { public static void main(String[] args) { System.out.println("Hi, ABC, good".matches("ABC ")); System.out.println("Hi, ABC, good".matches(".*ABC.*")); System.out.println("A,B;C".replaceAll(",;", "#")); System.out.println("A,B;C".replaceAll("[,;]", "#"));

String[] tokens = "A,B;C".split("[,;]"); for (int i = 0; i < tokens.length; i++) System.out.print(tokens[i] + " "); } }

10.10.11 Show the output of the following code:

public class Test { public static void main(String[] args) { String s = "Hi, Good Morning"; System.out.println(m(s)); }

public static int m(String s) { int count = 0; for (int i = 0; i < s.length(); i++) if (Character.isUpperCase(s.charAt(i))) count++;

return count; } }

10.11 The StringBuilder and StringBuffer Classes The StringBuilder and StringBuffer classes are similar to the String class except that the String class is immutable.

In general, the StringBuilder and StringBuffer classes can be used wherever a string is used. StringBuilder and StringBuffer are more flexible than String. You can add, insert, or append new contents into StringBuilder and StringBuffer objects, whereas the value of a String object is fixed once the string is created.

The StringBuilder class is similar to StringBuffer except that the methods for modi- fying the buffer in StringBuffer are synchronized, which means that only one task is allowed to execute the methods. Use StringBuffer if the class might be accessed by multiple tasks concurrently, because synchronization is needed in this case to prevent corruptions to

Point Key

StringBuilder

M10_LIAN1878_11_GE_C10.indd 416 12/29/17 4:15 PM

10.11 The StringBuilder and StringBuffer Classes 417

StringBuffer. Concurrent programming will be introduced in Chapter 32. Using String- Builder is more efficient if it is accessed by just a single task, because no synchronization is needed in this case. The constructors and methods in StringBuffer and StringBuilder are almost the same. This section covers StringBuilder. You can replace StringBuilder in all occurrences in this section by StringBuffer. The program can compile and run without any other changes.

The StringBuilder class has three constructors and more than 30 methods for managing the builder and modifying strings in the builder. You can create an empty string builder or a string builder from a string using the constructors, as shown in Figure 10.18.

StringBuilder constructors

Figure 10.18 The StringBuilder class contains the constructors for creating instances of StringBuilder.

java.lang.StringBuilder

+StringBuilder()

+StringBuilder(capacity: int)

+StringBuilder(s: String)

Constructs an empty string builder with capacity 16.

Constructs a string builder with the speci�ed capacity.

Constructs a string builder with the speci�ed string.

10.11.1 Modifying Strings in the StringBuilder You can append new contents at the end of a string builder, insert new contents at a speci- fied position in a string builder, and delete or replace characters in a string builder, using the methods listed in Figure 10.19.

Figure 10.19 The StringBuilder class contains the methods for modifying string builders.

java.lang.StringBuilder

+append(data: char[]): StringBuilder

+append(data: char[], o�set: int, len: int): StringBuilder

+append(v: aPrimitiveType): StringBuilder

+append(s: String): StringBuilder

+delete(startIndex: int, endIndex: int): StringBuilder

+deleteCharAt(index: int): StringBuilder

+insert(index: int, data: char[], o�set: int, len: int): StringBuilder

+insert(o�set: int, data: char[]): StringBuilder

+insert(o�set: int, b: aPrimitiveType): StringBuilder

+insert(o�set: int, s: String): StringBuilder

+replace(startIndex: int, endIndex: int, s: String): StringBuilder

+reverse(): StringBuilder

+setCharAt(index: int, ch: char): void

Appends a char array into this string builder. Appends a subarray in data into this string builder.

Appends a primitive-type value as a string to this builder.

Appends a string to this string builder.

Deletes characters from startIndex to endIndex–1.

Deletes a character at the speci�ed index.

Inserts a subarray of the data in the array into the builder at the speci�ed index.

Inserts data into this builder at the position offset.

Inserts a value converted to a string into this builder.

Inserts a string into this builder at the position offset.

Replaces the characters in this builder from startIndex to endIndex – 1 with the speci�ed string.

Reverses the characters in the builder.

Sets a new character at the speci�ed index in this builder.

M10_LIAN1878_11_GE_C10.indd 417 12/29/17 4:15 PM

418 Chapter 10 Object-Oriented Thinking

The StringBuilder class provides several overloaded methods to append boolean, char, char[], double, float, int, long, and String into a string builder. For example, the following code appends strings and characters into stringBuilder to form a new string, Welcome to Java:

StringBuilder stringBuilder = new StringBuilder(); stringBuilder.append("Welcome"); stringBuilder.append(' '); stringBuilder.append("to"); stringBuilder.append(' '); stringBuilder.append("Java");

The StringBuilder class also contains overloaded methods to insert boolean, char, char array, double, float, int, long, and String into a string builder. Consider the following code:

stringBuilder.insert(11, "HTML and ");

Suppose stringBuilder contains Welcome to Java before the insert method is applied. This code inserts "HTML and " at position 11 in stringBuilder (just before the J). The new stringBuilder is Welcome to HTML and Java.

You can also delete characters from a string in the builder using the two delete methods, reverse the string using the reverse method, replace characters using the replace method, or set a new character in a string using the setCharAt method.

For example, suppose stringBuilder contains Welcome to Java before each of the following methods is applied:

stringBuilder.delete(8, 11) changes the builder to Welcome Java. stringBuilder.deleteCharAt(8) changes the builder to Welcome o Java. stringBuilder.reverse() changes the builder to avaJ ot emocleW. stringBuilder.replace(11, 15, "HTML") changes the builder to Welcome to HTML. stringBuilder.setCharAt(0, 'w') sets the builder to welcome to Java.

All these modification methods except setCharAt do two things:

1. Change the contents of the string builder

2. Return the reference of the string builder

For example, the following statement:

StringBuilder stringBuilder1 = stringBuilder.reverse();

reverses the string in the builder and assigns the builder’s reference to stringBuilder1. Thus, stringBuilder and stringBuilder1 both point to the same StringBuilder object. Recall that a value-returning method can be invoked as a statement, if you are not interested in the return value of the method. In this case, the return value is simply ignored. For example, in the following statement:

stringBuilder.reverse():

the return value is ignored. Returning the reference of a StringBuilder enables the StringBuilder methods to be invoked in a chain such as the following:

stringBuilder.reverse().delete(8, 11).replace(11, 15, "HTML");

Tip If a string does not require any change, use String rather than StringBuilder. String is more efficient than StringBuilder.

append

insert

delete

deleteCharAt

reverse

replace

setCharAt

ignore return value

String or StringBuilder?

M10_LIAN1878_11_GE_C10.indd 418 12/29/17 4:15 PM

10.11 The StringBuilder and StringBuffer Classes 419

10.11.2 The toString, capacity, length, setLength, and charAt Methods

The StringBuilder class provides the additional methods for manipulating a string builder and obtaining its properties, as shown in Figure 10.20.

Figure 10.20 The StringBuilder class contains the methods for modifying string builders.

java.lang.StringBuilder

+toString(): String

+capacity(): int

+charAt(index: int): char

+length(): int

+setLength(newLength: int): void

+substring(startIndex: int): String

+substring(startIndex: int, endIndex: int): String

+trimToSize(): void

Returns a string object from the string builder.

Returns the capacity of this string builder.

Returns the character at the speci�ed index.

Returns the number of characters in this builder.

Sets a new length in this builder.

Returns a substring starting at startIndex.

Returns a substring from startIndex to endIndex – 1.

Reduces the storage size used for the string builder.

The capacity() method returns the current capacity of the string builder. The capacity is the number of characters the string builder is able to store without having to increase its size.

The length() method returns the number of characters actually stored in the string builder. The setLength(newLength) method sets the length of the string builder. If the newLength argument is less than the current length of the string builder, the string builder is truncated to contain exactly the number of characters given by the newLength argument. If the newLength argument is greater than or equal to the current length, sufficient null characters (\u0000) are appended to the string builder so length becomes the newLength argument. The newLength argument must be greater than or equal to 0.

The charAt(index) method returns the character at a specific index in the string builder. The index is 0 based. The first character of a string builder is at index 0, the next at index 1, and so on. The index argument must be greater than or equal to 0, and less than the length of the string builder.

Note The length of the string builder is always less than or equal to the capacity of the builder. The length is the actual size of the string stored in the builder, and the capacity is the current size of the builder. The builder’s capacity is automatically increased if more characters are added to exceed its capacity. Internally, a string builder is an array of characters, so the builder’s capacity is the size of the array. If the builder’s capacity is exceeded, the array is replaced by a new array. The new array size is 2 * (the previ- ous array size + 1).

Tip You can use new StringBuilder(initialCapacity) to create a String- Builder with a specified initial capacity. By carefully choosing the initial capacity, you can make your program more efficient. If the capacity is always larger than the actual length of the builder, the JVM will never need to reallocate memory for the builder. On the other hand, if the capacity is too large, you will waste memory space. You can use the trimToSize() method to reduce the capacity to the actual size.

capacity()

length()

setLength(int)

charAt(int)

length and capacity

initial capacity

trimToSize()

M10_LIAN1878_11_GE_C10.indd 419 12/29/17 4:15 PM

420 Chapter 10 Object-Oriented Thinking

10.11.3 Case Study: Ignoring Nonalphanumeric Characters When Checking Palindromes

Listing 5.14, Palindrome.java, considered all the characters in a string to check whether it is a palindrome. Write a new program that ignores nonalphanumeric characters in checking whether a string is a palindrome.

Here are the steps to solve the problem:

1. Filter the string by removing the nonalphanumeric characters. This can be done by creating an empty string builder, adding each alphanumeric character in the string to a string builder, and returning the string from the string builder. You can use the isLetterOrDigit(ch) method in the Character class to check whether character ch is a letter or a digit.

2. Obtain a new string that is the reversal of the filtered string. Compare the reversed string with the filtered string using the equals method.

The complete program is shown in Listing 10.10.

Listing 10.10 PalindromeIgnoreNonAlphanumeric.java 1 import java.util.Scanner; 2 3 public class PalindromeIgnoreNonAlphanumeric { 4 /** Main method */ 5 public static void main(String[] args) { 6 // Create a Scanner 7 Scanner input = new Scanner(System.in); 8 9 // Prompt the user to enter a string 10 System.out.print("Enter a string: "); 11 String s = input.nextLine(); 12 13 // Display result 14 System.out.println("Ignoring nonalphanumeric characters, \nis " 15 + s + " a palindrome? " + isPalindrome(s)); 16 } 17 18 /** Return true if a string is a palindrome */ 19 public static boolean isPalindrome(String s) { 20 // Create a new string by eliminating nonalphanumeric chars 21 String s1 = filter(s); 22 23 // Create a new string that is the reversal of s1 24 String s2 = reverse(s1); 25 26 // Check if the reversal is the same as the original string 27 return s2.equals(s1); 28 } 29 30 /** Create a new string by eliminating nonalphanumeric chars */ 31 public static String filter(String s) { 32 // Create a string builder 33 StringBuilder stringBuilder = new StringBuilder(); 34 35 // Examine each char in the string to skip alphanumeric char 36 for (int i = 0; i < s.length(); i++) { 37 if (Character.isLetterOrDigit(s.charAt(i))) { 38 stringBuilder.append(s.charAt(i)); 39 }

add letter or digit

check palindrome

M10_LIAN1878_11_GE_C10.indd 420 12/29/17 4:15 PM

10.11 The StringBuilder and StringBuffer Classes 421

40 } 41 42 // Return a new filtered string 43 return stringBuilder.toString(); 44 } 45 46 /** Create a new string by reversing a specified string */ 47 public static String reverse(String s) { 48 StringBuilder stringBuilder = new StringBuilder(s); 49 stringBuilder.reverse(); // Invoke reverse in StringBuilder 50 return stringBuilder.toString(); 51 } 52 }

Enter a string: ab<c>cb?a

Ignoring nonalphanumeric characters,

is ab<c>cb?a a palindrome? true

Enter a string: abcc><?cab

Ignoring nonalphanumeric characters,

is abcc><?cab a palindrome? false

The filter(String s) method (lines 31–44) examines each character in string s and cop- ies it to a string builder if the character is a letter or a numeric character. The filter method returns the string in the builder. The reverse(String s) method (lines 47–51) creates a new string that reverses the specified string s. The filter and reverse methods both return a new string. The original string is not changed.

The program in Listing 5.14 checks whether a string is a palindrome by comparing pairs of characters from both ends of the string. Listing 10.10 uses the reverse method in the StringBuilder class to reverse the string, then compares whether the two strings are equal to determine whether the original string is a palindrome.

10.11.1 What is the difference between StringBuilder and StringBuffer? 10.11.2 How do you create a string builder from a string? How do you return a string

from a string builder?

10.11.3 Write three statements to reverse a string s using the reverse method in the StringBuilder class.

10.11.4 Write three statements to delete a substring from a string s of 20 characters, starting at index 4 and ending with index 10. Use the delete method in the StringBuilder class.

10.11.5 What is the internal storage for characters in a string and a string builder? 10.11.6 Suppose s1 and s2 are given as follows:

StringBuilder s1 = new StringBuilder("Java"); StringBuilder s2 = new StringBuilder("HTML");

Show the value of s1 after each of the following statements. Assume the statements are independent.

a. s1.append(" is fun");

b. s1.append(s2);

Point Check

M10_LIAN1878_11_GE_C10.indd 421 12/29/17 4:15 PM

422 Chapter 10 Object-Oriented Thinking

c. s1.insert(2, "is fun");

d. s1.insert(1, s2);

e. s1.charAt(2);

f. s1.length();

g. s1.deleteCharAt(3);

h. s1.delete(1, 3);

i. s1.reverse();

j. s1.replace(1, 3, "Computer");

k. s1.substring(1, 3);

l. s1.substring(2);

10.11.7 Show the output of the following program:

public class Test { public static void main(String[] args) { String s = "Java"; StringBuilder builder = new StringBuilder(s); change(s, builder);

System.out.println(s); System.out.println(builder); }

private static void change(String s, StringBuilder builder) { s = s + " and HTML"; builder.append(" and HTML"); } }

Key Terms abstract data type (ADT) 390 aggregation 398 boxing 407 class abstraction 390 class encapsulation 390 class’s contract 390

composition 398 has-a relationship 398 multiplicity 397 stack 402 unboxing 407

ChapTer summary 1. The procedural paradigm focuses on designing methods. The object-oriented paradigm

couples data and methods together into objects. Software design using the object-oriented paradigm focuses on objects and operations on objects. The object-oriented approach combines the power of the procedural paradigm with an added dimension that integrates data with operations into objects.

2. Many Java methods require the use of objects as arguments. Java offers a convenient way to incorporate, or wrap, a primitive data type into an object (e.g., wrapping int into the Integer class, and wrapping double into the Double class).

M10_LIAN1878_11_GE_C10.indd 422 12/29/17 4:15 PM

3. Java can automatically convert a primitive-type value to its corresponding wrapper object in the context and vice versa.

4. The BigInteger class is useful for computing and processing integers of any size. The BigDecimal class can be used to compute and process floating-point numbers with any arbitrary precision.

5. A String object is immutable; its contents cannot be changed. To improve efficiency and save memory, the JVM stores two literal strings that have the same character sequence in a unique object. This unique object is called an interned string object.

6. A regular expression (abbreviated regex) is a string that describes a pattern for matching a set of strings. You can match, replace, or split a string by specifying a pattern.

7. The StringBuilder and StringBuffer classes can be used to replace the String class. The String object is immutable, but you can add, insert, or append new contents into StringBuilder and StringBuffer objects. Use String if the string contents do not require any change and use StringBuilder or StringBuffer if they might change.

Quiz Answer the quiz for this chapter online at the book Companion Website.

programming exerCises

Sections 10.2 and 10.3 *10.1 (The Time class) Design a class named Time. The class contains:

■■ The data fields hour, minute, and second that represent a time. ■■ A no-arg constructor that creates a Time object for the current time. (The

values of the data fields will represent the current time.) ■■ A constructor that constructs a Time object with a specified elapsed time

since midnight, January 1, 1970, in milliseconds. (The values of the data fields will represent this time.)

■■ A constructor that constructs a Time object with the specified hour, minute, and second.

■■ Three getter methods for the data fields hour, minute, and second, respectively.

■■ A method named setTime(long elapseTime) that sets a new time for the object using the elapsed time. For example, if the elapsed time is 555550000 milliseconds, the hour is 10, the minute is 19, and the second is 10.

Draw the UML diagram for the class then implement the class. Write a test program that creates three Time objects (using new Time(), new Time(555550000), and new Time(5, 23, 55)) and displays their hour, minute, and second in the format hour:minute:second.

(Hint: The first two constructors will extract the hour, minute, and second from the elapsed time. For the no-arg constructor, the current time can be obtained using System.currentTimeMillis(), as shown in Listing 2.7, Show CurrentTime.java. Assume the time is in GMT.)

Programming Exercises 423

M10_LIAN1878_11_GE_C10.indd 423 12/29/17 4:15 PM

424 Chapter 10 Object-Oriented Thinking

10.2 (The BMI class) Add the following new constructor in the BMI class:

/** Construct a BMI with the specified name, age, weight, * feet, and inches */ public BMI(String name, int age, double weight, double feet, double inches)

10.3 (The MyInteger class) Design a class named MyInteger. The class contains: ■■ An int data field named value that stores the int value represented by

this object. ■■ A constructor that creates a MyInteger object for the specified int value. ■■ A getter method that returns the int value. ■■ The methods isEven(), isOdd(), and isPrime() that return true if the

value in this object is even, odd, or prime, respectively. ■■ The static methods isEven(int), isOdd(int), and isPrime(int) that

return true if the specified value is even, odd, or prime, respectively. ■■ The static methods isEven(MyInteger), isOdd(MyInteger), and isPrime(MyInteger) that return true if the specified value is even, odd, or prime, respectively.

■■ The methods equals(int) and equals(MyInteger) that return true if the value in this object is equal to the specified value.

■■ A static method parseInt(char[]) that converts an array of numeric characters to an int value.

■■ A static method parseInt(String) that converts a string into an int value.

Draw the UML diagram for the class then implement the class. Write a client program that tests all methods in the class.

10.4 (The MyPoint class) Design a class named MyPoint to represent a point with x- and y-coordinates. The class contains:

■■ The data fields x and y that represent the coordinates with getter methods. ■■ A no-arg constructor that creates a point (0, 0). ■■ A constructor that constructs a point with specified coordinates. ■■ A method named distance that returns the distance from this point to a

specified point of the MyPoint type. ■■ A method named distance that returns the distance from this point to

another point with specified x- and y-coordinates. ■■ A static method named distance that returns the distance from two MyPoint

objects.

Draw the UML diagram for the class then implement the class. Write a test program that creates the two points (0, 0) and (10, 30.5) and displays the dis- tance between them.

Sections 10.4–10.8 *10.5 (Display the prime factors) Write a program that prompts the user to enter a

positive integer and displays all its smallest factors in decreasing order. For example, if the integer is 120, the smallest factors are displayed as 5, 3, 2, 2, 2. Use the StackOfIntegers class to store the factors (e.g., 2, 2, 2, 3, 5) and retrieve and display them in reverse order.

*10.6 (Display the prime numbers) Write a program that displays all the prime num- bers less than 120 in decreasing order. Use the StackOfIntegers class to store the prime numbers (e.g., 2, 3, 5, . . . ) and retrieve and display them in reverse order.

VideoNote

The MyPoint class

M10_LIAN1878_11_GE_C10.indd 424 12/29/17 4:15 PM

**10.7 (Game: ATM machine) Use the Account class created in Programming Exer- cise 9.7 to simulate an ATM machine. Create 10 accounts in an array with id 0, 1, . . . , 9, and an initial balance of $100. The system prompts the user to enter an id. If the id is entered incorrectly, ask the user to enter a correct id. Once an id is accepted, the main menu is displayed as shown in the sample run. You can enter choice 1 for viewing the current balance, 2 for withdrawing money, 3 for depositing money, and 4 for exiting the main menu. Once you exit, the system will prompt for an id again. Thus, once the system starts, it will not stop.

Enter an id: 4

Main menu 1: check balance 2: withdraw 3: deposit 4: exit Enter a choice: 1 The balance is 100.0

Main menu 1: check balance 2: withdraw 3: deposit 4: exit Enter a choice: 2 Enter an amount to withdraw: 3

Main menu 1: check balance 2: withdraw 3: deposit 4: exit Enter a choice: 1 The balance is 97.0

Main menu 1: check balance 2: withdraw 3: deposit 4: exit Enter a choice: 3 Enter an amount to deposit: 10

Main menu 1: check balance 2: withdraw 3: deposit 4: exit Enter a choice: 1 The balance is 107.0

Main menu 1: check balance 2: withdraw 3: deposit 4: exit Enter a choice: 4

Enter an id:

Programming Exercises 425

M10_LIAN1878_11_GE_C10.indd 425 12/29/17 4:15 PM

426 Chapter 10 Object-Oriented Thinking

***10.8 (Financial: the Tax class) Programming Exercise 8.12 writes a program for computing taxes using arrays. Design a class named Tax to contain the follow- ing instance data fields:

■■ int filingStatus: One of the four tax-filing statuses: 0—single filer, 1—married filing jointly or qualifying widow(er), 2—married filing separately, and 3—head of household. Use the public static constants SINGLE_FILER (0), MARRIED_JOINTLY_OR_QUALIFYING_WIDOW(ER) (1), MARRIED_ SEPARATELY (2), HEAD_OF_HOUSEHOLD (3) to represent the statuses.

■■ int[][] brackets: Stores the tax brackets for each filing status. ■■ double[] rates: Stores the tax rates for each bracket. ■■ double taxableIncome: Stores the taxable income.

Provide the getter and setter methods for each data field and the getTax() method that returns the tax. Also, provide a no-arg constructor and the construc- tor Tax(filingStatus, brackets, rates, taxableIncome).

Draw the UML diagram for the class and then implement the class. Write a test program that uses the Tax class to print the 2001 and 2009 tax tables for taxable income from $50,000 to $60,000 with intervals of $1,000 for all four statuses. The tax rates for the year 2009 were given in Table 3.2. The tax rates for 2001 are shown in Table 10.1.

Tax Rate Single Filers Married—Filing Jointly or Qualifying Widow(er)

Married—Filing Separately Head of Household

15% Up to $27,050 Up to $45,200 Up to $22,600 Up to $36,250

27.5% $27,051–$65,550 $45,201–$109,250 $22,601–$54,625 $36,251–$93,650

30.5% $65,551–$136,750 $109,251–$166,500 $54,626–$83,250 $93,651–$151,650

35.5% $136,751–$297,350 $166,501–$297,350 $83,251–$148,675 $151,651–$297,350

39.1% $297,351 or more $297,351 or more $ 148,676 or more $297,351 or more

tabLe 10.1 2001 U.S. Federal Personal Tax Rates

**10.9 (The Course class) Revise the Course class as follows: ■■ Revise the getStudents() method to return an array whose length is the

same as the number of students in the course. (Hint: create a new array and copy students to it.)

■■ The array size is fixed in Listing 10.6. Revise the addStudent method to automatically increase the array size if there is no room to add more students. This is done by creating a new larger array and copying the contents of the current array to it.

■■ Implement the dropStudent method. ■■ Add a new method named clear() that removes all students from the course.

Write a test program that creates a course, adds three students, removes one, and displays the students in the course.

*10.10 (The Queue class) Section 10.6 gives a class for Stack. Design a class named Queue for storing integers. Like a stack, a queue holds elements. In a stack, the elements are retrieved in a last-in first-out fashion. In a queue, the elements are retrieved in a first-in first-out fashion. The class contains:

■■ An int[] data field named elements that stores the int values in the queue. ■■ A data field named size that stores the number of elements in the queue. ■■ A constructor that creates a Queue object with default capacity 8. ■■ The method enqueue(int v) that adds v into the queue.

M10_LIAN1878_11_GE_C10.indd 426 12/29/17 4:15 PM

■■ The method dequeue() that removes and returns the element from the queue.

■■ The method empty() that returns true if the queue is empty. ■■ The method getSize() that returns the size of the queue.

Draw an UML diagram for the class. Implement the class with the initial array size set to 8. The array size will be doubled once the number of the elements exceeds the size. After an element is removed from the beginning of the array, you need to shift all elements in the array one position the left. Write a test program that adds 20 numbers from 1 to 20 into the queue then removes these numbers and displays them.

*10.11 (Geometry: the Circle2D class) Define the Circle2D class that contains: ■■ Two double data fields named x and y that specify the center of the circle

with getter methods. ■■ A data field radius with a getter method. ■■ A no-arg constructor that creates a default circle with (0, 0) for (x, y) and 1

for radius. ■■ A constructor that creates a circle with the specified x, y, and radius. ■■ A method getArea() that returns the area of the circle. ■■ A method getPerimeter() that returns the perimeter of the circle. ■■ A method contains(double x, double y) that returns true if the

specified point (x, y) is inside this circle (see Figure 10.21a). ■■ A method contains(Circle2D circle) that returns true if the specified

circle is inside this circle (see Figure 10.21b). ■■ A method overlaps(Circle2D circle) that returns true if the specified

circle overlaps with this circle (see Figure 10.21c).

Figure 10.21 (a) A point is inside the circle. (b) A circle is inside another circle. (c) A circle overlaps another circle.

(a) (b) (c)

Draw the UML diagram for the class then implement the class. Write a test program that creates a Circle2D object c1 (new Circle2D(2, 2, 5.5)), displays its area and perimeter, and displays the result of c1.contains(3, 3), c1.contains(new Circle2D(4, 5, 10.5)), and c1.overlaps(new Circle2D(3, 5, 2.3)).

***10.12 (Geometry: the Triangle2D class) Define the Triangle2D class that contains: ■■ Three points named p1, p2, and p3 of the type MyPoint with getter and

setter methods. MyPoint is defined in Programming Exercise 10.4. ■■ A no-arg constructor that creates a default triangle with the points (0, 0),

(1, 1), and (2, 5). ■■ A constructor that creates a triangle with the specified points. ■■ A method getArea() that returns the area of the triangle. ■■ A method getPerimeter() that returns the perimeter of the triangle. ■■ A method contains(MyPoint p) that returns true if the specified point p is inside this triangle (see Figure 10.22a).

Programming Exercises 427

M10_LIAN1878_11_GE_C10.indd 427 12/29/17 4:15 PM

428 Chapter 10 Object-Oriented Thinking

■■ A method contains(Triangle2D t) that returns true if the specified triangle is inside this triangle (see Figure 10.22b).

■■ A method overlaps(Triangle2D t) that returns true if the specified triangle overlaps with this triangle (see Figure 10.22c).

Figure 10.22 (a) A point is inside the triangle. (b) A triangle is inside another triangle. (c) A triangle overlaps another triangle.

(a) (b) (c)

Draw the UML diagram for the class and then implement the class. Write a test program that creates a Triangle2D object t1 using the constructor new Triangle2D(new MyPoint(2.5, 2), new MyPoint(4.2, 3), new MyPoint(5, 3.5)), displays its area and perimeter, and displays the result of t1.contains(3, 3), r1.contains(new Triangle2D(new MyPoint(2.9, 2), new MyPoint(4, 1), MyPoint(1, 3.4))), and t1 .overlaps(new Triangle2D(new MyPoint(2, 5.5), new MyPoint (4, –3), MyPoint(2, 6.5))).

(Hint: For the formula to compute the area of a triangle, see Programming Exer- cise 2.19. To detect whether a point is inside a triangle, draw three dashed lines, as shown in Figure 10.23. If the point is inside a triangle, each dashed line should intersect a side only once. If a dashed line intersects a side twice, then the point must be outside the triangle. For the algorithm of finding the intersecting point of two lines, see Programming Exercise 3.25.)

Figure 10.23 (a) A point is inside the triangle. (b) A point is outside the triangle.

(a) (b)

*10.13 (Geometry: the MyRectangle2D class) Define the MyRectangle2D class that contains:

■■ Two double data fields named x and y that specify the center of the rectangle with getter and setter methods. (Assume the rectangle sides are parallel to x- or y-axis.)

■■ The data fields width and height with getter and setter methods. ■■ A no-arg constructor that creates a default rectangle with (0, 0) for (x, y) and 1 for both width and height.

■■ A constructor that creates a rectangle with the specified x, y, width, and height.

M10_LIAN1878_11_GE_C10.indd 428 12/29/17 4:15 PM

■■ A method getArea() that returns the area of the rectangle. ■■ A method getPerimeter() that returns the perimeter of the rectangle. ■■ A method contains(double x, double y) that returns true if the

specified point (x, y) is inside this rectangle (see Figure 10.24a). ■■ A method contains(MyRectangle2D r) that returns true if the specified

rectangle is inside this rectangle (see Figure 10.24b). ■■ A method overlaps(MyRectangle2D r) that returns true if the specified

rectangle overlaps with this rectangle (see Figure 10.24c).

Figure 10.24 A point is inside the rectangle. (b) A rectangle is inside another rectangle. (c) A rectangle overlaps another rectangle. (d) Points are enclosed inside a rectangle.

(a) (b) (c) (d)

Draw the UML diagram for the class then implement the class. Write a test program that creates a MyRectangle2D object r1 (new MyRectangle2D (2, 2, 5.5, 4.9)), displays its area and perimeter, and displays the result of r1.contains(3, 3), r1.contains(new MyRectangle2D(4, 5, 10.5, 3.2)), and r1.overlaps(new MyRectangle2D(3, 5, 2.3, 5.4)).

*10.14 (The MyDate class) Design a class named MyDate. The class contains:

■■ The data fields year, month, and day that represent a date. month is 0-based, i.e., 0 is for January.

■■ A no-arg constructor that creates a MyDate object for the current date. ■■ A constructor that constructs a MyDate object with a specified elapsed time

since midnight, January 1, 1970, in milliseconds. ■■ A constructor that constructs a MyDate object with the specified year, month,

and day. ■■ Three getter methods for the data fields year, month, and day, respectively. ■■ A method named setDate(long elapsedTime) that sets a new date for

the object using the elapsed time.

Draw the UML diagram for the class then implement the class. Write a test program that creates two MyDate objects (using new MyDate() and new MyDate(34355555133101L)) and displays their year, month, and day.

(Hint: The first two constructors will extract the year, month, and day from the elapsed time. For example, if the elapsed time is 561555550000 milliseconds, the year is 1987, the month is 9, and the day is 18. You may use the GregorianCalendar class discussed in Programming Exercise 9.5 to simplify coding.)

*10.15 (Geometry: the bounding rectangle) A bounding rectangle is the minimum rec- tangle that encloses a set of points in a two-dimensional plane, as shown in Figure 10.24d. Write a method that returns a bounding rectangle for a set of points in a two-dimensional plane, as follows:

public static MyRectangle2D getRectangle(double[][] points)

Programming Exercises 429

M10_LIAN1878_11_GE_C10.indd 429 12/29/17 4:15 PM

430 Chapter 10 Object-Oriented Thinking

The Rectangle2D class is defined in Programming Exercise 10.13. Write a test program that prompts the user to enter five points and displays the bounding rectangle’s center, width, and height. Here is a sample run:

Enter five points: 1.0 2.5 3 4 5 6 7 8 9 10

The bounding rectangle's center (5.0, 6.25), width 8.0, height 7.5

Section 10.9 *10.16 (Divisible by 2 or 3) Find the first 10 numbers with 50 decimal digits that are

divisible by 2 or 3.

*10.17 (Square numbers) Find the first 10 square numbers that are greater than Long. MAX_VALUE. A square number is a number in the form of n2. For example, 4, 9, and 16 are square numbers. Find an efficient approach to run your program fast.

*10.18 (Large prime numbers) Write a program that finds five prime numbers larger than Long.MAX_VALUE.

*10.19 (Mersenne prime) A prime number is called a Mersenne prime if it can be writ- ten in the form 2p - 1 for some positive integer p. Write a program that finds all Mersenne primes with p … 100 and displays the output as shown below. (Hint: You have to use BigInteger to store the number because it is too big to be stored in long. Your program may take several hours to run.)

p 2^p – 1 ---------------------

2 3 3 7 5 31 ...

*10.20 (Approximate e) Programming Exercise 5.26 approximates e using the following series:

e = 1 + 1 1!

+ 1 2!

+ 1 3!

+ 1 4!

+ g + 1 i!

In order to get better precision, use BigDecimal with 25 digits of precision in the computation. Write a program that displays the e value for i = 100, 200, . . . , and 1000.

10.21 (Divisible by 5 or 6) Find the first 10 numbers greater than Long.MAX_VALUE that are divisible by 5 or 6.

Sections 10.10 and 10.11 **10.22 (Implement the String class) The String class is provided in the Java library.

Provide your own implementation for the following methods (name the new class MyString1):

public MyString1(char[] chars); public char charAt(int index); public int length(); public MyString1 substring(int begin, int end); public MyString1 toLowerCase(); public boolean equals(MyString1 s); public static MyString1 valueOf(int i);

M10_LIAN1878_11_GE_C10.indd 430 12/29/17 4:15 PM

**10.23 (Implement the String class) The String class is provided in the Java library. Provide your own implementation for the following methods (name the new class MyString2):

public MyString2(String s); public int compare(String s); public MyString2 substring(int begin); public MyString2 toUpperCase(); public char[] toChars(); public static MyString2 valueOf(boolean b);

10.24 (Implement the Character class) The Character class is provided in the Java library. Provide your own implementation for this class. Name the new class MyCharacter.

**10.25 (New string split method) The split method in the String class returns an array of strings consisting of the substrings split by the delimiters. However, the delimiters are not returned. Implement the following new method that returns an array of strings consisting of the substrings split by the matching delimiters, including the matching delimiters.

public static String[] split(String s, String regex)

For example, split("ab#12#453", "#") returns ab, #, 12, #, and 453 in an array of String and split("a?b?gf#e", "[?#]") returns a, ?, b, ?, gf, #, and e in an array of String.

*10.26 (Calculator) Revise Listing 7.9, Calculator.java, to accept an expression as a string in which the operands and operator are separated by zero or more spaces. For example, 3+4 and 3 + 4 are acceptable expressions. Here is a sample run:

**10.27 (Implement the StringBuilder class) The StringBuilder class is provided in the Java library. Provide your own implementation for the following methods (name the new class MyStringBuilder1):

public MyStringBuilder1(String s); public MyStringBuilder1 append(MyStringBuilder1 s); public MyStringBuilder1 append(int i); public int length(); public char charAt(int index); public MyStringBuilder1 toLowerCase(); public MyStringBuilder1 substring(int begin, int end); public String toString();

Programming Exercises 431

M10_LIAN1878_11_GE_C10.indd 431 12/29/17 4:15 PM

432 Chapter 10 Object-Oriented Thinking

**10.28 (Implement the StringBuilder class) The StringBuilder class is provided in the Java library. Provide your own implementation for the following methods (name the new class MyStringBuilder2):

public MyStringBuilder2(); public MyStringBuilder2(char[] chars); public MyStringBuilder2(String s); public MyStringBuilder2 insert(int offset, MyStringBuilder2 s); public MyStringBuilder2 reverse(); public MyStringBuilder2 substring(int begin); public MyStringBuilder2 toUpperCase();

M10_LIAN1878_11_GE_C10.indd 432 12/29/17 4:15 PM

Objectives ■■ To define a subclass from a superclass through inheritance (§11.2).

■■ To invoke the superclass’s constructors and methods using the super keyword (§11.3).

■■ To override instance methods in the subclass (§11.4).

■■ To distinguish differences between overriding and overloading (§11.5).

■■ To explore the toString() method in the Object class (§11.6).

■■ To discover polymorphism and dynamic binding (§§11.7 and 11.8).

■■ To describe casting and explain why explicit downcasting is necessary (§11.9).

■■ To explore the equals method in the Object class (§11.10).

■■ To store, retrieve, and manipulate objects in an ArrayList (§11.11).

■■ To construct an array list from an array, to sort and shuffle a list, and to obtain max and min element from a list (§11.12).

■■ To implement a Stack class using ArrayList (§11.13).

■■ To enable data and methods in a superclass accessible from subclasses using the protected visibility modifier (§11.14).

■■ To prevent class extending and method overriding using the final modifier (§11.15).

Inheritance and Polymorphism

Chapter

M11_LIAN1878_11_GE_C11.indd 433 12/29/17 4:18 PM

434 Chapter 11 Inheritance and Polymorphism

11.1 Introduction Object-oriented programming allows you to define new classes from existing classes. This is called inheritance.

As discussed in the preceding chapter, the procedural paradigm focuses on designing methods, and the object-oriented paradigm couples data and methods together into objects. Software design using the object-oriented paradigm focuses on objects and operations on objects. The object-oriented approach combines the power of the procedural paradigm with an added dimen- sion that integrates data with operations into objects.

Inheritance is an important and powerful feature for reusing software. Suppose you need to define classes to model circles, rectangles, and triangles. These classes have many common fea- tures. What is the best way to design these classes so as to avoid redundancy and make the system easy to comprehend and easy to maintain? The answer is to use inheritance.

11.2 Superclasses and Subclasses Inheritance enables you to define a general class (i.e., a superclass) and later extend it to more specialized classes (i.e., subclasses).

You use a class to model objects of the same type. Different classes may have some common properties and behaviors, which can be generalized in a class that can be shared by other classes. You can define a specialized class that extends the generalized class. The specialized classes inherit the properties and methods from the general class.

Consider geometric objects. Suppose you want to design the classes to model geometric objects such as circles and rectangles. Geometric objects have many common properties and behaviors. They can be drawn in a certain color and be filled or unfilled. Thus, a general class GeometricObject can be used to model all geometric objects. This class contains the proper- ties color and filled and their appropriate getter and setter methods. Assume this class also contains the dateCreated property, and the getDateCreated() and toString() methods. The toString() method returns a string representation of the object. Since a circle is a spe- cial type of geometric object, it shares common properties and methods with other geometric objects. Thus, it makes sense to define the Circle class that extends the GeometricObject class. Likewise, Rectangle can also be defined as a special type of GeometricObject. Figure 11.1 shows the relationship among these classes. A triangular arrow pointing to the gen- eralized class is used to denote the inheritance relationship between the two classes involved.

In Java terminology, a class C1 extended from another class C2 is called a subclass, and C2 is called a superclass. A superclass is also referred to as a parent class or a base class, and a subclass as a child class, an extended class, or a derived class. A subclass inherits accessible data fields and methods from its superclass and may also add new data fields and methods. Therefore, Circle and Rectangle are subclasses of GeometricObject, and GeometricObject is the superclass for Circle and Rectangle. A class defines a type. A type defined by a subclass is called a subtype, and a type defined by its superclass is called a supertype. Therefore, you can say that Circle is a subtype of GeometricObject, and GeometricObject is a supertype for Circle.

The subclass and its superclass are said to form a is-a relationship. A Circle object is a special type of general GeometricObject. The Circle class inherits all accessible data fields and methods from the GeometricObject class. In addition, it has a new data field, radius, and its associated getter and setter methods. The Circle class also contains the getArea(), getPerimeter(), and getDiameter() methods for returning the area, perimeter, and diam- eter of the circle.

The Rectangle class inherits all accessible data fields and methods from the GeometricObject class. In addition, it has the data fields width and height and their associated getter and setter methods. It also contains the getArea() and getPerimeter()

Point Key

inheritance

why inheritance?

Point Key

VideoNote

Geometric class hierarchy

subclass

superclass

subtype supertype

is-a relationship

width and height

M11_LIAN1878_11_GE_C11.indd 434 12/29/17 4:18 PM

11.2 Superclasses and Subclasses 435

methods for returning the area and perimeter of the rectangle. Note that you may have used the terms width and length to describe the sides of a rectangle in geometry. The common terms used in computer science are width and height, where width refers to the horizontal length, and height to the vertical length.

The GeometricObject, Circle, and Rectangle classes are shown in Listings 11.1, 11.2, and 11.3, respectively.

lisTing 11.1 GeometricObject.java 1 public class GeometricObject { 2 private String color = "white"; 3 private boolean filled; 4 private java.util.Date dateCreated; 5 6 /** Construct a default geometric object */ 7 public GeometricObject() { 8 dateCreated = new java.util.Date(); 9 } 10

data fields

constructor date constructed

Figure 11.1 The GeometricObject class is the superclass for Circle and Rectangle.

The color of the object (default: white).

Indicates whether the object is �lled with a color (default: false).

The date when the object was created.

Creates a GeometricObject.

Returns the color.

Sets a new color.

Returns the filled property.

Sets a new filled property.

Returns the dateCreated.

Returns a string representation of this object.

GeometricObject

–color: String

–filled: boolean

–dateCreated: java.util.Date

+GeometricObject()

Creates a GeometricObject with the speci�ed color and �lled values.

+GeometricObject(color: String, filled: boolean) +getColor(): String

+setColor(color: String): void

+isFilled(): boolean

+setFilled(filled: boolean): void

+getDateCreated(): java.util.Date

+toString(): String

Circle

–radius: double

+Circle()

+Circle(radius: double)

+Circle(radius: double, color: String, filled: boolean)

+getRadius(): double

+getArea(): double

+setRadius(radius: double): void

+printCircle(): void

+getPerimeter(): double

+getDiameter(): double

Rectangle

–height: double

+Rectangle()

+Rectangle(width: double, height: double)

+Rectangle(width: double, height: double color: String, filled: boolean)

+getWidth(): double

+getHeight(): double

+setHeight(height: double): void

+setWidth(width: double): void

+getArea(): double

+getPerimeter(): double

–width: double

M11_LIAN1878_11_GE_C11.indd 435 12/29/17 4:18 PM

436 Chapter 11 Inheritance and Polymorphism

11 /** Construct a geometric object with the specified color 12 * and filled value */ 13 public GeometricObject(String color, boolean filled) { 14 dateCreated = new java.util.Date(); 15 this.color = color; 16 this.filled = filled; 17 } 18 19 /** Return color */ 20 public String getColor() { 21 return color; 22 } 23 24 /** Set a new color */ 25 public void setColor(String color) { 26 this.color = color; 27 } 28 29 /** Return filled. Since filled is boolean, 30 its getter method is named isFilled */ 31 public boolean isFilled() { 32 return filled; 33 } 34 35 /** Set a new filled */ 36 public void setFilled(boolean filled) { 37 this.filled = filled; 38 } 39 40 /** Get dateCreated */ 41 public java.util.Date getDateCreated() { 42 return dateCreated; 43 } 44 45 /** Return a string representation of this object */ 46 public String toString() { 47 return "created on " + dateCreated + "\ncolor: " + color + 48 " and filled: " + filled; 49 } 50 }

Listing 11.2 Circle.java 1 public class Circle extends GeometricObject { 2 private double radius; 3 4 public Circle() { 5 } 6 7 public Circle(double radius) { 8 this.radius = radius; 9 } 10 11 public Circle(double radius, 12 String color, boolean filled) { 13 this.radius = radius; 14 setColor(color); 15 setFilled(filled); 16 } 17

data fields extends superclass

constructor

M11_LIAN1878_11_GE_C11.indd 436 12/29/17 4:18 PM

11.2 Superclasses and Subclasses 437

18 /** Return radius */ 19 public double getRadius() { 20 return radius; 21 } 22 23 /** Set a new radius */ 24 public void setRadius(double radius) { 25 this.radius = radius; 26 } 27 28 /** Return area */ 29 public double getArea() { 30 return radius * radius * Math.PI; 31 } 32 33 /** Return diameter */ 34 public double getDiameter() { 35 return 2 * radius; 36 } 37 38 /** Return perimeter */ 39 public double getPerimeter() { 40 return 2 * radius * Math.PI; 41 } 42 43 /** Print the circle info */ 44 public void printCircle() { 45 System.out.println("The circle is created " + getDateCreated() + 46 " and the radius is " + radius); 47 } 48 }

The Circle class (Listing 11.2) extends the GeometricObject class (Listing 11.1) using the following syntax:

methods

Subclass Superclass

public class Circle extends GeometricObject

The keyword extends (line 1) tells the compiler that the Circle class extends the GeometricObject class, thus inheriting the methods getColor, setColor, isFilled, setFilled, and toString.

The overloaded constructor Circle(double radius, String color, boolean filled) is implemented by invoking the setColor and setFilled methods to set the color and filled properties (lines 14 and 15). The public methods defined in the superclass GeometricObject are inherited in Circle, so they can be used in the Circle class.

You might attempt to use the data fields color and filled directly in the constructor as follows:

public Circle(double radius, String color, boolean filled) { this.radius = radius; this.color = color; // Illegal this.filled = filled; // Illegal }

private member in superclass

M11_LIAN1878_11_GE_C11.indd 437 12/29/17 4:18 PM

438 Chapter 11 Inheritance and Polymorphism

This is wrong because the private data fields color and filled in the GeometricObject class cannot be accessed in any class other than in the GeometricObject class itself. The only way to read and modify color and filled is through their getter and setter methods.

The Rectangle class (Listing 11.3) extends the GeometricObject class (Listing 11.1) using the following syntax:

Subclass Superclass

public class Rectangle extends GeometricObject

The keyword extends (line 1) tells the compiler the Rectangle class extends the GeometricObject class, thus inheriting the methods getColor, setColor, isFilled, setFilled, and toString.

Listing 11.3 Rectangle.java 1 public class Rectangle extends GeometricObject { 2 private double width; 3 private double height; 4

5 public Rectangle() { 6 } 7 8 public Rectangle(double width, double height) { 9 this.width = width; 10 this.height = height; 11 } 12 13 public Rectangle( 14 double width, double height, String color, boolean filled) { 15 this.width = width; 16 this.height = height; 17 setColor(color); 18 setFilled(filled); 19 } 20 21 /** Return width */

22 public double getWidth() { 23 return width; 24 } 25 26 /** Set a new width */ 27 public void setWidth(double width) { 28 this.width = width; 29 } 30 31 /** Return height */ 32 public double getHeight() { 33 return height; 34 } 35 36 /** Set a new height */ 37 public void setHeight(double height) { 38 this.height = height; 39 } 40

methods

constructor

data fields extends superclass

M11_LIAN1878_11_GE_C11.indd 438 12/29/17 4:18 PM

11.2 Superclasses and Subclasses 439

41 /** Return area */ 42 public double getArea() { 43 return width * height; 44 } 45 46 /** Return perimeter */ 47 public double getPerimeter() { 48 return 2 * (width + height); 49 } 50 }

The code in Listing 11.4 creates objects of Circle and Rectangle and invokes the methods on these objects. The toString() method is inherited from the GeometricObject class and is invoked from a Circle object (line 4) and a Rectangle object (line 11).

Listing 11.4 TestCircleRectangle.java 1 public class TestCircleRectangle { 2 public static void main(String[] args) { 3 Circle circle = new Circle(1); 4 System.out.println("A circle " + circle.toString()); 5 System.out.println("The color is " + circle.getColor()); 6 System.out.println("The radius is " + circle.getRadius()); 7 System.out.println("The area is " + circle.getArea()); 8 System.out.println("The diameter is " + circle.getDiameter()); 9 10 Rectangle rectangle = new Rectangle(2, 4); 11 System.out.println("\nA rectangle " + rectangle.toString()); 12 System.out.println("The area is " + rectangle.getArea()); 13 System.out.println("The perimeter is " + 14 rectangle.getPerimeter()); 15 } 16 }

invoke toString Rectangle object

invoke getColor invoke toString Circle object

A circle created on Thu Feb 10 19:54:25 EST 2011

color: white and filled: false The color is white The radius is 1.0 The area is 3.141592653589793 The diameter is 2.0 A rectangle created on Thu Feb 10 19:54:25 EST 2011 color: white and filled: false The area is 8.0 The perimeter is 12.0

Note the following points regarding inheritance:

■■ Contrary to the conventional interpretation, a subclass is not a subset of its superclass. In fact, a subclass usually contains more information and methods than its superclass.

■■ Private data fields in a superclass are not accessible outside the class. Therefore, they cannot be used directly in a subclass. They can, however, be accessed/mutated through public accessors/mutators if defined in the superclass.

private data fields

directory

. . . d2d1 f1 f2 . . .fm dn

The size of the directory can be defined recursively as follows:

size(d) = size(f1) + size(f2) + g + size(fm) + size(d1) + size(d2) + g + size(dn)

The File class, introduced in Section 12.10, can be used to represent a file or a directory and obtain the properties for files and directories. Two methods in the File class are useful for this problem:

■■ The length() method returns the size of a file.

■■ The listFiles() method returns an array of File objects under a directory.

Listing 18.7 gives a program that prompts the user to enter a directory or a file and displays its size.

Listing 18.7 DirectorySize.java 1 import java.io.File; 2 import java.util.Scanner; 3 4 public class DirectorySize { 5 public static void main(String[] args) { 6 // Prompt the user to enter a directory or a file 7 System.out.print("Enter a directory or a file: "); 8 Scanner input = new Scanner(System.in); 9 String directory = input.nextLine(); 10 11 // Display the size 12 System.out.println(getSize(new File(directory)) + " bytes"); 13 } 14 15 public static long getSize(File file) { 16 long size = 0; // Store the total size of all files 17

invoke method

getSize method

M18_LIAN1878_11_GE_C18.indd 753 12/29/17 8:00 PM

754 Chapter 18 Recursion

18 if (file.isDirectory()) { 19 File[] files = file.listFiles(); // All files and subdirectories 20 for (int i = 0; files != null && i < files.length; i++) { 21 size += getSize(files[i]); // Recursive call 22 } 23 } 24 else { // Base case 25 size += file.length(); 26 } 27 28 return size; 29 } 30 }

base case

recursive call

all subitems is directory?

Enter a directory or a file: c:\book

48619631 bytes

Enter a directory or a file: c:\book\Welcome.java

172 bytes

Enter a directory or a file: c:\book\NonExistentFile

0 bytes

If the file object represents a directory (line 18), each subitem (file or subdirectory) in the directory is recursively invoked to obtain its size (line 21). If the file object represents a file (line 24), the file size is obtained and added to the total size (line 25).

What happens if an incorrect or a nonexistent directory is entered? The program will detect that it is not a directory and invoke file.length() (line 25), which returns 0. Thus, in this case, the getSize method will return 0.

Tip To avoid mistakes, it is a good practice to test all cases. For example, you should test the program for an input of file, an empty directory, a nonexistent directory, and a nonexistent file.

18.6.1 What is the base case for the getSize method? 18.6.2 How does the program get all files and directories under a given directory? 18.6.3 How many times will the getSize method be invoked for a directory if the

directory has three subdirectories and each subdirectory has four files?

18.6.4 Will the program work if the directory is empty (i.e., it does not contain any files)? 18.6.5 Will the program work if line 20 is replaced by the following code?

for (int i = 0; i < files.length; i++)

18.6.6 Will the program work if lines 20 and 21 are replaced by the following code? for (File file: files) size += getSize(file); // Recursive call

testing all cases

Point Check

M18_LIAN1878_11_GE_C18.indd 754 12/29/17 8:00 PM

18.7 Case Study: Tower of Hanoi 755

18.7 Case Study: Tower of Hanoi The Tower of Hanoi problem is a classic problem that can be solved easily using recursion, but it is difficult to solve otherwise.

The problem involves moving a specified number of disks of distinct sizes from one tower to another while observing the following rules:

■■ There are n disks labeled 1, 2, 3, . . . , n and three towers labeled A, B, and C.

■■ No disk can be on top of a smaller disk at any time.

■■ All the disks are initially placed on tower A.

■■ Only one disk can be moved at a time and it must be the smallest disk on a tower.

The objective of the problem is to move all the disks from A to B with the assistance of C. For example, if you have three disks, the steps to move all of the disks from A to B are shown in Figure 18.6. For an interactive demo, see liveexample.pearsoncmg.com/dsanimation/ TowerOfHanoieBook.html.

Point Key

Figure 18.6 The goal of the Tower of Hanoi problem is to move disks from tower A to tower B without breaking the rules.

A B C

A B CA C

Original position

Step 1: Move disk 1 from A to B

Step 2: Move disk 2 from A to C

Step 3: Move disk 1 from B to C

Step 4: Move disk 3 from A to B

Step 5: Move disk 1 from C to A

Step 7: Move disk 1 from A to B

Step 6: Move disk 2 from C to B

A A B C

0 4

1 5

2 6

3 7

1 2 3

2 3

3 1 2

23 1

3 2

1 3 2

1 23

In the case of three disks, you can find the solution manually. For a larger number of disks, however—even for four—the problem is quite complex. Fortunately, the problem has an inher- ently recursive nature, which leads to a straightforward recursive solution.

The base case for the problem is n = 1. If n == 1, you could simply move the disk from A to B. When n > 1, you could split the original problem into the following three subproblems and solve them sequentially.

M18_LIAN1878_11_GE_C18.indd 755 12/29/17 8:00 PM

756 Chapter 18 Recursion

1. Move the first n − 1 disks from A to C recursively with the assistance of tower B, as shown in Step 1 in Figure 18.7.

2. Move disk n from A to B, as shown in Step 2 in Figure 18.7.

3. Move n − 1 disks from C to B recursively with the assistance of tower A, as shown in Step 3 in Figure 18.7.

Figure 18.7 The Tower of Hanoi problem can be decomposed into three subproblems.

A B Original position

A B Step 1: Move the �rst n – 1 disks from

A to C recursively

A B Step 2: Move disk n from A to B

A B

Step 3: Move n – 1 disks from C to B recursively

n – 1 disks

The following method moves n disks from the fromTower to the toTower with the assistance of the auxTower:

void moveDisks(int n, char fromTower, char toTower, char auxTower)

The algorithm for the method can be described as:

if (n == 1) // Stopping condition Move disk 1 from the fromTower to the toTower; else { moveDisks(n − 1, fromTower, auxTower, toTower); Move disk n from the fromTower to the toTower; moveDisks(n − 1, auxTower, toTower, fromTower); }

Listing 18.8 gives a program that prompts the user to enter the number of disks and invokes the recursive method moveDisks to display the solution for moving the disks.

Listing 18.8 TowerOfHanoi.java 1 import java.util.Scanner; 2 3 public class TowerOfHanoi { 4 /** Main method */

M18_LIAN1878_11_GE_C18.indd 756 12/29/17 8:00 PM

18.7 Case Study: Tower of Hanoi 757

5 public static void main(String[] args) { 6 // Create a Scanner 7 Scanner input = new Scanner(System.in); 8 System.out.print("Enter number of disks: "); 9 int n = input.nextInt(); 10 11 // Find the solution recursively 12 System.out.println("The moves are:"); 13 moveDisks(n, 'A', 'B', 'C'); 14 } 15 16 /** The method for finding the solution to move n disks 17 from fromTower to toTower with auxTower */ 18 public static void moveDisks(int n, char fromTower, 19 char toTower, char auxTower) { 20 if (n == 1) // Stopping condition 21 System.out.println("Move disk " + n + " from " + 22 fromTower + " to " + toTower); 23 else { 24 moveDisks(n − 1, fromTower, auxTower, toTower); 25 System.out.println("Move disk " + n + " from " + 26 fromTower + " to " + toTower); 27 moveDisks(n − 1, auxTower, toTower, fromTower); 28 } 29 } 30 }

recursion

base case

Enter number of disks: 4 The moves are: Move disk 1 from A to C Move disk 2 from A to B Move disk 1 from C to B Move disk 3 from A to C Move disk 1 from B to A Move disk 2 from B to C Move disk 1 from A to C Move disk 4 from A to B Move disk 1 from C to B Move disk 2 from C to A Move disk 1 from B to A Move disk 3 from C to B Move disk 1 from A to C Move disk 2 from A to B Move disk 1 from C to B

This problem is inherently recursive. Using recursion makes it possible to find a natural, simple solution. It would be difficult to solve the problem without using recursion.

Consider tracing the program for n = 3. The successive recursive calls are shown in Figure 18.8. As you can see, writing the program is easier than tracing the recursive calls. The system uses stacks to manage the calls behind the scenes. To some extent, recursion provides a level of abstraction that hides iterations and other details from the user.

18.7.1 How many times is the moveDisks method in Listing 18.8 invoked for moveDisks(5, 'A', 'B', 'C')? Point

Check

M18_LIAN1878_11_GE_C18.indd 757 12/29/17 8:00 PM

758 Chapter 18 Recursion

18.8 Case Study: Fractals Using recursion is ideal for displaying fractals because fractals are inherently recursive.

A fractal is a geometrical figure, but unlike triangles, circles, and rectangles, fractals can be divided into parts, each of which is a reduced-size copy of the whole. There are many interest- ing examples of fractals. This section introduces a simple fractal, the Sierpinski triangle, named after a famous Polish mathematician.

A Sierpinski triangle is created as follows:

1. Begin with an equilateral triangle, which is considered to be a Sierpinski fractal of order (or level) 0, as shown in Figure 18.9a.

2. Connect the midpoints of the sides of the triangle of order 0 to create a Sierpinski triangle of order 1 (see Figure 18.9b).

3. Leave the center triangle intact. Connect the midpoints of the sides of the three other triangles to create a Sierpinski triangle of order 2 (see Figure 18.9c).

4. You can repeat the same process recursively to create a Sierpinski triangle of order 3, 4, . . . , and so on (see Figure 18.9d). For an interactive demo, see liveexample.pearsoncmg. com/dsanimation/SierpinskiTriangleUsingHTML.html.

The problem is inherently recursive. How do you develop a recursive solution for it? Con- sider the base case when the order is 0. It is easy to draw a Sierpinski triangle of order 0. How do you draw a Sierpinski triangle of order 1? The problem can be reduced to drawing three Sierpinski triangles of order 0. How do you draw a Sierpinski triangle of order 2? The problem can be reduced to drawing three Sierpinski triangles of order 1, so the problem of drawing a Sierpinski triangle of order n can be reduced to drawing three Sierpinski triangles of order n − 1.

Listing 18.9 gives a program that displays a Sierpinski triangle of any order, as shown in Figure 18.9. You can enter an order in a text field to display a Sierpinski triangle of the speci- fied order.

Listing 18.9 SierpinskiTriangle.java 1 import javafx.application.Application; 2 import javafx.geometry.Point2D; 3 import javafx.geometry.Pos;

Point Key

VideoNote

Fractal (Sierpinski triangle)

Figure 18.8 Invoking moveDisks(3, 'A', 'B', 'C') spawns calls to moveDisks recursively.

moveDisks(2,'A','C','B') move disk 3 from A to B moveDisks(2,'C','B','A')

moveDisks(3,'A','B','C')

moveDisks(1,'A','B','C') move disk 2 from A to C moveDisks(1,'B','C','A')

moveDisks(2,'A','C','B')

moveDisks(1,'C','A','B') move disk 2 from C to B moveDisks(1,'A','B','C')

moveDisks(2,'C','B','A')

moveDisks(1,'A','B','C')

move disk 1 from A to B

moveDisks(1,'B','C','A')

move disk 1 from B to C

moveDisks(1,'C','A','B')

move disk 1 from C to A

moveDisks(1,'A','B','C')

move disk 1 from A to B

M18_LIAN1878_11_GE_C18.indd 758 12/29/17 8:00 PM

18.8 Case Study: Fractals 759

4 import javafx.scene.Scene; 5 import javafx.scene.control.Label; 6 import javafx.scene.control.TextField; 7 import javafx.scene.layout.BorderPane; 8 import javafx.scene.layout.HBox; 9 import javafx.scene.layout.Pane; 10 import javafx.scene.paint.Color; 11 import javafx.scene.shape.Polygon; 12 import javafx.stage.Stage; 13 14 public class SierpinskiTriangle extends Application { 15 @Override // Override the start method in the Application class 16 public void start(Stage primaryStage) { 17 SierpinskiTrianglePane pane = new SierpinskiTrianglePane(); 18 TextField tfOrder = new TextField(); 19 tfOrder.setOnAction( 20 e −> pane.setOrder(Integer.parseInt(tfOrder.getText()))); 21 tfOrder.setPrefColumnCount(4); 22 tfOrder.setAlignment(Pos.BOTTOM_RIGHT); 23 24 // Pane to hold label, text field, and a button 25 HBox hBox = new HBox(10); 26 hBox.getChildren().addAll(new Label("Enter an order: "), tfOrder); 27 hBox.setAlignment(Pos.CENTER); 28 29 BorderPane borderPane = new BorderPane(); 30 borderPane.setCenter(pane); 31 borderPane.setBottom(hBox); 32

(a) Order 0 (b) Order 1

recursive triangle pane

listener for text field

hold label and text field

M18_LIAN1878_11_GE_C18.indd 759 12/29/17 8:00 PM

760 Chapter 18 Recursion

33 // Create a scene and place it in the stage 34 Scene scene = new Scene(borderPane, 200, 210); 35 primaryStage.setTitle("SierpinskiTriangle"); // Set the stage title 36 primaryStage.setScene(scene); // Place the scene in the stage 37 primaryStage.show(); // Display the stage 38 39 pane.widthProperty().addListener(ov −> pane.paint()); 40 pane.heightProperty().addListener(ov −> pane.paint()); 41 } 42 43 /** Pane for displaying triangles */ 44 static class SierpinskiTrianglePane extends Pane { 45 private int order = 0; 46 47 /** Set a new order */ 48 public void setOrder(int order) { 49 this.order = order; 50 paint(); 51 } 52 53 SierpinskiTrianglePane() { 54 } 55 56 protected void paint() { 57 // Select three points in proportion to the pane size 58 Point2D p1 = new Point2D(getWidth() / 2, 10); 59 Point2D p2 = new Point2D(10, getHeight() − 10); 60 Point2D p3 = new Point2D(getWidth() − 10, getHeight() − 10); 61 62 this.getChildren().clear(); // Clear the pane before redisplay 63 64 displayTriangles(order, p1, p2, p3); 65 } 66 67 private void displayTriangles(int order, Point2D p1, 68 Point2D p2, Point2D p3) { 69 if (order == 0) { 70 // Draw a triangle to connect three points 71 Polygon triangle = new Polygon(); 72 triangle.getPoints().addAll(p1.getX(), p1.getY(), p2.getX(), 73 p2.getY(), p3.getX(), p3.getY()); 74 triangle.setStroke(Color.BLACK); 75 triangle.setFill(Color.WHITE); 76 77 this.getChildren().add(triangle); 78 } 79 else { 80 // Get the midpoint on each edge in the triangle 81 Point2D p12 = p1.midpoint(p2); 82 Point2D p23 = p2.midpoint(p3); 83 Point2D p31 = p3.midpoint(p1); 84 85 // Recursively display three triangles 86 displayTriangles(order − 1, p1, p12, p31); 87 displayTriangles(order − 1, p12, p2, p23); 88 displayTriangles(order − 1, p31, p23, p3); 89 } 90 } 91 } 92 }

listener for resizing

three initial points

clear the pane

draw a triangle

create a triangle

top subtriangle left subtriangle right subtriangle

M18_LIAN1878_11_GE_C18.indd 760 12/29/17 8:00 PM

18.8 Case Study: Fractals 761

The initial triangle has three points set in proportion to the pane size (lines 58–60). If order == 0, the displayTriangles(order, p1, p2, p3) method displays a triangle that connects the three points p1, p2, and p3 (lines 71–77), as shown in Figure 18.10a. Otherwise, it performs the following tasks:

1. Obtain the midpoint between p1 and p2 (line 81), the midpoint between p2 and p3 (line 82), and the midpoint between p3 and p1 (line 83), as shown in Figure 18.10b.

2. Recursively invoke displayTriangles with a reduced order to display three smaller Sierpinski triangles (lines 86–88). Note each small Sierpinski triangle is structurally identical to the original big Sierpinski triangle except that the order of a small triangle is one less, as shown in Figure 18.10b.

displayTriangle method

Figure 18.10 Drawing a Sierpinski triangle spawns calls to draw three small Sierpinski triangles recursively.

p3p2

Draw the Sierpinski triangle. displayTriangles(order, p1, p2, p3)

(a)

p3p2

p12 p31

p23

Recursively draw the small Sierpinski triangle. displayTriangles( order – 1, p1, p12, p31)

Recursively draw the small Sierpinski triangle. displayTriangles( order – 1, p31, p23, p3)

Recursively draw the small Sierpinski triangle. displayTriangles( order – 1, p12, p2, p23)

(b)

A Sierpinski triangle is displayed in a SierpinskiTrianglePane. The order property in the inner class SierpinskiTrianglePane specifies the order for the Sierpinski triangle. The Point2D class, introduced in Section 9.6.3 The Point2D Class, represents a point with x- and y-coordinates. Invoking p1.midpoint(p2) returns a new Point2D object that is the midpoint between p1 and p2 (lines 81–83).

18.8.1 How do you obtain the midpoint between two points? 18.8.2 What is the base case for the displayTriangles method? 18.8.3 How many times is the displayTriangles method invoked for a Sierpinski tri-

angle of order 0, order 1, order 2, and order n?

18.8.4 What happens if you enter a negative order? How do you fix this problem in the code?

18.8.5 Instead of drawing a triangle using a polygon, rewrite the code to draw a triangle by drawing three lines to connect the points in lines 71–77.

Point Check

M18_LIAN1878_11_GE_C18.indd 761 12/29/17 8:00 PM

762 Chapter 18 Recursion

18.9 Recursion vs. Iteration Recursion is an alternative form of program control. It is essentially repetition without a loop.

When you use loops, you specify a loop body. The repetition of the loop body is controlled by the loop control structure. In recursion, the method itself is called repeatedly. A selection statement must be used to control whether to call the method recursively or not.

Recursion bears substantial overhead. Each time the program calls a method, the system must allocate memory for all of the method’s local variables and parameters. This can consume considerable memory and requires extra time to manage the memory.

Any problem that can be solved recursively can be solved nonrecursively with iterations. Recursion has some negative aspects: It uses up too much time and too much memory. Why, then, should you use it? In some cases, using recursion enables you to specify a clear, simple solution for an inherently recursive problem that would otherwise be difficult to obtain. Exam- ples are the directory-size problem, the Tower of Hanoi problem, and the fractal problem, which are rather difficult to solve without using recursion.

The decision whether to use recursion or iteration should be based on the nature of, and your understanding of, the problem you are trying to solve. The rule of thumb is to use which- ever approach can best develop an intuitive solution that naturally mirrors the problem. If an iterative solution is obvious, use it. It will generally be more efficient than the recursive option.

Note Recursive programs can run out of memory, causing a StackOverflowError.

Tip If you are concerned about your program’s performance, avoid using recursion because it takes more time and consumes more memory than iteration. In general, recursion can be used to solve the inherent recursive problems such as Tower of Hanoi, recursive directories, and Sierpinski triangles.

18.9.1 Which of the following statements are true?

a. Any recursive method can be converted into a nonrecursive method.

b. Recursive methods take more time and memory to execute than nonrecursive methods.

c. Recursive methods are always simpler than nonrecursive methods.

d. There is always a selection statement in a recursive method to check whether a base case is reached.

18.9.2 What is a cause for a stack-overflow exception?

18.10 Tail Recursion A tail-recursive method is efficient.

A recursive method is said to be tail recursive if there are no pending operations to be performed on return from a recursive call, as illustrated in Figure 18.11a. However, method B in Figure 18.11b is not tail recursive because there are pending operations after a method call is returned.

For example, the recursive isPalindrome method (lines 6–13) in Listing 18.4 is tail recursive because there are no pending operations after recursively invoking isPalindrome in line 12. However, the recursive factorial method (lines 16–21) in Listing 18.1 is not tail recursive because there is a pending operation, namely multiplication, to be performed on return from each recursive call.

Point Key

recursion overhead

recursion advantages

recursion or iteration?

StackOverflowError

performance concern

Point Check

Point Keytail recursion

M18_LIAN1878_11_GE_C18.indd 762 12/29/17 8:00 PM

18.10 Tail Recursion 763

Tail recursion is desirable because the method ends when the last recursive call ends, and there is no need to store the intermediate calls in the stack. Compilers can optimize tail recur- sion to reduce stack size.

A nontail-recursive method can often be converted to a tail-recursive method by using auxiliary parameters. These parameters are used to contain the result. The idea is to incorpo- rate the pending operations into the auxiliary parameters in such a way that the recursive call no longer has a pending operation. You can define a new auxiliary recursive method with the auxiliary parameters. This method may overload the original method with the same name but a different signature. For example, the factorial method in Listing 18.1 is written in a tail- recursive way in Listing 18.10.

Listing 18.10 ComputeFactorialTailRecursion.java 1 public class ComputeFactorialTailRecursion { 2 /** Return the factorial for a specified number */ 3 public static long factorial(int n) { 4 return factorial(n, 1); // Call auxiliary method 5 } 6 7 /** Auxiliary tail−recursive method for factorial */ 8 private static long factorial(int n, int result) { 9 if (n == 0) 10 return result; 11 else 12 return factorial(n − 1, n * result); // Recursive call 13 } 14 }

The first factorial method (line 3) simply invokes the second auxiliary method (line 4). The second method contains an auxiliary parameter result that stores the result for the factorial of n. This method is invoked recursively in line 12. There is no pending operation after a call is returned. The final result is returned in line 10, which is also the return value from invoking factorial(n, 1) in line 4.

18.10.1 Identify tail-recursive methods in this chapter. 18.10.2 Rewrite the fib method in Listing 18.2 using tail recursion.

invoke auxiliary method

auxiliary method

original method

recursive call

Point Check

Recursive method A . . . . . .

. . . Invoke method A recursively

Recursive method B . . . . . . Invoke method B recursively . . . . . .

(a) Tail recursion (b) Nontail recursion

Figure 18.11 A tail-recursive method has no pending operations after a recursive call.

Key terms base case 742 direct recursion 745 indirect recursion 745 infinite recursion 745

recursive helper method 751 recursive method 742 stopping condition 742 tail recursion 762

M18_LIAN1878_11_GE_C18.indd 763 12/29/17 8:00 PM

764 Chapter 18 Recursion

ChapTer summary 1. A recursive method is one that directly or indirectly invokes itself. For a recursive

method to terminate, there must be one or more base cases.

2. Recursion is an alternative form of program control. It is essentially repetition without a loop control. It can be used to write simple, clear solutions for inherently recursive problems that would otherwise be difficult to solve.

3. Sometimes the original method needs to be modified to receive additional parameters in order to be invoked recursively. A recursive helper method can be defined for this purpose.

4. Recursion bears substantial overhead. Each time the program calls a method, the system must allocate memory for all of the method’s local variables and param- eters. This can consume considerable memory and requires extra time to manage the memory.

5. A recursive method is said to be tail recursive if there are no pending operations to be performed on return from a recursive call. Some compilers can optimize tail recursion to reduce stack size.

Quiz Answer the quiz for this chapter online at the book Companion Website.

programming exerCises

Sections 18.2 and 18.3 *18.1 (Identifying anagrams) Two words are anagrams of each other if they contain the

same letters that are arranged in different orders. Write a recursive method that can identify if two given words are anagrams of each other.

*18.2 (Fibonacci numbers) Rewrite the fib method in Listing 18.2 using iterations. Hint: To compute fib(n) without recursion, you need to obtain fib(n − 2) and fib(n − 1) first. Let f0 and f1 denote the two previous Fibonacci num- bers. The current Fibonacci number would then be f0 + f1. The algorithm can be described as follows:

f0 = 0; // For fib(0) f1 = 1; // For fib(1)

for (int i = 1; i <= n; i++) { currentFib = f0 + f1; f0 = f1; f1 = currentFib; } // After the loop, currentFib is fib(n)

Write a test program that prompts the user to enter an index and displays its Fibonacci number.

M18_LIAN1878_11_GE_C18.indd 764 12/29/17 8:00 PM

Programming Exercises 765

*18.3 Write a recursive method public static void dec2b(double x,int b, int n) that displays x, where 0 ≤ x < 1 in base b with at most n digits after the decimal point. For instance, dec2b(0.625,2,10) should return 0.101, and dec2b(0.625,3,10) should return 0.1212121212.

18.4 Combinations refer to the combination of n things taken p at a time without repetition. A recursive definition of C(n, p) is C(n, 0) = C(n, n) = 1, otherwise C(n, p) = C(n - 1, p) + C(n - 1, p - 1).

Write a method public static long C(long n, long p) that computes and returns C(n, p).

18.5 The previous method is probably not efficient if you wrote it exactly as its recur- sive definition. It is not because the method is recursive but because it has not been well-written. It is quite easy to verify that there is another recurrent defini- tion of C(n, p), which is C(n, 0) = C(n, n) = 1, otherwise C(n, p) = C(n, p - 1) * (n - p + 1) /p. Keeping this in mind, write the previous method.

*18.6 Write a recursive method public static void randomFillSortedArray (int[] x, int l, int r, int a, int b) that fills the array x between l and r with random values between a and b such that x is sorted. Here is a sample run:

Enter the array size: 10

Enter the limits: 0 1000

[235, 280, 382, 428, 458, 462, 484, 495, 536, 850]

Enter the array size: 10 Enter the limits: 0 9

[0, 0, 3, 3, 4, 6, 6, 8, 8, 9]

*18.7 (Fibonacci series) Modify Listing 18.2, ComputeFibonacci.java, so that the pro- gram finds the number of times the fib method is called. (Hint: Use a static variable and increment it every time the method is called.)

Section 18.4 *18.8 Let f be a function. Let l, r, and e be real numbers. Write a recursive method that

computes and returns a value v such that |f(v)| 6 = e, and l 6 v < r. That is, a method that finds an approximation of a zero of f between l and r. We assume that f(l )f(r) 6 0.

*18.9 (Print the characters in a string reversely) Write a recursive method that dis- plays a string reversely on the console using the following header:

public static void reverseDisplay(String value)

For example, reverseDisplay("abcd") displays dcba. Write a test program that prompts the user to enter a string and displays its reversal.

*18.10 (Occurrences of a specified character in a string) Write a recursive method that finds the number of occurrences of a specified letter in a string using the follow- ing method header:

public static int count(String str, char a)

For example, count("Welcome", 'e') returns 2. Write a test program that prompts the user to enter a string and a character, and displays the number of occurrences for the character in the string.

M18_LIAN1878_11_GE_C18.indd 765 12/29/17 8:00 PM

766 Chapter 18 Recursion

*18.11 Write a recursive method that displays all permutations of a given array of integers. Here is a sample run:

Enter the array size : 3

[1, 2, 3]

[1, 3, 2]

[2, 1, 3]

[2, 3, 1]

[3, 1, 2]

[3, 2, 1]

Section 18.5 **18.12 (Print the characters in a string reversely) Rewrite Programming Exercise 18.9

using a helper method to pass the substring high index to the method. The helper method header is

public static void reverseDisplay(String value, int high)

*18.13 (Find the largest number in an array) Write a recursive method that returns the largest integer in an array. Write a test program that prompts the user to enter a list of eight integers and displays the largest element.

*18.14 Assume that you are not able to perform other operations on an int value than to decrement it and compare it with zero and one. Write two methods odd(int x) and even(int x) that respectively return if x is odd or even. We assume that x 7= 0.

*18.15 (Occurrences of a specified character in a string) Rewrite Programming Exer- cise 18.10 using a helper method to pass the substring high index to the method. The helper method header is

public static int count(String str, char a, int high)

*18.16 Write a recursive method public static int pos(int[] a, int l, int r) that positions a[l] at its rank if the array a were sorted between l and r, and that returns this rank. That is, we assume that a is an unsorted array of int, l and r are int such that 0 6= l 6= r 6 a.length, after this call : k = pos(a, l, r); a contains the same elements before and after this call, but it is such that a[k] 7= a[i] for all i in [l, k-1] and a[k] 6 a[j] for all j in [k+1, r]. For instance, consider this fragment:

int[] a = {8,7,4,1,9,6,2,5,3,0};

System.out.println(Arrays.toString(a)); // [8, 7, 4, 1, 9, 6, 2, 5, 3, 0]

pos(a, 2, 7); // a[2..7] = {4, 1, 9, 6, 2, 5}

System.out.println(Arrays.toString(a));

/* [8, 7, 1, 2, 4, 6, 5, 9, 3, 0] is possible because 4 >= 2, 4 >= 1, 4 < 6, 4 < 5, 4 < 9.

But It could also be [8, 7, 2, 1, 4, 6, 5, 9, 3, 0], or [8, 7, 2, 1, 4, 9, 5, 6, 3, 0] */

*18.17 Write a recursive method that finds the kth-smallest value of a given array. Use the method pos of Programming Exercise 18.16.

M18_LIAN1878_11_GE_C18.indd 766 12/29/17 8:00 PM

Programming Exercises 767

Sections 18.6–18.10 *18.18 (Tower of Hanoi) Modify Listing 18.8, TowerOfHanoi.java, so the program finds

the number of moves needed to move n disks from tower A to tower B. (Hint: Use a static variable and increment it every time the method is called.)

*18.19 (Sierpinski triangle) Revise Listing 18.9 to develop a program that lets the user use the + /- buttons, primary/secondary mouse buttons, and UP/ DOWN arrow keys to increase or decrease the current order by 1, as shown in Figure 18.12a. The initial order is 0. If the current order is 0, the Decrease button is ignored.

(a) (b)

Figure 18.12 (a) Programming Exercise 18.19 uses the + or - buttons to increase or decrease the current order by 1. Source: Copyright © 1995–2016 Oracle and/or its affiliates. All rights reserved. Used with permission. (b) Programming Exercise 18.20 draws ovals using a recursive method.

*18.20 (Display circles) Write a Java program that displays ovals, as shown in Figure 18.12b. The circles are centered in the pane. The gap between two adja- cent circles is 10 pixels, and the gap between the border of the pane and the largest circle is also 10.

*18.21 (Decimal to binary) Write a recursive method that converts a decimal number into a binary number as a string. The method header is

public static String dec2Bin(int value)

Write a test program that prompts the user to enter a decimal number and displays its binary equivalent.

*18.22 (Decimal to hex) Write a recursive method that converts a decimal number into a hex number as a string. The method header is

public static String dec2Hex(int value)

Write a test program that prompts the user to enter a decimal number and displays its hex equivalent.

*18.23 (Binary to decimal) Write a recursive method that parses a binary number as a string into a decimal integer. The method header is

public static int bin2Dec(String binaryString)

Write a test program that prompts the user to enter a binary string and displays its decimal equivalent.

M18_LIAN1878_11_GE_C18.indd 767 12/29/17 8:00 PM

768 Chapter 18 Recursion

*18.24 Chess is played on a board with 64 squares arranged in a 8 * 8grid. The knight, a piece in this game, can move only in an L-shaped pattern, that is, two squares sideways and then one up or two squares up and one sideways. Can a knight access every square of the chessboard from any other square using this pattern? Write a recursive program that is able to show this.

**18.25 (String permutation) Write a recursive method to print all the permutations of a string. For example, for the string abc, the permutation is abc acb bac bca cab cba

(Hint: Define the following two methods. The second is a helper method.)

public static void displayPermutation(String s) public static void displayPermutation(String s1, String s2)

The first method simply invokes displayPermutation(" ", s). The second method uses a loop to move a character from s2 to s1 and recursively invokes it with new s1 and s2. The base case is that s2 is empty and prints s1 to the console.

Write a test program that prompts the user to enter a string and displays all its permutations.

**18.26 (Create a maze) Write a program that will find a path in a maze, as shown in Figure 18.13a. The maze is represented by a 9 * 9 board. The path must meet the following conditions:

■■ The program enables the user to place or remove a mark on a cell. A path consists of adjacent unmarked cells. Two cells are said to be adjacent if they are horizontal or vertical neighbors, but not if they are diagonal neighbors.

Figure 18.13 The program finds a path from the upper-left corner to the bottom-right corner.

(a) Correct path (b) Illegal path

■■ The path does not contain cells that form a square. The path in Figure 18.13b, for example, does not meet this condition. (The condition makes a path easy to identify on the board.)

M18_LIAN1878_11_GE_C18.indd 768 12/29/17 8:00 PM

Programming Exercises 769

**18.27 (Koch snowflake fractal) The text presented the Sierpinski triangle fractal. In this exercise, you will write a program to display another fractal, called the Koch snowflake, named after a famous Swedish mathematician. A Koch snowflake is created as follows:

1. Begin with an equilateral triangle, which is considered to be the Koch fractal of order (or level) 0, as shown in Figure 18.14a.

2. Divide each line in the shape into three equal line segments and draw an out- ward equilateral triangle with the middle line segment as the base to create a Koch fractal of order 1, as shown in Figure 18.14b.

3. Repeat Step 2 to create a Koch fractal of order 2, 3, . . . , and so on, as shown in Figures 18.14c and d.

(a) (b) (c) (d)

**18.28 (Nonrecursive directory size) Rewrite Listing 18.7, DirectorySize.java, without using recursion.

*18.29 Write a program that prompts the user for a path and displays the directory struc- ture on the command tool as follows:

VideoNote

Search a string in a directory

**18.30 (Find words) Write a program that finds all occurrences of a word in all the files under a directory, recursively. Pass the parameters from the command line as follows:

java Exercise18_30 dirName word

testdir

+--f1

+--d2

+--d22

+--f221

+--f21

+--d21

+--f212

+--f211

+--f2

+--d1

+--f12

An example of a testdir directory with two directories d1 and d2 and two files f1 and f2. The directory d1 has one file f12, and so on.

Hint: To do this, you will have to use java.io.File.

M18_LIAN1878_11_GE_C18.indd 769 12/29/17 8:00 PM

770 Chapter 18 Recursion

**18.31 (Replace words) Write a program that replaces all occurrences of a word with a new word in all the files under a directory, recursively. Pass the parameters from the command line as follows:

java Exercise18_31 dirName oldWord newWord

***18.32 (Game: Knight’s Tour) The Knight’s Tour is an ancient puzzle. The objective is to move a knight, starting from any square on a chessboard, to every other square once, as shown in Figure 18.15a. Note the knight makes only L-shaped moves (two spaces in one direction and one space in a perpendicular direction). As shown in Figure 18.15b, the knight can move to eight squares. Write a program that displays the moves for the knight, as shown in Figure 18.15c. When you click a cell, the knight is placed at the cell. This cell will be the starting point for the knight. Click the Solve button to display the path for a solution.

(Hint: A brute-force approach for this problem is to move the knight from one square to another available square arbitrarily. Using such an approach, your program will take a long time to finish. A better approach is to employ some heuristics. A knight has two, three, four, six, or eight possible moves, depending on its location. Intuitively, you should attempt to move the knight to the least accessible squares first and leave those more accessible squares open, so there will be a better chance of success at the end of the search.)

Figure 18.15 (a) A knight traverses all squares once. (b) A knight makes an L-shaped move. (c) A program displays a Knight’s Tour path. Source: Copyright © 1995–2016 Oracle and/or its affiliates. All rights reserved. Used with permission.

(a) (b) (c)

0 1 2 3 4 5 6 7 0

***18.33 (Game: Knight’s Tour animation) Write a program for the Knight’s Tour problem. Your program should let the user move a knight to any starting square and click the Solve button to animate a knight moving along the path, as shown in Figure 18.16.

M18_LIAN1878_11_GE_C18.indd 770 12/29/17 8:00 PM

Programming Exercises 771

**18.34 (Game: Eight Queens) The Eight Queens problem is to find a solution to place a queen in each row on a chessboard such that no two queens can attack each other. Write a program to solve the Eight Queens problem using recursion and display the result as shown in Figure 18.17.

**18.35 (H-tree fractal) An H-tree (introduced at the beginning of this chapter in Fig ure 18.1) is a fractal defined as follows:

1. Begin with a letter H. The three lines of H are of the same length, as shown in Figure 18.1a.

2. The letter H (in its sans-serif form, H) has four endpoints. Draw an H centered at each of the four endpoints to an H-tree of order 1, as shown in Figure 18.1b. These Hs are half the size of the H that contains the four endpoints.

3. Repeat Step 2 to create an H-tree of order 2, 3, . . . , and so on, as shown in Figures 18.1c and d.

Write a program that draws an H-tree, as shown in Figure 18.1.

18.36 (Sierpinski triangle) Write a program that lets the user to enter the order and display the filled Sierpinski triangles as shown in Figure 18.18.

**18.37 (Hilbert curve) The Hilbert curve, first described by German mathematician David Hilbert in 1891, is a space-filling curve that visits every point in a square grid with a size of 2 * 2, 4 * 4, 8 * 8, 16 * 16, or any other power of 2. Write a program that displays a Hilbert curve for the specified order, as shown in Figure 18.19.

M18_LIAN1878_11_GE_C18.indd 771 12/29/17 8:00 PM

772 Chapter 18 Recursion

**18.38 (Recursive tree) Write a program to display a recursive tree as shown in Figure 18.20.

VideoNote

Recursive tree

(a) (b) (c) (d)

**18.39 (Drag the tree) Revise Programming Exercise 18.38 to move the tree to where the mouse is dragged.

(a) (b) (c) (d)

M18_LIAN1878_11_GE_C18.indd 772 12/29/17 8:00 PM

Objectives ■■ To describe the benefits of generics (§19.2).

■■ To use generic classes and interfaces (§19.2).

■■ To define generic classes and interfaces (§19.3).

■■ To explain why generic types can improve reliability and readability (§19.3).

■■ To define and use generic methods and bounded generic types (§19.4).

■■ To develop a generic sort method to sort an array of Comparable objects (§19.5).

■■ To use raw types for backward compatibility (§19.6).

■■ To explain why wildcard generic types are necessary (§19.7).

■■ To describe generic-type erasure and list certain restrictions and l imitations on generic types caused by type erasure (§19.8).

■■ To design and implement generic matrix classes (§19.9).

Generics

Chapter

M19_LIAN1878_11_GE_C19.indd 773 12/29/17 4:31 PM

774 Chapter 19 Generics

Point Key

19.1 Introduction Generics enable you to detect errors at compile time rather than at runtime.

You have used a generic class ArrayList in Chapter 11, and generic interface Comparable in Chapter 13. Generics let you parameterize types. With this capability, you can define a class or a method with generic types that the compiler can replace with concrete types. For example, Java defines a generic ArrayList class for storing the elements of a generic type. From this generic class, you can create an ArrayList object for holding strings, and an ArrayList object for holding numbers. Here, strings and numbers are concrete types that replace the generic type.

The key benefit of generics is to enable errors to be detected at compile time rather than at runtime. A generic class or method permits you to specify allowable types of objects that the class or method can work with. If you attempt to use an incompatible object, the compiler will detect that error.

This chapter explains how to define and use generic classes, interfaces, and methods and demonstrates how generics can be used to improve software reliability and readability. It can be intertwined with Chapter 13, Abstract Classes and Interfaces.

19.2 Motivations and Benefits The motivation for using Java generics is to detect errors at compile time.

Java has allowed you to define generic classes, interfaces, and methods since JDK 1.5. Several interfaces and classes in the Java API were modified using generics. For example, prior to JDK 1.5, the java.lang.Comparable interface was defined as shown in Figure 19.1a, but since JDK 1.5, it has been modified as shown in Figure 19.1b.

what is generics?

why generics?

Point Key

Figure 19.1 The java.lang.Comparable interface was modified in JDK 1.5 with a generic type.

package java.lang;

public interface Comparable { public int compareTo(Object o) }

(a) Prior to JDK 1.5

package java.lang;

public interface Comparable<T> { public int compareTo(T o) }

(b) JDK 1.5

Here, <T> represents a formal generic type, which can be replaced later with an actual concrete type. Replacing a generic type is called a generic instantiation. By convention, a single capital letter such as E or T is used to denote a formal generic type.

To see the benefits of using generics, let us examine the code in Figure 19.2. The statement in Figure 19.2a declares that c is a reference variable whose type is Comparable and invokes the compareTo method to compare a Date object with a string. The code compiles fine, but it has a runtime error because a string cannot be compared with a date.

formal generic type actual concrete type generic instantiation

Figure 19.2 The new generic type detects possible errors at compile time.

Comparable c = new Date(); System.out.println(c.compareTo("red"));

(a) Prior to JDK 1.5

Comparable<Date> c = new Date(); System.out.println(c.compareTo("red"));

(b) JDK 1.5

The statement in Figure 19.2b declares that c is a reference variable whose type is Comparable<Date> and invokes the compareTo method to compare a Date object with a string. This code has a compile error because the argument passed to the compareTo method

M19_LIAN1878_11_GE_C19.indd 774 12/29/17 4:31 PM

19.2 Motivations and Benefits 775

must be of the Date type. Since the errors can be detected at compile time rather than at runt- ime, the generic type makes the program more reliable.

ArrayList was introduced in Section 11.11, The ArrayList Class. This class has been a generic class since JDK 1.5. Figure 19.3 shows the class diagram for ArrayList before and since JDK 1.5, respectively.

reliable

Figure 19.3 ArrayList is a generic class since JDK 1.5.

(a) ArrayList before JDK 1.5

+ArrayList() +add(o: Object): void +add(index: int, o: Object): void +clear(): void +contains(o: Object): boolean

+get(index:int): Object +indexOf(o: Object): int +isEmpty(): boolean +lastIndexOf(o: Object): int +remove(o: Object): boolean +size(): int +remove(index: int): boolean +set(index: int, o: Object): Object

java.util.ArrayList

(b) ArrayList since JDK 1.5

+ArrayList()

+clear(): void +contains(o: Object): boolean

+indexOf(o: Object): int +isEmpty(): boolean +lastIndexOf(o: Object): int +remove(o: Object): boolean +size(): int +remove(index: int): boolean

java.util.ArrayList<E>

+add(o: E): void +add(index: int, o: E): void

+get(index: int): E

+set(index: int, o: E): E

For example, the following statement creates a list for strings:

ArrayList<String> list = new ArrayList<>();

You can now add only strings into the list. For instance,

list.add("Red");

If you attempt to add a nonstring, a compile error will occur. For example, the following state- ment is now illegal because list can contain only strings.

list.add(new Integer(1));

Generic types must be reference types. You cannot replace a generic type with a primitive type such as int, double, or char. For example, the following statement is wrong:

ArrayList<int> intList = new ArrayList<>();

To create an ArrayList object for int values, you have to use

ArrayList<Integer> intList = new ArrayList<>();

You can add an int value to intList. For example,

intList.add(5);

Java automatically wraps 5 into new Integer(5). This is called autoboxing, as introduced in Section 10.8, Automatic Conversion between Primitive Types and Wrapper Class Types.

only strings allowed

generic reference type

autoboxing

M19_LIAN1878_11_GE_C19.indd 775 12/29/17 4:31 PM

776 Chapter 19 Generics

Casting is not needed to retrieve a value from a list with a specified element type because the compiler already knows the element type. For example, the following statements create a list that contains strings, add strings to the list, and retrieve strings from the list.

1 ArrayList<String> list = new ArrayList<>(); 2 list.add("Red"); 3 list.add("White"); 4 String s = list.get(0); // No casting is needed

Prior to JDK 1.5, without using generics, you would have had to cast the return value to String as:

String s = (String)(list.get(0)); // Casting needed prior to JDK 1.5

If the elements are of wrapper types, such as Integer, Double, and Character, you can directly assign an element to a primitive-type variable. This is called autounboxing, as intro- duced in Section 10.8. For example, see the following code:

1 ArrayList<Double> list = new ArrayList<>(); 2 list.add(5.5); // 5.5 is automatically converted to new Double(5.5) 3 list.add(3.0); // 3.0 is automatically converted to new Double(3.0) 4 Double doubleObject = list.get(0); // No casting is needed 5 double d = list.get(1); // Automatically converted to double

In lines 2 and 3, 5.5 and 3.0 are automatically converted into Double objects and added to list. In line 4, the first element in list is assigned to a Double variable. No casting is necessary because list is declared for Double objects. In line 5, the second element in list is assigned to a double variable. The object in list.get(1) is automatically converted into a primitive-type value.

19.2.1 Are there any compile errors in (a) and (b)?

no casting needed

autounboxing

Point Check

ArrayList dates = new ArrayList(); dates.add(new Date()); dates.add(new String());

(a) Prior to JDK 1.5

ArrayList<Date> dates = new ArrayList<>(); dates.add(new Date()); dates.add(new String());

(b) Since JDK 1.5

19.2.2 What is wrong in (a)? Is the code in (b) correct?

ArrayList dates = new ArrayList(); dates.add(new Date()); Date date = dates.get(0);

(a) Prior to JDK 1.5

ArrayList<Date> dates = new ArrayList<>(); dates.add(new Date()); Date date = dates.get(0);

(b) Since JDK 1.5

19.2.3 What are the benefits of using generic types?

19.3 Defining Generic Classes and Interfaces A generic type can be defined for a class or interface. A concrete type must be specified when using the class to create an object or using the class or interface to declare a reference variable.

Point Key

M19_LIAN1878_11_GE_C19.indd 776 12/29/17 4:31 PM

19.3 Defining Generic Classes and Interfaces 777

Let us revise the stack class in Section 11.13, Case Study: A Custom Stack Class, to generalize the element type with a generic type. The new stack class, named GenericStack, is shown in Figure 19.4 and is implemented in Listing 19.1.

Figure 19.4 The GenericStack class encapsulates the stack storage and provides the operations for manipulating the stack.

Creates an empty stack.

Returns the number of elements in this stack.

Returns the top element in this stack.

Returns and removes the top element in this stack.

Adds a new element to the top of this stack.

Returns true if the stack is empty.

An array list to store elements.

GenericStack<E>

–list: java.util.ArrayList<E>

+GenericStack()

+getSize(): int

+peek(): E

+pop(): E

+push(o: E): void

+isEmpty(): boolean

Listing 19.1 GenericStack.java 1 public class GenericStack<E> { 2 private java.util.ArrayList<E> list = new java.util.ArrayList<>(); 3 4 public int getSize() { 5 return list.size(); 6 } 7 8 public E peek() { 9 return list.get(getSize() − 1); 10 } 11 12 public void push(E o) { 13 list.add(o); 14 } 15 16 public E pop() { 17 E o = list.get(getSize() − 1); 18 list.remove(getSize() − 1); 19 return o; 20 } 21 22 public boolean isEmpty() { 23 return list.isEmpty(); 24 } 25 26 @Override 27 public String toString() { 28 return "stack: " + list.toString(); 29 } 30 }

The following example creates a stack to hold strings and adds three strings to the stack:

GenericStack<String> stack1 = new GenericStack<>(); stack1.push("London"); stack1.push("Paris"); stack1.push("Berlin");

push

peek

pop

generic array list

getSize

isEmpty

generic type E declared

M19_LIAN1878_11_GE_C19.indd 777 12/29/17 4:31 PM

778 Chapter 19 Generics

This example creates a stack to hold integers and adds three integers to the stack:

GenericStack<Integer> stack2 = new GenericStack<>(); stack2.push(1); // autoboxing 1 to new Integer(1) stack2.push(2); stack2.push(3);

Instead of using a generic type, you could simply make the type element Object, which can accommodate any object type. However, using a specific concrete type can improve software reliability and readability because certain errors can be detected at compile time rather than at runtime. For example, because stack1 is declared GenericStack<String>, only strings can be added to the stack. It would be a compile error if you attempted to add an integer to stack1.

Caution To create a stack of strings, you use new GenericStack<String>() or new GenericStack<>(). This could mislead you into thinking that the constructor of GenericStack should be defined as

public GenericStack<E>()

This is wrong. It should be defined as

public GenericStack()

Note Occasionally, a generic class may have more than one parameter. In this case, place the parameters together inside the brackets, separated by commas—for example, <E1, E2, E3>.

Note You can define a class or an interface as a subtype of a generic class or interface. For example, the java.lang.String class is defined to implement the Comparable interface in the Java API as follows:

public class String implements Comparable<String>

19.3.1 What is the generic definition for java.lang.Comparable in the Java API? 19.3.2 Since you create an instance of ArrayList of strings using new ArrayList

<String>(), should the constructor in the ArrayList class be defined as

public ArrayList<E>()

19.3.3 Can a generic class have multiple generic parameters? 19.3.4 How do you declare a generic type in a class?

19.4 Generic Methods A generic type can be defined for a static method.

You can define generic interfaces (e.g., the Comparable interface in Figure 19.1b) and classes (e.g., the GenericStack class in Listing 19.1). You can also use generic types to define generic methods. For example, Listing 19.2 defines a generic method print (lines 10–14) to print an array of objects. Line 6 passes an array of integer objects to invoke the generic print method. Line 7 invokes print with an array of strings.

benefits of using generic types

generic class constructor

multiple generic parameters

inheritance with generics

Point Check

Point Key

generic method

M19_LIAN1878_11_GE_C19.indd 778 12/29/17 4:31 PM

19.4 Generic Methods 779

Listing 19.2 GenericMethodDemo.java 1 public class GenericMethodDemo { 2 public static void main(String[] args ) { 3 Integer[] integers = {1, 2, 3, 4, 5}; 4 String[] strings = {"London", "Paris", "New York", "Austin"}; 5 6 GenericMethodDemo.<Integer>print(integers); 7 GenericMethodDemo.<String>print(strings); 8 } 9 10 public static <E> void print(E[] list) { 11 for (int i = 0; i < list.length; i++) 12 System.out.print(list[i] + " "); 13 System.out.println(); 14 } 15 }

To declare a generic method, you place the generic type <E> immediately after the keyword static in the method header. For example,

public static <E> void print(E[] list)

To invoke a generic method, prefix the method name with the actual type in angle brackets. For example,

GenericMethodDemo.<Integer>print(integers); GenericMethodDemo.<String>print(strings);

or simply invoke it as follows:

print(integers); print(strings);

In the latter case, the actual type is not explicitly specified. The compiler automatically discov- ers the actual type.

A generic type can be specified as a subtype of another type. Such a generic type is called bounded. For example, Listing 19.3 revises the equalArea method in Listing 13.4, TestGeometricObject.java, to test whether two geometric objects have the same area. The bounded generic type <E extends GeometricObject> (line 10) specifies that E is a generic subtype of GeometricObject. You must invoke equalArea by passing two instances of GeometricObject.

Listing 19.3 BoundedTypeDemo.java 1 public class BoundedTypeDemo { 2 public static void main(String[] args ) { 3 Rectangle rectangle = new Rectangle(2, 2); 4 Circle circle = new Circle(2); 5 6 System.out.println("Same area? " + 7 equalArea(rectangle, circle)); 8 } 9 10 public static <E extends GeometricObject> boolean equalArea( 11 E object1, E object2) { 12 return object1.getArea() == object2.getArea(); 13 } 14 }

generic method

declare a generic method

invoke generic method

bounded generic type

Rectangle in Listing 13.3 Circle in Listing 13.2

bounded generic type

M19_LIAN1878_11_GE_C19.indd 779 12/29/17 4:31 PM

780 Chapter 19 Generics

Note An unbounded generic type <E> is the same as <E extends Object>.

Note To define a generic type for a class, place it after the class name, such as GenericStack<E>. To define a generic type for a method, place the generic type before the method return type, such as <E> void max(E o1, E o2).

19.4.1 How do you declare a generic method? How do you invoke a generic method? 19.4.2 What is a bounded generic type?

19.5 Case Study: Sorting an Array of Objects You can develop a generic method for sorting an array of Comparable objects.

This section presents a generic method for sorting an array of Comparable objects. The objects are instances of the Comparable interface and they are compared using the compareTo method. To test the method, the program sorts an array of integers, an array of double numbers, an array of characters, and an array of strings. The program is shown in Listing 19.4.

Listing 19.4 GenericSort.java 1 public class GenericSort { 2 public static void main(String[] args) { 3 // Create an Integer array 4 Integer[] intArray = {new Integer(2), new Integer(4), 5 new Integer(3)}; 6 7 // Create a Double array 8 Double[] doubleArray = {new Double(3.4), new Double(1.3), 9 new Double(−22.1)}; 10 11 // Create a Character array 12 Character[] charArray = {new Character('a'), 13 new Character('J'), new Character('r')}; 14 15 // Create a String array 16 String[] stringArray = {"Tom", "Susan", "Kim"}; 17 18 // Sort the arrays 19 sort(intArray); 20 sort(doubleArray); 21 sort(charArray); 22 sort(stringArray); 23 24 // Display the sorted arrays 25 System.out.print("Sorted Integer objects: "); 26 printList(intArray); 27 System.out.print("Sorted Double objects: "); 28 printList(doubleArray); 29 System.out.print("Sorted Character objects: "); 30 printList(charArray); 31 System.out.print("Sorted String objects: "); 32 printList(stringArray); 33 } 34

generic class parameter vs. generic method parameter

Point Check

Point Key

sort Integer objects

sort Character objects sort Double objects

sort String objects

M19_LIAN1878_11_GE_C19.indd 780 12/29/17 4:31 PM

19.5 Case Study: Sorting an Array of Objects 781

35 /** Sort an array of comparable objects */ 36 public static <E extends Comparable<E>> void sort(E[] list) { 37 E currentMin; 38 int currentMinIndex; 39 40 for (int i = 0; i < list.length − 1; i++) { 41 // Find the minimum in the list[i+1..list.length−2] 42 currentMin = list[i]; 43 currentMinIndex = i; 44 45 for (int j = i + 1; j < list.length; j++) { 46 if (currentMin.compareTo(list[j]) > 0) { 47 currentMin = list[j]; 48 currentMinIndex = j; 49 } 50 } 51 52 // Swap list[i] with list[currentMinIndex] if necessary; 53 if (currentMinIndex != i) { 54 list[currentMinIndex] = list[i]; 55 list[i] = currentMin; 56 } 57 } 58 } 59 60 /** Print an array of objects */ 61 public static void printList(Object[] list) { 62 for (int i = 0; i < list.length; i++) 63 System.out.print(list[i] + " "); 64 System.out.println(); 65 } 66 }

compareTo

generic sort method

Sorted Integer objects: 2 3 4 Sorted Double objects: −22.1 1.3 3.4 Sorted Character objects: J a r Sorted String objects: Kim Susan Tom

The algorithm for the sort method is the same as in Listing 7.8, SelectionSort.java. The sort method in that program sorts an array of double values. The sort method in this example can sort an array of any object type, provided that the objects are also instances of the Comparable interface. The generic type is defined as <E extends Comparable<E>> (line 36). This has two meanings. First, it specifies that E is a subtype of Comparable. Second, it specifies that the elements to be compared are of the E type as well.

The sort method uses the compareTo method to determine the order of the objects in the array (line 46). Integer, Double, Character, and String implement Comparable, so the objects of these classes can be compared using the compareTo method. The program creates arrays of Integer objects, Double objects, Character objects, and String objects (lines 4–16) and invokes the sort method to sort these arrays (lines 19–22).

19.5.1 Given int[] list = {1, 2, −1}, can you invoke sort(list) using the sort method in Listing 19.4?

19.5.2 Given int[] list = {new Integer(1), new Integer(2), new Integer (−1)}, can you invoke sort(list) using the sort method in Listing 19.4?

Point Check

M19_LIAN1878_11_GE_C19.indd 781 12/29/17 4:31 PM

782 Chapter 19 Generics

19.6 Raw Types and Backward Compatibility A generic class or interface used without specifying a concrete type, called a raw type, enables backward compatibility with earlier versions of Java.

You can use a generic class without specifying a concrete type such as the following:

GenericStack stack = new GenericStack(); // raw type

This is roughly equivalent to

GenericStack<Object> stack = new GenericStack<Object>();

A generic class such as GenericStack and ArrayList used without a type parameter is called a raw type. Using raw types allows for backward compatibility with earlier versions of Java. For example, a generic type has been used in java.lang.Comparable since JDK 1.5, but a lot of code still uses the raw type Comparable, as given in Listing 19.5:

Listing 19.5 Max.java 1 public class Max { 2 /** Return the maximum of two objects */ 3 public static Comparable max(Comparable o1, Comparable o2) { 4 if (o1.compareTo(o2) > 0) 5 return o1; 6 else 7 return o2; 8 } 9 }

Comparable o1 and Comparable o2 are raw type declarations. Be careful: raw types are unsafe. For example, you might invoke the max method using

Max.max("Welcome", 23); // 23 is autoboxed into new Integer(23)

This would cause a runtime error because you cannot compare a string with an integer object. The Java compiler displays a warning on line 3 when compiled with the option –Xlint:unchecked, as shown in Figure 19.5.

Point Key

raw type backward compatibility

–Xlint:unchecked

A better way to write the max method is to use a generic type, as given in Listing 19.6.

Listing 19.6 MaxUsingGenericType.java 1 public class MaxUsingGenericType { 2 /** Return the maximum of two objects */ 3 public static <E extends Comparable<E>> E max(E o1, E o2) { 4 if (o1.compareTo(o2) > 0) 5 return o1;

bounded type

raw type

M19_LIAN1878_11_GE_C19.indd 782 12/29/17 4:31 PM

19.7 Wildcard Generic Types 783

6 else 7 return o2; 8 } 9 }

If you invoke the max method using

// 23 is autoboxed into new Integer(23) MaxUsingGenericType.max("Welcome", 23);

a compile error will be displayed because the two arguments of the max method in MaxUsingGenericType must have the same type (e.g., two strings or two integer objects). Furthermore, the type E must be a subtype of Comparable<E>.

As another example, in the following code you can declare a raw type stack in line 1, assign new GenericStack<String> to it in line 2, and push a string and an integer object to the stack in lines 3 and 4:

1 GenericStack stack; 2 stack = new GenericStack<String>(); 3 stack.push("Welcome to Java"); 4 stack.push(new Integer(2));

However, line 4 is unsafe because the stack is intended to store strings, but an Integer object is added into the stack. Line 3 should be okay, but the compiler will show warnings for both line 3 and line 4, because it cannot follow the semantic meaning of the program. All the com- piler knows is that stack is a raw type, and performing certain operations is unsafe. Therefore, warnings are displayed to alert potential problems.

Tip Since raw types are unsafe, this book will not use them from here on.

19.6.1 What is a raw type? Why is a raw type unsafe? Why is the raw type allowed in Java? 19.6.2 What is the syntax to declare an ArrayList reference variable using the raw type

and assign a raw type ArrayList object to it?

19.7 Wildcard Generic Types You can use unbounded wildcards, bounded wildcards, or lower bound wildcards to specify a range for a generic type.

What are wildcard generic types, and why are they needed? Listing 19.7 gives an example to demonstrate the needs. The example defines a generic max method for finding the maximum in a stack of numbers (lines 12–22). The main method creates a stack of integer objects, adds three integers to the stack, and invokes the max method to find the maximum number in the stack.

Listing 19.7 WildCardNeedDemo.java 1 public class WildCardNeedDemo { 2 public static void main(String[] args ) { 3 GenericStack<Integer> intStack = new GenericStack<>(); 4 intStack.push(1); // 1 is autoboxed into new Integer(1) 5 intStack.push(2); 6 intStack.push(−2); 7 8 System.out.print("The max number is " + max(intStack)); 9 } 10

Point Check

Point Key

GenericStack<Integer> type

M19_LIAN1878_11_GE_C19.indd 783 12/29/17 4:31 PM

784 Chapter 19 Generics

11 /** Find the maximum in a stack of numbers */ 12 public static double max(GenericStack<Number> stack) { 13 double max = stack.pop().doubleValue(); // Initialize max 14 15 while (!stack.isEmpty()) { 16 double value = stack.pop().doubleValue(); 17 if (value > max) 18 max = value; 19 } 20 21 return max; 22 } 23 }

The program in Listing 19.7 has a compile error in line 8 because intStack is not an instance of GenericStack<Number>. Thus, you cannot invoke max(intStack).

The fact is Integer is a subtype of Number, but GenericStack<Integer> is not a sub- type of GenericStack<Number>. To circumvent this problem, use wildcard generic types. A wildcard generic type has three forms: ?, ? extends T, and ? super T, where T is a generic type.

The first form, ?, called an unbounded wildcard, is the same as ? extends Object. The second form, ? extends T, called a bounded wildcard, represents T or a subtype of T. The third form, ? super T, called a lower bound wildcard, denotes T or a supertype of T.

You can fix the error by replacing line 12 in Listing 19.7 as follows:

public static double max(GenericStack<? extends Number> stack) {

<? extends Number> is a wildcard type that represents Number or a subtype of Number, so it is legal to invoke max(new GenericStack<Integer>()) or max(new GenericStack<Double>()).

Listing 19.8 shows an example of using the ? wildcard in the print method that prints objects in a stack and empties the stack. <?> is a wildcard that represents any object type. It is equivalent to <? extends Object>. What happens if you replace GenericStack<?> with GenericStack<Object>? It would be wrong to invoke print(intStack) because intStack is not an instance of GenericStack<Object>. Note that GenericStack<Integer> is not a subtype of GenericStack<Object> even though Integer is a subtype of Object.

Listing 19.8 AnyWildCardDemo.java 1 public class AnyWildCardDemo { 2 public static void main(String[] args) { 3 GenericStack<Integer> intStack = new GenericStack<>(); 4 intStack.push(1); // 1 is autoboxed into new Integer(1) 5 intStack.push(2); 6 intStack.push(−2); 7 8 print(intStack); 9 } 10 11 /** Prints objects and empties the stack */ 12 public static void print(GenericStack<?> stack) { 13 while (!stack.isEmpty()) { 14 System.out.print(stack.pop() + " "); 15 } 16 } 17 }

GenericStack<Number> type

unbounded wildcard bounded wildcard lower bound wildcard

wildcard type

GenericStack<Integer> type

M19_LIAN1878_11_GE_C19.indd 784 12/29/17 4:31 PM

19.7 Wildcard Generic Types 785

When is the wildcard <? super T> needed? Consider the example in Listing 19.9. The example creates a stack of strings in stack1 (line 3) and a stack of objects in stack2 (line 4) and invokes add(stack1, stack2) (line 8) to add the strings in stack1 into stack2. GenericStack<? super T> is used to declare stack2 in line 13. If <? super T> is replaced by <T>, a compile error will occur on add(stack1, stack2) in line 8 because stack1’s type is GenericStack<String> and stack2’s type is GenericStack<Object>. <? super T> represents type T or a supertype of T. Object is a supertype of String.

Listing 19.9 SuperWildCardDemo.java 1 public class SuperWildCardDemo { 2 public static void main(String[] args) { 3 GenericStack<String> stack1 = new GenericStack<>(); 4 GenericStack<Object> stack2 = new GenericStack<>(); 5 stack2.push("Java"); 6 stack2.push(2); 7 stack1.push("Sun"); 8 add(stack1, stack2); 9 AnyWildCardDemo.print(stack2); 10 } 11 12 public static <T> void add(GenericStack<T> stack1, 13 GenericStack<? super T> stack2) { 14 while (!stack1.isEmpty()) 15 stack2.push(stack1.pop()); 16 } 17 }

This program will also work if the method header in lines 12 and 13 is modified as follows:

public static <T> void add(GenericStack<? extends T> stack1, GenericStack<T> stack2)

The inheritance relationship involving generic types and wildcard types is summarized in Figure 19.6. In this figure, A and B represent classes or interfaces, and E is a generic-type parameter.

why <? Super T>

GenericStack<String> type

<? Super T> type

Figure 19.6 The relationship between generic types and wildcard types.

? super E

Object

E’s subclass ? extends E

A<? extends B> A<? super B>

A<?>

Object

A<B>A<B’s subclass> A<B’s superclass>

? E’s superclass

19.7.1 Is GenericStack the same as GenericStack<Object>? 19.7.2 What is an unbounded wildcard, a bounded wildcard, and a lower bound wildcard? 19.7.3 What happens if lines 12 and 13 in Listing 19.9 are changed to

public static <T> void add(GenericStack<T> stack1, GenericStack<T> stack2)

Point Check

M19_LIAN1878_11_GE_C19.indd 785 12/29/17 4:31 PM

786 Chapter 19 Generics

19.7.4 What happens if lines 12 and 13 in Listing 19.9 are changed to

public static <T> void add(GenericStack<? extends T> stack1, GenericStack<T> stack2)

19.8 Erasure and Restrictions on Generics The information on generics is used by the compiler but is not available at runtime. This is called type erasure.

Generics are implemented using an approach called type erasure: The compiler uses the generic-type information to compile the code, but erases it afterward. Thus, the generic infor- mation is not available at runtime. This approach enables the generic code to be backward compatible with the legacy code that uses raw types.

The generics are present at compile time. Once the compiler confirms that a generic type is used safely, it converts the generic type to a raw type. For example, the compiler checks whether the following code in (a) uses generics correctly, then translates it into the equivalent code in (b) for runtime use. The code in (b) uses the raw type.

Point Key

type erasure

erase generics

replace generic type

ArrayList<String> list = new ArrayList<>(); list.add("Oklahoma"); String state = list.get(0);

(a)

ArrayList list = new ArrayList(); list.add("Oklahoma"); String state = (String)(list.get(0));

(b)

When generic classes, interfaces, and methods are compiled, the compiler replaces the generic type with the Object type. For example, the compiler would convert the following method in (a) into (b).

public static <E> void print(E[] list) { for (int i = 0; i < list.length; i++) System.out.print(list[i] + " "); System.out.println(); }

public static void print(Object[] list) { for (int i = 0; i < list.length; i++) System.out.print(list[i] + " "); System.out.println(); }

(a) (b)

If a generic type is bounded, the compiler replaces it with the bounded type. For example, the compiler would convert the following method in (a) into (b).

replace bounded type

public static <E extends GeometricObject> boolean equalArea( E object1, E object2) { return object1.getArea() == object2.getArea(); }

public static boolean equalArea( GeometricObject object1, GeometricObject object2) { return object1.getArea() == object2.getArea(); }

(a) (b)

It is important to note a generic class is shared by all its instances regardless of its actual con- crete type. Suppose list1 and list2 are created as follows:

ArrayList<String> list1 = new ArrayList<>(); ArrayList<Integer> list2 = new ArrayList<>();

important fact

M19_LIAN1878_11_GE_C19.indd 786 12/29/17 4:31 PM

19.8 Erasure and Restrictions on Generics 787

Although ArrayList<String> and ArrayList<Integer> are two types at compile time, only one ArrayList class is loaded into the JVM at runtime. list1 and list2 are both instances of ArrayList, so the following statements display true:

System.out.println(list1 instanceof ArrayList); System.out.println(list2 instanceof ArrayList);

However, the expression list1 instanceof ArrayList<String> is wrong. Since ArrayList<String> is not stored as a separate class in the JVM, using it at runtime makes no sense.

Because generic types are erased at runtime, there are certain restrictions on how generic types can be used. Here are some of the restrictions:

Restriction 1: Cannot Use new E()

You cannot create an instance using a generic-type parameter. For example, the following statement is wrong:

E object = new E();

The reason is new E() is executed at runtime, but the generic type E is not available at runtime.

Restriction 2: Cannot Use new E[]

You cannot create an array using a generic type parameter. For example, the following statement is wrong:

E[] elements = new E[capacity];

You can circumvent this limitation by creating an array of the Object type then casting it to E[], as follows:

E[] elements = (E[])new Object[capacity];

However, casting to (E[]) causes an unchecked compile warning. The warning occurs because the compiler is not certain that casting will succeed at runtime. For example, if E is String and new Object[] is an array of Integer objects, (String[])(new Object[]) will cause a ClassCastException. This type of compile warning is a limita- tion of Java generics and is unavoidable.

Generic array creation using a generic class is not allowed, either. For example, the fol- lowing code is wrong:

ArrayList<String>[] list = new ArrayList<String>[10];

You can use the following code to circumvent this restriction:

Ar rayList<String>[] list = (ArrayList<String>[])new ArrayList[10];

However, you will still get a compile warning.

Restriction 3: A Generic Type Parameter of a Class Is Not Allowed in a Static Context

Since all instances of a generic class have the same runtime class, the static variables and methods of a generic class are shared by all its instances. Therefore, it is illegal to refer to a

no new E()

no new E[capacity]

unavoidable compile warning

M19_LIAN1878_11_GE_C19.indd 787 12/29/17 4:31 PM

788 Chapter 19 Generics

generic-type parameter for a class in a static method, field, or initializer. For example, the following code is illegal:

public class Test<E> { public static void m(E o1) { // Illegal }

public static E o1; // Illegal

static { E o2; // Illegal } }

Restriction 4: Exception Classes Cannot Be Generic

A generic class may not extend java.lang.Throwable, so the following class declara- tion would be illegal:

public class MyException<T> extends Exception { }

Why? If it were allowed, you would have a catch clause for MyException<T> as follows:

try { ... } catch (MyException<T> ex) { ... }

The JVM has to check the exception thrown from the try clause to see if it matches the type specified in a catch clause. This is impossible, because the type information is not present at runtime.

19.8.1 What is erasure? Why are Java generics implemented using erasure? 19.8.2 If your program uses ArrayList<String> and ArrayList<Date>, does the

JVM load both of them?

19.8.3 Can you create an instance using new E() for a generic type E? Why? 19.8.4 Can a method that uses a generic class parameter be static? Why? 19.8.5 Can you define a custom generic exception class? Why?

19.9 Case Study: Generic Matrix Class This section presents a case study on designing classes for matrix operations using generic types.

The addition and multiplication operations for all matrices are similar except that their element types differ. Therefore, you can design a superclass that describes the common operations shared by matrices of all types regardless of their element types, and you can define subclasses tailored to specific types of matrices. This case study gives implementations for two types: int and Rational. For the int type, the wrapper class Integer should be used to wrap an int value into an object, so the object is passed in the methods for operations.

The class diagram is shown in Figure 19.7. The methods addMatrix and multiplyMatrix add and multiply two matrices of a generic type E[][]. The static method printResult displays the matrices, the operator, and their result. The methods add, multiply, and zero are abstract because their implementations depend on the specific type of the array elements. For example, the

Point Check

Point Key

M19_LIAN1878_11_GE_C19.indd 788 12/29/17 4:31 PM

19.9 Case Study: Generic Matrix Class 789

zero() method returns 0 for the Integer type and 0/1 for the Rational type. These methods will be implemented in the subclasses in which the matrix element type is specified.

IntegerMatrix and RationalMatrix are concrete subclasses of GenericMatrix. These two classes implement the add, multiply, and zero methods defined in the GenericMatrix class.

Listing 19.10 implements the GenericMatrix class. <E extends Number> in line 1 specifies the generic type is a subtype of Number. Three abstract methods—add, multiply, and zero—are defined in lines 3, 6, and 9. These methods are abstract because we cannot imple- ment them without knowing the exact type of the elements. The addMaxtrix (lines 12–30) and multiplyMatrix (lines 33–57) methods implement the methods for adding and multiplying two matrices. All these methods must be nonstatic because they use generic-type E for the class. The printResult method (lines 60–84) is static because it is not tied to specific instances.

The matrix element type is a generic subtype of Number. This enables you to use an object of any subclass of Number as long as you can implement the abstract add, multiply, and zero methods in subclasses.

The addMatrix and multiplyMatrix methods (lines 12–57) are concrete methods. They are ready to use as long as the add, multiply, and zero methods are implemented in the subclasses.

The addMatrix and multiplyMatrix methods check the bounds of the matrices before performing operations. If the two matrices have incompatible bounds, the program throws an exception (lines 16 and 36).

Listing 19.10 GenericMatrix.java 1 public abstract class GenericMatrix<E extends Number> { 2 /** Abstract method for adding two elements of the matrices */ 3 protected abstract E add(E o1, E o2); 4 5 /** Abstract method for multiplying two elements of the matrices */ 6 protected abstract E multiply(E o1, E o2); 7 8 /** Abstract method for defining zero for the matrix element */ 9 protected abstract E zero(); 10 11 /** Add two matrices */ 12 public E[][] addMatrix(E[][] matrix1, E[][] matrix2) { 13 // Check bounds of the two matrices 14 if ((matrix1.length != matrix2.length) || 15 (matrix1[0].length != matrix2[0].length)) { 16 throw new RuntimeException( 17 "The matrices do not have the same size"); 18 } 19

abstract method

bounded generic type

add two matrices

Figure 19.7 The GenericMatrix class is an abstract superclass for IntegerMatrix and RationalMatrix.

#add(element1: E, element2: E): E #multiply(element1: E, element2: E): E #zero(): E +addMatrix(matrix1: E[][], matrix2: E[][]): E[][] +multiplyMatrix(matrix1: E[][], matrix2: E[][]): E[][] +printResult(m1: Number[][], m2: Number[][],

GenericMatrix<E extends Number> IntegerMatrix

RationalMatrixm3: Number[][], op: char): void

M19_LIAN1878_11_GE_C19.indd 789 12/29/17 4:31 PM

790 Chapter 19 Generics

20 E[][] result = 21 (E[][])new Number[matrix1.length][matrix1[0].length]; 22 23 // Perform addition 24 for (int i = 0; i < result.length; i++) 25 for (int j = 0; j < result[i].length; j++) { 26 result[i][j] = add(matrix1[i][j], matrix2[i][j]); 27 } 28 29 return result; 30 } 31 32 /** Multiply two matrices */ 33 public E[][] multiplyMatrix(E[][] matrix1, E[][] matrix2) { 34 // Check bounds 35 if (matrix1[0].length != matrix2.length) { 36 throw new RuntimeException( 37 "The matrices do not have compatible size"); 38 } 39 40 // Create result matrix 41 E[][] result = 42 (E[][])new Number[matrix1.length][matrix2[0].length]; 43 44 // Perform multiplication of two matrices 45 for (int i = 0; i < result.length; i++) { 46 for (int j = 0; j < result[0].length; j++) { 47 result[i][j] = zero(); 48 49 for (int k = 0; k < matrix1[0].length; k++) { 50 result[i][j] = add(result[i][j], 51 multiply(matrix1[i][k], matrix2[k][j])); 52 } 53 } 54 } 55 56 return result; 57 } 58 59 /** Print matrices, the operator, and their operation result */ 60 public static void printResult( 61 Number[][] m1, Number[][] m2, Number[][] m3, char op) { 62 for (int i = 0; i < m1.length; i++) { 63 for (int j = 0; j < m1[0].length; j++) 64 System.out.print(" " + m1[i][j]); 65 66 if (i == m1.length / 2) 67 System.out.print(" " + op + " "); 68 else 69 System.out.print(" "); 70 71 for (int j = 0; j < m2.length; j++) 72 System.out.print(" " + m2[i][j]); 73 74 if (i == m1.length / 2) 75 System.out.print(" = "); 76 else 77 System.out.print(" "); 78 79 for (int j = 0; j < m3.length; j++)

multiply two matrices

display result

M19_LIAN1878_11_GE_C19.indd 790 12/29/17 4:31 PM

19.9 Case Study: Generic Matrix Class 791

80 System.out.print(m3[i][j] + " "); 81 82 System.out.println(); 83 } 84 } 85 }

Listing 19.11 implements the IntegerMatrix class. The class extends GenericMatrix<Integer> in line 1. After the generic instantiation, the add method in GenericMatrix<Integer> is now Integer add(Integer o1, Integer o2). The add, multiply, and zero methods are implemented for Integer objects. These methods are still protected because they are invoked only by the addMatrix and multiplyMatrix methods.

Listing 19.11 IntegerMatrix.java 1 public class IntegerMatrix extends GenericMatrix<Integer> { 2 @Override /** Add two integers */ 3 protected Integer add(Integer o1, Integer o2) { 4 return o1 + o2; 5 } 6 7 @Override /** Multiply two integers */ 8 protected Integer multiply(Integer o1, Integer o2) { 9 return o1 * o2; 10 } 11 12 @Override /** Specify zero for an integer */ 13 protected Integer zero() { 14 return 0; 15 } 16 }

Listing 19.12 implements the RationalMatrix class. The Rational class was introduced in Listing 13.13, Rational.java. Rational is a subtype of Number. The RationalMatrix class extends GenericMatrix<Rational> in line 1. After the generic instantiation, the add method in GenericMatrix<Rational> is now Rational add(Rational r1, Rational r2). The add, multiply, and zero methods are implemented for Rational objects. These methods are still protected because they are invoked only by the addMatrix and multiplyMatrix methods.

Listing 19.12 RationalMatrix.java 1 public class RationalMatrix extends GenericMatrix<Rational> { 2 @Override /** Add two rational numbers */ 3 protected Rational add(Rational r1, Rational r2) { 4 return r1.add(r2); 5 } 6 7 @Override /** Multiply two rational numbers */ 8 protected Rational multiply(Rational r1, Rational r2) { 9 return r1.multiply(r2); 10 } 11 12 @Override /** Specify zero for a Rational number */ 13 protected Rational zero() { 14 return new Rational(0, 1); 15 } 16 }

implement zero

implement multiply

implement add

extends generic type

implement multiply

implement add

implement zero

M19_LIAN1878_11_GE_C19.indd 791 12/29/17 4:31 PM

792 Chapter 19 Generics

Listing 19.13 gives a program that creates two Integer matrices (lines 4 and 5) and an IntegerMatrix object (line 8), and adds and multiplies two matrices in lines 12 and 16.

Listing 19.13 TestIntegerMatrix.java 1 public class TestIntegerMatrix { 2 public static void main(String[] args) { 3 // Create Integer arrays m1, m2 4 Integer[][] m1 = new Integer[][]{{1, 2, 3}, {4, 5, 6}, {1, 1, 1}}; 5 Integer[][] m2 = new Integer[][]{{1, 1, 1}, {2, 2, 2}, {0, 0, 0}}; 6 7 // Create an instance of IntegerMatrix 8 IntegerMatrix integerMatrix = new IntegerMatrix(); 9 10 System.out.println("\nm1 + m2 is "); 11 GenericMatrix.printResult( 12 m1, m2, integerMatrix.addMatrix(m1, m2), '+'); 13 14 System.out.println("\nm1 * m2 is "); 15 GenericMatrix.printResult( 16 m1, m2, integerMatrix.multiplyMatrix(m1, m2), '*'); 17 } 18 }

add two matrices

create matrices

create IntegerMatrix

multiply two matrices

m1 + m2 is 1 2 3 1 1 1 2 3 4 4 5 6 + 2 2 2 = 6 7 8 1 1 1 0 0 0 1 1 1

m1 * m2 is 1 2 3 1 1 1 5 5 5 4 5 6 * 2 2 2 = 14 14 14 1 1 1 0 0 0 3 3 3

Listing 19.14 gives a program that creates two Rational matrices (lines 4–10) and a RationalMatrix object (line 13) and adds and multiplies two matrices in lines 17 and 19.

Listing 19.14 TestRationalMatrix.java 1 public class TestRationalMatrix { 2 public static void main(String[] args) { 3 // Create two Rational arrays m1 and m2 4 Rational[][] m1 = new Rational[3][3]; 5 Rational[][] m2 = new Rational[3][3]; 6 for (int i = 0; i < m1.length; i++) 7 for (int j = 0; j < m1[0].length; j++) { 8 m1[i][j] = new Rational(i + 1, j + 5); 9 m2[i][j] = new Rational(i + 1, j + 6); 10 } 11 12 // Create an instance of RationalMatrix 13 RationalMatrix rationalMatrix = new RationalMatrix(); 14 15 System.out.println("\nm1 + m2 is "); 16 GenericMatrix.printResult( 17 m1, m2, rationalMatrix.addMatrix(m1, m2), '+'); 18 19 System.out.println("\nm1 * m2 is ");

create RationalMatrix

create matrices

add two matrices

M19_LIAN1878_11_GE_C19.indd 792 12/29/17 4:31 PM

20 GenericMatrix.printResult( 21 m1, m2, rationalMatrix.multiplyMatrix(m1, m2), '*'); 22 } 23 }

multiply two matrices

m1 + m2 is 1/5 1/6 1/7 1/6 1/7 1/8 11/30 13/42 15/56 2/5 1/3 2/7 + 1/3 2/7 1/4 = 11/15 13/21 15/28 3/5 1/2 3/7 1/2 3/7 3/8 11/10 13/14 45/56

m1 * m2 is 1/5 1/6 1/7 1/6 1/7 1/8 101/630 101/735 101/840 2/5 1/3 2/7 * 1/3 2/7 1/4 = 101/315 202/735 101/420 3/5 1/2 3/7 1/2 3/7 3/8 101/210 101/245 101/280

19.9.1 Why are the add, multiple, and zero methods defined abstract in the GenericMatrix class?

19.9.2 How are the add, multiple, and zero methods implemented in the IntegerMatrix class?

19.9.3 How are the add, multiple, and zero methods implemented in the RationalMatrix class?

19.9.4 What would be wrong if the printResult method is defined as follows?

public static void printResult( E[][] m1, E[][] m2, E[][] m3, char op)

Point Check

Key terms actual concrete type 774 bounded generic type 779 bounded wildcard

(<? extends E>) 784 formal generic type 774 generic instantiation 774

lower bound wildcard (<? super E>) 784

raw type 782 unbounded wildcard (<?>) 784 type erasure (<?>) 786

ChapTer summary 1. Generics give you the capability to parameterize types. You can define a class or a

method with generic types, which are substituted with concrete types.

2. The key benefit of generics is to enable errors to be detected at compile time rather than at runtime.

3. A generic class or method permits you to specify allowable types of objects that the class or method can work with. If you attempt to use a class or method with an incompatible object, the compiler will detect the error.

4. A generic type defined in a class, interface, or a static method is called a formal generic type, which can be replaced later with an actual concrete type. Replacing a generic type is called a generic instantiation.

Chapter Summary 793

M19_LIAN1878_11_GE_C19.indd 793 12/29/17 4:31 PM

794 Chapter 19 Generics

5. A generic class such as ArrayList used without a type parameter is called a raw type. Use of raw types allows for backward compatibility with the earlier versions of Java.

6. A wildcard generic type has three forms: ?, ? extends T, and ? super T, where T is a generic type. The first form, ?, called an unbounded wildcard, is the same as ? extends Object. The second form, ? extends T, called a bounded wildcard, rep- resents T or a subtype of T. The third form, ? super T, called a lower bound wildcard, denotes T or a supertype of T.

7. Generics are implemented using an approach called type erasure. The compiler uses the generic-type information to compile the code but erases it afterward, so the generic information is not available at runtime. This approach enables the generic code to be backward compatible with the legacy code that uses raw types.

8. You cannot create an instance using a generic-type parameter such as new E().

9. You cannot create an array using a generic-type parameter such as new E[10].

10. You cannot use a generic-type parameter of a class in a static context.

11. Generic-type parameters cannot be used in exception classes.

Quiz Answer the quiz for this chapter online at the book Companion Website.

programming exerCiSeS 19.1 (Revising Listing 19.1) Revise the GenericStack class in Listing 19.1 to imple-

ment it using an array rather than an ArrayList. You should check the array size before adding a new element to the stack. If the array is full, create a new array that doubles the current array size and copy the elements from the current array to the new array.

19.2 (Implement GenericStack using inheritance) In Listing 19.1, GenericStack is implemented using composition. Define a new stack class that extends ArrayList.

Draw the UML diagram for the classes then implement GenericStack. Write a test program that prompts the user to enter five strings and displays them in reverse order.

19.3 (Pair of objects of the same type) Create a Pair class that encapsulates two objects of the same data type in an instance of Pair.

19.4 (Using wildcards) Write a generic static method that returns the smallest value in an instance of Pair from Programming Exercise 19.3.

19.5 (Inheritance between generic classes) Create a Triplet class that encapsulates three objects of the same data type in a given instance of Triplet.

M19_LIAN1878_11_GE_C19.indd 794 12/29/17 4:31 PM

19.6 (Several types) Create an Association class that encapsulates two objects of different types. Similar to Programming Exercise 19.5, create a Transition class that does the same of Association class with three objects.

19.7 (Sum of an association) Knowing that any object of type java.lang.Number can be evaluated as a double with its doubleValue() method, write a method that computes and returns the sum of the three numbers in an instance of Transition as defined in Programming Exercise 19.6.

19.8 (Shuffle ArrayList) Write the following method that shuffles an ArrayList:

public static <E> void shuffle(ArrayList<E> list)

19.9 (Sort ArrayList) Write the following method that sorts an ArrayList:

public static <E extends Comparable<E>> void sort(ArrayList<E> list)

19.10 (Smallest element in ArrayList) Write the following method that returns the smallest element in an ArrayList:

public static <E extends Comparable<E>> E min(ArrayList<E> list)

19.11 (ComplexMatrix) Use the Complex class introduced in Programming Exercise 13.17 to develop the ComplexMatrix class for performing matrix opera tions involving complex numbers. The ComplexMatrix class should extend the GenericMatrix class and implement the add, multiple, and zero methods. You need to modify GenericMatrix and replace every occurrence of Number by Object because Com- plex is not a subtype of Number. Write a test program that creates the following two matrices and displays the result of addition and multiplication of the matrices by invoking the printResul19.

*19.12 (Revising Listing 23.4) to make it be generic bubble sort.

Programming Exercises 795

M19_LIAN1878_11_GE_C19.indd 795 12/29/17 4:31 PM

A01_GIPE3898_09_SE_FM.indd 3 12/20/17 5:56 PM

This page intentionally left blank

Objectives ■■ To explore the relationship between interfaces and classes in the Java

Collections Framework hierarchy (§20.2).

■■ To use the common methods defined in the Collection interface for operating collections (§20.2).

■■ To use the Iterator interface to traverse the elements in a collection (§20.3).

■■ To use a foreach loop to traverse the elements in a collection (§20.3).

■■ To use a forEach method to perform an action on each element in a collection (§20.4).

■■ To explore how and when to use ArrayList or LinkedList to store a list of elements (§20.5).

■■ To compare elements using the Comparable interface and the Comparator interface (§20.6).

■■ To use the static utility methods in the Collections class for sorting, searching, shuffling lists, and finding the largest and smallest element in collections (§20.7).

■■ To develop a multiple bouncing balls application using ArrayList (§20.8).

■■ To distinguish between Vector and ArrayList and to use the Stack class for creating stacks (§20.9).

■■ To explore the relationships among Collection , Queue , LinkedList, and PriorityQueue and to create priority queues using the PriorityQueue class (§20.10).

■■ To use stacks to write a program to evaluate expressions (§20.11).

Lists, Stacks, Queues, and Priority Queues

Chapter

M20_LIAN1878_11_GE_C20.indd 797 12/29/17 4:33 PM

798 Chapter 20 Lists, Stacks, Queues, and Priority Queues

20.1 Introduction Choosing the best data structures and algorithms for a particular task is one of the keys to developing high-performance software.

Chapters 18–29 are typically taught in a data structures course. A data structure is a collection of data organized in some fashion. The structure not only stores data, but also supports opera- tions for accessing and manipulating the data. Without knowing data structures, you can still write programs, but your program may not be efficient. With a good knowledge of data structures, you can build efficient programs, which are important for practical applications.

In object-oriented thinking, a data structure, also known as a container or container object, is an object that stores other objects, referred to as data or elements. To define a data structure is essentially to define a class. The class for a data structure should use data fields to store data and provide meth- ods to support such operations as search, insertion, and deletion. To create a data structure is therefore to create an instance from the class. You can then apply the methods on the instance to manipulate the data structure, such as inserting an element into or deleting an element from the data structure.

Section 11.11 introduced the ArrayList class, which is a data structure to store elements in a list. Java provides several more data structures (lists, vectors, stacks, queues, priority queues, sets, and maps) that can be used to organize and manipulate data efficiently. These are commonly known as Java Collections Framework. We will introduce the applications of lists, vectors, stacks, queues, and priority queues in this chapter, and sets and maps in the next chapter. The implementation of these data structures will be discussed in Chapters 24–27. Through implementation, students gain a deep understanding on the efficiency of data struc- tures and on how and when to use certain data structures. Finally, we will introduce design and implement data structures and algorithms for graphs in Chapters 28 and 29.

20.2 Collections The Collection interface defines the common operations for lists, vectors, stacks, queues, priority queues, and sets.

The Java Collections Framework supports two types of containers:

■■ One for storing a collection of elements is simply called a collection.

■■ The other, for storing key/value pairs, is called a map.

Maps are efficient data structures for quickly searching an element using a key. We will introduce maps in the next chapter. Now we turn our attention to the following collections.

■■ Sets store a group of nonduplicate elements.

■■ Lists store an ordered collection of elements.

■■ Stacks store objects that are processed in a last-in, first-out fashion.

■■ Queues store objects that are processed in a first-in, first-out fashion.

■■ PriorityQueues store objects that are processed in the order of their priorities.

The common operations of these collections are defined in the interfaces, and implementa- tions are provided in concrete classes, as shown in Figure 20.1.

Note All the interfaces and classes defined in the Java Collections Framework are grouped in the java.util package.

Design Guide The design of the Java Collections Framework is an excellent example of using interfaces, abstract classes, and concrete classes. The interfaces define the common operations.

Point Key

data structure

why learning data structure

container

Java Collections Framework

Point Key

collection

map

Set

List

Stack

Queue

PrioriryQueue

M20_LIAN1878_11_GE_C20.indd 798 12/29/17 4:33 PM

20.2 Collections 799

The abstract classes provide partial implementation. The concrete classes implement the interfaces with concrete data structures. Providing an abstract class that partially imple- ments an interface makes it convenient for the user to write the code. The user can simply define a concrete class that extends the abstract class rather than implementing all the methods in the interface. The abstract classes such as AbstractCollection are provided for convenience. For this reason, they are called convenience abstract classes.

The Collection interface is the root interface for manipulating a collection of objects. Its public methods are listed in Figure 20.2. The AbstractCollection class provides partial implementation for the Collection interface. It implements all the methods in Collection except the add, size, and iterator methods. These are implemented in the concrete subclasses.

The Collection interface provides the basic operations for adding and removing elements in a collection. The add method adds an element to the collection. The addAll method adds all the elements in the specified collection to this collection. The remove method removes an element from the collection. The removeAll method removes the elements from this collec- tion that are present in the specified collection. The retainAll method retains the elements in this collection that are also present in the specified collection. All these methods return boolean. The return value is true if the collection is changed as a result of the method execu- tion. The clear() method simply removes all the elements from the collection.

Note The methods addAll, removeAll, and retainAll are similar to the set union, difference, and intersection operations.

The Collection interface provides various query operations. The size method returns the number of elements in the collection. The contains method checks whether the collection contains the specified element. The containsAll method checks whether the collection con- tains all the elements in the specified collection. The isEmpty method returns true if the collection is empty.

The Collection interface provides the toArray() method, which returns an array of Object for the collection. It also provides the toArray(T[]) method, which returns an array of the T[] type.

convenience abstract class

basic operations

set operations

query operations

Figure 20.1 A collection is a container that stores objects.

SortedSet

Collection

TreeSet

Vector

LinkedList

PriorityQueueAbstractQueue

AbstractCollection Stack

LinkedHashSet

Interfaces Abstract Classes Concrete Classes

ArrayList

NavigableSet

Deque

HashSetSet

List

Queue

AbstractList

AbstractSet

AbstractSequentialList

M20_LIAN1878_11_GE_C20.indd 799 12/29/17 4:33 PM

800 Chapter 20 Lists, Stacks, Queues, and Priority Queues

Design Guide Some of the methods in the Collection interface cannot be implemented in the concrete subclass. In this case, the method would throw java.lang. UnsupportedOperation Exception, a subclass of RuntimeException. This is a good design you can use in your project. If a method has no meaning in the subclass, you can implement it as follows:

public void someMethod() { throw new UnsupportedOperationException ("Method not supported"); }

Listing 20.1 gives an example to use the methods defined in the Collection interface.

Listing 20.1 TestCollection.java 1 import java.util.*; 2 3 public class TestCollection { 4 public static void main(String[] args) { 5 ArrayList<String> collection1 = new ArrayList<>();

unsupported operations

create an array list

Figure 20.2 The Collection interface contains the methods for manipulating the elements in a collection, and you can obtain an iterator object for traversing elements in the collection.

Returns true if this iterator has more elements to traverse.

Returns the next element from this iterator.

Removes the last element obtained using the next method.

+hasNext(): boolean

+next(): E

+remove(): void

java.util.Iterator<E> «interface»

+iterator(): Iterator<E>

+forEach(action: Consumer<? super E>):

default void

java.lang.Iterable<E> «interface»

Returns an iterator for the elements in this collection.

Performs an action for each element in this iterator.

+add(e: E): boolean

+addAll(c: Collection<? extends E>): boolean

+clear(): void

+contains(o: Object): boolean

+containsAll(c: Collection<?>): boolean

+equals(o: Object): boolean

+isEmpty(): boolean

+remove(o: Object): boolean

+removeAll(c: Collection<?>): boolean

+retainAll(c: Collection<?>): boolean

+size(): int

+toArray(): Object[]

+toArray(a: T[]): T[]

Adds a new element e to this collection.

Adds all the elements in the collection c to this collection.

Removes all the elements from this collection.

Returns true if this collection contains the element o.

Returns true if this collection contains all the elements in c.

Returns true if this collection contains no elements.

Removes the element o from this collection.

Removes all the elements in c from this collection.

Retains the elements that are both in c and in this collection.

Returns the number of elements in this collection.

Returns an array of Object for the elements in this collection.

Returns an array of the T[] type.

java.util.Collection<E> «interface»

M20_LIAN1878_11_GE_C20.indd 800 12/29/17 4:33 PM

20.2 Collections 801

6 collection1.add("New York"); 7 collection1.add("Atlanta"); 8 collection1.add("Dallas"); 9 collection1.add("Madison"); 10 11 System.out.println("A list of cities in collection1:"); 12 System.out.println(collection1); 13 14 System.out.println("\nIs Dallas in collection1? " 15 + collection1.contains("Dallas")); 16 17 collection1.remove("Dallas"); 18 System.out.println("\n" + collection1.size() + 19 " cities are in collection1 now"); 20 21 Collection<String> collection2 = new ArrayList<>(); 22 collection2.add("Seattle"); 23 collection2.add("Portland"); 24 collection2.add("Los Angeles"); 25 collection2.add("Atlanta"); 26 27 System.out.println("\nA list of cities in collection2:"); 28 System.out.println(collection2); 29 30 ArrayList<String> c1 = (ArrayList<String>)(collection1.clone()); 31 c1.addAll(collection2); 32 System.out.println("\nCities in collection1 or collection2: "); 33 System.out.println(c1); 34 35 c1 = (ArrayList<String>)(collection1.clone()); 36 c1.retainAll(collection2); 37 System.out.print("\nCities in collection1 and collection2: "); 38 System.out.println(c1); 39 40 c1 = (ArrayList<String>)(collection1.clone()); 41 c1.removeAll(collection2); 42 System.out.print("\nCities in collection1, but not in 2: "); 43 System.out.println(c1); 44 } 45 }

add elements

contains?

size?

clone addAll

retainAll

removeAll

A list of cities in collection1:

[New York, Atlanta, Dallas, Madison]

Is Dallas in collection1? true

3 cities are in collection1 now

A list of cities in collection2:

[Seattle, Portland, Los Angeles, Atlanta]

Cities in collection1 or collection2:

[New York, Atlanta, Madison, Seattle, Portland, Los Angeles, Atlanta]

Cities in collection1 and collection2: [Atlanta]

Cities in collection1, but not in 2: [New York, Madison]

The program creates a concrete collection object using ArrayList (line 5) and invokes the Collection interface’s contains method (line 15), remove method (line 17), size method (line 18), addAll method (line 31), retainAll method (line 36), and removeAll method (line 41).

M20_LIAN1878_11_GE_C20.indd 801 12/29/17 4:33 PM

802 Chapter 20 Lists, Stacks, Queues, and Priority Queues

For this example, we use ArrayList. You can use any concrete class of Collection such as HashSet and LinkedList to replace ArrayList to test these methods defined in the Collection interface.

The program creates a copy of an array list (lines 30, 35, and 40). The purpose of this is to keep the original array list intact and use its copy to perform addAll, retainAll, and removeAll operations.

Note All the concrete classes in the Java Collections Framework implement the java.lang .Cloneable and java.io.Serializable interfaces except that java.util .PriorityQueue does not implement the Cloneable interface. Thus, all instances of Collection except priority queues can be cloned and all instances of Collection can be serialized.

20.2.1 What is a data structure? 20.2.2 Describe the Java Collections Framework. List the interfaces, convenience abstract

classes, and concrete classes under the Collection interface.

20.2.3 Can a collection object be cloned and serialized? 20.2.4 What method do you use to add all the elements from one collection to another

collection?

20.2.5 When should a method throw an UnsupportedOperationException?

20.3 Iterators Each collection is Iterable. You can obtain its Iterator object to traverse all the elements in the collection.

Iterator is a classic design pattern for walking through a data structure without having to expose the details of how data is stored in the data structure.

The Collection interface extends the Iterable interface. The Iterable interface defines the iterator method, which returns an iterator. The Iterator interface provides a uniform way for traversing elements in various types of collections. The iterator() method in the Iterable interface returns an instance of Iterator, as shown in Figure 20.2, which provides sequential access to the elements in the collection using the next() method. You can also use the hasNext() method to check whether there are more elements in the iterator, and the remove() method to remove the last element returned by the iterator.

Listing 20.2 gives an example that uses the iterator to traverse all the elements in an array list.

Listing 20.2 TestIterator.java 1 import java.util.*; 2 3 public class TestIterator { 4 public static void main(String[] args) { 5 Collection<String> collection = new ArrayList<>(); 6 collection.add("New York"); 7 collection.add("Atlanta"); 8 collection.add("Dallas"); 9 collection.add("Madison"); 10 11 Iterator<String> iterator = collection.iterator(); 12 while (iterator.hasNext()) { 13 System.out.print(iterator.next().toUpperCase() + " "); 14 } 15 System.out.println(); 16 } 17 }

Cloneable

Serializable

Point Check

Point Key

create an array list add elements

iterator hasNext() next()

M20_LIAN1878_11_GE_C20.indd 802 12/29/17 4:33 PM

20.4 Using the forEach Method 803

The program creates a concrete collection object using ArrayList (line 5) and adds four strings into the list (lines 6–9). The program then obtains an iterator for the collection (line 11) and uses the iterator to traverse all the strings in the list and displays the strings in uppercase (lines 12–14).

Tip You can simplify the code in lines 11–14 using a foreach loop without using an iterator, as follows:

for (String element: collection) System.out.print(element.toUpperCase() + " ");

This loop is read as “for each element in the collection, do the following.” The foreach loop can be used for arrays (see Section 7.2.7) as well as any instance of Iterable.

20.3.1 How do you obtain an iterator from a collection object? 20.3.2 What method do you use to obtain an element in the collection from an iterator? 20.3.3 Can you use a foreach loop to traverse the elements in any instance of

Collection?

20.3.4 When using a foreach loop to traverse all elements in a collection, do you need to use the next() or hasNext() methods in an iterator?

20.4 Using the forEach Method You can use the forEach method to perform an action for each element in a collection.

Java 8 added a new default method forEach in the Iterable interface. The method takes an argument for specifying the action, which is an instance of a functional interface Consumer<? super E>. The Consumer interface defines the accept(E e) method for performing an action on the element e. You can rewrite the preceding example using a forEach method in Listing 20.3.

Listing 20.3 TestForEach.java 1 import java.util.*; 2 3 public class TestForEach { 4 public static void main(String[] args) { 5 Collection<String> collection = new ArrayList<>(); 6 collection.add("New York"); 7 collection.add("Atlanta"); 8 collection.add("Dallas"); 9 collection.add("Madison"); 10 11 collection.forEach(e −> System.out.print(e.toUpperCase() + " ")); 12 } 13 }

foreach loop

Point Check

Point Key

create an array list add elements

forEach method

NEW YORK ATLANTA DALLAS MADISON

M20_LIAN1878_11_GE_C20.indd 803 12/29/17 4:33 PM

804 Chapter 20 Lists, Stacks, Queues, and Priority Queues

You can write the code using a foreach loop or using a forEach method. Using a forEach is simpler in most cases.

20.4.1 Can you use the forEach method on any instance of Collection? Where is the forEach method defined?

20.4.2 Suppose each element in list is a StringBuilder, write a statement using a forEach method to change the first character to uppercase for each element in list.

20.5 Lists The List interface extends the Collection interface and defines a collection for storing elements in a sequential order. To create a list, use one of its two concrete classes: ArrayList or LinkedList.

We used ArrayList to test the methods in the Collection interface in the preceding sections. Now, we will examine ArrayList in more depth. We will also introduce another useful list, LinkedList, in this section.

20.5.1 The Common Methods in the List Interface ArrayList and LinkedList are defined under the List interface. The List interface extends Collection to define an ordered collection with duplicates allowed. The List interface adds position-oriented operations as well as a new list iterator that enables a list to be traversed bidirectionally. The methods introduced in the List interface are shown in Figure 20.3.

The add(index, element) method is used to insert an element at a specified index and the addAll(index, collection) method to insert a collection of elements at a specified index. The remove(index) method is used to remove an element at the specified index from the list. A new element can be set at the specified index using the set(index, element) method.

The indexOf(element) method is used to obtain the index of the specified element’s first occurrence in the list and the lastIndexOf(element) method to obtain the index of its last occurrence. A sublist can be obtained by using the subList(fromIndex, toIndex) method.

Point Check

Point Key

The statement in line 11 uses a lambda expression in (a), which is equivalent to using an anonymous inner class as shown in (b). Using a lambda expression not only simplifies the syntax but also simplifies the semantics.

forEach(e −>

System.out.print(e.toUppserCase() + " "))

forEach(

new java.util.function.Consumer<String>() {

public void accept(String e) {

System.out.print(e.toUpperCase() + " ");

}

)

(a) Use a lambda expression (b) Use an anonymous inner class

M20_LIAN1878_11_GE_C20.indd 804 12/29/17 4:33 PM

20.5 Lists 805

The add(element) method inserts the specified element into the list. The element is inserted immediately before the next element that would be returned by the next() method defined in the Iterator interface, if any, and after the element that would be returned by the previous() method, if any. If the list doesn’t contain any elements, the new element becomes the sole element in the list. The set(element) method can be used to replace the last element returned by the next method, or the previous method with the specified element.

Figure 20.3 The List interface stores elements in sequence and permits duplicates.

«interface» java.util.List<E>

+add(index: int, element: E): void

+addAll(index: int, c: Collection<? extends E>) : boolean +get(index: int): E

+indexOf(element: Object): int

+lastIndexOf(element: Object): int

+listIterator(): ListIterator<E>

+listIterator(startIndex: int): ListIterator<E>

+remove(index: int) : E

+set(index: int, element: E): E

+subList(fromIndex: int, toIndex: int) : List<E>

«interface» java.util.Collection<E>

Adds a new element at the speci�ed index.

Adds all the elements in c to this list at the speci�ed index. Returns the element in this list at the speci�ed index.

Returns the index of the �rst matching element.

Returns the index of the last matching element.

Returns the list iterator for the elements in this list.

Returns the iterator for the elements from startIndex.

Removes the element at the speci�ed index and returns the removed element.

Sets the element at the speci�ed index and returns the old element.

Returns a sublist from fromIndex to toIndex-1.

Figure 20.4 ListIterator enables traversal of a list bidirectionally.

+add(o: E): void +hasPrevious(): boolean

+nextIndex(): int +previous(): E +previousIndex(): int +set(o: E): void

Adds the speci�ed object to the list. Returns true if this list iterator has more elements when traversing backward. Returns the index of the next element. Returns the previous element in this list iterator. Returns the index of the previous element. Replaces the last element returned by the previous or next method with the speci�ed element.

«interface» java.util.Iterator<E>

«interface» java.util.ListIterator<E>

The listIterator() or listIterator(startIndex) method returns an instance of ListIterator. The ListIterator interface extends the Iterator interface to add bidirectional traversal of the list. The methods in ListIterator are listed in Figure 20.4.

M20_LIAN1878_11_GE_C20.indd 805 12/29/17 4:33 PM

806 Chapter 20 Lists, Stacks, Queues, and Priority Queues

The hasNext() method defined in the Iterator interface is used to check whether the iterator has more elements when traversed in the forward direction, and the hasPrevious() method to check whether the iterator has more elements when traversed in the backward direction.

The next() method defined in the Iterator interface returns the next element in the itera- tor, and the previous() method returns the previous element in the iterator. The nextIndex() method returns the index of the next element in the iterator, and the previousIndex() returns the index of the previous element in the iterator.

The AbstractList class provides a partial implementation for the List interface. The AbstractSequentialList class extends AbstractList to provide support for linked lists.

20.5.2 The ArrayList and LinkedList Classes The ArrayList class and the LinkedList class are two concrete implementations of the List interface. ArrayList stores elements in an array. The array is dynamically created. If the capacity of the array is exceeded, a larger new array is created and all the elements from the current array are copied to the new array. LinkedList stores elements in a linked list. Which of the two classes you use depends on your specific needs. If you need to support random access through an index without inserting or removing elements at the beginning of the list, ArrayList is the most efficient. If, however, your application requires the insertion or deletion of elements at the beginning of the list, you should choose LinkedList. A list can grow or shrink dynamically. Once it is created, an array is fixed. If your application does not require the insertion or deletion of elements, an array is the most efficient data structure.

ArrayList is a resizable-array implementation of the List interface. It also provides methods for manipulating the size of the array used internally to store the list, as shown in Figure 20.5. Each ArrayList instance has a capacity, which is the size of the array used to store the elements in the list. It is always at least as large as the list size. As elements are added to an ArrayList, its capacity grows automatically. An ArrayList does not automatically shrink. You can use the trimToSize() method to reduce the array capacity to the size of the list. An ArrayList can be constructed using its no-arg constructor, ArrayList(Collection), or ArrayList(initialCapacity).

ArrayList vs. LinkedList

linked list

trimToSize()

Figure 20.5 ArrayList implements List using an array.

java.util.ArrayList<E>

+ArrayList() +ArrayList(c: Collection<? extends E>)

+ArrayList(initialCapacity: int)

+trimToSize(): void

java.util.AbstractList<E>

Creates an empty list with the default initial capacity. Creates an array list from an existing collection. Creates an empty list with the speci�ed initial capacity.

Trims the capacity of this ArrayList instance to be the list’s current size.

LinkedList is a linked list implementation of the List interface. In addition to imple- menting the List interface, this class provides the methods for retrieving, inserting, and removing elements from both ends of the list, as shown in Figure 20.6. A LinkedList can be constructed using its no-arg constructor or LinkedList(Collection).

M20_LIAN1878_11_GE_C20.indd 806 12/29/17 4:33 PM

20.5 Lists 807

Listing 20.4 gives a program that creates an array list filled with numbers and inserts new elements into specified locations in the list. The example also creates a linked list from the array list and inserts and removes elements from the list. Finally, the example traverses the list forward and backward.

Listing 20.4 TestArrayAndLinkedList.java 1 import java.util.*; 2 3 public class TestArrayAndLinkedList { 4 public static void main(String[] args) { 5 List<Integer> arrayList = new ArrayList<>(); 6 arrayList.add(1); // 1 is autoboxed to new Integer(1) 7 arrayList.add(2); 8 arrayList.add(3); 9 arrayList.add(1); 10 arrayList.add(4); 11 arrayList.add(0, 10); 12 arrayList.add(3, 30); 13 14 System.out.println("A list of integers in the array list:"); 15 System.out.println(arrayList); 16 17 LinkedList<Object> linkedList = new LinkedList<Object>(arrayList); 18 linkedList.add(1, "red"); 19 linkedList.removeLast(); 20 linkedList.addFirst("green"); 21 22 System.out.println("Display the linked list forward:"); 23 ListIterator<Object> listIterator = linkedList.listIterator(); 24 while (listIterator.hasNext()) { 25 System.out.print(listIterator.next() + " "); 26 } 27 System.out.println(); 28 29 System.out.println("Display the linked list backward:"); 30 listIterator = linkedList.listIterator(linkedList.size()); 31 while (listIterator.hasPrevious()) { 32 System.out.print(listIterator.previous() + " "); 33 } 34 } 35 }

array list

linked list

list iterator

Figure 20.6 LinkedList provides methods for adding and inserting elements at both ends of the list.

Creates a default empty linked list.

Creates a linked list from an existing collection.

Adds the element to the head of this list.

Adds the element to the tail of this list.

Returns the �rst element from this list.

Returns the last element from this list.

Returns and removes the �rst element from this list.

Returns and removes the last element from this list.

+LinkedList()

+LinkedList(c: Collection<? extends E>)

+addFirst(element: E): void

+addLast(element: E): void

+getFirst(): E

+getLast(): E

+removeFirst(): E

+removeLast(): E

java.util.LinkedList<E>

java.util.AbstractSequentialList<E>

M20_LIAN1878_11_GE_C20.indd 807 12/29/17 4:33 PM

808 Chapter 20 Lists, Stacks, Queues, and Priority Queues

A list can hold identical elements. Integer 1 is stored twice in the list (lines 6 and 9). ArrayList and LinkedList operate similarly. The critical difference between them pertains to internal implementation, which affects their performance. LinkedList is efficient for inserting and removing elements at the beginning of the list, and ArrayList is more efficient for all other operations. For examples of demonstrating the performance differences between ArrayList and LinkedList, see liveexample.pearsoncmg.com/supplement/ArrayListvsLinkedList.pdf.

The get(i) method is available for a linked list, but it is a time-consuming operation. Do not use it to traverse all the elements in a list as shown in (a). Instead, you should use a foreach loop as shown in (b) or a forEach method as shown in (c). Note (b) and (c) use an iterator implicitly. You will know the reason when you learn how to implement a linked list in Chapter 24.

A list of integers in the array list: [10, 1, 2, 30, 3, 1, 4] Display the linked list forward: green 10 red 1 2 30 3 1 Display the linked list backward: 1 3 30 2 1 red 10 green

for (int i = 0; i < list.size(); i++)

process list.get(i);

}

for (listElementType e: list) {

process e;

}

list.forEach(e −>

process e

)

(a) Very inefficient (b) Efficient (c) Efficient

Tip Java provides the static asList method for creating a list from a variable-length list of arguments. Thus, you can use the following code to create a list of strings and a list of integers:

List<String> list1 = Arrays.asList("red", "green", "blue"); List<Integer> list2 = Arrays.asList(10, 20, 30, 40, 50);

20.5.1 How do you add and remove elements from a list? How do you traverse a list in both directions?

20.5.2 Suppose list1 is a list that contains the strings red, yellow, and green and list2 is another list that contains the strings red, yellow, and blue. Answer the follow- ing questions:

a. What are list1 and list2 after executing list1.addAll(list2)?

b. What are list1 and list2 after executing list1.add(list2)?

c. What are list1 and list2 after executing list1.removeAll(list2)?

d. What are list1 and list2 after executing list1.remove(list2)?

e. What are list1 and list2 after executing list1.retainAll(list2)?

f. What is list1 after executing list1.clear()?

20.5.3 What are the differences between ArrayList and LinkedList? Which list should you use to insert and delete elements at the beginning of a list?

20.5.4 Are all the methods in ArrayList also in LinkedList? What methods are in LinkedList but not in ArrayList?

20.5.5 How do you create a list from an array of objects?

Arrays.asList(T... a) method

Point Check

M20_LIAN1878_11_GE_C20.indd 808 12/29/17 4:33 PM

20.6 The Comparator Interface 809

20.6 The Comparator Interface Comparator can be used to compare the objects of a class that doesn’t implement Comparable or define a new criteria for comparing objects.

You have learned how to compare elements using the Comparable interface (introduced in Section 13.6). Several classes in the Java API, such as String, Date, Calendar, BigInteger, BigDecimal, and all the numeric wrapper classes for the primitive types, implement the Com- parable interface. The Comparable interface defines the compareTo method, which is used to compare two elements of the same class that implements the Comparable interface.

What if the elements’ classes do not implement the Comparable interface? Can these elements be compared? You can define a comparator to compare the elements of different classes. To do so, define a class that implements the java.util.Comparator<T> interface and overrides its compare method.

public int compare(T element1, T element2)

Returns a negative value if element1 is less than element2, a positive value if element1 is greater than element2, and zero if they are equal.

The GeometricObject class was introduced in Section 13.2, Abstract Classes. The GeometricObject class does not implement the Comparable interface. To compare the objects of the GeometricObject class, you can define a comparator class, as given in Listing 20.5.

Listing 20.5 GeometricObjectComparator.java 1 import java.util.Comparator; 2 3 public class GeometricObjectComparator 4 implements Comparator<GeometricObject>, java.io.Serializable { 5 public int compare(GeometricObject o1, GeometricObject o2) { 6 double area1 = o1.getArea(); 7 double area2 = o2.getArea(); 8 9 if (area1 < area2) 10 return −1; 11 else if (area1 == area2) 12 return 0; 13 else 14 return 1; 15 } 16 }

Line 4 implements Comparator<GeometricObject>. Line 5 overrides the compare method to compare two geometric objects. The class also implements Serializable. It is generally a good idea for comparators to implement Serializable so they can be serialized.

Listing 20.6 gives a method that returns a larger object between two geometric objects. The objects are compared using the GeometricObjectComparator.

Listing 20.6 TestComparator.java 1 import java.util.Comparator; 2 3 public class TestComparator { 4 public static void main(String[] args) { 5 GeometricObject g1 = new Rectangle(5, 5); 6 GeometricObject g2 = new Circle(5); 7

Point Key

comparator

implements Comparator implements compare

M20_LIAN1878_11_GE_C20.indd 809 12/29/17 4:33 PM

810 Chapter 20 Lists, Stacks, Queues, and Priority Queues

8 GeometricObject g = 9 max(g1, g2, new GeometricObjectComparator()); 10 11 System.out.println("The area of the larger object is " + 12 g.getArea()); 13 } 14 15 public static GeometricObject max(GeometricObject g1, 16 GeometricObject g2, Comparator<GeometricObject> c) { 17 if (c.compare(g1, g2) > 0) 18 return g1; 19 else 20 return g2; 21 } 22 }

invoke max

the max method

invoke compare

The area of the larger object is 78.53981633974483

The program creates a Rectangle and a Circle object in lines 5 and 6 (the Rectangle and Circle classes were defined in Section 13.2, Abstract Classes). They are all subclasses of GeometricObject. The program invokes the max method to obtain the geometric object with the larger area (lines 8 and 9).

The GeometricObjectComparator is created and passed to the max method (line 9) and this comparator is used in the max method to compare the geometric objects in line 17.

Since the Comparator interface is a single abstract method interface, you can use a lambda expression to simplify the program by replacing line 9 with the following code:

max(g1, g2, (o1, o2) −> o1.getArea() > o2.getArea() ? 1 : o1.getArea() == o2.getArea() ? 0 : −1);

Here, o1 and o2 are two parameters in the compare method in the Comparator interface. The method returns 1 if o1.getArea() > o2.getArea(), 0 if o1.getArea() == o2.getArea(), and −1 otherwise.

Note Comparing elements using the Comparable interface is referred to as comparing using natural order, and comparing elements using the Comparator interface is referred to as comparing using comparator.

The preceding example defines a comparator for comparing two geometric objects since the GeometricObject class does not implement the Comparable interface. Sometimes a class implements the Comparable interface, but if you would like to compare their objects using a different criteria, you can define a custom comparator. Listing 20.7 gives an example that compares string by their length.

Listing 20.7 SortStringByLength.java 1 public class SortStringByLength { 2 public static void main(String[] args) { 3 String[] cities = {"Atlanta", "Savannah", "New York", "Dallas"}; 4 java.util.Arrays.sort(cities, new MyComparator()); 5 6 for (String s : cities) { 7 System.out.print(s + " "); 8 } 9 }

Comparable vs. Comparator natural order using comparator

sort using comparator

M20_LIAN1878_11_GE_C20.indd 810 12/29/17 4:33 PM

10 11 public static class MyComparator implements 12 java.util.Comparator<String> { 13 @Override 14 public int compare(String s1, String s2) { 15 return s1.length() − s2.length(); 16 } 17 } 18 }

define custom comparator

override compare method

Atlanta dallas new York Savannah

Dallas Atlanta Savannah New York

The program defines a comparator class by implementing the Comparator interface (lines 11 and 12). The compare method is implemented to compare two strings by their lengths (lines 14–16). The program invokes the sort method to sort an array of strings using a com- parator (line 4).

Since Comparator is a functional interface, the code can be simplified using a lambda expression as follows:

java.util.Arrays.sort(cities, (s1, s2) −> {return s1.length() − s2.length();});

or simply

java.util.Arrays.sort(cities, (s1, s2) −> s1.length() − s2.length());

The List interface defines the sort(comparator) method that can be used to sort the ele- ments in a list using a specified comparator. Listing 20.8 gives an example of using a compara- tor to sort strings in a list by ignoring cases.

Listing 20.8 SortStringIgnoreCase.java 1 public class SortStringIgnoreCase { 2 public static void main(String[] args) { 3 java.util.List<String> cities = java.util.Arrays.asList 4 ("Atlanta", "Savannah", "New York", "Dallas"); 5 cities.sort((s1, s2) −> s1.compareToIgnoreCase(s2)); 6 7 for (String s: cities) { 8 System.out.print(s + " "); 9 } 10 } 11 }

lambda comparator

The program sorts a list of strings using a comparator that compares strings ignoring case (line 5). If you invoke list.sort(null), the list will be sorted using its natural order.

The comparator is created using a lambda expression. Note the lambda expression here does nothing but simply invokes the compareToIgnoreCase method. In the case like this, you can use a simpler and clearer syntax to replace the lambda expression as follows:

cities.sort(String::compareToIgnoreCase);

Here String::compareToIgnoreCase is known as method reference, which is equivalent to a lambda expression. The compiler automatically translates a method reference to an equivalent lambda expression.

method reference

20.6 The Comparator Interface 811

M20_LIAN1878_11_GE_C20.indd 811 12/29/17 4:33 PM

812 Chapter 20 Lists, Stacks, Queues, and Priority Queues

The Comparator interface also contains several useful static methods and default methods. You can use the static comparing(Function<? sup T, ? sup R> keyExtracter) method to create a Comparator<T> that compares the elements using the key extracted from a Function object. The Function object’s apply(T) method returns the key of type R for the object T. For example, the following code in (a) creates a Comparator that compares strings by their length using a lambda expression, which is equivalent to the code using an anonymous inner class in (b) and a method reference in (c).

Comparator.comparing method

Comparator.comparing(e −> e.length()) Comparator.comparing(

(a) Use a lambda expression new java.util.function.Function<String, Integer>() {

public Integer apply(String s) {

return s.length();

}

Comparator.comparing(String::length) })

The comparing method in the Comparator interface is implemented essentially as follows for the preceding example:

// comparing returns a Comparator public static Comparator<String> comparing(Function<String, Integer> f) { return (s1, s2) −> f.apply(s1).compareTo(f.apply(s2)); }

You can replace the comparator in Listing 20.7 using the following code:

java.util.Arrays.sort(cities, Comparator.comparing(String::length));

The Comparator.comparing method is particularly useful to create a Comparator using a property from an object. For example, the following code sorts a list of Loan objects (see Listing 10.2) based on their loanAmount property.

Loan[] list = {new Loan(5.5, 10, 2323), new Loan(5, 10, 1000)}; Arrays.sort(list, Comparator.comparing(Loan::getLoanAmount));

You can sort using a primary criteria, second, third, and so on using the Comparator’s default thenComparing method. For example, the following code sorts a list of Loan objects first on their loanAmount then on annualInterestRate.

Loan[] list = {new Loan(5.5, 10, 100), new Loan(5, 10, 1000)}; Arrays.sort(list, Comparator.comparing(Loan::getLoanAmount) .thenComparing(Loan::getAnnualInterestRate));

The default reverse() method can be used to reverse the order for a comparator. For example, the following code sorts a list of Loan objects on their loanAmount property in a decreasing order.

Arrays.sort(list, Comparator.comparing(Loan::getLoanAmount). reverse());

20.6.1 What are the differences between the Comparable interface and the Comparator interface? In which package is Comparable, and in which package is Comparator?

20.6.2 How do you define a class A that implements the Comparable interface? Are two instances of class A comparable? How do you define a class B that implements the Comparator interface, and override the compare method to compare two objects of type B1? How do you invoke the sort method to sort a list of objects of the type B1 using a comparator?

thenComparing method

Point Check

M20_LIAN1878_11_GE_C20.indd 812 12/29/17 4:33 PM

20.7 Static Methods for Lists and Collections 813

20.6.3 Write a lambda expression to create a comparator that compares two Loan objects by their annualInterestRate. Create a comparator using the Comparator. comparing method to compare Loan objects on annualInterestRate. Create a comparator to compare Loan objects first on annualInterestRate then on loanAmount.

20.6.4 Create a comparator using a lambda expression and the Comparator. comparing method, respectively, to compare Collection objects on their size.

20.6.5 Write a statement that sorts an array of Point2D objects on their y values and then on their x values.

20.6.6 Write a statement that sorts an ArrayList of strings named list in increasing order of their last character.

20.6.7 Write a statement that sorts a two-dimensional array of double[][] in increasing order of their second column. For example, if the array is double[][] x = {{3, 1}, {2, −1}, {2, 0}}, the sorted array will be {{2, −1}, {2, 0}, {3, 1}}.

20.6.8 Write a statement that sorts a two-dimensional array of double[][] in increasing order of their second column as the primary order and the first column as the second- ary order. For example, if the array is double[][] x = {{3, 1}, {2, −1}, {2, 0}, {1, −1}}, the sorted array will be {{1, −1}, {2, −1}, {2, 0}, {3, 1}}.

20.7 Static Methods for Lists and Collections The Collections class contains static methods to perform common operations in a collection and a list.

Section 11.12 introduced several static methods in the Collections class for array lists. The Collections class contains the sort, binarySearch, reverse, shuffle, copy, and fill methods for lists and max, min, disjoint, and frequency methods for collections, as shown in Figure 20.7.

Point Key

Figure 20.7 The Collections class contains static methods for manipulating lists and collections.

java.util.Collections

+sort(list: List): void

+sort(list: List, c: Comparator): void

+binarySearch(list: List, key: Object): int

+binarySearch(list: List, key: Object, c: Comparator): int

+reverse(list: List): void

+reverseOrder(): Comparator

+shu�le(list: List): void

+shu�le(list: List, rmd: Random): void

+copy(des: List, src: List): void

+nCopies(n: int, o: Object): List

+�ll(list: List, o: Object): void

+max(c: Collection): Object

+max(c: Collection, c: Comparator): Object

+min(c: Collection): Object

+min(c: Collection, c: Comparator): Object

+disjoint(c1: Collection, c2: Collection): boolean

+frequency(c: Collection, o: Object): int

Sorts the speci�ed list.

Sorts the speci�ed list with the comparator.

Searches the key in the sorted list using binary search.

Searches the key in the sorted list using binary search

with the comparator. Reverses the speci�ed list.

Returns a comparator with the reverse ordering.

Shuf�es the speci�ed list randomly.

Shuf�es the speci�ed list with a random object. Copies from the source list to the destination list.

Returns a list consisting of n copies of the object. Fills the list with the object.

Returns the max object in the collection.

Returns the max object using the comparator.

Returns the min object in the collection.

Returns the min object using the comparator.

Returns true if c1 and c2 have no elements in common.

Returns the number of occurrences of the speci�ed

element in the collection.

List

Collection

M20_LIAN1878_11_GE_C20.indd 813 12/29/17 4:33 PM

814 Chapter 20 Lists, Stacks, Queues, and Priority Queues

You can sort the comparable elements in a list in its natural order with the compareTo method in the Comparable interface. You may also specify a comparator to sort elements. For example, the following code sorts strings in a list:

List<String> list = Arrays.asList("red", "green", "blue"); Collections.sort(list); System.out.println(list);

The output is [blue, green, red]. The preceding code sorts a list in ascending order. To sort it in descending order, you can

simply use the Collections.reverseOrder() method to return a Comparator object that orders the elements in reverse of natural order. For example, the following code sorts a list of strings in descending order:

List<String> list = Arrays.asList("yellow", "red", "green", "blue"); Collections.sort(list, Collections.reverseOrder()); System.out.println(list);

The output is [yellow, red, green, blue]. You can use the binarySearch method to search for a key in a list. To use this method,

the list must be sorted in increasing order. If the key is not in the list, the method returns - (insertion point + 1). Recall that the insertion point is where the item would fall in the list if it were present. For example, the following code searches the keys in a list of integers and a list of strings:

List<Integer> list1 = Arrays.asList(2, 4, 7, 10, 11, 45, 50, 59, 60, 66); System.out.println("(1) Index: " + Collections.binarySearch(list1, 7)); System.out.println("(2) Index: " + Collections.binarySearch(list1, 9));

List<String> list2 = Arrays.asList("blue", "green", "red"); System.out.println("(3) Index: " + Collections.binarySearch(list2, "red")); System.out.println("(4) Index: " + Collections.binarySearch(list2, "cyan"));

The output of the preceding code is:

sort list

ascending order descending order

binarySearch

(1) Index: 2 (2) Index: −4 (3) Index: 2 (4) Index: −2

You can use the reverse method to reverse the elements in a list. For example, the following code displays [blue, green, red, yellow]:

List<String> list = Arrays.asList("yellow", "red", "green", "blue"); Collections.reverse(list); System.out.println(list);

You can use the shuffle(List) method to randomly reorder the elements in a list. For example, the following code shuffles the elements in list:

List<String> list = Arrays.asList("yellow", "red", "green", "blue"); Collections.shuffle(list); System.out.println(list);

reverse

shuffle

M20_LIAN1878_11_GE_C20.indd 814 12/29/17 4:33 PM

20.7 Static Methods for Lists and Collections 815

You can also use the shuffle(List, Random) method to randomly reorder the elements in a list with a specified Random object. Using a specified Random object is useful to generate a list with identical sequences of elements for the same original list. For example, the following code shuffles the elements in list:

List<String> list1 = Arrays.asList("yellow", "red", "green", "blue"); List<String> list2 = Arrays.asList("yellow", "red", "green", "blue"); Collections.shuffle(list1, new Random(20)); Collections.shuffle(list2, new Random(20)); System.out.println(list1); System.out.println(list2);

You will see that list1 and list2 have the same sequence of elements before and after the shuffling.

You can use the copy(det, src) method to copy all the elements from a source list to a destination list on the same index. The destination list must be as long as the source list. If it is longer, the remaining elements in the source list are not affected. For example, the following code copies list2 to list1:

List<String> list1 = Arrays.asList("yellow", "red", "green", "blue"); List<String> list2 = Arrays.asList("white", "black"); Collections.copy(list1, list2); System.out.println(list1);

The output for list1 is [white, black, green, blue]. The copy method performs a shallow copy: Only the references of the elements from the source list are copied.

You can use the nCopies(int n, Object o) method to create an immutable list that consists of n copies of the specified object. For example, the following code creates a list with five Calendar objects:

List<GregorianCalendar> list1 = Collections.nCopies (5, new GregorianCalendar(2005, 0, 1));

The list created from the nCopies method is immutable, so you cannot add, remove, or update elements in the list. All the elements have the same references.

You can use the fill(List list, Object o) method to replace all the elements in the list with the specified element. For example, the following code displays [black, black, black]:

List<String> list = Arrays.asList("red", "green", "blue"); Collections.fill(list, "black"); System.out.println(list);

You can use the max and min methods for finding the maximum and minimum elements in a collection. The elements must be comparable using the Comparable interface or the Comparator interface. See the following code for examples:

Collection<String> collection = Arrays.asList("red", "green", "blue"); System.out.println(Collections.max(collection)); // Use Comparable System.out.println(Collections.min(collection, Comparator.comparing(String::length))); // Use Comparator

The disjoint(collection1, collection2) method returns true if the two collections have no elements in common. For example, in the following code, disjoint(collection1, collection2) returns false, but disjoint(collection1, collection3) returns true:

Collection<String> collection1 = Arrays.asList("red", "cyan"); Collection<String> collection2 = Arrays.asList("red", "blue"); Collection<String> collection3 = Arrays.asList("pink", "tan"); System.out.println(Collections.disjoint(collection1, collection2)); System.out.println(Collections.disjoint(collection1, collection3));

copy

nCopies

fill

max and min methods

disjoint method

M20_LIAN1878_11_GE_C20.indd 815 12/29/17 4:33 PM

816 Chapter 20 Lists, Stacks, Queues, and Priority Queues

The frequency(collection, element) method finds the number of occurrences of the element in the collection. For example, frequency(collection, "red") returns 2 in the following code:

Collection<String> collection = Arrays.asList("red", "cyan", "red"); System.out.println(Collections.frequency(collection, "red"));

20.7.1 Are all the methods in the Collections class static? 20.7.2 Which of the following static methods in the Collections class are for lists and

which are for collections?

sort, binarySearch, reverse, shuffle, max, min, disjoint, frequency

20.7.3 Show the output of the following code: import java.util.*;

public class Test { public static void main(String[] args) { List<String> list = Arrays.asList("yellow", "red", "green", "blue"); Collections.reverse(list); System.out.println(list);

List<String> list1 = Arrays.asList("yellow", "red", "green", "blue"); List<String> list2 = Arrays.asList("white", "black"); Collections.copy(list1, list2); System.out.println(list1);

Collection<String> c1 = Arrays.asList("red", "cyan"); Collection<String> c2 = Arrays.asList("red", "blue"); Collection<String> c3 = Arrays.asList("pink", "tan"); System.out.println(Collections.disjoint(c1, c2)); System.out.println(Collections.disjoint(c1, c3));

Collection<String> collection = Arrays.asList("red", "cyan", "red"); System.out.println(Collections.frequency(collection, "red")); } }

20.7.4 Which method can you use to sort the elements in an ArrayList or a LinkedList? Which method can you use to sort an array of strings?

20.7.5 Which method can you use to perform binary search for elements in an ArrayList or a LinkedList? Which method can you use to perform binary search for an array of strings?

20.7.6 Write a statement to find the largest element in an array of comparable objects.

20.8 Case Study: Bouncing Balls This section presents a program that displays bouncing balls and enables the user to add and remove balls.

Section 15.12 presents a program that displays one bouncing ball. This section presents a pro- gram that displays multiple bouncing balls. You can use two buttons to suspend and resume the movement of the balls, a scroll bar to control the ball speed, and the + or - button to add or remove a ball, as shown in Figure 20.8.

frequency method

Point Check

Point Key

M20_LIAN1878_11_GE_C20.indd 816 12/29/17 4:33 PM

20.8 Case Study: Bouncing Balls 817

The example in Section 15.12 only had to store one ball. How do you store multiple balls in this example? The Pane’s getChildren() method returns an ObservableList<Node>, a subtype of List<Node>, for storing the nodes in the pane. Initially, the list is empty. When a new ball is created, add it to the end of the list. To remove a ball, simply remove the last one in the list.

Each ball has its state: the x-, y-coordinates, color, and direction to move. You can define a class named Ball that extends javafx.scene.shape.Circle. The x-, y-coordinates and the color are already defined in Circle. When a ball is created, it starts from the upper-left corner and moves downward to the right. A random color is assigned to a new ball.

The MultiplBallPane class is responsible for displaying the ball and the MultipleBounceBall class places the control components and implements the control. The relationship of these classes is shown in Figure 20.9. Listing 20.9 gives the program.

Figure 20.9 MultipleBounceBall contains MultipleBallPane and MultipleBallPane contains Ball.

m 1 1 1 Ball

dx: double

dy: double

+Ball (x: double, y: double radius: double, color: Color)

javafx.scene.shape.Circle javafx.scene.layout.Pane javafx.application.Application

MultipleBallPane

–animation: Timeline

+MultipleBallPane()

+play(): void

+pause(): void

+increaseSpeed(): void

+decreaseSpeed(): void

+rateProperty():

DoubleProperty

+moveBall(): void

MultipleBounceBall

Listing 20.9 MultipleBounceBall.java 1 import javafx.animation.KeyFrame; 2 import javafx.animation.Timeline; 3 import javafx.application.Application; 4 import javafx.beans.property.DoubleProperty; 5 import javafx.geometry.Pos; 6 import javafx.scene.Node; 7 import javafx.stage.Stage; 8 import javafx.scene.Scene; 9 import javafx.scene.control.Button; 10 import javafx.scene.control.ScrollBar; 11 import javafx.scene.layout.BorderPane; 12 import javafx.scene.layout.HBox;

M20_LIAN1878_11_GE_C20.indd 817 12/29/17 4:33 PM

818 Chapter 20 Lists, Stacks, Queues, and Priority Queues

13 import javafx.scene.layout.Pane; 14 import javafx.scene.paint.Color; 15 import javafx.scene.shape.Circle; 16 import javafx.util.Duration; 17 18 public class MultipleBounceBall extends Application { 19 @Override // Override the start method in the Application class 20 public void start(Stage primaryStage) { 21 MultipleBallPane ballPane = new MultipleBallPane(); 22 ballPane.setStyle("−fx−border−color: yellow"); 23 24 Button btAdd = new Button("+"); 25 Button btSubtract = new Button("−"); 26 HBox hBox = new HBox(10); 27 hBox.getChildren().addAll(btAdd, btSubtract); 28 hBox.setAlignment(Pos.CENTER); 29 30 // Add or remove a ball 31 btAdd.setOnAction(e −> ballPane.add()); 32 btSubtract.setOnAction(e −> ballPane.subtract()); 33 34 // Pause and resume animation 35 ballPane.setOnMousePressed(e −> ballPane.pause()); 36 ballPane.setOnMouseReleased(e −> ballPane.play()); 37 38 // Use a scroll bar to control animation speed 39 ScrollBar sbSpeed = new ScrollBar(); 40 sbSpeed.setMax(20); 41 sbSpeed.setValue(10); 42 ballPane.rateProperty().bind(sbSpeed.valueProperty()); 43 44 BorderPane pane = new BorderPane(); 45 pane.setCenter(ballPane); 46 pane.setTop(sbSpeed); 47 pane.setBottom(hBox); 48 49 // Create a scene and place the pane in the stage 50 Scene scene = new Scene(pane, 250, 150); 51 primaryStage.setTitle("MultipleBounceBall"); // Set the stage title 52 primaryStage.setScene(scene); // Place the scene in the stage 53 primaryStage.show(); // Display the stage 54 } 55 56 private class MultipleBallPane extends Pane { 57 private Timeline animation; 58 59 public MultipleBallPane() { 60 // Create an animation for moving the ball 61 animation = new Timeline( 62 new KeyFrame(Duration.millis(50), e −> moveBall())); 63 animation.setCycleCount(Timeline.INDEFINITE); 64 animation.play(); // Start animation 65 } 66 67 public void add() { 68 Color color = new Color(Math.random(), 69 Math.random(), Math.random(), 0.5); 70 getChildren().add(new Ball(30, 30, 20,color)); 71 } 72

create a ball pane set ball pane border

create buttons

add buttons to HBox

add a ball remove a ball

pause animation resume animation

create a scroll bar

bind animation rate

add a ball to pane

M20_LIAN1878_11_GE_C20.indd 818 12/29/17 4:33 PM

20.8 Case Study: Bouncing Balls 819

73 public void subtract() { 74 if (getChildren().size() > 0) { 75 getChildren().remove(getChildren().size() − 1); 76 } 77 } 78 79 public void play() { 80 animation.play(); 81 } 82 83 public void pause() { 84 animation.pause(); 85 } 86 87 public void increaseSpeed() { 88 animation.setRate(animation.getRate() + 0.1); 89 } 90 91 public void decreaseSpeed() { 92 animation.setRate( 93 animation.getRate() > 0 ? animation.getRate() − 0.1 : 0); 94 } 95 96 public DoubleProperty rateProperty() { 97 return animation.rateProperty(); 98 } 99 100 protected void moveBall() { 101 for (Node node: this.getChildren()) { 102 Ball ball = (Ball)node; 103 // Check boundaries 104 if (ball.getCenterX() < ball.getRadius() || 105 ball.getCenterX() > getWidth() − ball.getRadius()) { 106 ball.dx *= −1; // Change ball move direction 107 } 108 if (ball.getCenterY() < ball.getRadius() || 109 ball.getCenterY() > getHeight() − ball.getRadius()) { 110 ball.dy *= −1; // Change ball move direction 111 } 112 113 // Adjust ball position 114 ball.setCenterX(ball.dx + ball.getCenterX()); 115 ball.setCenterY(ball.dy + ball.getCenterY()); 116 } 117 } 118 } 119 120 class Ball extends Circle { 121 private double dx = 1, dy = 1; 122 123 Ball(double x, double y, double radius, Color color) { 124 super(x, y, radius); 125 setFill(color); // Set ball color 126 } 127 } 128 }

The add() method creates a new ball with a random color and adds it to the pane (line 70). The pane stores all the balls in a list. The subtract() method removes the last ball in the list (line 75).

remove a ball

move all balls

change x-direction

change y-direction

adjust ball positions

declare dx and dy

create a ball

M20_LIAN1878_11_GE_C20.indd 819 12/29/17 4:33 PM

820 Chapter 20 Lists, Stacks, Queues, and Priority Queues

When the user clicks the + button, a new ball is added to the pane (line 31). When the user clicks the - button, the last ball in the array list is removed (line 32).

The moveBall() method in the MultipleBallPane class gets every ball in the pane’s list and adjusts the balls’ positions (lines 114 and 115).

20.8.1 What is the return value from invoking pane.getChildren() for a pane? 20.8.2 How do you modify the code in the MutilpleBallApp program to remove the

first ball in the list when the - button is clicked? 20.8.3 How do you modify the code in the MutilpleBallApp program so each ball will

get a random radius between 10 and 20?

20.9 Vector and Stack Classes Vector is a subclass of AbstractList and Stack is a subclass of Vector in the Java API.

The Java Collections Framework was introduced in Java 2. Several data structures were supported earlier, among them the Vector and Stack classes. These classes were redesigned to fit into the Java Collections Framework, but all their old-style methods are retained for compatibility.

Vector is the same as ArrayList, except that it contains synchronized methods for access- ing and modifying the vector. Synchronized methods can prevent data corruption when a vector is accessed and modified by two or more threads concurrently. We will discuss synchronization in Chapter 32, Multithreading and Parallel Programming. For the many applications that do not require synchronization, using ArrayList is more efficient than using Vector.

The Vector class extends the AbstractList class. It also has the methods contained in the original Vector class defined prior to Java 2, as shown in Figure 20.10.

Point Check

Point Key

Figure 20.10 Starting in Java 2, the Vector class extends AbstractList and also retains all the methods in the original Vector class.

java.util.Vector<E>

+Vector()

+Vector(c: Collection<? extends E>)

+Vector(initialCapacity: int)

+Vector(initCapacity: int, capacityIncr: int)

+addElement(o: E): void

+capacity(): int

+copyInto(anArray: Object[]): void

+elementAt(index: int): E

+elements(): Enumeration<E>

+ensureCapacity(): void

+�rstElement(): E

+insertElementAt(o: E, index: int): void

+lastElement(): E

+removeAllElements(): void

+removeElement(o: Object): boolean

+removeElementAt(index: int): void

+setElementAt(o: E, index: int): void

+setSize(newSize: int): void

+trimToSize(): void

Creates a default empty vector with initial capacity 10.

Creates a vector from an existing collection.

Creates a vector with the speci�ed initial capacity.

Creates a vector with the speci�ed initial capacity and increment.

Appends the element to the end of this vector.

Returns the current capacity of this vector.

Copies the elements in this vector to the array.

Returns the object at the speci�ed index.

Returns an enumeration of this vector.

Increases the capacity of this vector.

Returns the �rst element in this vector.

Inserts o into this vector at the speci�ed index.

Returns the last element in this vector.

Removes all the elements in this vector.

Removes the �rst matching element in this vector.

Removes the element at the speci�ed index.

Sets a new element at the speci�ed index.

Sets a new size in this vector.

Trims the capacity of this vector to its size.

java.util.AbstractList<E>

M20_LIAN1878_11_GE_C20.indd 820 12/29/17 4:33 PM

20.10 Queues and Priority Queues 821

Most of the methods in the Vector class listed in the UML diagram in Figure 20.10 are similar to the methods in the List interface. These methods were introduced before the Java Collections Framework. For example, addElement(Object element) is the same as the add(Object element) method, except that the addElement method is synchronized. Use the ArrayList class if you don’t need synchronization. It works much faster than Vector.

Note The elements() method returns an Enumeration. The Enumeration interface was introduced prior to Java 2 and was superseded by the Iterator interface.

Note Vector is widely used in Java legacy code because it was the Java resizable-array imple- mentation before Java 2.

In the Java Collections Framework, Stack is implemented as an extension of Vector, as illustrated in Figure 20.11.

Figure 20.11 The Stack class extends Vector to provide a last-in, first-out data structure.

java.util.Stack<E>

+Stack()

+empty(): boolean

+peek(): E

+pop(): E

+push(o: E): E

+search(o: Object): int

java.util.Vector<E>

Creates an empty stack.

Returns true if this stack is empty.

Returns the top element in this stack.

Returns and removes the top element in this stack.

Adds a new element to the top of this stack.

Returns the position of the speci�ed element in this stack.

The Stack class was introduced prior to Java 2. The methods shown in Figure 20.11 were used before Java 2. The empty() method is the same as isEmpty(). The peek() method looks at the element at the top of the stack without removing it. The pop() method removes the top element from the stack and returns it. The push(Object element) method adds the specified element to the stack. The search(Object element) method checks whether the specified element is in the stack.

20.9.1 How do you create an instance of Vector? How do you add or insert a new element into a vector? How do you remove an element from a vector? How do you find the size of a vector?

20.9.2 How do you create an instance of Stack? How do you add a new element to a stack? How do you remove an element from a stack? How do you find the size of a stack?

20.9.3 Does Listing 20.1, TestCollection.java, compile and run if all the occurrences of ArrayList are replaced by LinkedList, Vector, or Stack?

20.10 Queues and Priority Queues In a priority queue, the element with the highest priority is removed first.

A queue is a first-in, first-out data structure. Elements are appended to the end of the queue and are removed from the beginning of the queue. In a priority queue, elements are assigned priorities. When accessing elements, the element with the highest priority is removed first. This section introduces queues and priority queues in the Java API.

Point Check

Point Key

queue priority queue

M20_LIAN1878_11_GE_C20.indd 821 12/29/17 4:33 PM

822 Chapter 20 Lists, Stacks, Queues, and Priority Queues

20.10.1 The Queue Interface The Queue interface extends java.util.Collection with additional insertion, extraction, and inspection operations, as shown in Figure 20.12.

Queue interface

Figure 20.12 The Queue interface extends Collection to provide additional insertion, extraction, and inspection operations.

«interface» java.util.Queue<E>

+o�er(element: E): boolean

+poll(): E

+remove(): E

+peek(): E

+element(): E

«interface» java.util.Collection<E>

Inserts an element into the queue.

Retrieves and removes the head of this queue, or null if this queue is empty.

Retrieves and removes the head of this queue and throws an exception if this queue is empty.

Retrieves, but does not remove, the head of this queue, returning null if this queue is empty.

Retrieves, but does not remove, the head of this queue, throws an exception if this queue is empty.

The offer method is used to add an element to the queue. This method is similar to the add method in the Collection interface, but the offer method is preferred for queues. The poll and remove methods are similar, except that poll() returns null if the queue is empty, whereas remove() throws an exception. The peek and element methods are similar, except that peek() returns null if the queue is empty, whereas element() throws an exception.

20.10.2 Deque and LinkedList The LinkedList class implements the Deque interface, which extends the Queue interface, as shown in Figure 20.13. Therefore, you can use LinkedList to create a queue. LinkedList is ideal for queue operations because it is efficient for inserting and removing elements from both ends of a list.

queue operations

Figure 20.13 LinkedList implements List and Deque.

«interface» java.util.Deque<E>

java.util.LinkedList<E>

«interface» java.util.Collection<E>

«interface» java.util.Queue<E>

«interface» java.util.List<E>

M20_LIAN1878_11_GE_C20.indd 822 12/29/17 4:33 PM

20.10 Queues and Priority Queues 823

Deque supports element insertion and removal at both ends. The name deque is short for “double-ended queue” and is usually pronounced “deck.” The Deque interface extends Queue with additional methods for inserting and removing elements from both ends of the queue. The methods addFirst(e), removeFirst(), addLast(e), removeLast(), getFirst(), and getLast() are defined in the Deque interface.

Listing 20.10 shows an example of using a queue to store strings. Line 3 creates a queue using LinkedList. Four strings are added to the queue in lines 4–7. The size() method defined in the Collection interface returns the number of elements in the queue (line 9). The remove() method retrieves and removes the element at the head of the queue (line 10).

Listing 20.10 TestQueue.java 1 public class TestQueue { 2 public static void main(String[] args) { 3 java.util.Queue<String> queue = new java.util.LinkedList<>(); 4 queue.offer("Oklahoma"); 5 queue.offer("Indiana"); 6 queue.offer("Georgia"); 7 queue.offer("Texas"); 8 9 while (queue.size() > 0) 10 System.out.print(queue.remove() + " "); 11 } 12 }

creates a queue inserts an element

queue size remove element

Figure 20.14 The PriorityQueue class implements a priority queue.

java.util.PriorityQueue<E>

+PriorityQueue()

+PriorityQueue(initialCapacity: int)

+PriorityQueue(c: Collection<? extends E>)

+PriorityQueue(initialCapacity: int, comparator: Comparator<? super E>)

Creates a default priority queue with initial capacity 11.

Creates a default priority queue with the speci�ed initial capacity.

Creates a priority queue with the speci�ed collection.

Creates a priority queue with the speci�ed initial capacity and the comparator.

«interface» java.util.Queue<E>

Oklahoma Indiana Georgia Texas

The PriorityQueue class implements a priority queue, as shown in Figure 20.14. By default, the priority queue orders its elements according to their natural ordering using Comparable. The element with the least value is assigned the highest priority, and thus is removed from the queue first. If there are several elements with the same highest priority, the tie is broken arbitrarily. You can also specify an ordering using Comparator in the constructor PriorityQueue(initialCapacity, comparator).

PriorityQueue class

M20_LIAN1878_11_GE_C20.indd 823 12/29/17 4:33 PM

824 Chapter 20 Lists, Stacks, Queues, and Priority Queues

Listing 20.11 shows an example of using a priority queue to store strings. Line 5 cre- ates a priority queue for strings using its no-arg constructor. This priority queue orders the strings using their natural order, so the strings are removed from the queue in increasing order. Lines 16 and 17 create a priority queue using the comparator obtained from Collections. reverseOrder(), which orders the elements in reverse order, so the strings are removed from the queue in decreasing order.

Listing 20.11 PriorityQueueDemo.java 1 import java.util.*; 2 3 public class PriorityQueueDemo { 4 public static void main(String[] args) { 5 PriorityQueue<String> queue1 = new PriorityQueue<>(); 6 queue1.offer("Oklahoma"); 7 queue1.offer("Indiana"); 8 queue1.offer("Georgia"); 9 queue1.offer("Texas"); 10 11 System.out.println("Priority queue using Comparable:"); 12 while (queue1.size() > 0) { 13 System.out.print(queue1.remove() + " "); 14 } 15 16 PriorityQueue<String> queue2 = new PriorityQueue<>( 17 4, Collections.reverseOrder()); 18 queue2.offer("Oklahoma"); 19 queue2.offer("Indiana"); 20 queue2.offer("Georgia"); 21 queue2.offer("Texas"); 22 23 System.out.println("\nPriority queue using Comparator:"); 24 while (queue2.size() > 0) { 25 System.out.print(queue2.remove() + " "); 26 } 27 } 28 }

a default queue inserts an element

a queue with comparator

Priority queue using Comparable: Georgia Indiana Oklahoma Texas Priority queue using Comparator: Texas Oklahoma Indiana Georgia

20.10.1 Is java.util.Queue a subinterface of java.util.Collection, java.util. Set, or java.util.List? Does LinkedList implement Queue?

20.10.2 How do you create a priority queue for integers? By default, how are elements ordered in a priority queue? Is the element with the least value assigned the highest priority in a priority queue?

20.10.3 How do you create a priority queue that reverses the natural order of the elements?

Point Check

M20_LIAN1878_11_GE_C20.indd 824 12/29/17 4:33 PM

20.11 Case Study: Evaluating Expressions 825

How does Google evaluate an expression? This section presents a program that evaluates a compound expression with multiple operators and parentheses (e.g., (15 + 2) * 34 − 2). For simplicity, assume the operands are integers, and the operators are of four types: +, −, *, and /.

The problem can be solved using two stacks, named operandStack and operator- Stack, for storing operands and operators, respectively. Operands and operators are pushed into the stacks before they are processed. When an operator is processed, it is popped from operatorStack and applied to the first two operands from operandStack (the two operands are popped from operandStack). The resultant value is pushed back to operandStack.

The algorithm proceeds in two phases:

Phase 1: Scanning the expression The program scans the expression from left to right to extract operands, operators, and the parentheses.

1.1. If the extracted item is an operand, push it to operandStack.

1.2. If the extracted item is a + or − operator, process all the operators at the top of operatorStack and push the extracted operator to operatorStack.

1.3. If the extracted item is a * or / operator, process the * or / operators at the top of operatorStack and push the extracted operator to operatorStack.

1.4. If the extracted item is a ( symbol, push it to operatorStack.

1.5. If the extracted item is a ) symbol, repeatedly process the operators from the top of operatorStack until seeing the ( symbol on the stack.

compound expression

process an operator

Figure 20.15 You can evaluate an arithmetic expression using a Google search engine. Source: Google and the Google logo are registered trademarks of Google Inc., used with permission.

20.11 Case Study: Evaluating Expressions Stacks can be used to evaluate expressions.

Stacks and queues have many applications. This section gives an application that uses stacks to evaluate expressions. You can enter an arithmetic expression from Google to evaluate the expression, as shown in Figure 20.15.

Point Key

M20_LIAN1878_11_GE_C20.indd 825 12/29/17 4:33 PM

826 Chapter 20 Lists, Stacks, Queues, and Priority Queues

Listing 20.12 gives the program, and Figure 20.16 shows some sample output.

Expression Scan Action operandStack operatorStack

(1 + 2) * 4 - 3 c

( Phase 1.4 (

(1 + 2) * 4 - 3 c

1 Phase 1.1 1 (

(1 + 2) * 4 - 3 c

+ Phase 1.2 1 + (

(1 + 2) * 4 - 3 c

2 Phase 1.1 2 1

(

(1 + 2) * 4 - 3 c

) Phase 1.5 3

(1 + 2) * 4 - 3 c

* Phase 1.3 3 *

(1 + 2) * 4 - 3 c

4 Phase 1.1 4 3

(1 + 2) * 4 - 3 c

- Phase 1.2 12 -

(1 + 2) * 4 - 3 c

3 Phase 1.1 3 12

(1 + 2) * 4 - 3 c

none Phase 2 9

tabLe 20.1 Evaluating an Expression

Phase 2: Clearing the stack

Repeatedly process the operators from the top of operatorStack until operatorStack is empty.

Table 20.1 shows how the algorithm is applied to evaluate the expression (1 + 2) * 4 − 3.

Listing 20.12 EvaluateExpression.java 1 import java.util.Stack; 2 3 public class EvaluateExpression { 4 public static void main(String[] args) {

M20_LIAN1878_11_GE_C20.indd 826 12/29/17 4:33 PM

20.11 Case Study: Evaluating Expressions 827

5 // Check number of arguments passed 6 if (args.length != 1) { 7 System.out.println( 8 "Usage: java EvaluateExpression \"expression\""); 9 System.exit(1); 10 } 11 12 try { 13 System.out.println(evaluateExpression(args[0])); 14 } 15 catch (Exception ex) { 16 System.out.println("Wrong expression: " + args[0]); 17 } 18 } 19 20 /** Evaluate an expression */ 21 public static int evaluateExpression(String expression) { 22 // Create operandStack to store operands 23 Stack<Integer> operandStack = new Stack<>(); 24 25 // Create operatorStack to store operators 26 Stack<Character> operatorStack = new Stack<>(); 27 28 // Insert blanks around (, ), +, −, /, and * 29 expression = insertBlanks(expression); 30 31 // Extract operands and operators 32 String[] tokens = expression.split(" "); 33 34 // Phase 1: Scan tokens 35 for (String token: tokens) { 36 if (token.length() == 0) // Blank space 37 continue; // Back to the while loop to extract the next token 38 else if (token.charAt(0) == '+' || token.charAt(0) == '−') { 39 // Process all +, −, *, / in the top of the operator stack 40 while (!operatorStack.isEmpty() && 41 (operatorStack.peek() == '+' || 42 operatorStack.peek() == '−' || 43 operatorStack.peek() == '*' || 44 operatorStack.peek() == '/')) { 45 processAnOperator(operandStack, operatorStack); 46 } 47 48 // Push the + or − operator into the operator stack 49 operatorStack.push(token.charAt(0)); 50 } 51 else if (token.charAt(0) == '*' || token.charAt(0) == '/') { 52 // Process all *, / in the top of the operator stack 53 while (!operatorStack.isEmpty() && 54 (operatorStack.peek() == '*' || 55 operatorStack.peek() == '/')) { 56 processAnOperator(operandStack, operatorStack); 57 } 58 59 // Push the * or / operator into the operator stack 60 operatorStack.push(token.charAt(0)); 61 } 62 else if(token.trim().charAt(0) =='(') { 63 operatorStack.push('('); // Push '(' to stack 64 }

check usage

evaluate expression

exception

operandStack

operatorStack

prepare for extraction

extract tokens

process tokens

* or / scanned

( scanned

+ or - scanned

M20_LIAN1878_11_GE_C20.indd 827 12/29/17 4:33 PM

828 Chapter 20 Lists, Stacks, Queues, and Priority Queues

65 else if (token.trim().charAt(0) ==')') { 66 // Process all the operators in the stack until seeing '(' 67 while (operatorStack.peek() != '(') { 68 processAnOperator(operandStack, operatorStack); 69 } 70 71 operatorStack.pop(); // Pop the '(' symbol from the stack 72 } 73 else { // An operand scanned 74 // Push an operand to the stack 75 operandStack.push(new Integer(token)); 76 } 77 } 78 79 // Phase 2: Process all the remaining operators in the stack 80 while (!operatorStack.isEmpty()) { 81 processAnOperator(operandStack, operatorStack); 82 } 83 84 // Return the result 85 return operandStack.pop(); 86 } 87 88 /** Process one operator: Take an operator from operatorStack and 89 * apply it on the operands in the operandStack */ 90 public static void processAnOperator( 91 Stack<Integer> operandStack, Stack<Character> operatorStack) { 92 char op = operatorStack.pop(); 93 int op1 = operandStack.pop(); 94 int op2 = operandStack.pop(); 95 if (op == '+') 96 operandStack.push(op2 + op1); 97 else if (op == '−') 98 operandStack.push(op2 − op1); 99 else if (op == '*') 100 operandStack.push(op2 * op1); 101 else if (op == '/') 102 operandStack.push(op2 / op1); 103 } 104 105 public static String insertBlanks(String s) { 106 String result = ""; 107 108 for (int i = 0; i < s.length(); i++) { 109 if (s.charAt(i) == '(' || s.charAt(i) == ')' || 110 s.charAt(i) == '+' || s.charAt(i) == '−' || 111 s.charAt(i) == '*' || s.charAt(i) == '/') 112 result += " " + s.charAt(i) + " "; 113 else 114 result += s.charAt(i); 115 } 116 117 return result; 118 } 119 }

You can use the GenericStack class provided by the book, or the java.util.Stack class defined in the Java API for creating stacks. This example uses the java.util.Stack class. The program will work if it is replaced by GenericStack.

) scanned

an operand scanned

clear operatorStack

return result

process +

process -

process *

process /

insert blanks

M20_LIAN1878_11_GE_C20.indd 828 12/29/17 4:33 PM

The program takes an expression as a command-line argument in one string. The evaluateExpression method creates two stacks, operandStack and

operatorStack (lines 23 and 26), and extracts operands, operators, and parentheses delim- ited by space (lines 29–32). The insertBlanks method is used to ensure that operands, operators, and parentheses are separated by at least one blank (line 29).

The program scans each token in the for loop (lines 35–77). If a token is empty, skip it (line 37). If a token is an operand, push it to operandStack (line 75). If a token is a + or − operator (line 38), process all the operators from the top of operatorStack, if any (lines 40–46), and push the newly scanned operator into the stack (line 49). If a token is a * or / operator (line 51), process all the * and / operators from the top of operatorStack, if any (lines 53–57), and push the newly scanned operator to the stack (line 60). If a token is a ( symbol (line 62), push it into operatorStack. If a token is a ) symbol (line 65), process all the operators from the top of operatorStack until seeing the ) symbol (lines 67–69) and pop the ) symbol from the stack.

After all tokens are considered, the program processes the remaining operators in operatorStack (lines 80–82).

The processAnOperator method (lines 90–103) processes an operator. The method pops the operator from operatorStack (line 92) and pops two operands from operandStack (lines 93 and 94). Depending on the operator, the method performs an operation and pushes the result of the operation back to operandStack (lines 96, 98, 100, and 102).

20.11.1 Can the EvaluateExpression program evaluate the following expressions "1 + 2", "1 + 2", "(1) + 2", "((1)) + 2", and "(1 + 2)"?

20.11.2 Show the change of the contents in the stacks when evaluating "3 + (4 + 5) * (3 + 5) + 4 * 5" using the EvaluateExpression program.

20.11.3 If you enter an expression "4 + 5 5 5", the program will display 10. How do you fix this problem?

Point Check

Key terms collection 798 comparator 809 convenience abstract class 799 data structure 798

linked list 806 list 798 priority queue 798 queue 798

ChapTer summary 1. The Collection interface defines the common operations for lists, vectors, stacks,

queues, priority queues, and sets.

2. Each collection is Iterable. You can obtain its Iterator object to traverse all the elements in the collection.

3. All the concrete classes except PriorityQueue in the Java Collections Framework implement the Cloneable and Serializable interfaces. Thus, their instances can be cloned and serialized.

4. A list stores an ordered collection of elements. To allow duplicate elements to be stored in a collection, you need to use a list. A list not only can store duplicate elements but also allows the user to specify where they are stored. The user can access elements by an index.

Chapter Summary 829

M20_LIAN1878_11_GE_C20.indd 829 12/29/17 4:33 PM

830 Chapter 20 Lists, Stacks, Queues, and Priority Queues

5. Two types of lists are supported: ArrayList and LinkedList. ArrayList is a resiz- able-array implementation of the List interface. All the methods in ArrayList are defined in List. LinkedList is a linked-list implementation of the List interface. In addition to implementing the List interface, this class provides the methods for retriev- ing, inserting, and removing elements from both ends of the list.

6. Comparator can be used to compare the objects of a class that doesn’t implement Comparable.

7. The Vector class extends the AbstractList class. Starting with Java 2, Vector has been the same as ArrayList, except that the methods for accessing and modifying the vector are synchronized. The Stack class extends the Vector class and provides several methods for manipulating the stack.

8. The Queue interface represents a queue. The PriorityQueue class implements Queue for a priority queue.

Quiz Answer the quiz for this chapter online at the book Companion Website.

programming exerCiSeS

Sections 20.2–20.7 *20.1 (Display words in descending alphabetical order) Write a program that reads

words from a text file and displays all the words (duplicates allowed) in descending alphabetical order. The words must start with a letter. The text file is passed as a command-line argument.

*20.2 (Store numbers in a linked list) Write a program that lets the user enter num- bers from a graphical user interface and displays them in a text area, as shown in Figure 20.17a. Use a linked list to store the numbers. Do not store dupli- cate numbers. Add the buttons Sort, Shuffle, and Reverse to sort, shuffle, and reverse the list.

Figure 20.17 (a) The numbers are stored in a list and displayed in the text area. Source: Copyright © 1995–2016 Oracle and/or its affiliates. All rights reserved. Used with permission. (b) The colliding balls are combined.

(a) (b)

M20_LIAN1878_11_GE_C20.indd 830 12/29/17 4:33 PM

*20.3 (Guessing the capitals) Rewrite Programming Exercise 8.37 to store the pairs of states and capitals so that the questions are displayed randomly.

*20.4 (Implement Comparable) Implement a new class that implements Comparator of GregorianCalendar class to be able sort the calendar in increasing order based on day, month, and year, in that order. Write a method to dis- play the GregorianCalendar instance in “dd-MMM-yyyy” format using SimpleDateFormat class. Write a test program with 10 GregorianCalendar instances and display the results after the sort.

***20.5 (Combine colliding bouncing balls) The example in Section 20.8 displays multiple bouncing balls. Extend the example to detect collisions. Once two balls collide, remove the later ball that was added to the pane and add its radius to the other ball, as shown in Figure 20.17b. Use the Suspend button to suspend the animation, and the Resume button to resume the anima- tion. Add a mouse-pressed handler that removes a ball when the mouse is pressed on the ball.

20.6 (Use iterators on ArrayList) Write a test program that stores 10 million integers in ascending order (i.e., 1,2, . . ., 10m) in an ArrayList, displays the execution time taken to traverse the list, and searches for the 10 millionth element using the get(index) vs using the iterator method.

***20.7 (Game: hangman) Programming Exercise 7.35 presents a console version of the popular hangman game. Write a GUI program that lets a user play the game. The user guesses a word by entering one letter at a time, as shown in Figure 20.18. If the user misses seven times, a hanging man swings. Once a word is finished, the user can press the Enter key to continue to guess another word.

**20.8 (Game: lottery) Revise Programming Exercise 3.15 to add an additional $2,000 award if two digits from the user input are in the lottery number. (Hint: Sort the three digits in the lottery number and three digits in the user input into two lists, and use the Collection’s containsAll method to check whether the two digits in the user input are in the lottery number.)

Sections 20.8–20.10 ***20.9 (Remove the largest ball first) Modify Listing 20.10, MultipleBallApp

.java to assign a random radius between 2 and 20 when a ball is created. When the - button is clicked, one of largest balls is removed.

20.10 (Perform set operations on priority queues) Given two stacks of textbooks of the following subjects {“Chemistry”, “Mathematics”, “Biology”, “ English”} and {“Biology”, “English”, “Geography”, “Physics”}, find the subjects that are (1) only present in the first stack; (2) only present in the second stack; (3) present in both stacks.

*20.11 (Remove Consecutive Integers) Write a program that reads 10 integers and displays them in the reverse of the order in which they were read. If two con- secutive numbers are identical, then only display one of them. Implement your program using only stack and not arrays or queues.

20.12 (Create Stack from list) Define a class MyStack that extends Stack to be able to have its constructor use a list of objects instead of pushing each object individually.

Programming Exercises 831

M20_LIAN1878_11_GE_C20.indd 831 12/29/17 4:33 PM

832 Chapter 20 Lists, Stacks, Queues, and Priority Queues

M20_LIAN1878_11_GE_C20.indd 832 12/29/17 4:33 PM

Write a program to evaluate postfix expressions. Pass the expression as a command-line argument in one string.

**20.13 (Postfix notation) Postfix notation is a way of writing expressions without using parentheses. For example, the expression (1 + 2) * 3 would be written as 1 2 + 3 *. A postfix expression is evaluated using a stack. Scan a postfix expression from left to right. A variable or constant is pushed into the stack. When an operator is encountered, apply the operator with the top two operands in the stack and replace the two operands with the result. The following diagram shows how to evaluate 1 2 + 3 *.

1 1 3

1 2 + 3 * 1 2 + 3 *

scanned scanned

1 2 + 3 *

scanned

1 2 + 3 *

scanned

1 2 + 3 *

scanned

Programming Exercises 833

**20.14 (Postfix notation) Postfix notation is a way of writing expressions without using parentheses. For example, the expression (1 + 2) * 3 would be written as 1 2 + 3 *. A postfix expression is evaluated using a stack. Scan a postfix expression from left to right. A variable or constant is pushed into the stack. When an operator is encountered, apply the operator with the top two oper- ands in the stack and replace the two operands with the result. The following diagram shows how to evaluate 1 2 + 3 *:

1 1 3

1 2 + 3 * 1 2 + 3 *

scanned scanned

1 2 + 3 *

scanned

1 2 + 3 *

scanned

1 2 + 3 *

scanned

Write a program to evaluate postfix expressions. Pass the expression as a com- mand-line argument in one string.

M20_LIAN1878_11_GE_C20.indd 833 12/29/17 4:33 PM

834 Chapter 20 Lists, Stacks, Queues, and Priority Queues

***20.15 (Game: the 24-point card game) Improve Programming Exercise 20.13 to enable the computer to display the expression if one exists, as shown in Figure 20.19. Otherwise, report that the expression does not exist. Place the label for verification result at the bottom of the UI. The expression must use all four cards and evaluated to 24.

Figure 20.19 The program can automatically find a solution if one exists. Source: Fotolia.

20.16 (Game: the 24-point card game) Improve Programming Exercise 20.15 by adding another expression verifier text box below the card images that accept prefix notations instead of infix notations. You have to write your own prefix evaluator.

***20.17 (Game: the 24-point card game) This exercise is a variation of the 24-point card game described in Programming Exercise 20.13. Write a program to check whether there is a 24-point solution for the four specified numbers. The program lets the user enter four values, each between 1 and 13, as shown in Figure 20.20. The user can then click the Solve button to display the solution or display “No solution” if none exists:

M20_LIAN1878_11_GE_C20.indd 834 12/29/17 4:33 PM

*20.18 (Directory size) Listing 18.10, DirectorySize.java, gives a recursive method for finding a directory size. Rewrite this method without using recursion. Your program should use a queue to store the subdirectories under a directory. The algorithm can be described as follows:

long getSize(File directory) { long size = 0; add directory to the queue;

while (queue is not empty) { Remove an item from the queue into t; if (t is a file) size += t.length(); else add all the files and subdirectories under t into the queue; }

return size; }

*20.19 (Use Comparator) Write the following generic method using selection sort and a comparator:

public static <E> void selectionSort(E[] list, Comparator<? super E> comparator)

Write a test program that creates an array of 10 GeometricObjects and invokes this method using the GeometricObjectComparator introduced in Listing 20.4 to sort the elements. Display the sorted elements. Use the following statement to create the array:

GeometricObject[] list1 = {new Circle(5), new Rectangle(4, 5), new Circle(5.5), new Rectangle(2.4, 5), new Circle(0.5), new Rectangle(4, 65), new Circle(4.5), new Rectangle(4.4, 1), new Circle(6.5), new Rectangle(4, 5)};

*20.20 (Nonrecursive Tower of Hanoi) Implement the moveDisks method in Listing 18.8 using a stack instead of using recursion.

*20.21 (Use Comparator) Write the following generic method using selection sort and a comparator:

public static <E> void selectionSort(E[] list, Comparator<? super E> comparator)

Write a test program that creates an array of 10 GeometricObjects and invokes this method using the GeometricObjectComparator introduced in

Programming Exercises 835

M20_LIAN1878_11_GE_C20.indd 835 12/29/17 4:33 PM

836 Chapter 20 Lists, Stacks, Queues, and Priority Queues

Listing 20.5 to sort the elements. Display the sorted elements. Use the following statement to create the array:

Also in the same program, write the code that sorts six strings by their last character. Use the following statement to create the array:

String[] list2 = {"red", "blue", "green", "yellow", "orange", "pink"};

*20.22 (Nonrecursive Tower of Hanoi) Implement the moveDisks method in Listing 18.8 using a stack instead of using recursion.

20.23 (Evaluate expression) Modify Listing 20.12, EvaluateExpression class to make all operators (+−*/) have equal precedence. For example, 4 + 3 − 2 * 10 is 50. Write a test program that prompts the user to enter an expression and displays the result.

M20_LIAN1878_11_GE_C20.indd 836 12/29/17 4:33 PM

Objectives ■■ To store unordered, nonduplicate elements using a set (§21.2).

■■ To explore how and when to use HashSet (§21.2.1), LinkedHashSet (§21.2.2), or TreeSet (§21.2.3) to store a set of elements.

■■ To compare the performance of sets and lists (§21.3).

■■ To use sets to develop a program that counts the keywords in a Java source file (§21.4).

■■ To tell the differences between Collection and Map and describe when and how to use HashMap, LinkedHashMap, or TreeMap to store values associated with keys (§21.5).

■■ To use maps to develop a program that counts the occurrence of the words in a text (§21.6).

■■ To obtain singleton sets, lists, and maps and unmodifiable sets, lists, and maps, use the static methods in the Collections class (§21.7).

Sets and Maps

Chapter

M21_LIAN1878_11_GE_C21.indd 837 12/29/17 4:35 PM

838 Chapter 21 Sets and Maps

21.1 Introduction A set is an efficient data structure for storing and processing nonduplicate elements. A map is like a dictionary that provides a quick lookup to retrieve a value using a key.

The “No-Fly” list is a list, created and maintained by the U.S. government’s Terrorist Screen- ing Center, of people who are not permitted to board a commercial aircraft for travel in or out of the United States. Suppose we need to write a program that checks whether a person is on the No-Fly list. You can use a list to store names in the No-Fly list. However, a more efficient data structure for this application is a set.

Suppose your program also needs to store detailed information about terrorists in the No-Fly list. The detailed information such as gender, height, weight, and nationality can be retrieved using the name as the key. A map is an efficient data structure for such a task.

This chapter introduces sets and maps in the Java Collections Framework.

21.2 Sets You can create a set using one of its three concrete classes: HashSet, LinkedHashSet, or TreeSet.

The Set interface extends the Collection interface, as shown in Figure 20.1. It does not introduce new methods or constants, but it stipulates that an instance of Set contains no dupli- cate elements. The concrete classes that implement Set must ensure that no duplicate elements can be added to the set. That is, no two elements e1 and e2 can be in the set such that e1.equals(e2) is true.

The AbstractSet class extends AbstractCollection and partially implements Set. The AbstractSet class provides concrete implementations for the equals method and the hashCode method. The hash code of a set is the sum of the hash codes of all the elements in the set. Since the size method and iterator method are not implemented in the Abstract- Set class, AbstractSet is an abstract class.

Three concrete classes of Set are HashSet, LinkedHashSet, and TreeSet, as shown in Figure 21.1.

21.2.1 HashSet The HashSet class is a concrete class that implements Set. You can create an empty hash set using its no-arg constructor, or create a hash set from an existing collection. By default, the initial capacity is 16 and the load factor is 0.75. If you know the size of your set, you can specify the initial capacity and load factor in the constructor. Otherwise, use the default setting. The load factor is a value between 0.0 and 1.0.

The load factor measures how full the set is allowed to be before its capacity is increased. When the number of elements exceeds the product of the capacity and load factor, the capacity is auto- matically doubled. For example, if the capacity is 16 and load factor is 0.75, the capacity will be doubled to 32 when the size reaches 12 (16 * 0.75 = 12). A higher load factor decreases the space costs but increases the search time. Generally, the default load factor 0.75 is a good trade-off between time and space costs. We will discuss more on the load factor in Chapter 27, Hashing.

A HashSet can be used to store duplicate-free elements. For efficiency, objects added to a hash set need to implement the hashCode method in a manner that properly disperses the hash code. The hashCode method is defined in the Object class. The hash codes of two objects must be the same if the two objects are equal. Two unequal objects may have the same hash code, but you should implement the hashCode method to avoid too many such cases. Most of the classes in the Java API implement the hashCode method. For example, the hashCode in the Integer class returns its int value. The hashCode in the Character class returns the Unicode of the character. The hashCode in the String class returns s0 * 31

(n - 1) + s1 * 31(n - 2) + g + sn - 1, where s i is s.charAt(i).

Point Key

why set?

why map?

Point Key

set

no duplicates

AbstractSet

hash set

load factor

hashCode()

M21_LIAN1878_11_GE_C21.indd 838 12/29/17 4:35 PM

21.2 Sets 839

Figure 21.1 The Java Collections Framework provides three concrete set classes.

java.util.AbstractSet<E>

+HashSet()

+HashSet(c: Collection<? extends E>)

+HashSet(initialCapacity: int)

+HashSet(initialCapacity: int, loadFactor: �oat)

java.util.TreeSet<E>

+TreeSet()

+TreeSet(c: Collection<? extends E>)

+TreeSet(comparator: Comparator<?

super E>)

+TreeSet(s: SortedSet<E>)

+first(): E

+last(): E

+headSet(toElement: E): SortedSet<E>

+tailSet(fromElement: E): SortedSet<E>

«interface» java.util.Set<E>

java.util.HashSet<E>

«interface» java.util.SortedSet<E>

+pollFirst(): E

+pollLast(): E

+lower(e: E): E

+higher(e: E):E

+�oor(e: E): E

+ceiling(e: E): E

«interface» java.util.NavigableSet<E>

«interface» java.util.Collection<E>

java.util.LinkedHashSet<E>

+LinkedHashSet()

+LinkedHashSet(c: Collection<? extends E>)

+LinkedHashSet(initialCapacity: int)

+LinkedHashSet(initialCapacity: int, loadFactor: �oat)

Listing 21.1 gives a program that creates a hash set to store strings and uses a foreach loop and a forEach method to traverse the elements in the set.

lisTing 21.1 TestHashSet.java 1 import java.util.*; 2 3 public class TestHashSet { 4 public static void main(String[] args) { 5 // Create a hash set 6 Set<String> set = new HashSet<>(); 7 8 // Add strings to the set 9 set.add("London"); 10 set.add("Paris"); 11 set.add("New York"); 12 set.add("San Francisco"); 13 set.add("Beijing");

create a set

add element

M21_LIAN1878_11_GE_C21.indd 839 12/29/17 4:35 PM

840 Chapter 21 Sets and Maps

14 set.add("New York"); 15 16 System.out.println(set); 17 18 // Display the elements in the hash set 19 for (String s: set) { 20 System.out.print(s.toUpperCase() + " "); 21 } 22 23 // Process the elements using a forEach method 24 System.out.println(); 25 set.forEach(e −> System.out.print(e.toLowerCase() + " ")); 26 } 27 }

traverse elements

forEach method

[San Francisco, New York, Paris, Beijing, London] SAN FRANCISCO NEW YORK PARIS BEIJING LONDON

The strings are added to the set (lines 9–14). New York is added to the set more than once, but only one string is stored because a set does not allow duplicates.

As shown in the output, the strings are not stored in the order in which they are inserted into the set. There is no particular order for the elements in a hash set. To impose an order on them, you need to use the LinkedHashSet class, which is introduced in the next section.

Recall that the Collection interface extends the Iterable interface, so the elements in a set are iterable. A foreach loop is used to traverse all the elements in the set (lines 19–21). You can also use a forEach method to process each element in a set (line 25).

Since a set is an instance of Collection, all methods defined in Collection can be used for sets. Listing 21.2 gives an example that applies the methods in the Collection interface on sets.

Listing 21.2 TestMethodsInCollection.java 1 public class TestMethodsInCollection { 2 public static void main(String[] args) { 3 // Create set1 4 java.util.Set<String> set1 = new java.util.HashSet<>(); 5 6 // Add strings to set1 7 set1.add("London"); 8 set1.add("Paris"); 9 set1.add("New York"); 10 set1.add("San Francisco"); 11 set1.add("Beijing"); 12 13 System.out.println("set1 is " + set1); 14 System.out.println(set1.size() + " elements in set1"); 15 16 // Delete a string from set1 17 set1.remove("London"); 18 System.out.println("\nset1 is " + set1); 19 System.out.println(set1.size() + " elements in set1"); 20 21 // Create set2 22 java.util.Set<String> set2 = new java.util.HashSet<>(); 23 24 // Add strings to set2 25 set2.add("London"); 26 set2.add("Shanghai"); 27 set2.add("Paris");

create a set

add element

get size

remove element

create a set

add element

M21_LIAN1878_11_GE_C21.indd 840 12/29/17 4:35 PM

21.2 Sets 841

28 System.out.println("\nset2 is " + set2); 29 System.out.println(set2.size() + " elements in set2"); 30 31 System.out.println("\nIs Taipei in set2? " 32 + set2.contains("Taipei")); 33 34 set1.addAll(set2); 35 System.out.println("\nAfter adding set2 to set1, set1 is " 36 + set1); 37 38 set1.removeAll(set2); 39 System.out.println("After removing set2 from set1, set1 is " 40 + set1); 41 42 set1.retainAll(set2); 43 System.out.println("After retaining common elements in set2 " 44 + "and set2, set1 is " + set1); 45 } 46 }

contains element?

addAll

removeAll

retainAll

set1 is [San Francisco, New York, Paris, Beijing, London] 5 elements in set1

set1 is [San Francisco, New York, Paris, Beijing] 4 elements in set1

set2 is [Shanghai, Paris, London] 3 elements in set2

Is Taipei in set2? false

After adding set2 to set1, set1 is [San Francisco, New York, Shanghai, Paris, Beijing, London]

After removing set2 from set1, set1 is [San Francisco, New York, Beijing]

After retaining common elements in set1 and set2, set1 is []

The program creates two sets (lines 4 and 22). The size() method returns the number of the elements in a set (line 14). Line 17

set1.remove("London");

removes London from set1. The contains method (line 32) checks whether an element is in the set. Line 34

set1.addAll(set2);

adds set2 to set1. Therefore, set1 becomes [San Francisco, New York, Shanghai, Paris, Beijing, London].

Line 38

set1.removeAll(set2);

removes set2 from set1. Thus, set1 becomes [San Francisco, New York, Beijing]. Line 42

set1.retainAll(set2);

M21_LIAN1878_11_GE_C21.indd 841 12/29/17 4:35 PM

842 Chapter 21 Sets and Maps

retains the common elements in set1 and set2. Since set1 and set2 have no common elements, set1 becomes empty.

21.2.2 LinkedHashSet LinkedHashSet extends HashSet with a linked-list implementation that supports an ordering of the elements in the set. The elements in a HashSet are not ordered, but the elements in a LinkedHashSet can be retrieved in the order in which they were inserted into the set. A LinkedHashSet can be created by using one of its four constructors, as shown in Figure 21.1. These constructors are similar to the constructors for HashSet.

Listing 21.3 gives a test program for LinkedHashSet. The program simply replaces HashSet by LinkedHashSet in Listing 21.1.

Listing 21.3 TestLinkedHashSet.java 1 import java.util.*; 2 3 public class TestLinkedHashSet { 4 public static void main(String[] args) { 5 // Create a hash set 6 Set<String> set = new LinkedHashSet<>(); 7 8 // Add strings to the set 9 set.add("London"); 10 set.add("Paris"); 11 set.add("New York"); 12 set.add("San Francisco"); 13 set.add("Beijing"); 14 set.add("New York"); 15 16 System.out.println(set); 17 18 // Display the elements in the hash set 19 for (String element: set) 20 System.out.print(element.toLowerCase() + " "); 21 } 22 }

linked hash set

create linked hash set

add element

display elements

[London, Paris, New York, San Francisco, Beijing] london paris new york san francisco beijing

A LinkedHashSet is created in line 6. As shown in the output, the strings are stored in the order in which they are inserted. Since LinkedHashSet is a set, it does not store duplicate elements.

The LinkedHashSet maintains the order in which the elements are inserted. To impose a different order (e.g., increasing or decreasing order), you can use the TreeSet class, which is introduced in the next section.

Tip If you don’t need to maintain the order in which the elements are inserted, use HashSet, which is more efficient than LinkedHashSet.

21.2.3 TreeSet As shown in Figure 21.1, SortedSet is a subinterface of Set, which guarantees that the elements in the set are sorted. In addition, it provides the methods first() and last() for returning the first and last elements in the set, and headSet(toElement) and

M21_LIAN1878_11_GE_C21.indd 842 12/29/17 4:35 PM

21.2 Sets 843

tailSet(fromElement) for returning a portion of the set whose elements are less than toElement and greater than or equal to fromElement, respectively.

NavigableSet extends SortedSet to provide navigation methods lower(e), floor(e), ceiling(e), and higher(e) that return elements, respectively, less than, less than or equal, greater than or equal, and greater than a given element and return null if there is no such element. The pollFirst() and pollLast() methods remove and return the first and last element in the tree set, respectively.

TreeSet implements the SortedSet interface. To create a TreeSet, use a constructor, as shown in Figure 21.1. You can add objects into a tree set as long as they can be compared with each other.

As discussed in Section 20.5, the elements can be compared in two ways: using the Comparable interface or the Comparator interface.

Listing 21.4 gives an example of ordering elements using the Comparable interface. The preceding example in Listing 21.3 displays all the strings in their insertion order. This example rewrites the preceding example to display the strings in alphabetical order using the TreeSet class.

Listing 21.4 TestTreeSet.java 1 import java.util.*; 2 3 public class TestTreeSet { 4 public static void main(String[] args) { 5 // Create a hash set 6 Set<String> set = new HashSet<>(); 7 8 // Add strings to the set 9 set.add("London"); 10 set.add("Paris"); 11 set.add("New York"); 12 set.add("San Francisco"); 13 set.add("Beijing"); 14 set.add("New York"); 15 16 TreeSet<String> treeSet = new TreeSet<>(set); 17 System.out.println("Sorted tree set: " + treeSet); 18 19 // Use the methods in SortedSet interface 20 System.out.println("first(): " + treeSet.first()); 21 System.out.println("last(): " + treeSet.last()); 22 System.out.println("headSet(\"New York\"): " + 23 treeSet.headSet("New York")); 24 System.out.println("tailSet(\"New York\"): " + 25 treeSet.tailSet("New York")); 26 27 // Use the methods in NavigableSet interface 28 System.out.println("lower(\"P\"): " + treeSet.lower("P")); 29 System.out.println("higher(\"P\"): " + treeSet.higher("P")); 30 System.out.println("floor(\"P\"): " + treeSet.floor("P")); 31 System.out.println("ceiling(\"P\"): " + treeSet.ceiling("P")); 32 System.out.println("pollFirst(): " + treeSet.pollFirst()); 33 System.out.println("pollLast(): " + treeSet.pollLast()); 34 System.out.println("New tree set: " + treeSet); 35 } 36 }

tree set

create hash set

create tree set

display elements

Sorted tree set: [Beijing, London, New York, Paris, San Francisco]

first(): Beijing last(): San Francisco

M21_LIAN1878_11_GE_C21.indd 843 12/29/17 4:35 PM

844 Chapter 21 Sets and Maps

The example creates a hash set filled with strings, then creates a tree set for the same strings. The strings are sorted in the tree set using the compareTo method in the Comparable interface.

The elements in the set are sorted once you create a TreeSet object from a HashSet object using new TreeSet<>(set) (line 16). You may rewrite the program to cre- ate an instance of TreeSet using its no-arg constructor and add the strings into the TreeSet object.

treeSet.first() returns the first element in treeSet (line 20) and treeSet.last() returns the last element in treeSet (line 21). treeSet.headSet("New York") returns the elements in treeSet before New York (lines 22–23). treeSet.tailSet("New York") returns the elements in treeSet after New York, including New York (lines 24–25).

treeSet.lower("P") returns the largest element less than P in treeSet (line 28). treeSet.higher("P") returns the smallest element greater than P in treeSet (line 29). treeSet.floor("P") returns the largest element less than or equal to P in treeSet (line 30). treeSet.ceiling("P") returns the smallest element greater than or equal to P in treeSet (line 31). treeSet.pollFirst() removes the first element in treeSet and returns the removed element (line 32). treeSet.pollLast() removes the last element in treeSet and returns the removed element (line 33).

Note All the concrete classes in Java Collections Framework (see Figure 20.1) have at least two constructors. One is the no-arg constructor that constructs an empty collection. The other constructs instances from a collection. Thus the TreeSet class has the constructor TreeSet(Collection c) for constructing a TreeSet from a collection c. In this example, new TreeSet<>(set) creates an instance of TreeSet from the collection set.

Tip If you don’t need to maintain a sorted set when updating a set, you should use a hash set because it takes less time to insert and remove elements in a hash set. When you need a sorted set, you can create a tree set from the hash set.

If you create a TreeSet using its no-arg constructor, the compareTo method is used to com- pare the elements in the set, assuming the class of the elements implements the Comparable interface. To use a comparator, you have to use the constructor TreeSet(Comparator comparator) to create a sorted set that uses the compare method in the comparator to order the elements in the set.

Listing 21.5 gives a program that demonstrates how to sort elements in a tree set using the Comparator interface.

Listing 21.5 TestTreeSetWithComparator.java 1 import java.util.*; 2

headSet("New York"): [Beijing, London] tailSet("New York"): [New York, Paris, San Francisco] lower("P"): New York higher("P"): Paris floor("P"): New York ceiling("P"): Paris pollFirst(): Beijing pollLast(): San Francisco New tree set: [London, New York, Paris]

M21_LIAN1878_11_GE_C21.indd 844 12/29/17 4:35 PM

21.2 Sets 845

3 public class TestTreeSetWithComparator { 4 public static void main(String[] args) { 5 // Create a tree set for geometric objects using a comparator 6 Set<GeometricObject> set = 7 new TreeSet<>(new GeometricObjectComparator()); 8 set.add(new Rectangle(4, 5)); 9 set.add(new Circle(40)); 10 set.add(new Circle(40)); 11 set.add(new Rectangle(4, 1)); 12 13 // Display geometric objects in the tree set 14 System.out.println("A sorted set of geometric objects"); 15 for (GeometricObject element: set) 16 System.out.println("area = " + element.getArea()); 17 } 18 }

tree set

display elements

A sorted set of geometric objects area = 4.0 area = 20.0 area = 5021.548245743669

The GeometricObjectComparator class is defined in Listing 20.4. The program creates a tree set of geometric objects using the GeometricObjectComparator for comparing the elements in the set (lines 6 and 7).

The Circle and Rectangle classes were defined in Section 13.2, Abstract Classes. They are all subclasses of GeometricObject. They are added to the set (lines 8–11).

Two circles of the same radius are added to the tree set (lines 9 and 10), but only one is stored because the two circles are equal (determined by the comparator in this case) and the set does not allow duplicates.

21.2.1 How do you create an instance of Set? How do you insert a new element in a set? How do you remove an element from a set? How do you find the size of a set?

21.2.2 If two objects o1 and o2 are equal, what is o1.equals(o2) and o1.hashCode() == o2.hashCode()?

21.2.3 What are the differences among HashSet, LinkedHashSet, and TreeSet? 21.2.4 How do you traverse the elements in a set? 21.2.5 How do you sort the elements in a set using the compareTo method in the

Comparable interface? How do you sort the elements in a set using the Comparator interface? What would happen if you added an element that could not be compared with the existing elements in a tree set?

21.2.6 Suppose set1 is a set that contains the strings red, yellow, and green and that set2 is another set that contains the strings red, yellow, and blue. Answer the following questions:

■■ What are in set1 and set2 after executing set1.addAll(set2)?

■■ What are in set1 and set2 after executing set1.add(set2)?

■■ What are in set1 and set2 after executing set1.removeAll(set2)?

■■ What are in set1 and set2 after executing set1.remove(set2)?

Point Check

M21_LIAN1878_11_GE_C21.indd 845 12/29/17 4:35 PM

846 Chapter 21 Sets and Maps

■■ What are in set1 and set2 after executing set1.retainAll(set2)?

■■ What is in set1 after executing set1.clear()?

21.2.7 Show the output of the following code: import java.util.*;

public class Test { public static void main(String[] args) { LinkedHashSet<String> set1 = new LinkedHashSet<>(); set1.add("New York"); LinkedHashSet<String> set2 = set1; LinkedHashSet<String> set3 = (LinkedHashSet<String>)(set1.clone()); set1.add("Atlanta"); System.out.println("set1 is " + set1); System.out.println("set2 is " + set2); System.out.println("set3 is " + set3); set1.forEach(e −> System.out.print(e + " ")); } }

21.2.8 Show the output of the following code: Set<String> set = new LinkedHashSet<>(); set.add("ABC"); set.add("ABD"); System.out.println(set);

21.2.9 What will the output be if lines 6–7 in Listing 21.5 are replaced by the following code:

Set<GeometricObject> set = new HashSet<>();

21.2.10 Show the output of the following code: Set<String> set = new TreeSet<>( Comparator.comparing(String::length)); set.add("ABC"); set.add("ABD"); System.out.println(set);

21.3 Comparing the Performance of Sets and Lists Sets are more efficient than lists for storing nonduplicate elements. Lists are useful for accessing elements through the index.

The elements in a list can be accessed through the index. However, sets do not support indexing because the elements in a set are unordered. To traverse all elements in a set, use a foreach loop. We now conduct an interesting experiment to test the performance of sets and lists. Listing 21.6 gives a program that shows the execution time of (1) test- ing whether an element is in a hash set, linked hash set, tree set, array list, or linked list and (2) removing elements from a hash set, linked hash set, tree set, array list, and linked list.

Point Key

M21_LIAN1878_11_GE_C21.indd 846 12/29/17 4:35 PM

21.3 Comparing the Performance of Sets and Lists 847

Listing 21.6 SetListPerformanceTest.java 1 import java.util.*; 2 3 public class SetListPerformanceTest { 4 static final int N = 50000; 5 6 public static void main(String[] args) { 7 // Add numbers 0, 1, 2, ..., N − 1 to the array list 8 List<Integer> list = new ArrayList<>(); 9 for (int i = 0; i < N; i++) 10 list.add(i); 11 Collections.shuffle(list); // Shuffle the array list 12 13 // Create a hash set, and test its performance 14 Collection<Integer> set1 = new HashSet<>(list); 15 System.out.println("Member test time for hash set is " + 16 getTestTime(set1) + " milliseconds"); 17 System.out.println("Remove element time for hash set is " + 18 getRemoveTime(set1) + " milliseconds"); 19 20 // Create a linked hash set, and test its performance 21 Collection<Integer> set2 = new LinkedHashSet<>(list); 22 System.out.println("Member test time for linked hash set is " + 23 getTestTime(set2) + " milliseconds"); 24 System.out.println("Remove element time for linked hash set is " 25 + getRemoveTime(set2) + " milliseconds"); 26 27 // Create a tree set, and test its performance 28 Collection<Integer> set3 = new TreeSet<>(list); 29 System.out.println("Member test time for tree set is " + 30 getTestTime(set3) + " milliseconds"); 31 System.out.println("Remove element time for tree set is " + 32 getRemoveTime(set3) + " milliseconds"); 33 34 // Create an array list, and test its performance 35 Collection<Integer> list1 = new ArrayList<>(list); 36 System.out.println("Member test time for array list is " + 37 getTestTime(list1) + " milliseconds"); 38 System.out.println("Remove element time for array list is " + 39 getRemoveTime(list1) + " milliseconds"); 40 41 // Create a linked list, and test its performance 42 Collection<Integer> list2 = new LinkedList<>(list); 43 System.out.println("Member test time for linked list is " + 44 getTestTime(list2) + " milliseconds"); 45 System.out.println("Remove element time for linked list is " + 46 getRemoveTime(list2) + " milliseconds"); 47 } 48 49 public static long getTestTime(Collection<> c) { 50 long startTime = System.currentTimeMillis(); 51 52 // Test if a number is in the collection 53 for (int i = 0; i < N; i++) 54 c.contains((int)(Math.random() * 2 * N)); 55 56 return System.currentTimeMillis() − startTime;

create test data

shuffle

a hash set

a linked hash set

a tree set

an array list

a linked list

start time

test membership

return execution time

M21_LIAN1878_11_GE_C21.indd 847 12/29/17 4:35 PM

848 Chapter 21 Sets and Maps

57 } 58 59 public static long getRemoveTime(Collection<Integer> c) { 60 long startTime = System.currentTimeMillis(); 61 62 for (int i = 0; i < N; i++) 63 c.remove(i); 64 65 return System.currentTimeMillis() − startTime; 66 } 67 }

remove from container

return execution time

Member test time for hash set is 20 milliseconds Remove element time for hash set is 27 milliseconds Member test time for linked hash set is 27 milliseconds Remove element time for linked hash set is 26 milliseconds Member test time for tree set is 47 milliseconds Remove element time for tree set is 34 milliseconds Member test time for array list is 39802 milliseconds Remove element time for array list is 16196 milliseconds Member test time for linked list is 52197 milliseconds Remove element time for linked list is 14870 milliseconds

The program creates a list for numbers from 0 to N−1 (for N = 50000) (lines 8–10) and shuf- fles the list (line 11). From this list, the program creates a hash set (line 14), a linked hash set (line 21), a tree set (line 28), an array list (line 35), and a linked list (line 42). The program obtains the execution time for testing whether a number is in the hash set (line 16), linked hash set (line 23), tree set (line 30), array list (line 37), or linked list (line 44) and obtains the execution time for removing the elements from the hash set (line 18), linked hash set (line 25), tree set (line 32), array list (line 39), and linked list (line 46).

The getTestTime method invokes the contains method to test whether a number is in the container (line 54) and the getRemoveTime method invokes the remove method to remove an element from the container (line 63).

As these runtimes illustrate, sets are much more efficient than lists for testing whether an ele- ment is in a set or a list. Therefore, the No-Fly list should be implemented using a hash set instead of a list, because it is much faster to test whether an element is in a hash set than in a list.

You may wonder why sets are more efficient than lists. The questions will be answered in Chapters 24 and 27 when we introduce the implementations of lists and sets.

21.3.1 Suppose you need to write a program that stores unordered, nonduplicate ele- ments, what data structure should you use?

21.3.2 Suppose you need to write a program that stores nonduplicate elements in the order of insertion, what data structure should you use?

21.3.3 Suppose you need to write a program that stores nonduplicate elements in increas- ing order of the element values, what data structure should you use?

21.3.4 Suppose you need to write a program that stores a fixed number of the elements (possibly duplicates), what data structure should you use?

21.3.5 Suppose you need to write a program that stores the elements in a list with fre- quent operations to append and delete elements at the end of the list, what data structure should you use?

21.3.6 Suppose you need to write a program that stores the elements in a list with fre- quent operations to insert and delete elements at the beginning of the list, what data structure should you use?

sets are better

Point Check

M21_LIAN1878_11_GE_C21.indd 848 12/29/17 4:35 PM

21.4 Case Study: Counting Keywords 849

21.4 Case Study: Counting Keywords This section presents an application that counts the number of keywords in a Java source file.

For each word in a Java source file, we need to determine whether the word is a keyword. To handle this efficiently, store all the keywords in a HashSet and use the contains method to test if a word is in the keyword set. Listing 21.7 gives this program.

Listing 21.7 CountKeywords.java 1 import java.util.*; 2 import java.io.*; 3 4 public class CountKeywords { 5 public static void main(String[] args) throws Exception { 6 Scanner input = new Scanner(System.in); 7 System.out.print("Enter a Java source file: "); 8 String filename = input.nextLine(); 9 10 File file = new File(filename); 11 if (file.exists()) { 12 System.out.println("The number of keywords in " + filename 13 + " is " + countKeywords(file)); 14 } 15 else { 16 System.out.println("File " + filename + " does not exist"); 17 } 18 } 19 20 public static int countKeywords(File file) throws Exception { 21 // Array of all Java keywords + true, false and null 22 String[] keywordString = {"abstract", "assert", "boolean", 23 "break", "byte", "case", "catch", "char", "class", "const", 24 "continue", "default", "do", "double", "else", "enum", 25 "extends", "for", "final", "finally", "float", "goto", 26 "if", "implements", "import", "instanceof", "int", 27 "interface", "long", "native", "new", "package", "private", 28 "protected", "public", "return", "short", "static", 29 "strictfp", "super", "switch", "synchronized", "this", 30 "throw", "throws", "transient", "try", "void", "volatile", 31 "while", "true", "false", "null"}; 32 33 Set<String> keywordSet = 34 new HashSet<>(Arrays.asList(keywordString)); 35 int count = 0; 36 37 Scanner input = new Scanner(file); 38 39 while (input.hasNext()) { 40 String word = input.next(); 41 if (keywordSet.contains(word)) 42 count++; 43 } 44 45 return count; 46 } 47 }

Point Key

enter a filename

file exists?

count keywords

keywords

keyword set

is a keyword?

M21_LIAN1878_11_GE_C21.indd 849 12/29/17 4:35 PM

850 Chapter 21 Sets and Maps

The program prompts the user to enter a Java source filename (line 7) and reads the filename (line 8). If the file exists, the countKeywords method is invoked to count the keywords in the file (line 13).

The countKeywords method creates an array of strings for the keywords (lines 22–31) and creates a hash set from this array (lines 33–34). It then reads each word from the file and tests if the word is in the set (line 41). If so, the program increases the count by 1 (line 42).

You may rewrite the program to use a LinkedHashSet, TreeSet, ArrayList, or LinkedList to store the keywords. However, using a HashSet is the most efficient for this program.

21.4.1 Will the CountKeywords program work if lines 33–34 are changed to

Set<String> keywordSet = new LinkedHashSet<>(Arrays.asList(keywordString));

21.4.2 Will the CountKeywords program work if lines 33–34 are changed to

List<String> keywordSet = new ArrayList<>(Arrays.asList(keywordString));

21.5 Maps You can create a map using one of its three concrete classes: HashMap, LinkedHashMap, or TreeMap.

A map is a container object that stores a collection of key/value pairs. It enables fast retrieval, deletion, and updating of the pair through the key. A map stores the values along with the keys. The keys are like indexes. In List, the indexes are integers. In Map, the keys can be any objects. A map cannot contain duplicate keys. Each key maps to one value. A key and its corresponding value form an entry stored in a map, as shown in Figure 21.2a. Figure 21.2b shows a map in which each entry consists of a Social Security number as the key and a name as the value.

Point Check

Point Key

map

Enter a Java source file: c:\ Welcome.java

The number of keywords in c:\ Welcome.java is 5

Enter a Java source file: c:\ TTT.java

File c:\ TTT.java does not exist

Figure 21.2 The entries consisting of key/value pairs are stored in a map.

Corresponding element value

Entry

Search key

A map

Corresponding value

Entry

Search key

(b)(a)

111-34-3434 John

132-56-6290 Peter

M21_LIAN1878_11_GE_C21.indd 850 12/29/17 4:35 PM

21.5 Maps 851

There are three types of maps: HashMap, LinkedHashMap, and TreeMap. The common features of these maps are defined in the Map interface. Their relationship is shown in Figure 21.3.

Figure 21.3 A map stores key/value pairs.

SortedMap

HashMap LinkedHashMap

TreeMap

Interfaces Abstract Classes Concrete Classes

Map

NavigableMap

AbstractMap

The Map interface provides the methods for querying, updating, and obtaining a collection of values and a set of keys, as shown in Figure 21.4.

Figure 21.4 The Map interface maps keys to values.

+clear(): void +containsKey(key: Object): boolean

+containsValue(value: Object): boolean

+isEmpty(): boolean +keySet(): Set<K> +put(key: K, value: V): V +putAll(m: Map<? extends K,? extends V>): void

+get(key: Object): V +entrySet(): Set<Map.Entry<K,V>>

Removes all entries from this map. Returns true if this map contains an entry for the speci�ed key.

Returns true if this map maps one or more keys to the speci�ed value. Returns a set consisting of the entries in this map. Returns the value for the speci�ed key in this map. Returns true if this map contains no entries. Returns a set consisting of the keys in this map. Puts an entry into this map.

Removes the entries for the speci�ed key. Returns the number of entries in this map. Returns a collection consisting of the values in this map.

Adds all the entries from m to this map.

Performs an action for each entry in this map.

+remove(key: Object): V +size(): int +values(): Collection<V>

«interface» java.util.Map<K,V>

+forEach(action: Consumer<? super K, ? super V): default void

The update methods include clear, put, putAll, and remove. The clear() method removes all entries from the map. The put(K key, V value) method adds an entry for the specified key and value in the map. If the map formerly contained an entry for this key, the old value is replaced by the new value, and the old value associated with the key is returned. The putAll(Map m) method adds all entries in m to this map. The remove(Object key) method removes the entry for the specified key from the map.

The query methods include containsKey, containsValue, isEmpty, and size. The containsKey(Object key) method checks whether the map contains an entry for the speci- fied key. The containsValue(Object value) method checks whether the map contains an entry for this value. The isEmpty() method checks whether the map contains any entries. The size() method returns the number of entries in the map.

update methods

query methods

M21_LIAN1878_11_GE_C21.indd 851 12/29/17 4:35 PM

852 Chapter 21 Sets and Maps

You can obtain a set of the keys in the map using the keySet() method, and a collection of the values in the map using the values() method. The entrySet() method returns a set of entries. The entries are instances of the Map.Entry<K, V> interface, where Entry is an inner interface for the Map interface, as shown in Figure 21.5. Each entry in the set is a key/ value pair in the underlying map.

keySet()

values()

entrySet()

Figure 21.5 The Map.Entry interface operates on an entry in the map.

+getKey(): K

+getValue(): V

+setValue(value: V): void

Returns the key from this entry.

Returns the value from this entry.

Replaces the value in this entry with a new value.

«interface» java.util.Map.Entry<K,V>

Java 8 added a default forEach method in the Map interface for performing an action on each entry in the map. This method can be used like an iterator for traversing the entries in the map.

The AbstractMap class is a convenience abstract class that implements all the methods in the Map interface except the entrySet() method.

The HashMap, LinkedHashMap, and TreeMap classes are three concrete implementations of the Map interface, as shown in Figure 21.6.

forEach method

AbstractMap

concrete implementation

Figure 21.6 The Java Collections Framework provides three concrete map classes.

java.util.AbstractMap<K,V>

java.util.HashMap<K,V>

java.util.LinkedHashMap<K,V>

+LinkedHashMap()

+LinkedHashMap(m: Map<? extends K,? extends V>)

+LinkedHashMap(initialCapacity: int, loadFactor: oat, accessOrder: boolean)

java.util.TreeMap<K,V>

+TreeMap() +TreeMap(m: Map<? extends K,? extends V>) +TreeMap(c: Comparator<? super K>)

«interface» java.util.SortedMap<K,V>

+�rstKey(): K

+lastKey(): K

+comparator(): Comparator<? super K>)

+headMap(toKey: K): SortedMap<K,V>

+tailMap(fromKey: K): SortedMap<K,V>

«interface» java.util.Map<K, V>

«interface» java.util.NavigableMap<K, V>

+�oorKey(key: K): K +ceilingKey(key: K): K +lowerKey(key: K): K +higherKey(key: K): K +pollFirstEntry(): Map.EntrySet<K, V> +pollLastEntry(): Map.EntrySet<K, V>

+HashMap()

+HashMap(m: Map<? extends K, ? extends V>)

+HashMap(initialCapacity: int,loadFactor: oat)

M21_LIAN1878_11_GE_C21.indd 852 12/29/17 4:35 PM

21.5 Maps 853

The HashMap class is efficient for locating a value, inserting an entry, and deleting an entry. LinkedHashMap extends HashMap with a linked-list implementation that supports an

ordering of the entries in the map. The entries in a HashMap are not ordered, but the entries in a LinkedHashMap can be retrieved either in the order in which they were inserted into the map (known as the insertion order) or in the order in which they were last accessed, from least recently to most recently accessed (access order). The no-arg constructor constructs a LinkedHashMap with the insertion order. To construct a LinkedHashMap with the access order, use LinkedHashMap(initialCapacity, loadFactor, true).

The TreeMap class is efficient for traversing the keys in a sorted order. The keys can be sorted using the Comparable interface or the Comparator interface. If you create a TreeMap using its no-arg constructor, the compareTo method in the Comparable interface is used to compare the keys in the map, assuming the class for the keys implements the Comparable interface. To use a comparator, you have to use the TreeMap(Comparator comparator) constructor to create a sorted map that uses the compare method in the comparator to order the entries in the map based on the keys.

SortedMap is a subinterface of Map, which guarantees the entries in the map are sorted. In addition, it provides the methods firstKey() and lastKey() for returning the first and the last keys in the map, and headMap(toKey) and tailMap(fromKey) for returning a portion of the map whose keys are less than toKey and greater than or equal to fromKey, respectively.

NavigableMap extends SortedMap to provide the navigation methods lowerKey(key), floorKey(key), ceilingKey(key), and higherKey(key) that return keys, respectively, less than, less than or equal, greater than or equal, and greater than a given key and return null if there is no such key. The pollFirstEntry() and pollLastEntry() methods remove and return the first and the last entry in the tree map, respectively.

Note Prior to Java 2, java.util.Hashtable was used for mapping keys with values. Hashtable was redesigned to fit into the Java Collections Framework with all its methods retained for compatibility. Hashtable implements the Map interface and is used in the same way as HashMap, except that the update methods in Hashtable are synchronized.

Listing 21.8 gives an example that creates a hash map, a linked hash map, and a tree map for mapping students to ages. The program first creates a hash map with the student’s name as its key and the age as its value. The program then creates a tree map from the hash map and displays the entries in ascending order of the keys. Finally, the program creates a linked hash map, adds the same entries to the map, and displays the entries.

Listing 21.8 TestMap.java 1 import java.util.*; 2 3 public class TestMap { 4 public static void main(String[] args) { 5 // Create a HashMap 6 Map<String, Integer> hashMap = new HashMap<>(); 7 hashMap.put("Smith", 30); 8 hashMap.put("Anderson", 31); 9 hashMap.put("Lewis", 29); 10 hashMap.put("Cook", 29); 11 12 System.out.println("Display entries in HashMap"); 13 System.out.println(hashMap + "\n"); 14 15 // Create a TreeMap from the preceding HashMap 16 Map<String, Integer> treeMap = new TreeMap<>(hashMap); 17 System.out.println("Display entries in ascending order of key");

HashMap

LinkedHashMap

insertion order

access order

TreeMap

SortedMap

NavigableMap

Hashtable

hash map linked hash map tree map

create map add entry

tree map

M21_LIAN1878_11_GE_C21.indd 853 12/29/17 4:35 PM

854 Chapter 21 Sets and Maps

18 System.out.println(treeMap); 19 20 // Create a LinkedHashMap 21 Map<String, Integer> linkedHashMap = 22 new LinkedHashMap<>(16, 0.75f, true); 23 linkedHashMap.put("Smith", 30); 24 linkedHashMap.put("Anderson", 31); 25 linkedHashMap.put("Lewis", 29); 26 linkedHashMap.put("Cook", 29); 27 28 // Display the age for Lewis 29 System.out.println("\nThe age for " + "Lewis is " + 30 linkedHashMap.get("Lewis")); 31 32 System.out.println("Display entries in LinkedHashMap"); 33 System.out.println(linkedHashMap); 34 35 // Display each entry with name and age 36 System.out.print("\nNames and ages are "); 37 treeMap.forEach( 38 (name, age) −> System.out.print(name + ": " + age + " ")); 39 } 40 }

linked hash map

forEach method

Display entries in HashMap {Cook=29, Smith=30, Lewis=29, Anderson=31}

Display entries in ascending order of key {Anderson=31, Cook=29, Lewis=29, Smith=30}

The age for Lewis is 29 Display entries in LinkedHashMap {Smith=30, Anderson=31, Cook=29, Lewis=29}

Names and ages are Anderson: 31 Cook: 29 Lewis: 29 Smith: 30

As shown in the output, the entries in the HashMap are in random order. The entries in the TreeMap are in increasing order of the keys. The entries in the LinkedHashMap are in the order of their access, from least recently accessed to most recently.

All the concrete classes that implement the Map interface have at least two constructors. One is the no-arg constructor that constructs an empty map, and the other constructs a map from an instance of Map. Thus, new TreeMap<>(hashMap) (line 16) constructs a tree map from a hash map.

You can create an insertion- or access-ordered linked hash map. An access-ordered linked hash map is created in lines 21–22. The most recently accessed entry is placed at the end of the map. The entry with the key Lewis is last accessed in line 30, so it is displayed last in line 33.

It is convenient to process all the entries in the map using the forEach method. The program uses a forEach method to display a name and its age (lines 37–38).

Tip If you don’t need to maintain an order in a map when updating it, use a HashMap. When you need to maintain the insertion order or access order in the map, use a LinkedHashMap. When you need the map to be sorted on keys, use a TreeMap.

21.5.1 How do you create an instance of Map? How do you add an entry to a map consist- ing of a key and a value? How do you remove an entry from a map? How do you find the size of a map? How do you traverse entries in a map?

Point Check

M21_LIAN1878_11_GE_C21.indd 854 12/29/17 4:35 PM

21.6 Case Study: Occurrences of Words 855

21.5.2 Describe and compare HashMap, LinkedHashMap, and TreeMap. 21.5.3 Show the output of the following code:

import java.util.*; public class Test { public static void main(String[] args) { Map<String, String> map = new LinkedHashMap<>(); map.put("123", "John Smith"); map.put("111", "George Smith"); map.put("123", "Steve Yao"); map.put("222", "Steve Yao"); System.out.println("(1) " + map); System.out.println("(2) " + new TreeMap<String, String>(map)); map.forEach((k, v) −> { if (k.equals("123")) System.out.println(v);}); } }

21.6 Case Study: Occurrences of Words This case study writes a program that counts the occurrences of words in a text and displays the words and their occurrences in alphabetical order of the words.

The program uses a TreeMap to store an entry consisting of a word and its count. For each word, check whether it is already a key in the map. If not, add an entry to the map with the word as the key and value 1. Otherwise, increase the value for the word (key) by 1 in the map. Assume the words are case insensitive; for example, Good is treated the same as good.

Listing 21.9 gives the solution to the problem.

Listing 21.9 CountOccurrenceOfWords.java 1 import java.util.*; 2 3 public class CountOccurrenceOfWords { 4 public static void main(String[] args) { 5 // Set text in a string 6 String text = "Good morning. Have a good class. " + 7 "Have a good visit. Have fun!"; 8 9 // Create a TreeMap to hold words as key and count as value 10 Map<String, Integer> map = new TreeMap<>(); 11 12 String[] words = text.split("[\\s+\\p{P}]"); 13 for (int i = 0; i < words.length; i++) { 14 String key = words[i].toLowerCase(); 15 16 if (key.length() > 0) { 17 if (!map.containsKey(key)) { 18 map.put(key, 1); 19 } 20 else { 21 int value = map.get(key); 22 value++; 23 map.put(key, value); 24 } 25 } 26 } 27 28 // Display key and value for each entry

Point Key

tree map

split string

add entry

update entry

M21_LIAN1878_11_GE_C21.indd 855 12/29/17 4:35 PM

856 Chapter 21 Sets and Maps

29 map.forEach((k, v) −> System.out.println(k + "\t" + v)); 30 } 31 }

display entry

a 2 class 1 fun 1 good 3 have 3 morning 1 visit 1

The program creates a TreeMap (line 10) to store pairs of words and their occurrence counts. The words serve as the keys. Since all values in the map must be stored as objects, the count is wrapped in an Integer object.

The program extracts a word from a text using the split method (line 12) in the String class (see Section 10.10.4 and Appendix H). The text is split into words using a whitespace \s or punctuation \p{P} as a delimiter. For each word extracted, the program checks whether it is already stored as a key in the map (line 17). If not, a new pair consisting of the word and its initial count (1) is stored in the map (line 18). Otherwise, the count for the word is incremented by 1 (lines 21–23).

The program displays the count and the key in each entry using the forEach method in the Map class (line 29).

Since the map is a tree map, the entries are displayed in increasing order of words. To display them in ascending order of the occurrence counts, see Programming Exercise 21.8.

Now sit back and think how you would write this program without using map. Your new program would be longer and more complex. You will find that map is a very efficient and powerful data structure for solving problems such as this.

Java Collections Framework provides comprehensive support of organizing and manipulat- ing data. Suppose you wish to display the words in increasing order of their occurrence values, how do you modify the program? This can be done simply by creating a list of map entries and creating a Comparator for sorting the entries on their values as follows:

List<Map.Entry<String, Integer>> entries = new ArrayList<>(map.entrySet()); Collections.sort(entries, (entry1, entry2) −> { return entry1.getValue().compareTo(entry2.getValue()); }); for (Map.Entry<String, Integer> entry: entries) { System.out.println(entry.getKey() + "\t" + entry.getValue()); }

21.6.1 Will the CountOccurrenceOfWords program work if line 10 is changed to

Map<String, int> map = new TreeMap<>();

21.6.2 Will the CountOccurrenceOfWords program work if line 17 is changed to

if (map.get(key) == null) {

21.6.3 Will the CountOccurrenceOfWords program work if line 29 is changed to

for (String key: map) System.out.println(key + "\t" + map.getValue(key));

21.6.4 How do you simplify the code in lines 17–24 in Listing 21.9 in one line using a conditional expression?

Point Check

M21_LIAN1878_11_GE_C21.indd 856 12/29/17 4:35 PM

21.7 Singleton and Unmodifiable Collections and Maps 857

21.7 Singleton and Unmodifiable Collections and Maps You can create singleton sets, lists, and maps and unmodifiable sets, lists, and maps using the static methods in the Collections class.

The Collections class contains the static methods for lists and collections. It also contains the methods for creating immutable singleton sets, lists, and maps and for creating read-only sets, lists, and maps, as shown in Figure 21.7.

Point Key

Figure 21.7 The Collections class contains the static methods for creating singleton and read-only sets, lists, and maps.

java.util.Collections

+singleton(o: Object): Set

+singletonList(o: Object): List

+singletonMap(key: Object, value: Object): Map

+unmodifiableCollection(c: Collection): Collection

+unmodifiableList(list: List): List

+unmodifiableMap(m: Map): Map

+unmodifiableSet(s: Set): Set

+unmodifiableSortedMap(s: SortedMap): SortedMap

+unmodifiableSortedSet(s: SortedSet): SortedSet

Returns an immutable set containing the speci�ed object.

Returns an immutable list containing the speci�ed object.

Returns an immutable map with the key and value pair.

Returns a read-only view of the collection.

Returns a read-only view of the list.

Returns a read-only view of the map.

Returns a read-only view of the set.

Returns a read-only view of the sorted map.

Returns a read-only view of the sorted set.

The Collections class defines three constants—EMPTY_SET, EMPTY_LIST, and EMPTY_MAP—for an empty set, an empty list, and an empty map. These collections are immu- table. The class also provides the singleton(Object o) method for creating an immutable set containing only a single item, the singletonList(Object o) method for creating an immutable list containing only a single item, and the singletonMap(Object key, Object value) method for creating an immutable map containing only a single entry.

The Collections class also provides six static methods for returning read-only views for collections: unmodifiableCollection(Collection c), unmodifiableList(List list), unmodifiableMap(Map m), unmodifiableSet(Set set), unmodifiableSor tedMap(SortedMap m), and unmodifiableSortedSet(SortedSet s). This type of view is like a reference to the actual collection. However, you cannot modify the collection through a read-only view. Attempting to modify a collection through a read-only view will cause an UnsupportedOperationException.

21.7.1 What is wrong in the following code?

Set<String> set = Collections.singleton("Chicago"); set.add("Dallas");

21.7.2 What happens when you run the following code?

List list = Collections.unmodifiableList(Arrays.asList("Chicago", "Boston")); list.remove("Dallas");

read-only view

Point Check

Key terms hash map 853 hash set 838 linked hash map 853 linked hash set 842 map 850

read-only view 857 set 838 tree map 853 tree set 843

M21_LIAN1878_11_GE_C21.indd 857 12/29/17 4:35 PM

858 Chapter 21 Sets and Maps

ChapTer summary 1. A set stores nonduplicate elements. To allow duplicate elements to be stored in a

collection, you need to use a list.

2. A map stores key/value pairs. It provides a quick lookup for a value using a key.

3. Three types of sets are supported: HashSet, LinkedHashSet, and TreeSet. HashSet stores elements in an unpredictable order. LinkedHashSet stores elements in the order they were inserted. TreeSet stores elements sorted. HashSet, LinkedHashSet, and TreeSet are subtypes of Collection.

4. The Map interface maps keys to the elements. The keys are like indexes. In List, the indexes are integers. In Map, the keys can be any objects. A map cannot contain duplicate keys. Each key can map to at most one value. The Map interface provides the methods for querying, updating, and obtaining a collection of values and a set of keys.

5. Three types of maps are supported: HashMap, LinkedHashMap, and TreeMap. HashMap is efficient for locating a value, inserting an entry, and deleting an entry. LinkedHashMap supports ordering of the entries in the map. The entries in a HashMap are not ordered, but the entries in a LinkedHashMap can be retrieved either in the order in which they were inserted into the map (known as the insertion order) or in the order in which they were last accessed, from least recently accessed to most recently (access order). TreeMap is efficient for traversing the keys in a sorted order. The keys can be sorted using the Comparable interface or the Comparator interface.

Quiz Answer the quiz for this chapter online at the book Companion Website.

programming exerCises

Sections 21.2–21.4 21.1 (Perform set operations on hash sets) Given two stacks of textbooks of the fol-

lowing subjects 5"Chemistry", "Mathematics", "Biology", "English"6 and 5"Biology", "English", "Geography", "Physics"6 respectively, find the subjects that are (1) only present in the first stack; (2) only present in the second stack; (3) present in both stacks. (You may clone the sets to preserve the original sets from being changed by set methods.)

21.2 (Display nonduplicate names in ascending order) Given one or more text files, each representing a day’s attendance in a course and containing the names of the students who attended the course on that particular day, write a program that dis- plays, in ascending order, the names of those students who have attended at least one day of the course. The text file(s) is/are passed as command-line argument(s).

**21.3 (Checking whether a key exists in a set) Given a problem that is identical to the previ- ous Programming Exercise 21.2, write a program that displays, in ascending order, the first names of those students who have attended more than one day of the course. If there are students with the same first name, display their first and last names.

*21.4 (Count consonants and vowels) Write a program that prompts the user to enter a text file name and displays the number of vowels and consonants in the file. Use a set to store the vowels A, E, I, O, and U.

***21.5 (Syntax highlighting) Write a program that converts a Java file into an HTML file. In the HTML file, the keywords, comments, and literals are displayed in

M21_LIAN1878_11_GE_C21.indd 858 12/29/17 4:35 PM

bold navy, green, and blue, respectively. Use the command line to pass a Java file and an HTML file. For example, the following command

java Exercise21_05 Welcome.java Welcome.html

converts Welcome.java into Welcome.html. Figure 21.8a shows a Java file. The corresponding HTML file is shown in Figure 21.8b.

Figure 21.8 The Java code in plain text in (a) is displayed in HTML with syntax highlighted in (b). Source: Copyright © 1995–2016 Oracle and/or its affiliates. All rights reserved. Used with permission.

(a) (b)

Sections 21.5–21.7 *21.6 (Count the occurrences of words/names) Given a problem that is identical to

Programming Exercises 21.2 and 21.3, write a program that will count the num- ber of courses each student attended. In descending order of the number of days attended, display the names of students and the number of courses each student has attended. If the number of days attended by any two students is the same, display their names in ascending order. For example, if both Billy and John attended classes for 2 days, display Billy: 2 first, and then John: 2.

**21.7 (Revise Listing 21.9, CountOccurrenceOfWords.java) Rewrite Listing 21.9 to display the words in ascending order of occurrence counts.

**21.8 (Count the occurrences of words in a text file) Rewrite Listing 21.9 to read the text from a text file. The text file is passed as a command-line argument. Words are delimited by whitespace characters, punctuation marks (,;.:?), quotation marks ('"), and parentheses. Count words in case-insensitive fashion (e.g., consider Good and good to be the same word). The words must start with a let- ter. Display the output in alphabetical order of words, with each word preceded by its occurrence count.

**21.9 (Guess the capitals using maps) Rewrite Programming Exercise 8.37 to store pairs of each state and its capital in a map. Your program should prompt the user to enter a state, and should display the capital for the state.

*21.10 (Count the occurrences of each keyword) Rewrite Listing 21.7, CountKeywords. java to read in a Java source-code file and count the occurrence of each keyword in the file, but don’t count the keyword if it is in a comment or in a string literal.

**21.11 (Baby name popularity ranking) Use the data files from Programming Exercise 12.31 to write a program that enables the user to select a year, gender, and enter a name to display the ranking of the name for the selected year and gender, as shown in Figure 21.9. To achieve the best efficiency, create two arrays for boy’s names and girl’s names, respectively. Each array has 10 elements for 10 years. Each element is a map that stores a name and its ranking in a pair with the name as the key.

Programming Exercises 859

M21_LIAN1878_11_GE_C21.indd 859 12/29/17 4:35 PM

860 Chapter 21 Sets and Maps

**21.12 (Name for both genders) Write a program that prompts the user to enter one of the filenames described in Programming Exercise 12.31 and displays the names that are used for both genders in the file. Use sets to store names and find common names in two sets. Here is a sample run:

Enter the year: 2010

Enter the gender: M

Enter the name: Javier

Boy name Javier is ranked #190 in year 2010

Enter another inquiry? Y

Enter the year: 2001

Enter the gender: F

Enter the name: Emily

Girl name Emily is ranked #1 in year 2001

Enter another inquiry? N

Enter a file name for baby name ranking: babynamesranking2001.txt 69 names used for both genders They are Tyler Ryan Christian ...

Figure 21.9 The user selects a year and gender, enters a year, and clicks the Find Ranking button to display the ranking. Source: Copyright © 1995–2016 Oracle and/or its affiliates. All rights reserved. Used with permission.

** 21.13 (Baby name popularity ranking) Revise Programming Exercise 21.11 to prompt the user to enter year, gender, and name and display the ranking for the name. Prompt the user to enter another inquiry or exit the program. Here is a sample run:

**21.14 (Web crawler) Rewrite Listing 12.18, WebCrawler.java, to improve the perfor- mance by using appropriate new data structures for listOfPendingURLs and listofTraversedURLs.

**21.15 (Addition quiz) Rewrite Programming Exercise 11.16 to store the answers in a set rather than a list.

M21_LIAN1878_11_GE_C21.indd 860 12/29/17 4:35 PM

Objectives ■■ To estimate algorithm efficiency using the Big O notation (§22.2).

■■ To explain growth rates and why constants and nondominating terms can be ignored in the estimation (§22.2).

■■ To determine the complexity of various types of algorithms (§22.3).

■■ To analyze the binary search algorithm (§22.4.1).

■■ To analyze the selection sort algorithm (§22.4.2).

■■ To analyze the Tower of Hanoi algorithm (§22.4.3).

■■ To describe common growth functions (constant, logarithmic, log- linear, quadratic, cubic, and exponential) (§22.4.4).

■■ To design efficient algorithms for finding Fibonacci numbers using dynamic programming (§22.5).

■■ To find the GCD using Euclid’s algorithm (§22.6).

■■ To find prime numbers using the sieve of Eratosthenes (§22.7).

■■ To design efficient algorithms for finding the closest pair of points using the divide-and-conquer approach (§22.8).

■■ To solve the Eight Queens problem using the backtracking approach (§22.9).

■■ To design efficient algorithms for finding a convex hull for a set of points (§22.10).

Developing Efficient Algorithms

Chapter

M22_LIAN1878_11_GE_C22.indd 861 12/29/17 4:38 PM

862 Chapter 22 Developing Efficient Algorithms

22.1 Introduction Algorithm design is to develop a mathematical process for solving a problem. Algo- rithm analysis is to predict the performance of an algorithm.

The preceding two chapters introduced classic data structures (lists, stacks, queues, priority queues, sets, and maps) and applied them to solve problems. This chapter will use a variety of examples to introduce common algorithmic techniques (dynamic programming, divide-and-conquer, and backtracking) for developing efficient algorithms. Later in the book, we will introduce efficient algorithms in Chapters 23–29. Before introducing developing efficient algorithms, we need to address the question on how to measure algorithm efficiency.

22.2 Measuring Algorithm Efficiency Using Big O Notation

The Big O notation obtains a function for measuring algorithm time complexity based on the input size. You can ignore multiplicative constants and nondominating terms in the function.

Suppose two algorithms perform the same task, such as search (linear search vs. binary search). Which one is better? To answer this question, you might implement these algorithms and run the programs to get execution times. However, there are two problems with this approach:

■■ First, many tasks run concurrently on a computer. The execution time of a particular program depends on the system load.

■■ Second, the execution time depends on specific input. Consider, for example, linear search and binary search. If an element to be searched happens to be the first in the list, linear search will find the element quicker than binary search.

It is very difficult to compare algorithms by measuring their execution times. To overcome these problems, a theoretical approach was developed to analyze algorithms independent of computers and specific input. This approach approximates the effect of a change on the size of the input. In this way, you can see how fast an algorithm’s execution time increases as the input size increases, so you can compare two algorithms by examining their growth rates.

Consider linear search. The linear search algorithm compares the key with the elements in the array sequentially until the key is found or the array is exhausted. If the key is not in the array, it requires n comparisons for an array of size n. If the key is in the array, it requires n/2 comparisons on average. The algorithm’s execution time is proportional to the size of the array. If you double the size of the array, you will expect the number of comparisons to double. The algorithm grows at a linear rate. The growth rate has an order of magnitude of n. Computer scientists use the Big O notation to represent the “order of magnitude.” Using this notation, the complexity of the linear search algorithm is O(n), pronounced as “order of n.” We call an algorithm with a time complexity of O(n) linear algorithm, and it exhibits a linear growth rate.

For the same input size, an algorithm’s execution time may vary, depending on the input. An input that results in the shortest execution time is called the best-case input, and an input that results in the longest execution time is the worst-case input. Best-and worst- case analyses are to analyze the algorithms for their best- and worst-case inputs. Best- and worst-case analyses are not representative, but worst-case analysis is very useful. You can be assured that the algorithm will never be slower than the worst case. An average-case analysis attempts to determine the average amount of time among all possible inputs of the same size. Average-case analysis is ideal, but difficult to perform because for many problems it is hard to determine the relative probabilities and distributions of various input instances. Worst-case analysis is easier to perform, so the analysis is generally conducted for the worst case.

Point Key

what is algorithm efficiency?

growth rates

Big O notation

best-case input worst-case input

average-case analysis

Point Key

M22_LIAN1878_11_GE_C22.indd 862 12/29/17 4:38 PM

22.2 Measuring Algorithm Efficiency Using Big O Notation 863

The linear search algorithm requires n comparisons in the worst case and n/2 comparisons in the average case if you are nearly always looking for something known to be in the list. Using the Big O notation, both the cases require O(n) time. The multiplicative constant (1/2) can be omitted. Algorithm analysis is focused on growth rate. The multiplicative constants have no impact on growth rates. The growth rate for n/2 or 100n is the same as for n, as illustrated in Table 22.1. Therefore, O(n) = O(n/2) = O(100n).

ignoring multiplicative constants

Table 22.1 Growth Rates

f(n) n

100

200

100

200

100

10000

20000

2 2 2 f(200) / f(100)

n n/2 100n

Consider the algorithm for finding the maximum number in an array of n elements. To find the maximum number if n is 2, it takes one comparison and if n is 3, it takes two com- parisons. In general, it takes n - 1 comparisons to find the maximum number in a list of n elements. Algorithm analysis is for large input size. If the input size is small, there is no significance in estimating an algorithm’s efficiency. As n grows larger, the n part in the expression n - 1 dominates the complexity. The Big O notation allows you to ignore the nondominating part (e.g., -1 in the expression n - 1) and highlight the important part (e.g., n in the expression n - 1). Therefore, the complexity of this algorithm is O(n).

The Big O notation estimates the execution time of an algorithm in relation to the input size. If the time is not related to the input size, the algorithm is said to take constant time with the notation O(1). For example, a method that retrieves an element at a given index in an array takes constant time because the time does not grow as the size of the array increases.

The following mathematical summations are often useful in algorithm analysis:

1 + 2 + 3 + g+ (n - 2) + (n - 1) = n(n - 1)

2 = O(n2)

1 + 2 + 3 + g+ (n - 1) + n = n(n + 1)

2 = O(n2)

a0 + a1 + a2 + a3 + g+ a(n - 1) + an = an + 1 - 1

a - 1 = O(an)

20 + 21 + 22 + 23 + g+ 2(n - 1) + 2n = 2n + 1 - 1

2 - 1 = 2n + 1 - 1 = O(2n)

Note Time complexity is a measure of execution time using the Big O notation. Similarly, you can also measure space complexity using the Big O notation. Space complexity measures the amount of memory space used by an algorithm. The space complexity for most algorithms presented in this book is O(n), that is, they exhibit linear growth rate to the input size. For example, the space complexity for linear search is O(n).

22.2.1 Why is a constant factor ignored in the Big O notation? Why is a nondominating term ignored in the Big O notation?

22.2.2 What is the order of each of the following functions? (n2 + 1)2

n ,

(n2 + log2 n2) n

, n3 + 100n2 + n, 2n + 100n2 + 45n, n2n + n22n

large input size

ignoring nondominating terms

constant time

useful summations

time complexity space complexity

Point Check

M22_LIAN1878_11_GE_C22.indd 863 12/29/17 4:38 PM

864 Chapter 22 Developing Efficient Algorithms

22.3 Examples: Determining Big O This section gives several examples of determining Big O for repetition, sequence, and selection statements.

Example 1 Consider the time complexity for the following loop:

for (int i = 1; i <= n; i++) { k = k + 5; }

It is a constant time, c, for executing

k = k + 5;

Since the loop is executed n times, the time complexity for the loop is

T(n) = (a constant c) * n = O(n).

The theoretical analysis predicts the performance of the algorithm. To see how this algorithm performs, we run the code in Listing 22.1 to obtain the execution time for n = 1,000,000, 10,000,000, 100,000,000, and 1,000,000,000.

lisTing 22.1 PerformanceTest.java 1 public class PerformanceTest { 2 public static void main(String[] args) { 3 getTime(1000000); 4 getTime(10000000); 5 getTime(100000000); 6 getTime(1000000000); 7 } 8 9 public static void getTime(long n) { 10 long startTime = System.currentTimeMillis(); 11 long k = 0; 12 for (long i = 1; i <= n; i++) { 13 k = k + 5; 14 } 15 long endTime = System.currentTimeMillis(); 16 System.out.println("Execution time for n = " + n 17 + " is " + (endTime − startTime) + " milliseconds"); 18 } 19 }

Point Key

input size 1,000,000 input size 10,000,000 input size 100,000,000 input size 1,000,000,000

time before execution

time after execution

Execution time for n = 1,000,000 is 6 milliseconds Execution time for n = 10,000,000 is 61 milliseconds Execution time for n = 100,000,000 is 610 milliseconds Execution time for n = 1,000,000,000 is 6048 milliseconds

Our analysis predicts a linear time complexity for this loop. As shown in the sample output, when the input size increases 10 times, the runtime increases roughly 10 times. The execution confirms to the prediction.

Example 2 What is the time complexity for the following loop?

for (int i = 1; i <= n; i++) { for (int j = 1; j <= n; j++) {

M22_LIAN1878_11_GE_C22.indd 864 12/29/17 4:38 PM

22.3 Examples: Determining Big O 865

k = k + i + j; } }

It is a constant time, c, for executing

k = k + i + j;

The outer loop executes n times. For each iteration in the outer loop, the inner loop is executed n times. Thus, the time complexity for the loop is

T(n) = (a constant c) * n * n = O(n2)

An algorithm with the O(n2) time complexity is called a quadratic algorithm and it exhibits a quadratic growth rate. The quadratic algorithm grows quickly as the problem size increases. If you double the input size, the time for the algorithm is quadrupled. Algo- rithms with a nested loop are often quadratic.

Example 3 Consider the following loop:

for (int i = 1; i <= n; i++) { for (int j = 1; j <= i; j++) { k = k + i + j; } }

The outer loop executes n times. For i = 1, 2, c , the inner loop is executed one time, two times, and n times. Thus, the time complexity for the loop is

T(n) = c + 2c + 3c + 4c + g + nc

= cn(n + 1)/2

= (c/2) n2 + (c/2)n

= O(n2)

Example 4 Consider the following loop:

for (int i = 1; i <= n; i++) { for (int j = 1; j <= 20; j++) { k = k + i + j; } }

The inner loop executes 20 times and the outer loop n times. Therefore, the time complexity for the loop is

T(n) = 20 * c * n = O(n)

Example 5 Consider the following sequences:

for (int j = 1; j <= 10; j++) { k = k + 4; } for (int i = 1; i <= n; i++) { for (int j = 1; j <= 20; j++) {

quadratic time

M22_LIAN1878_11_GE_C22.indd 865 12/29/17 4:38 PM

866 Chapter 22 Developing Efficient Algorithms

k = k + i + j; } }

The first loop executes 10 times and the second loop 20 * n times. Thus, the time complexity for the loop is

T(n) = 10 * c + 20 * c * n = O(n)

Example 6 Consider the following selection statement:

if (list.contains(e)) { System.out.println(e); } else for (Object t: list) { System.out.println(t); }

Suppose the list contains n elements. The execution time for list.contains(e) is O(n). The loop in the else clause takes O(n) time. Hence, the time complexity for the entire statement is

T(n) = if test time + worst@case time (if clause, else clause)

= O(n) + O(n) = O(n)

Example 7 Consider the computation of an for an integer n. A simple algorithm would multiply a n times, as follows:

result = 1; for (int i = 1; i <= n; i++) result *= a;

The algorithm takes O(n) time. Without loss of generality, assume that n = 2k. You can improve the algorithm using the following scheme:

result = a; for (int i = 1; i <= k; i++) result = result * result;

The algorithm takes O(log n) time. For an arbitrary n, you can revise the algorithm and prove that the complexity is still O(log n). (See CheckPoint Question 22.3.5.)

Note An algorithm with the O(logn) time complexity is called a logarithmic algorithm and it exhibits a logarithmic growth rate. The base of the log is 2, but the base does not affect a logarithmic growth rate, so it can be omitted. In algorithm analysis, the base is usually 2.

22.3.1 Count the number of iterations in the following loops.

int count = 1; while (count < 30) { count = count * 2; }

int count = 15; while (count < 30) { count = count * 3; }

(a) (b)

omitting base

Point Check

M22_LIAN1878_11_GE_C22.indd 866 12/29/17 4:38 PM

int count = 1; while (count < n) { count = count * 2; }

int count = 15; while (count < n) { count = count * 3; }

22.3.2 How many stars are displayed in the following code if n is 10? How many if n is 20? Use the Big O notation to estimate the time complexity.

for (int i = 0; i < n; i++) { System.out.print('*'); }

for (int i = 0; i < n; i++) { for (int j = 0; j < n; j++) { System.out.print('*'); } }

(a) (b)

for (int k = 0; k < n; k++) { for (int i = 0; i < n; i++) { for (int j = 0; j < n; j++) { System.out.print('*'); } } }

for (int k = 0; k < 10; k++) { for (int i = 0; i < n; i++) { for (int j = 0; j < n; j++) { System.out.print('*'); } } }

22.3.3 Use the Big O notation to estimate the time complexity of the following methods:

public static void mA(int n) { for (int i = 0; i < n; i++) { System.out.print(Math.random()); } }

public static void mB(int n) { for (int i = 0; i < n; i++) { for (int j = 0; j < i; j++) System.out.print(Math.random()); } }

(a) (b)

public static void mC(int[ ] m) { for (int i = 0; i < m.length; i++) { System.out.print(m[i]); }

for (int i = m.length − 1; i >= 0; ) { System.out.print(m[i]); i−−; } }

public static void mD(int[] m) { for (int i = 0; i < m.length; i++) { for (int j = 0; j < i; j++) System.out.print(m[i] * m[j]); } }

22.3.4 Design an O(n) time algorithm for computing the sum of numbers from n1 to n2 for (n1 6 n2). Can you design an O(1) for performing the same task?

22.3.5 Example 7 in Section 22.3 assumes n = 2k. Revise the algorithm for an arbitrary n and prove that the complexity is still O(logn).

22.3 Examples: Determining Big O 867

M22_LIAN1878_11_GE_C22.indd 867 12/29/17 4:38 PM

868 Chapter 22 Developing Efficient Algorithms

22.4 Analyzing Algorithm Time Complexity This section analyzes the complexity of several well-known algorithms: binary search, selection sort, and Tower of Hanoi.

22.4.1 Analyzing Binary Search The binary search algorithm presented in Listing 7.7, BinarySearch.java, searches for a key in a sorted array. Each iteration in the algorithm contains a fixed number of operations, denoted by c. Let T(n) denote the time complexity for a binary search on a list of n elements. Without loss of generality, assume n is a power of 2 and k = logn. Since a binary search eliminates half of the input after two comparisons,

T(n) = Tan 2 b + c = Ta n

22 b + c + c = Ta n

2k b + kc

= T(1) + c logn = 1 + (logn)c

= O(logn)

Ignoring constants and nondominating terms, the complexity of the binary search algorithm is O(logn). This is a logarithmic algorithm. The logarithmic algorithm grows slowly as the prob- lem size increases. In the case of binary search, each time you double the array size, at most one more comparison will be required. If you square the input size of any logarithmic-time algo- rithm, you only double the time of execution. Therefore, a logarithmic-time algorithm is very efficient.

22.4.2 Analyzing Selection Sort The selection sort algorithm presented in Listing 7.8, SelectionSort.java, finds the smallest element in the list and swaps it with the first element. It then finds the smallest element remain- ing and swaps it with the first element in the remaining list, and so on until the remaining list contains only one element left to be sorted. The number of comparisons is n - 1 for the first iteration, n - 2 for the second iteration, and so on. Let T(n) denote the complexity for selection sort and c denote the total number of other operations such as assignments and additional comparisons in each iteration. Thus,

T(n) = (n - 1) + c + (n - 2) + c + g+ 2 + c + 1 + c

= (n - 1)(n - 1 + 1)

2 + c(n - 1) =

2 -

n 2

+ cn - c

= O(n2)

Therefore, the complexity of the selection sort algorithm is O(n2).

22.4.3 Analyzing the Tower of Hanoi Problem The Tower of Hanoi problem presented in Listing 18.8, TowerOfHanoi.java, recursively moves n disks from tower A to tower B with the assistance of tower C as follows:

1. Move the first n - 1 disks from A to C with the assistance of tower B.

Point Key

binary search animation on the Companion Website

logarithmic time

selection sort animation on the Companion Website

M22_LIAN1878_11_GE_C22.indd 868 12/29/17 4:38 PM

22.4 Analyzing Algorithm Time Complexity 869

2. Move disk n from A to B.

3. Move n - 1 disks from C to B with the assistance of tower A.

The complexity of this algorithm is measured by the number of moves. Let T(n) denote the number of moves for the algorithm to move n disks from tower A to tower B with T(1) = 1. Thus,

T(n) = T(n - 1) + 1 + T(n - 1)

= 2T(n - 1) + 1

= 2(2T(n - 2) + 1) + 1

= 2(2(2T(n - 3) + 1) + 1) + 1

= 2n - 1T(1) + 2n - 2 + g + 2 + 1

= 2n - 1 + 2n - 2 + g + 2 + 1 = (2n - 1) = O(2n)

An algorithm with O(2n) time complexity is called an exponential algorithm and it exhibits an exponential growth rate. As the input size increases, the time for the exponential algorithm grows exponentially. Exponential algorithms are not practical for large input size. Suppose the disk is moved at a rate of 1 per second. It would take 232/(365 * 24 * 60 * 60) = 136 years to move 32 disks and 264/(365 * 24 * 60 * 60) = 585 billion years to move 64 disks.

22.4.4 Common Recurrence Relations Recurrence relations are a useful tool for analyzing algorithm complexity. As shown in the preceding examples, the complexity for binary search, selection sort, and the Tower of Hanoi is

T(n) = T an 2 b + c, T(n) = T(n - 1) + O(n), and T(n) = 2T(n - 1) + O(1), respectively.

Table 22.2 summarizes the common recurrence relations.

O(2n) exponential time

Recurrence Relation Result Example

T(n) = T(n/2) + O(1) T(n) = O(logn) Binary search, Euclid’s GCD

T(n) = T(n - 1) + O(1) T(n) = O(n) Linear search

T(n) = 2T(n/2) + O(1) T(n) = O(n) CheckPoint Question 22.8.2

T(n) = 2T(n/2) + O(n) T(n) = O(n logn) Merge sort (Chapter 23) T(n) = T(n - 1) + O(n) T(n) = O(n2) Selection sort T(n) = 2T(n - 1) + O(1) T(n) = O(2n) Tower of Hanoi T(n) = T(n - 1) + T(n - 2) + O(1) T(n) = O(2n) Recursive Fibonacci algorithm

Table 22.2 Common Recurrence Functions

22.4.5 Comparing Common Growth Functions The preceding sections analyzed the complexity of several algorithms. Table 22.3 lists some common growth functions and shows how growth rates change as the input size doubles from n = 25 to n = 50.

M22_LIAN1878_11_GE_C22.indd 869 12/29/17 4:38 PM

870 Chapter 22 Developing Efficient Algorithms

These functions are ordered as follows, as illustrated in Figure 22.1.

O(1) 6 O(logn) 6 O(n) 6 O(n logn) 6 O(n2) 6 O(n3) 6 O(2n)

Function Name n = 25 n = 50 f(50) and f(25)

O(1) Constant time 1 1 1

O(logn) Logarithmic time 4.64 5.64 1.21

O(n) Linear time 25 50 2

O(n logn) Log-linear time 116 282 2.43

O(n2) Quadratic time 625 2,500 4

O(n3) Cubic time 15,625 125,000 8

O(2n) Exponential time 3.36 * 107 1.27 * 1015 3.35 * 107

Table 22.3 Change of Growth Rates

Figure 22.1 As the size n increases, the function grows.

O(1)

O(logn)

O(n)

O(n logn)

O(2n) O(n3) O(n2)

f(n)

22.4.1 Put the following growth functions in order:

5n3

4,032 , 44 logn, 10n logn, 500, 2n2,

45 , 3n

22.4.2 Estimate the time complexity for adding two n * m matrices and multiplying an n * m matrix by an m * k matrix.

22.4.3 Describe an algorithm for finding the occurrence of the max element in an array. Analyze the complexity of the algorithm.

22.4.4 Describe an algorithm for removing duplicates from an array. Analyze the complexity of the algorithm.

22.4.5 Analyze the following sorting algorithm:

for (int i = 0; i < list.length − 1; i++) { if (list[i] > list[i + 1]) { swap list[i] with list[i + 1]; i = −1; } }

22.4.6 Analyze the complexity for computing a polynomial f(x) of degree n for a given x value using a brute-force approach and the Horner’s approach, respectively. A brute- force approach is to compute each term in the polynomial and add them together. The Horner’s approach was introduced in Section 6.7.

f(x) = anxn + an - 1xn - 1 + an - 2xn - 2 + g+ a1x1 + a0

Point Check

M22_LIAN1878_11_GE_C22.indd 870 12/29/17 4:38 PM

22.5 Finding Fibonacci Numbers Using Dynamic Programming 871

22.5 Finding Fibonacci Numbers Using Dynamic Programming

This section analyzes and designs an efficient algorithm for finding Fibonacci numbers using dynamic programming.

Section 18.3, Case Study: Computing Fibonacci Numbers, gave a recursive method for finding the Fibonacci number, as follows:

/** The method for finding the Fibonacci number */ public static long fib(long index) { if (index == 0) // Base case return 0; else if (index == 1) // Base case return 1; else // Reduction and recursive calls return fib(index − 1) + fib(index − 2); }

We can now prove that the complexity of this algorithm is O(2n). For convenience, let the index be n. Let T(n) denote the complexity for the algorithm that finds fib(n), and c denote the constant time for comparing the index with 0 and 1. Thus,

T(n) = T(n - 1) + T(n - 2) + c

… 2T(n - 1) + c

… 2(2T(n - 2) + c) + c

= 22T(n - 2) + 2c + c

Similar to the analysis of the Tower of Hanoi problem, we can show that T(n) is O(2n). This algorithm is not efficient. Is there an efficient algorithm for finding a Fibonacci

number? The trouble with the recursive fib method is that the method is invoked redun- dantly with the same arguments. For example, to compute fib(4), fib(3) and fib(2) are invoked. To compute fib(3), fib(2) and fib(1) are invoked. Note fib(2) is redun- dantly invoked. We can improve it by avoiding repeatedly calling of the fib method with the same argument. Note a new Fibonacci number is obtained by adding the preceding two numbers in the sequence. If you use the two variables f0 and f1 to store the two preceding numbers, the new number, f2, can be immediately obtained by adding f0 with f1. Now you should update f0 and f1 by assigning f1 to f0 and assigning f2 to f1, as shown in Figure 22.2.

Point Key

Figure 22.2 Variables f0, f1, and f2 store three consecutive Fibonacci numbers in the series.

f0 f1 f2 Fibonacci series: 0 1 1 2 3 5 8 13 21 34 55 89 ... indices: 0 1 2 3 4 5 6 7 8 9 10 11

M22_LIAN1878_11_GE_C22.indd 871 12/29/17 4:38 PM

872 Chapter 22 Developing Efficient Algorithms

The new method is implemented in Listing 22.2.

lisTing 22.2 ImprovedFibonacci.java 1 import java.util.Scanner; 2 3 public class ImprovedFibonacci { 4 /** Main method */ 5 public static void main(String args[]) { 6 // Create a Scanner 7 Scanner input = new Scanner(System.in); 8 System.out.print("Enter an index for the Fibonacci number: "); 9 int index = input.nextInt(); 10 11 // Find and display the Fibonacci number 12 System.out.println( 13 "Fibonacci number at index " + index + " is " + fib(index)); 14 } 15 16 /** The method for finding the Fibonacci number */ 17 public static long fib(long n) { 18 long f0 = 0; // For fib(0) 19 long f1 = 1; // For fib(1) 20 long f2 = 1; // For fib(2) 21 22 if (n == 0) 23 return f0; 24 else if (n == 1) 25 return f1; 26 else if (n == 2) 27 return f2; 28 29 for (int i = 3; i <= n; i++) { 30 f0 = f1; 31 f1 = f2; 32 f2 = f0 + f1; 33 } 34 35 return f2; 36 } 37 }

input

invoke fib

f0 f1 f2

update f0, f1, f2

Enter an index for the Fibonacci number: 6 Fibonacci number at index 6 is 8

Enter an index for the Fibonacci number: 7 Fibonacci number at index 7 is 13

Obviously, the complexity of this new algorithm is O(n). This is a tremendous improvement over the recursive O(2n) algorithm.

The algorithm for computing Fibonacci numbers presented here uses an approach known as dynamic programming. Dynamic programming is the process of solving subproblems, then combining the solutions of the subproblems to obtain an overall solution. This naturally leads to a recursive solution. However, it would be inefficient to use recursion because the

O(n)

dynamic programming

M22_LIAN1878_11_GE_C22.indd 872 12/29/17 4:38 PM

22.6 Finding Greatest Common Divisors Using Euclid’s Algorithm 873

subproblems overlap. The key idea behind dynamic programming is to solve each subproblem only once and store the results for subproblems for later use to avoid redundant computing of the subproblems.

22.5.1 What is dynamic programming? Give an example of dynamic programming. 22.5.2 Why is the recursive Fibonacci algorithm inefficient, but the nonrecursive

Fibonacci algorithm efficient?

22.6 Finding Greatest Common Divisors Using Euclid’s Algorithm

This section presents several algorithms in the search for an efficient algorithm for finding the greatest common divisor of two integers.

The greatest common divisor (GCD) of two integers is the largest number that can evenly divide both integers. Listing 5.9, GreatestCommonDivisor.java, presented a brute-force algorithm for finding the GCD of two integers m and n. Brute force refers to an algorithmic approach that solves a problem in the simplest or most direct or obvious way. As a result, such an algorithm can end up doing far more work to solve a given problem than a cleverer or more sophisticated algorithm might do. On the other hand, a brute-force algorithm is often easier to implement than a more sophisticated one and, because of this simplicity, sometimes it can be more efficient.

The brute-force algorithm checks whether k (for k = 2, 3, 4, and so on) is a common divisor for n1 and n2, until k is greater than n1 or n2. The algorithm can be described as follows:

public static int gcd(int m, int n) { int gcd = 1;

for (int k = 2; k <= m && k <= n; k++) { if (m % k == 0 && n % k == 0) gcd = k; }

return gcd; }

Assuming m Ú n, the complexity of this algorithm is obviously O(n). Is there a better algorithm for finding the GCD? Rather than searching a possible divisor

from 1 up, it is more efficient to search from n down. Once a divisor is found, the divisor is the GCD. Therefore, you can improve the algorithm using the following loop:

for (int k = n; k >= 1; k−−) { if (m % k == 0 && n % k == 0) { gcd = k; break; } }

This algorithm is better than the preceding one, but its worst-case time complexity is still O(n). A divisor for a number n cannot be greater than n / 2, so you can further improve the

algorithm using the following loop:

for (int k = m / 2; k >= 1; k−−) { if (m % k == 0 && n % k == 0) { gcd = k; break; } }

Point Check

Point Key

GCD

brute force

assume m Ú n O(n)

improved solutions

M22_LIAN1878_11_GE_C22.indd 873 12/29/17 4:38 PM

874 Chapter 22 Developing Efficient Algorithms

However, this algorithm is incorrect because n can be a divisor for m. This case must be con- sidered. The correct algorithm is shown in Listing 22.3.

lisTing 22.3 GCD.java 1 import java.util.Scanner; 2 3 public class GCD { 4 /** Find GCD for integers m and n */ 5 public static int gcd(int m, int n) { 6 int gcd = 1; 7 8 if (m % n == 0) return n; 9 10 for (int k = n / 2; k >= 1; k−−) { 11 if (m % k == 0 && n % k == 0) { 12 gcd = k; 13 break; 14 } 15 } 16 17 return gcd; 18 } 19 20 /** Main method */ 21 public static void main(String[] args) { 22 // Create a Scanner 23 Scanner input = new Scanner(System.in); 24 25 // Prompt the user to enter two integers 26 System.out.print("Enter first integer: "); 27 int m = input.nextInt(); 28 System.out.print("Enter second integer: "); 29 int n = input.nextInt(); 30 31 System.out.println("The greatest common divisor for " + m + 32 " and " + n + " is " + gcd(m, n)); 33 } 34 }

check divisor

GCD found

input

Enter first integer: 2525

Enter second integer: 125 The greatest common divisor for 2525 and 125 is 25

Enter first integer: 3

Enter second integer: 3 The greatest common divisor for 3 and 3 is 3

Assuming m Ú n, the for loop is executed at most n/2 times, which cuts the time by half from the previous algorithm. The time complexity of this algorithm is still O(n), but practically, it is much faster than the algorithm in Listing 5.9.

O(n)

M22_LIAN1878_11_GE_C22.indd 874 12/29/17 4:38 PM

22.6 Finding Greatest Common Divisors Using Euclid’s Algorithm 875

Note The Big O notation provides a good theoretical estimate of algorithm efficiency. However, two algorithms of the same time complexity are not necessarily equally efficient. As shown in the preceding example, both algorithms in Listings 5.9 and 22.3 have the same complexity, but in practice, the one in Listing 22.3 is obviously better.

A more efficient algorithm for finding the GCD was discovered by Euclid around 300 b.c. This is one of the oldest known algorithms. It can be defined recursively as follows:

Let gcd(m, n) denote the GCD for integers m and n:

■■ If m % n is 0, gcd(m, n) is n.

■■ Otherwise, gcd(m, n) is gcd(n, m % n).

It is not difficult to prove the correctness of this algorithm. Suppose m % n = r. Thus, m = qn + r, where q is the quotient of m / n. Any number that divides m and n evenly must also divide r evenly. Therefore, gcd(m, n) is the same as gcd(n, r), where r = m % n. The algorithm can be implemented as in Listing 22.4.

lisTing 22.4 GCDEuclid.java 1 import java.util.Scanner; 2 3 public class GCDEuclid { 4 /** Find GCD for integers m and n */ 5 public static int gcd(int m, int n) { 6 if (m % n == 0) 7 return n; 8 else 9 return gcd(n, m % n); 10 } 11 12 /** Main method */ 13 public static void main(String[] args) { 14 // Create a Scanner 15 Scanner input = new Scanner(System.in); 16 17 // Prompt the user to enter two integers 18 System.out.print("Enter first integer: "); 19 int m = input.nextInt(); 20 System.out.print("Enter second integer: "); 21 int n = input.nextInt(); 22 23 System.out.println("The greatest common divisor for " + m + 24 " and " + n + " is " + gcd(m, n)); 25 } 26 }

practical consideration

Euclid’s algorithm

base case

reduction

input

Enter first integer: 2525 Enter second integer: 125 The greatest common divisor for 2525 and 125 is 25

Enter first integer: 3 Enter second integer: 3 The greatest common divisor for 3 and 3 is 3

M22_LIAN1878_11_GE_C22.indd 875 12/29/17 4:38 PM

876 Chapter 22 Developing Efficient Algorithms

In the best case when m % n is 0, the algorithm takes just one step to find the GCD. It is difficult to analyze the average case. However, we can prove that the worst-case time complex- ity is O(log n).

Assuming m Ú n, we can show that m % n < m / 2, as follows:

■■ If n <= m / 2, m % n < m / 2, since the remainder of m divided by n is always less than n.

■■ If n > m / 2, m % n = m – n < m / 2. Therefore, m % n < m / 2.

Euclid’s algorithm recursively invokes the gcd method. It first calls gcd(m, n), then calls gcd(n, m % n) and gcd(m % n, n % (m % n)), and so on, as follows:

gcd(m, n) = gcd(n, m % n) = gcd(m % n, n % (m % n)) = . . .

Since m % n < m / 2 and n % (m % n) < n / 2, the argument passed to the gcd method is reduced by half after every two iterations. After invoking gcd two times, the second parameter is less than n/2. After invoking gcd four times, the second parameter is less than n/4. After

invoking gcd six times, the second parameter is less than n

23 . Let k be the number of times the

gcd method is invoked. After invoking gcd k times, the second parameter is less than n

2(k/2) ,

which is greater than or equal to 1. That is,

2(k/2) Ú 1 1 n Ú 2(k/2) 1 logn Ú k/2 1 k … 2 logn

Therefore, k … 2 logn. Thus, the time complexity of the gcd method is O(log n). The worst case occurs when the two numbers result in the most divisions. It turns out that

two successive Fibonacci numbers will result in the most divisions. Recall that the Fibonacci series begins with 0 and 1, and each subsequent number is the sum of the preceding two numbers in the series, such as:

0 1 1 2 3 5 8 13 21 34 55 89 c

The series can be recursively defined as

fib(0) = 0; fib(1) = 1; fib(index) = fib(index − 2) + fib(index − 1); index >= 2

For two successive Fibonacci numbers fib(index) and fib(index − 1),

gcd(fib(index), fib(index − 1)) = gcd(fib(index − 1), fib(index − 2)) = gcd(fib(index − 2), fib(index − 3)) = gcd(fib(index − 3), fib(index − 4)) = . . . = gcd(fib(2), fib(1)) = 1

For example,

gcd(21, 13) = gcd(13, 8) = gcd(8, 5) = gcd(5, 3)

best case

average case

worst case

M22_LIAN1878_11_GE_C22.indd 876 12/29/17 4:38 PM

22.7 Efficient Algorithms for Finding Prime Numbers 877

= gcd(3, 2) = gcd(2, 1) = 1

Therefore, the number of times the gcd method is invoked is the same as the index. We can prove that index … 1.44 logn, where n = fib (index - 1). This is a tighter bound than index … 2 logn.

Table 22.4 summarizes the complexity of three algorithms for finding the GCD.

Algorithm Complexity Description

Listing 5.9 O(n) Brute-force, checking all possible divisors

Listing 22.3 O(n) Checking half of all possible divisors

Listing 22.4 O(log n) Euclid’s algorithm

Table 22.4 Comparisons of GCD Algorithms

22.6.1 Prove the following algorithm for finding the GCD of the two integers m and n is incorrect:

int gcd = 1; for (int k = Math.min(Math.sqrt(n), Math.sqrt(m)); k >= 1; k−−) { if (m % k == 0 && n % k == 0) { gcd = k; break; } }

22.7 Efficient Algorithms for Finding Prime Numbers This section presents several algorithms in the search for an efficient algorithm for finding prime numbers.

A $150,000 award awaits the first individual or group who discovers a prime number with at least 100,000,000 decimal digits (w2.eff.org/awards/coop-prime-rules.php).

Can you design a fast algorithm for finding prime numbers? An integer greater than 1 is prime if its only positive divisor is 1 or itself. For example, 2,

3, 5, and 7 are prime numbers, but 4, 6, 8, and 9 are not. How do you determine whether a number n is prime? Listing 5.15 presented a brute-force

algorithm for finding prime numbers. The algorithm checks whether 2, 3, 4, 5, . . . , or n − 1 divides n. If not, n is prime. This algorithm takes O(n) time to check whether n is prime. Note you need to check only whether 2, 3, 4, 5, . . . , and n/2 divides n. If not, n is prime. This algorithm is slightly improved, but it is still of O(n).

In fact, we can prove that if n is not a prime, n must have a factor that is greater than 1 and less than or equal to 2n. Here is the proof. Since n is not a prime, there exist two numbers p and q such that n = pq with 1 6 p … q. Note that n = 2n 2n. p must be less than or equal to 2n. Hence, you need to check only whether 2, 3, 4, 5, . . . , or 2n divides n. If not, n is prime. This significantly reduces the time complexity of the algorithm to O12n2 .

Now consider the algorithm for finding all the prime numbers up to n. A straightforward implementation is to check whether i is prime for i = 2, 3, 4, . . . , n. The program is given in Listing 22.5.

Point Check

Point Key

what is prime?

M22_LIAN1878_11_GE_C22.indd 877 12/29/17 4:38 PM

878 Chapter 22 Developing Efficient Algorithms

lisTing 22.5 PrimeNumbers.java 1 import java.util.Scanner; 2 3 public class PrimeNumbers { 4 public static void main(String[] args) { 5 Scanner input = new Scanner(System.in); 6 System.out.print("Find all prime numbers <= n, enter n: "); 7 int n = input.nextInt(); 8 9 final int NUMBER_PER_LINE = 10; // Display 10 per line 10 int count = 0; // Count the number of prime numbers 11 int number = 2; // A number to be tested for primeness 12 13 System.out.println("The prime numbers are:"); 14 15 // Repeatedly find prime numbers 16 while (number <= n) { 17 // Assume the number is prime 18 boolean isPrime = true; // Is the current number prime? 19 20 // Test if number is prime 21 for (int divisor = 2; divisor <= (int)(Math.sqrt(number)); 22 divisor++) { 23 if (number % divisor == 0) { // If true, number is not prime 24 isPrime = false; // Set isPrime to false 25 break; // Exit the for loop 26 } 27 } 28 29 // Print the prime number and increase the count 30 if (isPrime) { 31 count++; // Increase the count 32 33 if (count % NUMBER_PER_LINE == 0) { 34 // Print the number and advance to the new line 35 System.out.printf("%7d\n", number); 36 } 37 else 38 System.out.printf("%7d", number); 39 } 40 41 // Check if the next number is prime 42 number++; 43 } 44 45 System.out.println("\n" + count + 46 " prime(s) less than or equal to " + n); 47 } 48 }

check prime

increase count

check next number

Find all prime numbers <= n, enter n: 1000 The prime numbers are:

2 3 5 7 11 13 17 19 23 29

31 37 41 43 47 53 59 61 67 71

...

168 prime(s) less than or equal to 1000

M22_LIAN1878_11_GE_C22.indd 878 12/29/17 4:38 PM

22.7 Efficient Algorithms for Finding Prime Numbers 879

The program is not efficient if you have to compute Math.sqrt(number) for every iteration of the for loop (line 21). A good compiler should evaluate Math.sqrt(number) only once for the entire for loop. To ensure this happens, you can explicitly replace line 21 with the following two lines:

int squareRoot = (int)(Math.sqrt(number)); for (int divisor = 2; divisor <= squareRoot; divisor++) {

In fact, there is no need to actually compute Math.sqrt(number) for every number. You need to look only for the perfect squares such as 4, 9, 16, 25, 36, 49, and so on. Note for all the numbers between 36 and 48, inclusively, their (int)(Math.sqrt(number)) is 6. With this insight, you can replace the code in lines 16–26 with the following:

... int squareRoot = 1;

// Repeatedly find prime numbers while (number <= n) { // Assume the number is prime boolean isPrime = true; // Is the current number prime?

if (squareRoot * squareRoot < number) squareRoot++;

// Test if number is prime for (int divisor = 2; divisor <= squareRoot; divisor++) { if (number % divisor == 0) { // If true, number is not prime isPrime = false; // Set isPrime to false break; // Exit the for loop } } ...

Now we turn our attention to analyzing the complexity of this program. Since it takes 2i steps in the for loop (lines 21–27) to check whether number i is prime, the algorithm takes22 + 23 + 24 + g + 2n steps to find all the prime numbers less than or equal to n. Observe that 22 + 23 + 24 + g + 2n … n2n Therefore, the time complexity for this algorithm is O1n2n2 .

To determine whether i is prime, the algorithm checks whether 2, 3, 4, 5, . . . , and 2i are divisible by i. This algorithm can be further improved. In fact, you need to check only whether the prime numbers from 2 to 2i are possible divisors for i.

We can prove that if i is not prime, there must exist a prime number p such that i = pq and p … q. Here is the proof. Assume i is not prime; let p be the smallest factor of i. p must be prime, otherwise, p has a factor k with 2 … k 6 p. k is also a factor of i, which contradicts that p be the smallest factor of i. Therefore, if i is not prime, you can find a prime number from 2 to 2i that is divisible by i. This leads to a more efficient algorithm for finding all prime numbers up to n, as given in Listing 22.6.

lisTing 22.6 EfficientPrimeNumbers.java 1 import java.util.Scanner; 2 3 public class EfficientPrimeNumbers { 4 public static void main(String[] args) { 5 Scanner input = new Scanner(System.in); 6 System.out.print("Find all prime numbers <= n, enter n: ");

M22_LIAN1878_11_GE_C22.indd 879 12/29/17 4:38 PM

880 Chapter 22 Developing Efficient Algorithms

7 int n = input.nextInt(); 8 9 // A list to hold prime numbers 10 java.util.List<Integer> list = 11 new java.util.ArrayList<>(); 12 13 final int NUMBER_PER_LINE = 10; // Display 10 per line 14 int count = 0; // Count the number of prime numbers 15 int number = 2; // A number to be tested for primeness 16 int squareRoot = 1; // Check whether number <= squareRoot 17 18 System.out.println("The prime numbers are \n"); 19 20 // Repeatedly find prime numbers 21 while (number <= n) { 22 // Assume the number is prime 23 boolean isPrime = true; // Is the current number prime? 24 25 if (squareRoot * squareRoot < number) squareRoot++; 26 27 // Test whether number is prime 28 for (int k = 0; k < list.size() 29 && list.get(k) <= squareRoot; k++) { 30 if (number % list.get(k) == 0) { // If true, not prime 31 isPrime = false; // Set isPrime to false 32 break; // Exit the for loop 33 } 34 } 35 36 // Print the prime number and increase the count 37 if (isPrime) { 38 count++; // Increase the count 39 list.add(number); // Add a new prime to the list 40 if (count % NUMBER_PER_LINE == 0) { 41 // Print the number and advance to the new line 42 System.out.println(number); 43 } 44 else 45 System.out.print(number + " "); 46 } 47 48 // Check whether the next number is prime 49 number++; 50 } 51 52 System.out.println("\n" + count + 53 " prime(s) less than or equal to " + n); 54 } 55 }

check prime

increase count

check next number

Find all prime numbers <= n, enter n: 1000 The prime numbers are: 2 3 5 7 11 13 17 19 23 29 31 37 41 43 47 53 59 61 67 71 ... ... 168 prime(s) less than or equal to 1000

M22_LIAN1878_11_GE_C22.indd 880 12/29/17 4:38 PM

22.7 Efficient Algorithms for Finding Prime Numbers 881

Let p (i) denote the number of prime numbers less than or equal to i. The primes under 20 are 2, 3, 5, 7, 11, 13, 17, and 19. Therefore, p (2) is 1, p (3) is 2, p (6) is 3, and p (20) is 8.

It has been proved that p (i) is approximately i

logi (see primes.utm.edu/howmany.shtml).

For each number i, the algorithm checks whether a prime number less than or equal to 2i divides i. The number of the prime numbers less than or equal to 2i is2i

log2i = 22ilogi Thus, the complexity for finding all prime numbers up to n is

222 log 2

+ 223 log 3

+ 224 log 4

+ 225 log 5

+ 226 log 6

+ 227 log 7

+ 228 log 8

+ g + 22n logn

Since 2i

log i 6

2n logn

for i 6 n and n Ú 16,

222 log 2

+ 223 log 3

+ 224 log 4

+ 225 log 5

+ 226 log 6

+ 227 log 7

+ 228 log 8

+ g + 22n logn

6 2n2n logn

Therefore, the complexity of this algorithm is O¢ n2n logn

≤. This algorithm is another example of dynamic programming. The algorithm stores the

results of the subproblems in the array list and uses them later to check whether a new number is prime.

Is there any algorithm better than O¢ n2n logn

≤? Let us examine the well-known Eratosthenes algorithm for finding prime numbers. Eratosthenes (276–194 b.c.) was a Greek mathematician who devised a clever algorithm, known as the Sieve of Eratosthenes, for finding all prime numbers … n. His algorithm is to use an array named primes of n Boolean values. Initially, all elements in primes are set true. Since the multiples of 2 are not prime, set primes[2 * i] to false for all 2 … i … n/2, as shown in Figure 22.3. Since we don’t care about primes[0] and primes[1], these values are marked * in the figure.

dynamic programming

Sieve of Eratosthenes

Figure 22.3 The values in primes are changed with each prime number k.

3 3 T T T T T T T T T T T T T T T T T T T T T T T T T T initial

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27

k = 2

index

primes array

3 3 T T F T F T F T F T F T F T F T F T F T F T F T F T k = 3 3 3 T T F T F T F F F T F T F F F T F T F F F T F T F F

k = 5 3 3 T T F T F T F F F T F T F F F T F T F F F T F F F F

Since the multiples of 3 are not prime, set primes[3 * i] to false for all 3 … i … n/3. Because the multiples of 5 are not prime, set primes[5 * i] to false for all 5 … i … n/5. Note you don’t need to consider the multiples of 4 because the multiples of 4 are also the multiples of 2, which have already been considered. Similarly, multiples of 6, 8, and 9 need not be considered. You only need to consider the multiples of a prime number k = 2, 3, 5, 7, 11, . . . , and set the corresponding element in primes to false. Afterward, if primes[i] is

M22_LIAN1878_11_GE_C22.indd 881 12/29/17 4:38 PM

882 Chapter 22 Developing Efficient Algorithms

still true, then i is a prime number. As shown in Figure 22.3, 2, 3, 5, 7, 11, 13, 17, 19, and 23 are prime numbers. Listing 22.7 gives the program for finding the prime numbers using the Sieve of Eratosthenes algorithm.

lisTing 22.7 SieveOfEratosthenes.java 1 import java.util.Scanner; 2 3 public class SieveOfEratosthenes { 4 public static void main(String[] args) { 5 Scanner input = new Scanner(System.in); 6 System.out.print("Find all prime numbers <= n, enter n: "); 7 int n = input.nextInt(); 8 9 boolean[] primes = new boolean[n + 1]; // Prime number sieve 10 11 // Initialize primes[i] to true 12 for (int i = 0; i < primes.length; i++) { 13 primes[i] = true; 14 } 15 16 for (int k = 2; k <= n / k; k++) { 17 if (primes[k]) { 18 for (int i = k; i <= n / k; i++) { 19 primes[k * i] = false; // k * i is not prime 20 } 21 } 22 } 23 24 final int NUMBER_PER_LINE = 10; // Display 10 per line 25 int count = 0; // Count the number of prime numbers found so far 26 // Print prime numbers 27 for (int i = 2; i < primes.length; i++) { 28 if (primes[i]) { 29 count++; 30 if (count % NUMBER_PER_LINE == 0) 31 System.out.printf("%7d\n", i); 32 else 33 System.out.printf("%7d", i); 34 } 35 } 36 37 System.out.println("\n" + count + 38 " prime(s) less than or equal to " + n); 39 } 40 }

sieve

initialize sieve

nonprime

Find all prime numbers <= n, enter n: 1000 The prime numbers are: 2 3 5 7 11 13 17 19 23 29 31 37 41 43 47 53 59 61 67 71 ... ... 168 prime(s) less than or equal to 1000

Note k <= n / k (line 16). Otherwise, k * i would be greater than n (line 19). What is the time complexity of this algorithm?

M22_LIAN1878_11_GE_C22.indd 882 12/29/17 4:38 PM

22.8 Finding the Closest Pair of Points Using Divide-and-Conquer 883

For each prime number k (line 17), the algorithm sets primes[k * i] to false (line 19). This is performed n / k − k + 1 times in the for loop (line 18). Thus, the complexity for finding all prime numbers up to n is

n 2

- 2 + 1 + n 3

- 3 + 1 + n 5

- 5 + 1 + n 7

- 7 + 1 + n 11

- 11 + 1 c

= O an 2

+ n 3

+ n 5

+ n 7

+ n 11

+ g b 6 O(np(n))

= O ¢n 2n logn

≤

Algorithm Complexity Description

Listing 5.15 O(n2) Brute-force, checking all possible divisors

Listing 22.5 O1n2n2 Checking divisors up to 2n Listing 22.6

O¢n2n logn

≤ Checking prime divisors up to 2n Listing 22.7

O¢n2n logn

≤ Sieve of Eratosthenes Table 22.5 Comparisons of Prime-Number Algorithms

This upper bound O¢n2n logn

≤ is very loose. The actual time complexity is much better than O¢ n2n

logn ≤. The Sieve of Eratosthenes algorithm is good for a small n such that the array

primes can fit in the memory. Table 22.5 summarizes the complexity of these three algorithms for finding all prime

numbers up to n.

The number of items in the series is p(n).

22.7.1 Prove that if n is not prime, there must exist a prime number p such that p 6 = 2n and p is a factor of n.

22.7.2 Describe how the sieve of Eratosthenes is used to find the prime numbers.

Pedagogical Note The following sections present interesting and challenging problems. It is time that you begin to study advanced algorithms to become a proficient programmer. We recommend that you study the algorithms and implement them in the exercises.

22.8 Finding the Closest Pair of Points Using Divide-and-Conquer

This section presents efficient algorithms for finding the closest pair of points using divide-and-conquer.

Given a set of points, the closest-pair problem is to find the two points that are nearest to each other. As shown in Figure 22.4, a line is drawn to connect the two nearest points in the closest-pair animation.

Section 8.6, Case Study: Finding the Closest Pair, presented a brute-force algorithm for finding the closest pair of points. The algorithm computes the distances between all pairs of

Point Check

Point Key

closest-pair animation on Companion Website

M22_LIAN1878_11_GE_C22.indd 883 12/29/17 4:38 PM

884 Chapter 22 Developing Efficient Algorithms

points and finds the one with the minimum distance. Clearly, the algorithm takes O(n2) time. Can we design a more efficient algorithm?

We will use an approach called divide-and-conquer to solve this problem. The approach divides the problem into subproblems, solves the subproblems, then combines the solutions of the subproblems to obtain the solution for the entire problem. Unlike the dynamic programming approach, the subproblems in the divide-and-conquer approach don’t overlap. A subproblem is like the original problem with a smaller size, so you can apply recursion to solve the problem. In fact, all the solutions for recursive problems follow the divide-and-conquer approach.

Listing 22.8 describes how to solve the closest pair problem using the divide-and-conquer approach.

lisTing 22.8 Algorithm for Finding the Closest Pair Step 1: Sort the points in increasing order of x-coordinates. For the points with the same x-coordinates, sort on y-coordinates. This results in a sorted list S of points.

Step 2: Divide S into two subsets, S1 and S2, of equal size using the midpoint in the sorted list. Let the midpoint be in S1. Recursively find the closest pair in S1 and S2. Let d1 and d2 denote the distance of the closest pairs in the two subsets, respectively.

Step 3: Find the closest pair between a point in S1 and a point in S2 and denote their distance as d3. The closest pair is the one with the distance min(d1, d2, d3).

Selection sort takes O(n2) time. In Chapter 23, we will introduce merge sort and heap sort. These sorting algorithms take O(n log n) time. Step 1 can be done in O(n log n) time.

Step 3 can be done in O(n) time. Let d = min(d1, d2). We already know that the closest-pair distance cannot be larger than d. For a point in S1 and a point in S2 to form the closest pair in S, the left point must be in stripL and the right point in stripR, as illustrated in Figure 22.5a.

For a point p in stripL, you need only consider a right point within the d * 2d rectangle, as shown in 22.5b. Any right point outside the rectangle cannot form the closest pair with p. Since the closest-pair distance in S2 is greater than or equal to d, there can be at most six points in the rectangle. Thus, for each point in stripL, at most six points in stripR need to be considered.

For each point p in stripL, how do you locate the points in the corresponding d * 2d rectangle area in stripR? This can be done efficiently if the points in stripL and stripR are sorted in increasing order of their y-coordinates. Let pointsOrderedOnY be the list of the points sorted in increasing order of y-coordinates. pointsOrderedOnY can be obtained beforehand in the algorithm. stripL and stripR can be obtained from pointsOrderedOnY in Step 3 as given in Listing 22.9.

divide-and-conquer

Figure 22.4 The closet-pair animation draws a line to connect the closest pair of points dynamically as points are added and removed interactively. Source: Copyright © 1995–2016 Oracle and/or its affiliates. All rights reserved. Used with permission.

M22_LIAN1878_11_GE_C22.indd 884 12/29/17 4:38 PM

22.8 Finding the Closest Pair of Points Using Divide-and-Conquer 885

lisTing 22.9 Algorithm for Obtaining stripL and stripR 1 for each point p in pointsOrderedOnY 2 if (p is in S1 and mid.x − p.x <= d) 3 append p to stripL; 4 else if (p is in S2 and p.x − mid.x <= d) 5 append p to stripR;

Let the points in stripL and stripR be 5p0, p1, c , pk6 and 5q0, q1, c , qt6 , as shown in Figure 22.5c. The closest pair between a point in stripL and a point in stripR can be found using the algorithm described in Listing 22.10.

lisTing 22.10 Algorithm for Finding the Closest Pair in Step 3

1 d = min(d1, d2); 2 r = 0; // r is the index of a point in stripR 3 for (each point p in stripL) { 4 // Skip the points in stripR below p.y − d 5 while (r < stripR.length && q[r].y <= p.y − d) 6 r++; 7 8 let r1 = r; 9 while (r1 < stripR.length && |q[r1].y − p.y| <= d) { 10 // Check if (p, q[r1]) is a possible closest pair 11 if (distance(p, q[r1]) < d) { 12 d = distance(p, q[r1]); 13 (p, q[r1]) is now the current closest pair; 14 } 15 16 r1 = r1 + 1; 17 } 18 }

The points in stripL are considered from p0, p1, c , pk in this order. For a point p in stripL, skip the points in stripR that are below p.y − d (lines 5–6). Once a point is skipped, it will no longer be considered. The while loop (lines 9–17) checks whether (p, q[r1]) is a possible closest pair. There are at most six such q[r1] pairs, because the distance between two points in stripR cannot be less than d. Thus, the complexity for finding the closest pair in Step 3 is O(n).

stripL

stripR

update closest pair

Figure 22.5 The midpoint divides the points into two sets of equal size.

mid

(a) (b) (c)

stripL stripR

S1 S2

d d

stripL stripR

p d p

q[r]

stripL stripR

M22_LIAN1878_11_GE_C22.indd 885 12/29/17 4:38 PM

886 Chapter 22 Developing Efficient Algorithms

Note Step 1 in Listing 22.8 is performed only once to presort the points. Assume all the points are presorted. Let T(n) denote the time complexity for this algorithm. Thus, Step 2 Step 3

T(n) = 2T(n/2) + O(n) = O(n logn)

Therefore, the closest pair of points can be found in O(n log n) time. The complete implementa- tion of this algorithm is left as an exercise (see Programming Exercise 22.7).

22.8.1 What is divide-and-conquer approach? Give an example. 22.8.2 What is the difference between divide-and-conquer and dynamic programming? 22.8.3 Can you design an algorithm for finding the minimum element in a list using

divide-and-conquer? What is the complexity of this algorithm?

22.9 Solving the Eight Queens Problem Using Backtracking

This section solves the Eight Queens problem using the backtracking approach.

The Eight Queens problem is to find a solution to place a queen in each row on a chessboard such that no two queens can attack each other. The problem can be solved using recursion (see Programming Exercise 18.34). In this section, we will introduce a common algorithm design technique called backtracking for solving this problem. The backtracking approach searches for a candidate solution incrementally, abandoning that option as soon as it determines that the candidate cannot possibly be a valid solution, and then looks for a new candidate.

You can use a two-dimensional array to represent a chessboard. However, since each row can have only one queen, it is sufficient to use a one-dimensional array to denote the position of the queen in the row. Thus, you can define the queens array as

int[] queens = new int[8];

Assign j to queens[i] to denote that a queen is placed in row i and column j. Figure 22.6a shows the contents of the queens array for the chessboard in Figure 22.6b.

Point Check

Point Key

backtracking

(a) (b)

0 4 7 5 2 6

1 3

queens[0] queens[1] queens[2] queens[3] queens[4] queens[5] queens[6] queens[7]

The search starts from the first row with k = 0, where k is the index of the current row being considered. The algorithm checks whether a queen can be possibly placed in the jth column in the row for j = 0, 1, c , 7, in this order. The search is implemented as follows:

■■ If successful, it continues to search for a placement for a queen in the next row. If the current row is the last row, a solution is found.

■■ If not successful, it backtracks to the previous row and continues to search for a new placement in the next column in the previous row.

■■ If the algorithm backtracks to the first row and cannot find a new placement for a queen in this row, no solution can be found.

search algorithm

Eight Queens animation on the Companion Website

M22_LIAN1878_11_GE_C22.indd 886 12/29/17 4:38 PM

22.9 Solving the Eight Queens Problem Using Backtracking 887

To see how the algorithm works, go to http://liveexample.pearsoncmg.com/dsanimation/ EightQueens.html.

Listing 22.11 gives the program that displays a solution for the Eight Queens problem.

lisTing 22.11 EightQueens.java 1 import javafx.application.Application; 2 import javafx.geometry.Pos; 3 import javafx.stage.Stage; 4 import javafx.scene.Scene; 5 import javafx.scene.control.Label; 6 import javafx.scene.image.Image; 7 import javafx.scene.image.ImageView; 8 import javafx.scene.layout.GridPane; 9 10 public class EightQueens extends Application { 11 public static final int SIZE = 8; // The size of the chessboard 12 // queens are placed at (i, queens[i]) 13 // −1 indicates that no queen is currently placed in the ith row 14 // Initially, place a queen at (0, 0) in the 0th row 15 private int[] queens = {−1, −1, −1, −1, −1, −1, −1, −1}; 16 17 @Override // Override the start method in the Application class 18 public void start(Stage primaryStage) { 19 search(); // Search for a solution 20 21 // Display chessboard 22 GridPane chessBoard = new GridPane(); 23 chessBoard.setAlignment(Pos.CENTER); 24 Label[][] labels = new Label[SIZE][SIZE]; 25 for (int i = 0; i < SIZE; i++) 26 for (int j = 0; j < SIZE; j++) { 27 chessBoard.add(labels[i][j] = new Label(), j, i); 28 labels[i][j].setStyle("−fx−border−color: black"); 29 labels[i][j].setPrefSize(55, 55); 30 } 31 32 // Display queens 33 Image image = new Image("image/queen.jpg"); 34 for (int i = 0; i < SIZE; i++) 35 labels[i][queens[i]].setGraphic(new ImageView(image)); 36 37 // Create a scene and place it in the stage 38 Scene scene = new Scene(chessBoard, 55 * SIZE, 55 * SIZE); 39 primaryStage.setTitle("EightQueens"); // Set the stage title 40 primaryStage.setScene(scene); // Place the scene in the stage 41 primaryStage.show(); // Display the stage 42 } 43 44 /** Search for a solution */ 45 private boolean search() { 46 // k − 1 indicates the number of queens placed so far 47 // We are looking for a position in the kth row to place a queen 48 int k = 0; 49 while (k >= 0 && k < SIZE) { 50 // Find a position to place a queen in the kth row 51 int j = findPosition(k); 52 if (j < 0) { 53 queens[k] = −1; 54 k−−; // Backtrack to the previous row 55 } else {

queen positions

search for solution

create cells

set queen image

find a column

backtrack

M22_LIAN1878_11_GE_C22.indd 887 12/29/17 4:38 PM

888 Chapter 22 Developing Efficient Algorithms

56 queens[k] = j; 57 k++; 58 } 59 } 60 61 if (k == −1) 62 return false; // No solution 63 else 64 return true; // A solution is found 65 } 66 67 public int findPosition(int k) { 68 int start = queens[k] + 1; // Search for a new placement 69 70 for (int j = start; j < SIZE; j++) { 71 if (isValid(k, j)) 72 return j; // (k, j) is the place to put the queen now 73 } 74 75 return −1; 76 } 77 78 /** Return true if a queen can be placed at (row, column) */ 79 public boolean isValid(int row, int column) { 80 for (int i = 1; i <= row; i++) 81 if (queens[row − i] == column // Check column 82 || queens[row − i] == column − i // Check upleft diagonal 83 || queens[row − i] == column + i) // Check upright diagonal 84 return false; // There is a conflict 85 return true; // No conflict 86 } 87 }

The program invokes search() (line 19) to search for a solution. Initially, no queens are placed in any rows (line 15). The search now starts from the first row with k = 0 (line 48) and finds a place for the queen (line 51). If successful, place it in the row (line 56) and consider the next row (line 57). If not successful, backtrack to the previous row (lines 53–54).

The findPosition(k) method searches for a possible position to place a queen in row k starting from queen[k] + 1 (line 68). It checks whether a queen can be placed at start, start + 1, . . . , and 7, in this order (lines 70–73). If possible, return the column index (line 72); otherwise, return −1 (line 75).

The isValid(row, column) method is called to check whether placing a queen at the specified position causes a conflict with the queens placed earlier (line 71). It ensures that no queen is placed in the same column (line 81), in the upper-left diagonal (line 82), or in the upper-right diagonal (line 83), as shown in Figure 22.7.

place a queen search the next row

Figure 22.7 Invoking isValid(row, column) checks whether a queen can be placed at (row, column).

0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7

upright diagonal

check column

upleft (row, column)

M22_LIAN1878_11_GE_C22.indd 888 12/29/17 4:38 PM

22.10 Computational Geometry: Finding a Convex Hull 889

22.9.1 What is backtracking? Give an example. 22.9.2 If you generalize the Eight Queens problem to the n-Queens problem in an n-by-n

chessboard, what will be the complexity of the algorithm?

22.10 Computational Geometry: Finding a Convex Hull This section presents efficient geometric algorithms for finding a convex hull for a set of points.

Computational geometry is to study the algorithms for geometrical problems. It has appli- cations in computer graphics, games, pattern recognition, image processing, robotics, geo- graphical information systems, and computer-aided design and manufacturing. Section 22.8 presented a geometrical algorithm for finding the closest pair of points. This section introduces geometrical algorithms for finding a convex hull.

Given a set of points, a convex hull is the smallest convex polygon that encloses all these points, as shown in Figure 22.8a. A polygon is convex if every line connecting two vertices is inside the polygon. For example, the vertices v0, v1, v2, v3, v4, and v5 in Figure 22.8a form a convex polygon, but not in Figure 22.8b, because the line that connects v3 and v1 is not inside the polygon.

Point Check

Point Key

convex hull

(a) A convex hull (b) A nonconvex polygon (c) Convex hull animation

v5 v0

v0v5

A convex hull has many applications in game programming, pattern recognition, and image processing. Before we introduce the algorithms, it is helpful to get acquainted with the concept using an interactive tool from liveexample.pearsoncmg.com/dsanimation/ConvexHull.html, as shown in Figure 22.8c. This tool allows you to add and remove points and displays the convex hull dynamically.

Many algorithms have been developed to find a convex hull. This section introduces two popular algorithms: the gift-wrapping algorithm and Graham’s algorithm.

22.10.1 Gift-Wrapping Algorithm An intuitive approach called the gift-wrapping algorithm works as described in Listing 22.12.

lisTing 22.12 Finding a Convex Hull Using Gift- Wrapping Algorithm

Step 1: Given a list of points S, let the points in S be labeled s0, s1, . . . , sk. Select the rightmost lowest point S. As shown in Figure 22.9a, h0 is such a point. Add h0 to list H. (H is initially empty. H will hold all points in the convex hull after the algorithm is finished.) Let t0 be h0.

convex hull animation on the Companion Website

M22_LIAN1878_11_GE_C22.indd 889 12/29/17 4:38 PM

890 Chapter 22 Developing Efficient Algorithms

Step 2: Let t1 be s0. For every point p in S, if p is on the right side of the direct line from t0 to t1, then let t1 be p. (After Step 2, no points lie on the right side of the direct line from t0 to t1, as shown in Figure 22.9b.)

Step 3: If t1 is h0 (see Figure 22.9d), the points in H form a convex hull for S. Otherwise, add t1 to H, let t0 be t1, and go back to Step 2 (see Figure 22.9c).

The convex hull is expanded incrementally. The correctness is supported by the fact that no points lie on the right side of the direct line from t0 to t1 after Step 2. This ensures that every line segment with two points in S falls inside the polygon.

Finding the rightmost lowest point in Step 1 can be done in O(n) time. Whether a point is on the left side of a line, right side, or on the line can be determined in O(1) time (see Program- ming Exercise 3.32). Thus, it takes O(n) time to find a new point t1 in Step 2. Step 2 is repeated h times, where h is the size of the convex hull. Therefore, the algorithm takes O(hn) time. In the worst-case, h is n.

The implementation of this algorithm is left as an exercise (see Programming Exercise 22.9).

22.10.2 Graham’s Algorithm A more efficient algorithm was developed by Ronald Graham in 1972, as given in Listing 22.13.

lisTing 22.13 Finding a Convex Hull Using Graham’s Algorithm

Step 1: Given a list of points S, select the rightmost lowest point and name it p0. As shown in Figure 22.10a, p0 is such a point.

correctness of the algorithm

time complexity of the algorithm

Figure 22.9 (a) h0 is the rightmost lowest point in S. (b) Step 2 finds point t1. (c) A convex hull is expanded repeatedly. (d) A convex hull is found when t1 becomes h0.

(a) Step 1 (b) Step 2 (c) Repeat Step 2 (d) H is found

t1 = h0t0

t1 t0h0

Figure 22.10 (a) p0 is the rightmost lowest point in S. (b) Points are sorted by their angles. (c–d) A convex hull is discovered incrementally.

(a) Step 1 (b) Step 2 (c) p3 into H (d) p2 off H

p0 p0 p0 p0 XX

p1 p1

p1p2 p2 p2p3 p3

x-axis x-axis x-axis

Step 2: Sort the points in S angularly along the x-axis with p0 as the center, as shown in Figure 22.10b. If there is a tie and two points have the same angle, discard the one that is closer to p0. The points in S are now sorted as p0, p1, p2, . . . , pn−1.

M22_LIAN1878_11_GE_C22.indd 890 12/29/17 4:38 PM

Step 3: Push p0, p1, and p2 into stack H. (After the algorithm finishes, H contains all the points in the convex hull.)

Step 4: i = 3; while (i < n) { Let t1 and t2 be the top first and second element in stack H; if (pi is on the left side of the direct line from t2 to t1) { Push pi to H; i++; // Consider the next point in S. } else Pop the top element off stack H. }

Step 5: The points in H form a convex hull.

The convex hull is discovered incrementally. Initially, p0, p1, and p2 form a convex hull. Consider p3. p3 is outside of the current convex hull since points are sorted in increasing order of their angles. If p3 is strictly on the left side of the line from p1 to p2 (see Figure 22.10c), push p3 into H. Now p0, p1, p2, and p3 form a convex hull. If p3 is on the right side of the line from p1 to p2 (see Figure 22.10d), pop p2 out of H and push p3 into H. Now p0, p1, and p3 form a convex hull and p2 is inside this convex hull. You can prove by induction that all the points in H in Step 5 form a convex hull for all the points in the input list S.

Finding the rightmost lowest point in Step 1 can be done in O(n) time. The angles can be computed using trigonometry functions. However, you can sort the points without actually computing their angles. Observe p2 would make a greater angle than p1 if and only if p2 lies on the left side of the line from p0 to p1. Whether a point is on the left side of a line can be determined in O(1) time, as shown in Programming Exercise 3.32. Sorting in Step 2 can be done in O(n log n) time using the merge-sort or heap-sort algorithms that will be introduced in Chapter 23. Step 4 can be done in O(n) time. Therefore, the algorithm takes O(n logn) time.

The implementation of this algorithm is left as an exercise (see Programming Exercise 22.11).

22.10.1 What is a convex hull? 22.10.2 Describe the gift-wrapping algorithm for finding a convex hull. Should list H be

implemented using an ArrayList or a LinkedList?

22.10.3 Describe Graham’s algorithm for finding a convex hull. Why does the algorithm use a stack to store the points in a convex hull?

correctness of the algorithm

time complexity of the algorithm

Point Check

Key Terms average-case analysis 862 backtracking approach 886 best-case input 862 Big O notation 862 brute force 873 constant time 863 convex hull 889 divide-and-conquer approach 884

dynamic programming approach 872 exponential time 869 growth rate 862 logarithmic time 868 quadratic time 865 space complexity 863 time complexity 863 worst-case input 862

ChapTer summary 1. The Big O notation is a theoretical approach for analyzing the performance of an

algorithm. It estimates how fast an algorithm’s execution time increases as the input size increases, which enables you to compare two algorithms by examining their growth rates.

Chapter Summary 891

M22_LIAN1878_11_GE_C22.indd 891 12/29/17 4:38 PM

892 Chapter 22 Developing Efficient Algorithms

2. An input that results in the shortest execution time is called the best-case input, and one that results in the longest execution time is called the worst-case input. Best- and worst-case analyses are not representative, but worst-case analysis is very useful. You can be assured that the algorithm will never be slower than the worst case.

3. An average-case analysis attempts to determine the average amount of time among all possible input of the same size. Average-case analysis is ideal, but difficult to perform because for many problems, it is hard to determine the relative probabilities and distribu- tions of various input instances.

4. If the time is not related to the input size, the algorithm is said to take constant time with the notation O(1).

5. Linear search takes O(n) time. An algorithm with the O(n) time complexity is called a linear algorithm and it exhibits a linear growth rate. Binary search takes O(logn) time. An algorithm with the O(log n) time complexity is called a logarithmic algorithm and it exhibits a logarithmic growth rate.

6. The worst-time complexity for selection sort is O(n2). An algorithm with the O(n2) time complexity is called a quadratic algorithm and it exhibits a quadratic growth rate.

7. The time complexity for the Tower of Hanoi problem is O(2n). An algorithm with the O(2n) time complexity is called an exponential algorithm, and it exhibits an exponential growth rate.

8. A Fibonacci number at a given index can be found in O(n) time using dynamic program- ming approach.

9. Dynamic programming is the process of solving subproblems, then combining the solu- tions of the subproblems to obtain an overall solution. The key idea behind dynamic pro- gramming is to solve each subproblem only once and store the results for subproblems for later use to avoid redundant computing of the subproblems.

10. Euclid’s GCD algorithm takes O(logn) time.

11. All prime numbers less than or equal to n can be found in O¢ n2n logn

≤ time. 12. The closest pair can be found in O(n logn) time using the divide-and-conquer approach.

13. The divide-and-conquer approach divides the problem into subproblems, solves the sub- problems, and then combines the solutions of the subproblems to obtain the solution for the entire problem. Unlike the dynamic programming approach, the subproblems in the divide-and-conquer approach don’t overlap. A subproblem is like the original problem with a smaller size, so you can apply recursion to solve the problem.

14. The Eight Queens problem can be solved using backtracking.

15. The backtracking approach searches for a candidate solution incrementally, abandoning that option as soon as it determines the candidate cannot possibly be a valid solution, then looks for a new candidate.

16. A convex hull for a set of points can be found in O(n2) time using the gift-wrapping algorithm, and in O(n logn) time using the Graham’s algorithm.

M22_LIAN1878_11_GE_C22.indd 892 12/29/17 4:38 PM

Quiz Answer the quiz for this chapter online at the book Companion Website.

programming exerCiSeS *22.1 (Natural numbers and their squares) Knowing that (x + 1)2 = x2 + 2x + 1,

and that multiplication is more time consuming than addition, write an efficient program that displays the first ten natural numbers and their squares. Knowing that (x + 1)3 = x3 + 3x2 + 3x + 1, the same process can be followed to dis- play the cubes of these numbers.

Here is a sample run:

0^2 = 0

1^2 = 1

2^2 = 4

...

8^2 = 64

9^2 = 81

10^2 = 100

**22.2 (Efficient polynomial calculation) For a polynomial f(x) = anxn + an-1xn-1 + an - 2x

n - 2 + c + a1x + a0, write an efficient program that prompts the user to enter the degree n, the coefficients a0 to an, and the value for x, and computes and displays ƒ(x). Here is a sample run:

Enter n: 3

Enter x: 2

Enter a0: 1

Enter a1: 2

Enter a2: 3

Enter a3: 4

f(2.000000) = 49.000000

*22.3 (Pattern matching) Write an O(n) time program that prompts the user to enter two strings and tests whether the second string is a substring of the first string. Suppose the neighboring characters in the string are distinct. (Don’t use the indexOf method in the String class.) Here is a sample run of the program:

Enter a string s1: Welcome to Java

Enter a string s2: come

matched at index 3

*22.4 (Computing exponentials) Write a method public static double exp(double x, int n) that computes exp(x, n), which is an approximation of the exponential of x to the order n.

exp(x, n) = x0/0! + x1/1! + x2/2! + c + xn/n!

Programming Exercises 893

M22_LIAN1878_11_GE_C22.indd 893 12/29/17 4:38 PM

894 Chapter 22 Developing Efficient Algorithms

*22.5 (Fill and sort array) Write a program that randomly fills an array of integers and then sorts it. The array size is entered by the user and the values are chosen at random in {-1, 0, 1}. Analyze the time complexity of your program. Here is a sample run:

Enter the size: 10

1 0 1 −1 0 1 −1 0 0 0

−1 −1 0 0 0 0 0 1 1 1

*22.6 (Execution time for GCD) Write a program that obtains the execution time for finding the GCD of every two consecutive Fibonacci numbers from the index 46 to index 50 using the algorithms in Listings 22.3 and 22.4. Your program should print a table like this:

46 47 48 49 50

Listing 22.3 GCD

Listing 22.4 GCDEuclid

**22.7 (Closest pair of points) Section 22.8 introduced an algorithm for finding the closest pair of points using a divide-and-conquer approach. Implement the algorithm to meet the following requirements:

■■ Define a class named Pair with the data fields p1 and p2 to represent two points and a method named getDistance() that returns the distance between the two points.

■■ Implement the following methods:

/** Return the distance of the closest pair of points */ public static Pair getClosestPair(double[][] points)

/** Return the distance of the closest pair of points */ public static Pair getClosestPair(Point2D[] points)

/** Return the distance of the closest pair of points * in pointsOrderedOnX[low..high]. This is a recursive * method. pointsOrderedOnX and pointsOrderedOnY are * not changed in the subsequent recursive calls. */ public static Pair distance(Point2D[] pointsOrderedOnX, int low, int high, Point2D[] pointsOrderedOnY)

/** Compute the distance between two points p1 and p2 */ public static double distance(Point2D p1, Point2D p2)

/** Compute the distance between points (x1, y1) and (x2, y2) */ public static double distance(double x1, double y1, double x2, double y2)

exp(1.0, 0) = 1.0 exp(1.0, 2) = 2.5 exp(1.0, 4) = 2.708333333333333 exp(1.0, 6) = 2.7180555555555554 exp(1.0, 8) = 2.71827876984127 exp(1.0, 10) = 2.7182818011463845

Analyze the time complexity of your method. Here is a sample run:

M22_LIAN1878_11_GE_C22.indd 894 12/29/17 4:38 PM

**22.8 (All prime numbers up to 10,000,000,000) Write a program that finds all prime numbers up to 10,000,000,000. There are approximately 455,052,511 such prime numbers. Your program should meet the following requirements:

■■ Your program should store the prime numbers in a binary data file, named PrimeNumbers.dat. When a new prime number is found, the number is appended to the file.

■■ To find whether a new number is prime, your program should load the prime numbers from the file to an array of the long type of size 10000. If no number in the array is a divisor for the new number, continue to read the next 10000 prime numbers from the data file, until a divisor is found or all numbers in the file are read. If no divisor is found, the new number is prime.

■■ Since this program takes a long time to finish, you should run it as a batch job from a UNIX machine. If the machine is shut down and rebooted, your program should resume by using the prime numbers stored in the binary data file rather than start over from scratch.

**22.9 (Geometry: gift-wrapping algorithm for finding a convex hull) Section 22.10.1 introduced the gift-wrapping algorithm for finding a convex hull for a set of points. Assume Java’s coordinate system is used for the points. Implement the algorithm using the following method:

/** Return the points that form a convex hull */ public static ArrayList<Point2D> getConvexHull(double[][] s)

Point2D is defined in Section 9.6.3.

Write a test program that prompts the user to enter the set size and the points, and displays the points that form a convex hull. Here is a sample run:

Programming Exercises 895

How many points are in the set? 6

Enter 6 points: 1 2.4 2.5 2 1.5 34.5 5.5 6 6 2.4 5.5 9 The convex hull is (1.5, 34.5) (5.5, 9.0) (6.0, 2.4) (2.5, 2.0) (1.0, 2.4)

22.10 (Number of prime numbers) Programming Exercise 22.8 stores the prime numbers in a file named PrimeNumbers.dat. Write a program that finds the num- ber of prime numbers that are less than or equal to 12, 120, 1,200, 12,000, 100,000, 1,200,000, 12,000,000, 120,000,000, 1,200,000,000, and 12,000,000,000. Your program should read the data from PrimeNumbers.dat.

**22.11 (Geometry: Graham’s algorithm for finding a convex hull) Section 22.10.2 introduced Graham’s algorithm for finding a convex hull for a set of points. Assume Java’s coordinate system is used for the points. Implement the algo- rithm using the following method:

/** Return the points that form a convex hull */ public static ArrayList<MyPoint> getConvexHull(double[][] s)

MyPoint is a static inner class defined as follows: private static class MyPoint implements Comparable<MyPoint> { double x, y;

MyPoint rightMostLowestPoint;

M22_LIAN1878_11_GE_C22.indd 895 12/29/17 4:38 PM

896 Chapter 22 Developing Efficient Algorithms

MyPoint(double x, double y) { this.x = x; this.y = y; }

public void setRightMostLowestPoint(MyPoint p) { rightMostLowestPoint = p; }

@Override public int compareTo(MyPoint o) { // Implement it to compare this point with point o // angularly along the x-axis with rightMostLowestPoint // as the center, as shown in Figure 22.10b. By implementing // the Comparable interface, you can use the Array.sort // method to sort the points to simplify coding. } }

Write a test program that prompts the user to enter the set size and the points, and displays the points that form a convex hull. Here is a sample run:

How many points are in the set? 6

Enter six points: 1 2.4 2.5 2 1.5 34.5 5.5 6 6 2.4 5.5 9 The convex hull is (1.5, 34.5) (5.5, 9.0) (6.0, 2.4) (2.5, 2.0) (1.0, 2.4)

*22.12 (Last 100 prime numbers) Programming Exercise 22.8 stores the prime numbers in a file named PrimeNumbers.dat. Write an efficient program that reads the last 100 numbers in the file. (Hint: Don’t read all numbers from the file. Skip all numbers before the last 100 numbers in the file.)

*22.13 (Geometry: convex hull animation) Programming Exercise 22.11 finds a convex hull for a set of points entered from the console. Write a program that enables the user to add or remove points by clicking the left or right mouse button and displays a convex hull, as shown in Figure 22.8c.

*22.14 (Execution time for prime numbers) Write a program that obtains the execu- tion time for finding all the prime numbers less than 9,000,000, 11,000,000, 13,000,000, 15,000,000, 17,000,000, and 19,000,000 using the algorithms in Listings 22.5–22.7. Your program should print a table like this:

9000000 11000000 13000000 15000000 17000000 19000000

Listing 22.5

Listing 22.6

Listing 22.7

**22.15 (Geometry: noncrossed polygon) Write a program that enables the user to add or remove points by clicking the left or right mouse button and displays a non- crossed polygon that links all the points, as shown in Figure 22.11a. A polygon is crossed if two or more sides intersect, as shown in Figure 22.11b. Use the following algorithm to construct a polygon from a set of points:

M22_LIAN1878_11_GE_C22.indd 896 12/29/17 4:38 PM

Step 1: Given a set of points S, select the rightmost lowest point p0 in the set S.

Step 2: Sort the points in S angularly along the x-axis with p0 as the center. If there is a tie and two points have the same angle, the one that is closer to p0 is considered greater. The points in S are now sorted as p0, p1, p2, ..., pn−1.

Step 3: The sorted points form a noncrossed polygon.

**22.16 (Linear search animation) Write a program that animates the linear search algorithm. Create an array that consists of 25 distinct numbers from 1 to 25 in a random order. The array elements are displayed in a histogram, as shown in Figure 22.12. You need to enter a search key in the text field. Clicking the Step button causes the program to perform one comparison in the algorithm and repaints the histogram with a bar indicating the search position. This button also freezes the text field to prevent its value from being changed. When the algorithm is finished, display the status in the label at the top of the border pane to inform the user. Clicking the Reset button creates a new random array for a new start. This button also makes the text field editable.

Figure 22.11 (a) Programming Exercise 22.15 displays a noncrossed polygon for a set of points. Source: Copyright © 1995–2016 Oracle and/or its affiliates. All rights reserved. Used with permission. (b) Two or more sides intersect in a crossed polygon.

(a) (b) Crossed polygon

Programming Exercises 897

Figure 22.12 The program animates a linear search.

**22.17 (Closest-pair animation) Write a program that enables the user to add/remove points by clicking the left/right mouse button and displays a line that connects the pair of nearest points, as shown in Figure 22.4.

**22.18 (Binary search animation) Write a program that animates the binary search algorithm. Create an array with numbers from 1 to 25 in this order. The array elements are displayed in a histogram, as shown in Figure 22.13. You need to

M22_LIAN1878_11_GE_C22.indd 897 12/29/17 4:38 PM

898 Chapter 22 Developing Efficient Algorithms

enter a search key in the text field. Clicking the Step button causes the program to perform one comparison in the algorithm. Use a light-gray color to paint the bars for the numbers in the current search range and use a black color to paint the bar indicating the middle number in the search range. The Step button also freezes the text field to prevent its value from being changed. When the algorithm is finished, display the status in a label at the top of a border pane. Clicking the Reset button enables a new search to start. This button also makes the text field editable.

Figure 22.13 The program animates a binary search.

(a) (b)

*22.19 (Largest block) The problem for finding a largest block is described in Programming Exercise 8.35. Design a dynamic programming algorithm for solving this problem in O(n2) time. Write a test program that displays a 10- by- 10 square matrix, as shown in Figure 22.14a. Each element in the matrix is 0 or 1, randomly generated with a click of the Refresh button. Display each number centered in a text field. Use a text field for each entry. Allow the user to change the entry value. Click the Find Largest Block button to find a largest square submatrix that consists of 1s. Highlight the numbers in the block, as shown in Figure 22.14b. See liveexample.pearsoncmg.com/dsanimation/LargestBlock .html for an interactive test.

M22_LIAN1878_11_GE_C22.indd 898 12/29/17 4:38 PM

***22.20 (Game: multiple Sudoku solutions) The complete solution for the Sudoku problem is given in Supplement VI.A. A Sudoku problem may have multiple solutions. Modify Sudoku.java in Supplement VI.A to display the total number of solutions. Display two solutions if multiple solutions exist.

***22.21 (Game: Sudoku) The complete solution for the Sudoku problem is given in Supplement VI.C. Write a program that lets the user enter the input from the text fields, as shown in Figure 22.15a. Clicking the Solve button displays the result, as shown in Figures 22.15b and c.

Programming Exercises 899

(a) (b) (c)

***22.22 (Game: recursive Sudoku) Write a recursive solution for the Sudoku problem. ***22.23 (Game: multiple Eight Queens solution) Write a program to display all possible

solutions for the Eight Queens puzzle in a scroll pane, as shown in Figure 22.16. For each solution, put a label to denote the solution number. (Hint: Place all solution panes into an HBox and place this one pane into a ScrollPane. If you run into a StackOverflowError, run the program using java –Xss200m Exercise22_23 from the command window.)

M22_LIAN1878_11_GE_C22.indd 899 12/29/17 4:38 PM

900 Chapter 22 Developing Efficient Algorithms

*22.24 (Find the smallest number) Write a method that uses the divide-and-conquer approach to find the smallest number in a list.

***22.25 (Game: Sudoku) Revise Programming Exercise 22.21 to display all solutions for the Sudoku game, as shown in Figure 22.17a. When you click the Solve button, the program stores all solutions in an ArrayList. Each element in the list is a two-dimensional 9-by-9 grid. If the program has multiple solutions, the Next button appears as shown in Figure 22.17b. You can click the Next button to display the next solution and also add a label to show the solution count. When the Clear button is clicked, the cells are cleared and the Next button is hidden as shown in Figure 22.17c.

(a) (b) (c)

Enter the number of objects: 6 Enter the weights of the objects: 7 5 2 3 5 8 Container 1 contains objects with weight 2 3 5 Container 2 contains objects with weight 5 Container 3 contains objects with weight 7 Container 4 contains objects with weight 8

*22.26 (Bin packing with smallest object first) The bin packing problem is to pack the objects of various weights into containers. Assume each container can hold a maximum of 10 pounds. The program uses an algorithm that places an object with the smallest weight into the first bin in which it would fit. Your program should prompt the user to enter the total number of objects and the weight of each object. The program displays the total number of containers needed to pack the objects, and the contents of each container. Here is a sample run of the program:

Does this program produce an optimal solution, that is, finding the minimum number of containers to pack the objects?

M22_LIAN1878_11_GE_C22.indd 900 12/29/17 4:38 PM

**22.27 (Optimal bin packing) Rewrite the preceding program so that it finds an optimal solution that packs all objects using the smallest number of containers. Here is a sample run of the program:

Programming Exercises 901

Enter the number of objects: 6 Enter the weights of the objects: 7 5 2 3 5 8 Container 1 contains objects with weight 7 3 Container 2 contains objects with weight 5 5 Container 3 contains objects with weight 2 8 The optimal number of bins is 3

What is the time complexity of your program?

M22_LIAN1878_11_GE_C22.indd 901 12/29/17 4:38 PM

A01_GIPE3898_09_SE_FM.indd 3 12/20/17 5:56 PM

This page intentionally left blank

Objectives ■■ To study and analyze time complexity of various sorting algorithms

(§§23.2–23.7).

■■ To design, implement, and analyze insertion sort (§23.2).

■■ To design, implement, and analyze bubble sort (§23.3).

■■ To design, implement, and analyze merge sort (§23.4).

■■ To design, implement, and analyze quick sort (§23.5).

■■ To design and implement a binary heap (§23.6).

■■ To design, implement, and analyze heap sort (§23.6).

■■ To design, implement, and analyze bucket and radix sorts (§23.7).

■■ To design, implement, and analyze external sort for files that have a large amount of data (§23.8).

Sorting

Chapter

M23_LIAN1878_11_GE_C23.indd 903 12/29/17 4:40 PM

904 Chapter 23 Sorting

23.1 Introduction Sorting algorithms are good examples for studying algorithm design and analysis.

When president Barack Obama visited Google in 2007, Google as a candidate CEO Eric Schmidt asked Obama the most efficient way to sort a million 32-bit integers (www.youtube .com/watch?v=k4RRi_ntQc8). Obama answered that the bubble sort would be the wrong way to go. Was he right? We will examine different sorting algorithms in this chapter and see if he was correct.

Sorting is a classic subject in computer science. There are three reasons to study sorting algorithms.

■■ First, sorting algorithms illustrate many creative approaches to problem solving, and these approaches can be applied to solve other problems.

■■ Second, sorting algorithms are good for practicing fundamental programming tech- niques using selection statements, loops, methods, and arrays.

■■ Third, sorting algorithms are excellent examples to demonstrate algorithm performance.

The data to be sorted might be integers, doubles, characters, or objects. Section 7.11, Sorting Arrays, presented selection sort. The selection-sort algorithm was extended to sort an array of objects in Section 19.5, Case Study: Sorting an Array of Objects. The Java API contains several overloaded sort methods for sorting primitive-type values and objects in the java .util.Arrays and java.util.Collections classes. For simplicity, this chapter assumes

1. data to be sorted are integers,

2. data are stored in an array, and

3. data are sorted in ascending order.

The programs can be easily modified to sort other types of data, to sort in descending order, or to sort data in an ArrayList or a LinkedList.

There are many algorithms for sorting. You have already learned selection sort. This chap- ter introduces insertion sort, bubble sort, merge sort, quick sort, bucket sort, radix sort, and external sort.

23.2 Insertion Sort The insertion-sort algorithm sorts a list of values by repeatedly inserting a new ele- ment into a sorted sublist until the whole list is sorted.

Figure 23.1 shows how to sort the list {2, 9, 5, 4, 8, 1, 6} using insertion sort. For an interac- tive demo on how insertion sort works, go to liveexample.pearsoncmg.com/dsanimation/ InsertionSortNeweBook.html.

The algorithm can be described as follows:

for (int i = 1; i < list.length; i++) { insert list[i] into a sorted sublist list[0..i−1] so that list[0..i] is sorted. }

To insert list[i] into list[0..i−1], save list[i] into a temporary variable, say currentElement. Move list[i−1] to list[i] if list[i−1] > currentElement, move list[i−2] to list[i−1] if list[i−2] > currentElement, and so on, until list[i−k] <= currentElement or k > i (we pass the first element of the sorted list). Assign currentElement to list[i−k+1]. For example, to insert 4 into {2, 5, 9} in Step 4 in Figure 23.2, move list[2] (9) to list[3] since 9 > 4 and move list[1] (5) to list[2] since 5 > 4. Finally, move currentElement (4) to list[1].

Point Key

why study sorting?

what data to sort?

Point Key

insertion-sort animation on Companion Website

M23_LIAN1878_11_GE_C23.indd 904 12/29/17 4:40 PM

23.2 Insertion Sort 905

The algorithm can be expanded and implemented as in Listing 23.1.

lisTing 23.1 InsertionSort.java 1 public class InsertionSort { 2 /** The method for sorting the numbers */ 3 public static void insertionSort(int[] list) { 4 for (int i = 1; i < list.length; i++) { 5 /** Insert list[i] into a sorted sublist list[0..i−1] so that 6 list[0..i] is sorted. */ 7 int currentElement = list[i]; 8 int k; 9 for (k = i − 1; k >= 0 && list[k] > currentElement; k−−) { 10 list[k + 1] = list[k];

shift

Figure 23.1 Insertion sort repeatedly inserts a new element into a sorted sublist.

Step 1: Initially, the sorted sublist contains the �rst element in the list. Insert 9 into the sublist.

Step 2: The sorted sublist is {2, 9}. Insert 5 into the sublist.

Step 3: The sorted sublist is {2, 5, 9}. Insert 4 into the sublist.

Step 4: The sorted sublist is {2, 4, 5, 9}. Insert 8 into the sublist.

Step 5: The sorted sublist is {2, 4, 5, 8, 9}. Insert 1 into the sublist.

Step 6: The sorted sublist is {1, 2, 4, 5, 8, 9}. Insert 6 into the sublist.

Step 7: The entire list is now sorted.

Figure 23.2 A new element is inserted into a sorted sublist.

[0] [1] [2] [3] [4] [5]

2list 5 9 4

[6]

[0] [1] [2] [3] [4] [5]

2list 4 5 9

[6]

[6][0] [1] [2] [3] [4] [5]

2list 5 9

2 5 9list

Step 1: Save 4 to a temporary variable currentElement

currentElement: 4

Step 2: Move list[2] to list[3]

Step 3: Move list[1] to list[2]

Step 4: Assign currentElement to list[1]

M23_LIAN1878_11_GE_C23.indd 905 12/29/17 4:40 PM

906 Chapter 23 Sorting

11 } 12 13 // Insert the current element into list[k + 1] 14 list[k + 1] = currentElement; 15 } 16 } 17 }

The insertionSort(int[] list) method sorts an array of int elements. The method is implemented with a nested for loop. The outer loop (with the loop control variable i) (line 4) is iterated in order to obtain a sorted sublist, which ranges from list[0] to list[i]. The inner loop (with the loop control variable k) inserts list[i] into the sublist from list[0] to list[i−1].

To better understand this method, trace it with the following statements:

int[] list = {1, 9, 4, 6, 5, −4}; InsertionSort.insertionSort(list);

The insertion-sort algorithm presented here sorts a list of elements by repeatedly inserting a new element into a sorted partial array until the whole array is sorted. At the kth iteration, to insert an element into an array of size k, it may take k comparisons to find the insertion position and k moves to insert the element. Let T(n) denote the complexity for insertion sort, and c denote the total number of other operations such as assignments and additional comparisons in each iteration. Thus,

T(n) = (2 + c) + (2 * 2 + c) + g + (2 * (n - 1) + c)

= 2(1 + 2 + g + n - 1) + c(n - 1)

= 2 (n - 1)n

2 + cn - c = n2 - n + cn - c

= O(n2)

Therefore, the complexity of the insertion-sort algorithm is O(n2). Hence, the selection and insertion sorts are of the same time complexity.

23.2.1 Describe how an insertion sort works. What is the time complexity for an insertion sort?

23.2.2 Use Figure 23.1 as an example to show how to apply an insertion sort on {45, 11, 50, 59, 60, 2, 4, 7, 10}.

23.2.3 If a list is already sorted, how many comparisons will the insertionSort method perform?

23.3 Bubble Sort A bubble sort sorts the array in multiple passes. Each pass successively swaps the neighboring elements if the elements are not in order.

The bubble-sort algorithm makes several passes through the array. On each pass, successive neighboring pairs are compared. If a pair is in decreasing order, its values are swapped; other- wise, the values remain unchanged. The technique is called a bubble sort or sinking sort because the smaller values gradually “bubble” their way to the top and the larger values sink to the bottom. After the first pass, the last element becomes the largest in the array. After the second pass, the second-to-last element becomes the second largest in the array. This process is continued until all elements are sorted.

insert

insertion-sort time complexity

Point Check

Point Key

bubble sort

M23_LIAN1878_11_GE_C23.indd 906 12/29/17 4:40 PM

23.3 Bubble Sort 907

Figure 23.3a shows the first pass of a bubble sort on an array of six elements (2 9 5 4 8 1). Compare the elements in the first pair (2 and 9) and no swap is needed because they are already in order. Compare the elements in the second pair (9 and 5) and swap 9 with 5 because 9 is greater than 5. Compare the elements in the third pair (9 and 4) and swap 9 with 4. Compare the elements in the fourth pair (9 and 8) and swap 9 with 8. Compare the elements in the fifth pair (9 and 1) and swap 9 with 1. The pairs being compared are highlighted and the numbers already sorted are italicized in Figure 23.3. For an interactive demo on how bubble sort works, go to liveexample.pearsoncmg.com/dsanimation/BubbleSortNeweBook.html.

bubble-sort illustration

bubble sort animation on the Companion Website

Figure 23.3 Each pass compares and orders the pairs of elements sequentially.

(a) 1st pass (b) 2nd pass (c) 3rd pass (d) 4th pass (e) 5th pass

2 5 9 4 8 1

2 4 8 15 9

2 4 9 15 8

2 4 1 95 8

2 5 8 19 4

2 5 8 94 1

2 1 8 94 5

2 4 8 91 5

2 1 8 94 5 1 4 8 92 52 4 1 95 8

2 4 5 8 1 9

2 4 5 1 8 9

The first pass places the largest number (9) as the last in the array. In the second pass, as shown in Figure 23.3b, you compare and order pairs of elements sequentially. There is no need to consider the last pair because the last element in the array is already the largest. In the third pass, as shown in Figure 23.3c, you compare and order pairs of elements sequentially except the last two elements because they are already in order. Thus, in the kth pass, you don’t need to consider the last k - 1 elements because they are already ordered.

The algorithm for a bubble sort is described in Listing 23.2.

Listing 23.2 Bubble-Sort Algorithm 1 for (int k = 1; k < list.length; k++) { 2 // Perform the kth pass 3 for (int i = 0; i < list.length − k; i++) { 4 if (list[i] > list[i + 1]) 5 swap list[i] with list[i + 1]; 6 } 7 }

Note if no swap takes place in a pass, there is no need to perform the next pass because all the elements are already sorted. You can use this property to improve the algorithm in Listing 23.2, as in Listing 23.3.

Listing 23.3 Improved Bubble-Sort Algorithm 1 boolean needNextPass = true; 2 for (int k = 1; k < list.length && needNextPass; k++) { 3 // Array may be sorted and next pass not needed 4 needNextPass = false; 5 // Perform the kth pass 6 for (int i = 0; i < list.length – k; i++) { 7 if (list[i] > list[i + 1]) { 8 swap list[i] with list[i + 1]; 9 needNextPass = true; // Next pass still needed 10 } 11 } 12 }

algorithm

M23_LIAN1878_11_GE_C23.indd 907 12/29/17 4:40 PM

908 Chapter 23 Sorting

The algorithm can be implemented in Listing 23.4.

Listing 23.4 BubbleSort.java 1 public class BubbleSort { 2 /** Bubble sort method */ 3 public static void bubbleSort(int[] list) { 4 boolean needNextPass = true; 5 6 for (int k = 1; k < list.length && needNextPass; k++) { 7 // Array may be sorted and next pass not needed 8 needNextPass = false; 9 for (int i = 0; i < list.length − k; i++) { 10 if (list[i] > list[i + 1]) { 11 // Swap list[i] with list[i + 1] 12 int temp = list[i]; 13 list[i] = list[i + 1]; 14 list[i + 1] = temp; 15 16 needNextPass = true; // Next pass still needed 17 } 18 } 19 } 20 } 21 22 /** A test method */ 23 public static void main(String[] args) { 24 int[] list = {2, 3, 2, 5, 6, 1, −2, 3, 14, 12}; 25 bubbleSort(list); 26 for (int i = 0; i < list.length; i++) 27 System.out.print(list[i] + " "); 28 } 29 }

perform one pass

−2 1 2 2 3 3 5 6 12 14

In the best case, the bubble-sort algorithm needs just the first pass to find that the array is already sorted—no next pass is needed. Since the number of comparisons is n - 1 in the first pass, the best-case time for a bubble sort is O(n).

In the worst case, the bubble-sort algorithm requires n - 1 passes. The first pass makes n - 1 comparisons, the second pass makes n - 2 comparisons, and so on; the last pass makes 1 comparison. Thus, the total number of comparisons is as follows:

(n - 1) + (n - 2) + g + 2 + 1

= (n - 1)n

2 =

2 -

n 2

= O(n2)

Therefore, the worst-case time for a bubble sort is O(n2).

23.3.1 Describe how a bubble sort works. What is the time complexity for a bubble sort? 23.3.2 Use Figure 23.3 as an example to show how to apply a bubble sort on {45, 11, 50,

59, 60, 2, 4, 7, 10}.

23.3.3 If a list is already sorted, how many comparisons will the bubbleSort method perform?

bubble-sort time complexity

Point Check

M23_LIAN1878_11_GE_C23.indd 908 12/29/17 4:40 PM

23.4 Merge Sort 909

23.4 Merge Sort The merge-sort algorithm can be described recursively as follows: The algorithm divides the array into two halves and applies a merge sort on each half recursively. After the two halves are sorted, the algorithm then merges them.

The algorithm for a merge sort is given in Listing 23.5.

Listing 23.5 Merge-Sort Algorithm 1 public static void mergeSort(int[] list) { 2 if (list.length > 1) { 3 mergeSort(list[0 ... list.length / 2); 4 mergeSort(list[list.length / 2 + 1 ... list.length]); 5 merge list[0 ... list.length / 2] with 6 list[list.length / 2 + 1 ... list.length]; 7 } 8 }

Figure 23.4 illustrates a merge sort of an array of eight elements (2 9 5 4 8 1 6 7). The original array is split into (2 9 5 4) and (8 1 6 7). Apply a merge sort on these two subarrays recursively to split (2 9 5 4) into (2 9) and (5 4) and (8 1 6 7) into (8 1) and (6 7). This process continues until the subarray contains only one element. For example, array (2 9) is split into the subarrays (2) and (9). Since array (2) contains a single element, it cannot be further split. Now merge (2) with (9) into a new sorted array (2 9) and (5) with (4) into a new sorted array (4 5). Merge (2 9) with (4 5) into a new sorted array (2 4 5 9) and finally merge (2 4 5 9) with (1 6 7 8) into a new sorted array (1 2 4 5 6 7 8 9).

Point Key

merge sort

base condition sort first half sort second half merge two halves

merge-sort illustration

Figure 23.4 Merge sort employs a divide-and-conquer approach to sort the array.

split

9 5 4 8 1 6 7

merge

81 6 7

1 7 8

1 4 6 7 8 9

merge

divide

conquer

2 5 8 1 6 79 4

8 1 6 7

2 54 9

2 9

2 59 4

5 4

4 5

2 5

8 1 6 7

The recursive call continues dividing the array into subarrays until each subarray contains only one element. The algorithm then merges these small subarrays into larger sorted subarrays until one sorted array results.

The merge-sort algorithm is implemented in Listing 23.6.

Listing 23.6 MergeSort.java 1 public class MergeSort { 2 /** The method for sorting the numbers */ 3 public static void mergeSort(int[] list) { 4 if (list.length > 1) { base case

M23_LIAN1878_11_GE_C23.indd 909 12/29/17 4:40 PM

910 Chapter 23 Sorting

5 // Merge sort the first half 6 int[] firstHalf = new int[list.length / 2]; 7 System.arraycopy(list, 0, firstHalf, 0, list.length / 2); 8 mergeSort(firstHalf); 9 10 // Merge sort the second half 11 int secondHalfLength = list.length − list.length / 2; 12 int[] secondHalf = new int[secondHalfLength]; 13 System.arraycopy(list, list.length / 2, 14 secondHalf, 0, secondHalfLength); 15 mergeSort(secondHalf); 16 17 // Merge firstHalf with secondHalf into list 18 merge(firstHalf, secondHalf, list); 19 } 20 } 21 22 /** Merge two sorted lists */ 23 public static void merge(int[] list1, int[] list2, int[] temp) { 24 int current1 = 0; // Current index in list1 25 int current2 = 0; // Current index in list2 26 int current3 = 0; // Current index in temp 27 28 while (current1 < list1.length && current2 < list2.length) { 29 if (list1[current1] < list2[current2]) 30 temp[current3++] = list1[current1++]; 31 else 32 temp[current3++] = list2[current2++]; 33 } 34 35 while (current1 < list1.length) 36 temp[current3++] = list1[current1++]; 37 38 while (current2 < list2.length) 39 temp[current3++] = list2[current2++]; 40 } 41 42 /** A test method */ 43 public static void main(String[] args) { 44 int[] list = {2, 3, 2, 5, 6, 1, −2, 3, 14, 12}; 45 mergeSort(list); 46 for (int i = 0; i < list.length; i++) 47 System.out.print(list[i] + " "); 48 } 49 }

The mergeSort method (lines 3–20) creates a new array firstHalf, which is a copy of the first half of list (line 7). The algorithm invokes mergeSort recursively on firstHalf (line 8). The length of the firstHalf is list.length / 2 and the length of the second- Half is list.length − list.length / 2. The new array secondHalf was created to contain the second part of the original array list. The algorithm invokes mergeSort recursively on secondHalf (line 15). After firstHalf and secondHalf are sorted, they are merged to list (line 18). Thus, array list is now sorted.

The merge method (lines 23–40) merges two sorted arrays list1 and list2 into array temp. current1 and current2 point to the current element to be considered in list1 and list2 (lines 24–26). The method repeatedly compares the current elements from list1 and list2 and moves the smaller one to temp. current1 is increased by 1 (line 30) if the smaller one is in list1, and current2 is increased by 1 (line 32) if the smaller one is in list2.

sort first half

sort second half

merge two halves

list1 to temp

list2 to temp

rest of list1 to temp

rest of list2 to temp

M23_LIAN1878_11_GE_C23.indd 910 12/29/17 4:40 PM

23.4 Merge Sort 911

Finally, all the elements in one of the lists are moved to temp. If there are still unmoved elements in list1, copy them to temp (lines 35–36). If there are still unmoved elements in list2, copy them to temp (lines 38–39).

Figure 23.5 illustrates how to merge the two arrays list1 (2 4 5 9) and list2 (1 6 7 8). Initially, the current elements to be considered in the arrays are 2 and 1. Compare them and move the smaller element 1 to temp, as shown in Figure 23.5a. current2 and current3 are increased by 1. Continue to compare the current elements in the two arrays and move the smaller one to temp until one of the arrays is completely moved. As shown in Figure 23.5b, all the elements in list2 are moved to temp and current1 points to element 9 in list1. Copy 9 to temp, as shown in Figure 23.5c. For an interactive demo on how merge works, go to liveexample.pearsoncmg.com/dsanimation/MergeSortNeweBook.html.

merge animation on Companion Website

Figure 23.5 Two sorted arrays are merged into one sorted array.

(a) After moving 1 to temp

2 5 9

current1 current2

current3

4 1 7 86

current1 current2

current3

2 5

1 7

61 2

current1 current2

2 5 9

current3

4 1 7 86

54 7 8 961 2

(b) After moving all the elements in list2 to temp

Figure 23.6 Temporary arrays are created to support a merge sort.

(a)

Divide Copy second halfCopy �rst half

Merge to list

Recursive sort

Merge

secondHalf (temporary array)

�rstHalf (temporary array)

New sorted list

Original list

(b)

Divide

Copy this to the original list

Recursively sort

Merge

Sort second half of the original array

Sort �rst half of the original array

New sorted temporary list

Original list

The mergeSort method creates two temporary arrays (lines 6 and 12) during the dividing process, copies the first half and the second half of the array into the temporary arrays (lines 7 and 13), sorts the temporary arrays (lines 8 and 15), then merges them into the original array (line 18), as shown in Figure 23.6a. You can rewrite the code to recursively sort the first half of the array and the second half of the array without creating new temporary arrays, then merge the two arrays into a temporary array and copy its contents to the original array, as shown in Figure 23.6b. This is left for you to do in Programming Exercise 23.20.

Note A merge sort can be implemented efficiently using parallel processing. See Section 32.16, Parallel Programming, for a parallel implementation of a merge sort.

M23_LIAN1878_11_GE_C23.indd 911 12/29/17 4:40 PM

912 Chapter 23 Sorting

Let T(n) denote the time required for sorting an array of n elements using a merge sort. Without loss of generality, assume n is a power of 2. The merge-sort algorithm splits the array into two subarrays, sorts the subarrays using the same algorithm recursively, then merges the subarrays. Therefore,

T(n) = T an 2 b + T an

2 b + mergetime

The first T an 2 b is the time for sorting the first half of the array and the second T an

2 b is the

time for sorting the second half. To merge two subarrays, it takes at most n - 1 comparisons to compare the elements from the two subarrays, and n moves to move elements to the temporary array. Thus, the total time is 2n - 1. Therefore,

T(n) = T an 2 b + T an

2 b + 2n - 1 = O(n log n)

The complexity of a merge sort is O(n logn). This algorithm is better than selection sort, inser- tion sort, and bubble sort because the time complexity of these algorithms is O(n2). The sort method in the java.util.Arrays class is implemented using a variation of the merge-sort algorithm.

23.4.1 Describe how a merge sort works. What is the time complexity for a merge sort? 23.4.2 Use Figure 23.4 as an example to show how to apply a merge sort on {45, 11, 50,

59, 60, 2, 4, 7, 10}.

23.4.3 What is wrong if lines 6–15 in Listing 23.6, MergeSort.java, are replaced by the following code?

// Merge sort the first half int[] firstHalf = new int[list.length / 2 + 1]; System.arraycopy(list, 0, firstHalf, 0, list.length / 2 + 1); mergeSort(firstHalf);

// Merge sort the second half int secondHalfLength = list.length − list.length / 2 − 1; int[] secondHalf = new int[secondHalfLength]; System.arraycopy(list, list.length / 2 + 1, secondHalf, 0, secondHalfLength); mergeSort(secondHalf);

23.5 Quick Sort A quick sort works as follows: The algorithm selects an element, called the pivot, in the array. It divides the array into two parts so all the elements in the first part are less than or equal to the pivot, and all the elements in the second part are greater than the pivot. The quick-sort algorithm is then recursively applied to the first part and then the second part.

The quick-sort algorithm, developed by C.A.R. Hoare in 1962, is described in Listing 23.7.

Listing 23.7 Quick-Sort Algorithm 1 public static void quickSort(int[] list) { 2 if (list.length > 1) { 3 select a pivot; 4 partition list into list1 and list2 such that

merge-sort time complexity

O(n logn) merge sort

Point Check

Point Key

quick sort

base condition select the pivot partition the list

M23_LIAN1878_11_GE_C23.indd 912 12/29/17 4:40 PM

23.5 Quick Sort 913

5 all elements in list1 <= pivot and 6 all elements in list2 > pivot; 7 quickSort(list1); 8 quickSort(list2); 9 } 10 }

Each partition places the pivot in the right place. The selection of the pivot affects the perfor- mance of the algorithm. Ideally, the algorithm should choose the pivot that divides the two parts evenly. For simplicity, assume that the first element in the array is chosen as the pivot. (Programming Exercise 23.4 proposes an alternative strategy for selecting the pivot.)

Figure 23.7 illustrates how to sort an array (5 2 9 3 8 4 0 1 6 7) using quick sort. Choose the first element, 5, as the pivot. The array is partitioned into two parts, as shown in Figure 23.7b. The highlighted pivot is placed in the right place in the array. Apply quick sort on two subarrays (4 2 1 3 0) then (8 9 6 7). The pivot 4 partitions (4 2 1 3 0) into just one subarrays (0 2 1 3), as shown in Figure 23.7c. Apply quick sort on (0 2 1 3). The pivot 0 partitions it into just one subarrays (2 1 3), as shown in Figure 23.7d. Apply quick sort on (2 1 3). The pivot 2 partitions it into (1) and (3), as shown in Figure 23.7e. Apply quick sort on (1). Since the array contains just one element, no further partition is needed.

sort first part sort second part

how to partition

quick-sort illustration

list2

pivot

list1

Figure 23.7 The quick-sort algorithm is recursively applied to subarrays.

(a) The original array

pivot

5 761048392

(b) The original array is partitioned

pivotpivot

4 769850312

(d) The subarray (0 2 1 3) is partitioned

pivot

0 2 1 3

(e) The subarray (2 1 3) is partitioned1 2 3

pivot

0 2 1 3 4

The quick-sort algorithm is implemented in Listing 23.8. There are two overloaded quickSort methods in the class. The first method (line 2) is used to sort an array. The second is a helper method (line 6) that sorts a subarray with a specified range.

Listing 23.8 QuickSort.java 1 public class QuickSort { 2 public static void quickSort(int[] list) { 3 quickSort(list, 0, list.length − 1); 4 } 5 6 public static void quickSort(int[] list, int first, int last) { 7 if (last > first) { 8 int pivotIndex = partition(list, first, last); 9 quickSort(list, first, pivotIndex − 1); 10 quickSort(list, pivotIndex + 1, last); 11 }

sort method

helper method

recursive call

M23_LIAN1878_11_GE_C23.indd 913 12/29/17 4:40 PM

914 Chapter 23 Sorting

12 } 13 14 /** Partition the array list[first..last] */ 15 public static int partition(int[] list, int first, int last) { 16 int pivot = list[first]; // Choose the first element as the pivot 17 int low = first + 1; // Index for forward search 18 int high = last; // Index for backward search 19 20 while (high > low) { 21 // Search forward from left 22 while (low <= high && list[low] <= pivot) 23 low++; 24 25 // Search backward from right 26 while (low <= high && list[high] > pivot) 27 high−−; 28 29 // Swap two elements in the list 30 if (high > low) { 31 int temp = list[high]; 32 list[high] = list[low]; 33 list[low] = temp; 34 } 35 } 36 37 while (high > first && list[high] >= pivot) 38 high−−; 39 40 // Swap pivot with list[high] 41 if (pivot > list[high]) { 42 list[first] = list[high]; 43 list[high] = pivot; 44 return high; 45 } 46 else { 47 return first; 48 } 49 } 50 51 /** A test method */ 52 public static void main(String[] args) { 53 int[] list = {2, 3, 2, 5, 6, 1, −2, 3, 14, 12}; 54 quickSort(list); 55 for (int i = 0; i < list.length; i++) 56 System.out.print(list[i] + " "); 57 } 58 }

forward

backward

swap

place pivot pivot’s new index

pivot’s original index

−2 1 2 2 3 3 5 6 12 14

The partition method (lines 15–49) partitions the array list[first..last] using the pivot. The first element in the partial array is chosen as the pivot (line 16). Initially, low points to the second element in the subarrays (line 17) and high points to the last element in the subarrays (line 18).

Starting from the left, the method searches forward in the array for the first element that is greater than the pivot (lines 22–23), then searches from the right backward for the first ele- ment in the array that is less than or equal to the pivot (lines 26–27). It then swaps these two

M23_LIAN1878_11_GE_C23.indd 914 12/29/17 4:40 PM

23.5 Quick Sort 915

elements and repeats the same search and swap operations until all the elements are searched in a while loop (lines 20–35).

The method returns the new index for the pivot that divides the subarrays into two parts if the pivot has been moved (line 44). Otherwise, it returns the original index for the pivot (line 47).

Figure 23.8 illustrates how to partition an array (5 2 9 3 8 4 0 1 6 7). Choose the first element, 5, as the pivot. Initially, low is the index that points to element 2 and high points to element 7, as shown in Figure 23.8a. Advance index low forward to search for the first element (9) that is greater than the pivot, and move index high backward to search for the first element (1) that is less than or equal to the pivot, as shown in Figure 23.8b. Swap 9 with 1, as shown in Figure 23.8c. Continue the search and move low to point to element 8 and high to point to element 0, as shown in Figure 23.8d. Swap element 8 with 0, as shown in Figure 23.8e. Continue to move low until it passes high, as shown in Figure 23.8f. Now all the elements are examined. Swap the pivot with element 4 at index high. The final partition is shown in Figure 23.8g. The index of the pivot is returned when the method is finished. For an interactive demo on how partition works, go to liveexample.pearsoncmg.com/dsanimation/ QuickSortNeweBook.html.

partition illustration

partition animation on Companion Website

Figure 23.8 The partition method returns the index of the pivot after it is put in the cor- rect place.

pivot low high

(b) Search forward and backward

The index of the pivot is returned

2 9 3 8 4 0 1 6 7

pivot low high

(a) Initialize pivot, low, and high2 9 3 8 4 0 1 6 7

pivot low high

(d) Continue search2 1 3 8 4 0 9 6 7

pivot low high

(e) 8 is swapped with 02 1 3 0 4 8 9 6 7

pivot low high

(f) When high < low, search is over2 1 3 0 4 8 9 6 7

pivot

(g) Pivot is in the right place2 1 3 0 5 8 9 6 7

To partition an array of n elements, it takes n comparisons and n moves in the worst case. Thus, the time required for partition is O(n). O(n) partition time

M23_LIAN1878_11_GE_C23.indd 915 12/29/17 4:40 PM

916 Chapter 23 Sorting

In the worst case, the pivot divides the array each time into one big subarray with the other array empty. The size of the big subarray is one less than the one before divided. The algorithm requires (n - 1) + (n - 2) + g + 2 + 1 = O(n2) time.

In the best case, the pivot divides the array each time into two parts of about the same size. Let T(n) denote the time required for sorting an array of n elements using quick sort. Thus,

recursive quick sort on partition time two subarrays

T(n) = Tan 2 b + Tan

2 b + n.

Similar to the merge-sort analysis, T(n) = O(n log n). On the average, the pivot will not divide the array into two parts of the same size or one

empty part each time. Statistically, the sizes of the two parts are very close. Therefore, the average time is O(n logn). The exact average-case analysis is beyond the scope of this book.

Both merge and quick sorts employ the divide-and-conquer approach. For merge sort, the bulk of the work is to merge two sublists, which takes place after the sublists are sorted. For quick sort, the bulk of the work is to partition the list into two sublists, which takes place before the sublists are sorted. Merge sort is more efficient than quick sort in the worst case, but the two are equally efficient in the average case. Merge sort requires a temporary array for sorting two subarrays. Quick sort does not need additional array space. Thus, quick sort is more space efficient than merge sort.

23.5.1 Describe how quick sort works. What is the time complexity for a quick sort? 23.5.2 Why is quick sort more space efficient than merge sort? 23.5.3 Use Figure 23.7 as an example to show how to apply a quick sort on {45, 11, 50,

59, 60, 2, 4, 7, 10}.

23.5.4 If lines 37–38 in the QuickSort program is removed, will it still work? Give a counter example to show that it will not work.

23.6 Heap Sort A heap sort uses a binary heap. It first adds all the elements to a heap and then removes the largest elements successively to obtain a sorted list.

Heap sorts use a binary heap, which is a complete binary tree. A binary tree is a hierarchical structure. It either is empty or it consists of an element, called the root, and two distinct binary trees, called the left subtree and right subtree. The length of a path is the number of the edges in the path. The depth of a node is the length of the path from the root to the node. A node is called a leaf if it does not have subtrees.

A binary heap is a binary tree with the following properties:

■■ Shape property: It is a complete binary tree.

■■ Heap property: Each node is greater than or equal to any of its children.

O(n2) worst-case time

O(n logn) best-case time

O(n logn) average-case time

quick sort vs. merge sort

Point Check

Point Key

heap sort

root left subtree right subtree length depth leaf

Figure 23.9 A binary heap is a special complete binary tree.

(a) A heap (b) (c) (d) 22 29

14 3322 29

14 33

39 39

22 29

M23_LIAN1878_11_GE_C23.indd 916 12/29/17 4:40 PM

23.6 Heap Sort 917

A binary tree is complete if each of its levels is full, except that the last level may not be full and all the leaves on the last level are placed leftmost. For example, in Figure 23.9, the binary trees in (a) and (b) are complete, but the binary trees in (c) and (d) are not complete. Further, the binary tree in (a) is a heap, but the binary tree in (b) is not a heap because the root (39) is less than its right child (42).

Note Heap is a term with many meanings in computer science. In this chapter, heap means a binary heap.

Pedagogical Note A heap can be implemented efficiently for inserting keys and deleting the root. For an interactive demo on how a heap works, go to liveexample.pearsoncmg.com/ dsanimation/ HeapeBook.html, as shown in Figure 23.10.

complete binary tree

heap

heap animation on Companion Website

23.6.1 Storing a Heap A heap can be stored in an ArrayList or an array if the heap size is known in advance. The heap in Figure 23.11a can be stored using the array in Figure 23.11b. The root is at position 0, and its two children are at positions 1 and 2. For a node at position i, its left child is at position 2i + 1, its right child is at position 2i + 2, and its parent is (i - 1)/2. For example, the node for element 39 is at position 4, so its left child (element 14) is at 9 (2 * 4 + 1), its right child (element 33) is at 10 (2 * 4 + 2), and its parent (element 42) is at 1 ((4 - 1)/2).

Figure 23.11 A binary heap can be implemented using an array.

22 29

14 33

30 17

(a) A heap

[0]

42 59 32 39 44 13 22 29 14 33 30 17

[12]

[13][11][10][9][8][7][6][5][4][3][2][1]

(b) A heap stored in an array

parent left

right

M23_LIAN1878_11_GE_C23.indd 917 12/29/17 4:40 PM

918 Chapter 23 Sorting

23.6.2 Adding a New Node To add a new node to the heap, first add it to the end of the heap then rebuild the tree as follows:

Let the last node be the current node; while (the current node is greater than its parent) { Swap the current node with its parent; Now the current node is one level up; }

Suppose a heap is initially empty. That heap is shown in Figure 23.12, after adding numbers 3, 5, 1, 19, 11, and 22 in this order.

Figure 23.12 Elements 3, 5, 1, 19, 11, and 22 are inserted into the heap.

(a) After adding 3 (b) After adding 5 (c) After adding 1

(d) After adding 19 (e) After adding 11 (f) After adding 22 3

3 5

Figure 23.13 Rebuild the heap after adding a new node.

(a) Add 88 to a heap (b) After swapping 88 with 19 (c) After swapping 88 with 22

1 883 5 3 5

1 19

1 193 5

Now consider adding 88 into the heap. Place the new node 88 at the end of the tree, as shown in Figure 23.13a. Swap 88 with 19, as shown in Figure 23.13b. Swap 88 with 22, as shown in Figure 23.13c.

23.6.3 Removing the Root Often you need to remove the maximum element, which is the root in a heap. After the root is removed, the tree must be rebuilt to maintain the heap property. The algorithm for rebuilding the tree can be described as follows:

Move the last node to replace the root; Let the root be the current node; while (the current node has children and the current node is smaller than one of its children) { Swap the current node with the larger of its children; Now the current node is one level down; }

M23_LIAN1878_11_GE_C23.indd 918 12/29/17 4:40 PM

23.6 Heap Sort 919

Figure 23.14 shows the process of rebuilding a heap after the root 62 is removed from Figure 23.11a. Move the last node 9 to the root, as shown in Figure 23.14a. Swap 9 with 59, as shown in Figure 23.14b; swap 9 with 44, as shown in Figure 23.14c; and swap 9 with 30, as shown in Figure 23.14d.

Figure 23.14 Rebuild the heap after the root 62 is removed.

22 29 14 33 30 17

32 39 44 13

42 59

(a) After moving 9 to the root

22 29 14 33 30 17

32 39 44 13

42 9

(b) After swapping 9 with 59

22 29 14 33 30 17

32 39 9 13

42 44

22 29 14 33 179

32 39 30 13

42 44

(d) After swapping 9 with 30

Figure 23.15 shows the process of rebuilding a heap after the root 59 is removed from Figure 23.14d. Move the last node 17 to the root, as shown in Figure 23.15a. Swap 17 with 44, as shown in Figure 23.15b, then swap 17 with 30, as shown in Figure 23.15c.

Figure 23.15 Rebuild the heap after the root 59 is removed.

(a) After moving 17 to the root (b) After swapping 17 with 44 22 29 14 33 9

32 39 30 13

22 29 14 33 9

32 39 13

3042

22 29 14 33 9

32 39 30 13

42 44

M23_LIAN1878_11_GE_C23.indd 919 12/29/17 4:40 PM

920 Chapter 23 Sorting

23.6.4 The Heap Class Now you are ready to design and implement the Heap class. The class diagram is shown in Figure 23.16. Its implementation is given in Listing 23.9.

Figure 23.16 The Heap class provides operations for manipulating a heap.

Heap<E extends Comparable<E>>

–list: java.util.ArrayList<E>

+Heap() +Heap(objects: E[])

+remove(): E +add(newObject: E): void

+getSize(): int

Creates a default empty Heap. Creates a Heap with the speci�ed objects.

Removes the root from the heap and returns it. Adds a new object to the heap.

Returns the size of the heap. Returns true if the heap is empty.+isEmpty(): boolean

Listing 23.9 Heap.java 1 public class Heap<E extends Comparable<E>> { 2 private java.util.ArrayList<E> list = new java.util.ArrayList<>(); 3 4 /** Create a default heap */ 5 public Heap() { 6 } 7 8 /** Create a heap from an array of objects */ 9 public Heap(E[] objects) { 10 for (int i = 0; i < objects.length; i++) 11 add(objects[i]); 12 } 13 14 /** Add a new object into the heap */ 15 public void add(E newObject) { 16 list.add(newObject); // Append to the heap 17 int currentIndex = list.size() − 1; // The index of the last node 18 19 while (currentIndex > 0) { 20 int parentIndex = (currentIndex − 1) / 2; 21 // Swap if the current object is greater than its parent 22 if (list.get(currentIndex).compareTo( 23 list.get(parentIndex)) > 0) { 24 E temp = list.get(currentIndex); 25 list.set(currentIndex, list.get(parentIndex)); 26 list.set(parentIndex, temp); 27 } 28 else 29 break; // The tree is a heap now 30 31 currentIndex = parentIndex; 32 } 33 } 34 35 /** Remove the root from the heap */ 36 public E remove() {

internal heap representation

no-arg constructor

constructor

add a new object append the object

swap with parent

heap now

remove the root

M23_LIAN1878_11_GE_C23.indd 920 12/29/17 4:40 PM

23.6 Heap Sort 921

37 if (list.size() == 0) return null; 38 39 E removedObject = list.get(0); 40 list.set(0, list.get(list.size() − 1)); 41 list.remove(list.size() − 1); 42 43 int currentIndex = 0; 44 while (currentIndex < list.size()) { 45 int leftChildIndex = 2 * currentIndex + 1; 46 int rightChildIndex = 2 * currentIndex + 2; 47 48 // Find the maximum between two children 49 if (leftChildIndex >= list.size()) break; // The tree is a heap 50 int maxIndex = leftChildIndex; 51 if (rightChildIndex < list.size()) { 52 if (list.get(maxIndex).compareTo( 53 list.get(rightChildIndex)) < 0) { 54 maxIndex = rightChildIndex; 55 } 56 } 57 58 // Swap if the current node is less than the maximum 59 if (list.get(currentIndex).compareTo( 60 list.get(maxIndex)) < 0) { 61 E temp = list.get(maxIndex); 62 list.set(maxIndex, list.get(currentIndex)); 63 list.set(currentIndex, temp); 64 currentIndex = maxIndex; 65 } 66 else 67 break; // The tree is a heap 68 } 69 70 return removedObject; 71 } 72 73 /** Get the number of nodes in the tree */ 74 public int getSize() { 75 return list.size(); 76 } 77 }

A heap is represented using an array list internally (line 2). You can change the array list to other data structures, but the Heap class contract will remain unchanged.

The add(E newObject) method (lines 15–33) appends the object to the tree then swaps the object with its parent if the object is greater than its parent. This process continues until the new object becomes the root or is not greater than its parent.

The remove() method (lines 36–71) removes and returns the root. To maintain the heap property, the method moves the last object to the root position and swaps it with its larger child if it is less than the larger child. This process continues until the last object becomes a leaf or is not less than its children.

23.6.5 Sorting Using the Heap Class To sort an array using a heap, first create an object using the Heap class, add all the elements to the heap using the add method, and remove all the elements from the heap using the remove method. The elements are removed in descending order. Listing 23.10 gives a program for sorting an array using a heap.

empty heap

root new root remove the last

adjust the tree

compare two children

swap with the larger child

M23_LIAN1878_11_GE_C23.indd 921 12/29/17 4:40 PM

922 Chapter 23 Sorting

Listing 23.10 HeapSort.java 1 public class HeapSort { 2 /** Heap sort method */ 3 public static <E extends Comparable<E>> void heapSort(E[] list) { 4 // Create a Heap of integers 5 Heap<E> heap = new Heap<>(); 6 7 // Add elements to the heap 8 for (int i = 0; i < list.length; i++) 9 heap.add(list[i]); 10 11 // Remove elements from the heap 12 for (int i = list.length − 1; i >= 0; i−−) 13 list[i] = heap.remove(); 14 } 15 16 /** A test method */ 17 public static void main(String[] args) { 18 Integer[] list = {−44, −5, −3, 3, 3, 1, −4, 0, 1, 2, 4, 5, 53}; 19 heapSort(list); 20 for (int i = 0; i < list.length; i++) 21 System.out.print(list[i] + " "); 22 } 23 }

create a Heap

add element

remove element

invoke sort method

−44 −5 −4 −3 0 1 1 2 3 3 4 5 53

23.6.6 Heap Sort Time Complexity Let us turn our attention to analyzing the time complexity for the heap sort. Let h denote the height for a heap of n elements. The height of a nonempty tree is the length of the path from the root node to its furthest leaf. The height of a tree that contains a single node is 0. Conven- tionally, the height of an empty tree is −1. Since a heap is a complete binary tree, the first level has 1 (20) node, the second level has 2 (21) nodes, the kth level has 2k - 1 nodes, the h level has 2h - 1 nodes, and the last (h + 1)th level has at least 1 and at most 2h nodes. Therefore,

1 + 2 + g + 2h - 1 6 n … 1 + 2 + g + 2h - 1 + 2h

That is,

2h - 1 6 n … 2h + 1 - 1 2h 6 n + 1 … 2h + 1 h 6 log(n + 1) … h + 1

Thus, h 6 log(n + 1) and h Ú log(n + 1) - 1. Therefore, log(n + 1) - 1 … h 6 log(n + 1). Hence, the height of a heap is O(logn). More precisely, you can prove that h = :log n; . for a non-empty tree.

Since the add method traces a path from a leaf to a root, it takes at most h steps to add a new element to the heap. Thus, the total time for constructing an initial heap is O(n logn) for an array of n elements. Since the remove method traces a path from a root to a leaf, it takes at most h steps to rebuild a heap after removing the root from the heap. Since the remove method is invoked n times, the total time for producing a sorted array from a heap is O(n logn).

Both merge and heap sorts require O(n logn) time. A merge sort requires a temporary array for merging two subarrays; a heap sort does not need additional array space. Therefore, a heap sort is more space efficient than a merge sort.

height of a heap

O(n logn) worst-case time

heap sort vs. merge sort

M23_LIAN1878_11_GE_C23.indd 922 12/29/17 4:40 PM

23.7 Bucket and Radix Sorts 923

23.6.1 What is a complete binary tree? What is a heap? Describe how to remove the root from a heap and how to add a new object to a heap.

23.6.2 What is the return value from invoking the remove method if the heap is empty? 23.6.3 Add the elements 4, 5, 1, 2, 9, and 3 into a heap in this order. Draw the diagrams to

show the heap after each element is added.

23.6.4 Show the heap after the root in the heap in Figure 23.15c is removed. 23.6.5 What is the time complexity of inserting a new element into a heap, and what is the

time complexity of deleting an element from a heap?

23.6.6 Show the steps of creating a heap using {45, 11, 50, 59, 60, 2, 4, 7, 10}. 23.6.7 Given the following heap, show the steps of removing all nodes from the heap.

Point Check

22 29

14 33

17 30

23.6.8 Which of the following statements are wrong? 1 Heap<Object> heap1 = new Heap<>(); 2 Heap<Number> heap2 = new Heap<>(); 3 Heap<BigInteger> heap3 = new Heap<>(); 4 Heap<Calendar> heap4 = new Heap<>(); 5 Heap<String> heap5 = new Heap<>();

23.6.9 What is the height of a nonempty heap? What is the height of a heap with 16, 17, and 512 elements? If the height of a heap is 5, what is the maximum number of nodes in the heap?

23.7 Bucket and Radix Sorts Bucket and radix sorts are efficient for sorting integers.

All sort algorithms discussed so far are general sorting algorithms that work for any types of keys (e.g., integers, strings, and any comparable objects). These algorithms sort the elements by comparing their keys. It has been proven that no sorting algorithms based on comparisons can perform better than O(n logn). However, if the keys are integers, you can use a bucket sort without having to compare the keys.

The bucket sort algorithm works as follows. Assume the keys are in the range from 0 to t. We need t + 1 buckets labeled 0, 1, . . . , and t. If an element’s key is i, the element is put into the bucket i. Each bucket holds the elements with the same key value.

Elements with key 0

Elements with key 1

Elements with key 2 . . .

Elements with key t

bucket[0] bucket[1] bucket[2] bucket[t]

You can use an ArrayList to implement a bucket. The bucket-sort algorithm for sorting a list of elements can be described as follows:

Point Key

bucket sort

M23_LIAN1878_11_GE_C23.indd 923 12/29/17 4:40 PM

924 Chapter 23 Sorting

public static void bucketSort(E[] list) { E[] bucket = (E[])new java.util.ArrayList[t+1];

// Distribute the elements from list to buckets for (int i = 0; i < list.length; i++) { int key = list[i].getKey(); // Assume element has the getKey() method

if (bucket[key] == null) bucket[key] = new java.util.ArrayList<>();

bucket[key].add(list[i]); }

// Now move the elements from the buckets back to list int k = 0; // k is an index for list for (int i = 0; i < bucket.length; i++) { if (bucket[i] != null) { for (int j = 0; j < bucket[i].size(); j++) list[k++] = bucket[i].get(j); } } }

Clearly, it takes O(n + t) time to sort the list and uses O(n + t) space, where n is the list size. Note if t is too large, using the bucket sort is not desirable. Instead, you can use a radix sort.

The radix sort is based on the bucket sort, but a radix sort uses only 10 buckets. It is worthwhile to note a bucket sort is stable, meaning that if two elements in the original

list have the same key value, their order is not changed in the sorted list. That is, if element e1 and element e2 have the same key and e1 precedes e2 in the original list, e1 still precedes e2 in the sorted list.

Assume the keys are positive integers. The idea for the radix sort is to divide the keys into subgroups based on their radix positions. It applies a bucket sort repeatedly for the key values on radix positions, starting from the least-significant position.

Consider sorting the elements with the following keys:

331, 454, 230, 34, 343, 45, 59, 453, 345, 231, 9

Apply the bucket sort on the last radix position, and the elements are put into the buckets as follows:

stable

radix sort

radix sort on Companion Website

queue

230 331 231

343 453

454 34

45 345

59 9

bucket[0] bucket[1] bucket[2] bucket[3] bucket[4] bucket[5] bucket[6] bucket[7] bucket[8] bucket[9]

After collecting the elements from the buckets, the elements are in the following order:

230, 331, 231, 343, 453, 454, 34, 45, 345, 59, 9

Apply the bucket sort on the second-to-last radix position, and the elements are put into the buckets as follows:

queue

bucket[0] bucket[1] bucket[2]

230 331 231 34

bucket[3]

343 45

345

bucket[4]

453 454 59

bucket[5]

bucket[6] bucket[7] bucket[8] bucket[9]

M23_LIAN1878_11_GE_C23.indd 924 12/29/17 4:40 PM

23.8 External Sort 925

After collecting the elements from the buckets, the elements are in the following order:

9, 230, 331, 231, 34, 343, 45, 345, 453, 454, 59

(Note 9 is 009.) Apply the bucket sort on the third-to-last radix position, and the elements are put into the

buckets as follows: queue

9 34 45 59

bucket[0] bucket[1]

230 231

bucket[2]

331 343 345

bucket[3]

453 454

bucket[4] bucket[5] bucket[6] bucket[7] bucket[8] bucket[9]

After collecting the elements from the buckets, the elements are in the following order:

9, 34, 45, 59, 230, 231, 331, 343, 345, 453, 454

The elements are now sorted. Radix sort takes O(dn) time to sort n elements with integer keys, where d is the maximum

number of the radix positions among all keys.

23.7.1 Can you sort a list of strings using a bucket sort? 23.7.2 Show how the radix sort works using the numbers 454, 34, 23, 43, 74, 86, and 76.

23.8 External Sort You can sort a large amount of data using an external sort.

All the sort algorithms discussed in the preceding sections assume all the data to be sorted are available at one time in internal memory, such as in an array. To sort data stored in an external file, you must first bring the data to the memory then sort it internally. However, if the file is too large, all the data in the file cannot be brought to memory at one time. This section dis- cusses how to sort data in a large external file. This is called an external sort.

For simplicity, assume two million int values are stored in a binary file named largedata .dat. This file was created using the program in Listing 23.11.

Listing 23.11 CreateLargeFile.java 1 import java.io.*; 2 3 public class CreateLargeFile { 4 public static void main(String[] args) throws Exception { 5 DataOutputStream output = new DataOutputStream( 6 new BufferedOutputStream( 7 new FileOutputStream("largedata.dat")); 8 9 for (int i = 0; i < 2_000_000; i++) 10 output.writeInt((int)(Math.random() * 1000000))); 11 12 output.close(); 13 14 // Display first 100 numbers 15 DataInputStream input = new DataInputStream( 16 new BufferedInputStream(new FileInputStream("largedata.dat"))); 17 for (int i = 0; i < 100; i++)

Point Check

Point Key

external sort

a binary output stream

output an int value

close output file

M23_LIAN1878_11_GE_C23.indd 925 12/29/17 4:40 PM

926 Chapter 23 Sorting

18 System.out.print(input.readInt() + " "); 19 20 input.close(); 21 } 22 }

read an int value

close input file

Figure 23.17 The original file is sorted in segments.

Program

Array

……

Original �le

Temporary �le

S1 S2 Sk

Unsorted

Sorted segment

Figure 23.18 Sorted segments are merged iteratively.

S1 S2 S3 S4 S5 S6 S7 S8

S1, S2 merged

S1, S2, S3, S4 merged S5, S6, S7, S8 merged

S3, S4 merged S5, S6 merged S7, S8 merged

Merge step

S1, S2, S3, S4, S5, S6, S7, S8 merged Final sortedsegment

569193 131317 608695 776266 767910 624915 458599 5010 ... (omitted)

A variation of merge sort can be used to sort this file in two phases:

Phase I: Repeatedly bring data from the file to an array, sort the array using an internal sorting algorithm, and output the data from the array to a temporary file. This process is shown in Figure 23.17. Ideally, you want to create a large array, but its maximum size depends on how much memory is allocated to the JVM by the operating system. Assume the maximum array size is 100,000 int values. In the temporary file, every 100,000 int values are sorted. They are denoted as S1, S2, c , and Sk, where the last segment, Sk, may contain less than 100,000 values.

Phase II: Merge a pair of sorted segments (e.g., S1 with S2, S3 with S4, c , and so on) into a larger sorted segment and save the new segment into a new temporary file. Continue the same process until only one sorted segment results. Figure 23.18 shows how to merge eight segments.

Note It is not necessary to merge two successive segments. For example, you can merge S1 with S5 , S2 with S6 , S3 with S7 , and S4 with S8 , in the first merge step. This observa- tion is useful in implementing Phase II efficiently.

M23_LIAN1878_11_GE_C23.indd 926 12/29/17 4:40 PM

23.8 External Sort 927

23.8.1 Implementing Phase I Listing 23.12 gives the method that reads each segment of data from a file, sorts the segment, and stores the sorted segments into a new file. The method returns the number of segments.

Listing 23.12 Creating Initial Sorted Segments 1 /** Sort original file into sorted segments */ 2 private static int initializeSegments 3 (int segmentSize, String originalFile, String f1) 4 throws Exception { 5 int[] list = new int[segmentSize]; 6 DataInputStream input = new DataInputStream( 7 new BufferedInputStream(new FileInputStream(originalFile))); 8 DataOutputStream output = new DataOutputStream( 9 new BufferedOutputStream(new FileOutputStream(f1))); 10 11 int numberOfSegments = 0; 12 while (input.available() > 0) { 13 numberOfSegments++; 14 int i = 0; 15 for ( ; input.available() > 0 && i < segmentSize; i++) { 16 list[i] = input.readInt(); 17 } 18 19 // Sort an array list[0..i−1] 20 java.util.Arrays.sort(list, 0, i); 21 22 // Write the array to f1.dat 23 for (int j = 0; j < i; j++) { 24 output.writeInt(list[j]); 25 } 26 } 27 28 input.close(); 29 output.close(); 30 31 return numberOfSegments; 32 }

The method creates an array with the maximum size in line 5, a data input stream for the original file in line 6, and a data output stream for a temporary file in line 8. Buffered streams are used to improve performance.

Lines 14–17 read a segment of data from the file into the array. Line 20 sorts the array. Lines 23–25 write the data in the array to the temporary file.

The number of segments is returned in line 31. Note every segment has MAX_ARRAY_SIZE number of elements except the last segment, which may have fewer elements.

23.8.2 Implementing Phase II In each merge step, two sorted segments are merged to form a new segment. The size of the new segment is doubled. The number of segments is reduced by half after each merge step. A segment is too large to be brought to an array in memory. To implement a merge step, copy half the number of segments from the file f1.dat to a temporary file f2.dat. Then, merge the first remaining segment in f1.dat with the first segment in f2.dat into a temporary file named f3.dat, as shown in Figure 23.19.

original file

file with sorted segments

sort a segment

output to file

close file

return # of segments

M23_LIAN1878_11_GE_C23.indd 927 12/29/17 4:40 PM

928 Chapter 23 Sorting

Note f1.dat may have one segment more than f2.dat. If so, move the last segment into f3.dat after the merge.

Listing 23.13 gives a method that copies the first half of the segments in f1.dat to f2.dat. Listing 23.14 gives a method that merges a pair of segments in f1.dat and f2.dat. Listing 23.15 gives a method that merges two segments.

Listing 23.13 Copying First Half Segments 1 private static void copyHalfToF2(int numberOfSegments, 2 int segmentSize, DataInputStream f1, DataOutputStream f2) 3 throws Exception { 4 for (int i = 0; i < (numberOfSegments / 2) * segmentSize; i++) { 5 f2.writeInt(f1.readInt()); 6 } 7 }

Listing 23.14 Merging All Segments 1 private static void mergeSegments(int numberOfSegments, 2 int segmentSize, DataInputStream f1, DataInputStream f2, 3 DataOutputStream f3) throws Exception { 4 for (int i = 0; i < numberOfSegments; i++) { 5 mergeTwoSegments(segmentSize, f1, f2, f3); 6 } 7 8 // If f1 has one extra segment, copy it to f3 9 while (f1.available() > 0) { 10 f3.writeInt(f1.readInt()); 11 } 12 }

Listing 23.15 Merging Two Segments 1 private static void mergeTwoSegments(int segmentSize, 2 DataInputStream f1, DataInputStream f2, 3 DataOutputStream f3) throws Exception { 4 int intFromF1 = f1.readInt(); 5 int intFromF2 = f2.readInt(); 6 int f1Count = 1; 7 int f2Count = 1; 8 9 while (true) { 10 if (intFromF1 < intFromF2) { 11 f3.writeInt(intFromF1);

input stream f1 output stream f2

segments copied

input stream f1 and f2 output stream f3

merge two segments

extra segment in f1?

input stream f1 and f2 output stream f3 read from f1 read from f2

write to f3

Figure 23.19 Sorted segments are merged iteratively.

S1 S2 S3 S4 S5 S6 S7 S8 f1.dat

S1 S2 S3 S4 f2.dat

Copy to f2.dat

f3.datS1, S5 merged S2, S6 merged S3, S7 merged S4, S8 merged

M23_LIAN1878_11_GE_C23.indd 928 12/29/17 4:40 PM

23.8 External Sort 929

12 if (f1.available() == 0 || f1Count++ >= segmentSize) { 13 f3.writeInt(intFromF2); 14 break; 15 } 16 else { 17 intFromF1 = f1.readInt(); 18 } 19 } 20 else { 21 f3.writeInt(intFromF2); 22 if (f2.available() == 0 || f2Count++ >= segmentSize) { 23 f3.writeInt(intFromF1); 24 break; 25 } 26 else { 27 intFromF2 = f2.readInt(); 28 } 29 } 30 } 31 32 while (f1.available() > 0 && f1Count++ < segmentSize) { 33 f3.writeInt(f1.readInt()); 34 } 35 36 while (f2.available() > 0 && f2Count++ < segmentSize) { 37 f3.writeInt(f2.readInt()); 38 } 39 }

23.8.3 Combining Two Phases Listing 23.16 gives the complete program for sorting int values in largedata.dat and storing the sorted data in sortedfile.dat.

Listing 23.16 SortLargeFile.java 1 import java.io.*; 2 3 public class SortLargeFile { 4 public static final int MAX_ARRAY_SIZE = 100000; 5 public static final int BUFFER_SIZE = 100000; 6 7 public static void main(String[] args) throws Exception { 8 // Sort largedata.dat to sortedfile.dat 9 sort("largedata.dat", "sortedfile.dat"); 10 11 // Display the first 100 numbers in the sorted file 12 displayFile("sortedfile.dat"); 13 } 14 15 /** Sort data in source file and into target file */ 16 public static void sort(String sourcefile, String targetfile) 17 throws Exception { 18 // Implement Phase 1: Create initial segments 19 int numberOfSegments = 20 initializeSegments(MAX_ARRAY_SIZE, sourcefile, "f1.dat"); 21 22 // Implement Phase 2: Merge segments recursively 23 merge(numberOfSegments, MAX_ARRAY_SIZE, 24 "f1.dat", "f2.dat", "f3.dat", targetfile);

write to f3

segment in f2 finished

remaining f1 segment

remaining f2 segment

max array size I/O stream buffer size

create initial segments

merge recursively

segment in f1 finished

M23_LIAN1878_11_GE_C23.indd 929 12/29/17 4:40 PM

930 Chapter 23 Sorting

25 } 26 27 /** Sort original file into sorted segments */ 28 private static int initializeSegments 29 (int segmentSize, String originalFile, String f1) 30 throws Exception { 31 // Same as Listing 23.12, so omitted 32 } 33 34 private static void merge(int numberOfSegments, int segmentSize, 35 String f1, String f2, String f3, String targetfile) 36 throws Exception { 37 if (numberOfSegments > 1) { 38 mergeOneStep(numberOfSegments, segmentSize, f1, f2, f3); 39 merge((numberOfSegments + 1) / 2, segmentSize * 2, 40 f3, f1, f2, targetfile); 41 } 42 else { // Rename f1 as the final sorted file 43 File sortedFile = new File(targetfile); 44 if (sortedFile.exists()) sortedFile.delete(); 45 new File(f1).renameTo(sortedFile); 46 } 47 } 48 49 private static void mergeOneStep(int numberOfSegments, 50 int segmentSize, String f1, String f2, String f3) 51 throws Exception { 52 DataInputStream f1Input = new DataInputStream( 53 new BufferedInputStream(new FileInputStream(f1), BUFFER_SIZE)); 54 DataOutputStream f2Output = new DataOutputStream( 55 new BufferedOutputStream(new FileOutputStream(f2), BUFFER_SIZE)); 56 57 // Copy half number of segments from f1.dat to f2.dat 58 copyHalfToF2(numberOfSegments, segmentSize, f1Input, f2Output); 59 f2Output.close(); 60 61 // Merge remaining segments in f1 with segments in f2 into f3 62 DataInputStream f2Input = new DataInputStream( 63 new BufferedInputStream(new FileInputStream(f2), BUFFER_SIZE)); 64 DataOutputStream f3Output = new DataOutputStream( 65 new BufferedOutputStream(new FileOutputStream(f3), BUFFER_SIZE)); 66 67 mergeSegments(numberOfSegments / 2, 68 segmentSize, f1Input, f2Input, f3Output); 69 70 f1Input.close(); 71 f2Input.close(); 72 f3Output.close(); 73 } 74 75 /** Copy first half number of segments from f1.dat to f2.dat */ 76 private static void copyHalfToF2(int numberOfSegments, 77 int segmentSize, DataInputStream f1, DataOutputStream f2) 78 throws Exception { 79 // Same as Listing 23.13, so omitted 80 } 81 82 /** Merge all segments */ 83 private static void mergeSegments(int numberOfSegments, 84 int segmentSize, DataInputStream f1, DataInputStream f2,

merge one step merge recursively

final sorted file

input stream f1Input

output stream f2Output

copy half segments to f2 close f2Output

input stream f2Input

output stream f3Output

merge two segments

close streams

M23_LIAN1878_11_GE_C23.indd 930 12/29/17 4:40 PM

23.8 External Sort 931

85 DataOutputStream f3) throws Exception { 86 // Same as Listing 23.14, so omitted 87 } 88 89 /** Merges two segments */ 90 private static void mergeTwoSegments(int segmentSize, 91 DataInputStream f1, DataInputStream f2, 92 DataOutputStream f3) throws Exception { 93 // Same as Listing 23.15, so omitted 94 } 95 96 /** Display the first 100 numbers in the specified file */ 97 public static void displayFile(String filename) { 98 try { 99 DataInputStream input = 100 new DataInputStream(new FileInputStream(filename)); 101 for (int i = 0; i < 100; i++) 102 System.out.print(input.readInt() + " "); 103 input.close(); 104 } 105 catch (IOException ex) { 106 ex.printStackTrace(); 107 } 108 } 109 }

display file

0 1 1 1 2 2 2 3 3 4 5 6 8 8 9 9 9 10 10 11...(omitted)

Before you run this program, first run Listing 23.11, CreateLargeFile.java, to create the file largedata.dat. Invoking sort("largedata.dat", "sortedfile.dat") (line 9) reads data from largedata.dat and writes sorted data to sortedfile.dat. Invoking displayFile("sortedfile.dat") (line 12) displays the first 100 numbers in the speci- fied file. Note the files are created using binary I/O. You cannot view them using a text editor such as Notepad.

The sort method first creates initial segments from the original array and stores the sorted segments in a new file, f1.dat (lines 19–20), then produces a sorted file in targetfile (lines 23–24).

The merge method

merge(int numberOfSegments, int segmentSize, String f1, String f2, String f3, String targetfile)

merges the segments in f1 into f3 using f2 to assist the merge. The merge method is invoked recursively with many merge steps. Each merge step reduces the numberOfSegments by half and doubles the sorted segment size. After the completion of one merge step, the next merge step merges the new segments in f3 to f2 using f1 to assist the merge. The statement to invoke the new merge method is

merge((numberOfSegments + 1) / 2, segmentSize * 2, f3, f1, f2, targetfile);

The numberOfSegments for the next merge step is (numberOfSegments + 1) / 2. For example, if numberOfSegments is 5, numberOfSegments is 3 for the next merge step because every two segments are merged but one is left unmerged.

The recursive merge method ends when numberOfSegments is 1. In this case, f1 contains sorted data. File f1 is renamed to targetfile (line 45).

M23_LIAN1878_11_GE_C23.indd 931 12/29/17 4:40 PM

932 Chapter 23 Sorting

23.8.4 External Sort Complexity In the external sort, the dominating cost is that of I/O. Assume n is the number of elements to be sorted in the file. In Phase I, n number of elements are read from the original file and output to a temporary file. Therefore, the I/O for Phase I is O(n).

In Phase II, before the first merge step, the number of sorted segments is n c

, where c is

MAX_ARRAY_SIZE. Each merge step reduces the number of segments by half. Thus, after the

first merge step, the number of segments is n 2c

. After the second merge step, the number of

segments is n

22c , and after the third merge step the number of segments is

23c . After logan

c b

merge steps, the number of segments is reduced to 1. Therefore, the total number of merge

steps is logan c b .

In each merge step, half the number of segments are read from file f1 then written into a temporary file f2. The remaining segments in f1 are merged with the segments in f2. The

number of I/Os in each merge step is O(n). Since the total number of merge steps is logan c b ,

the total number of I/Os is

O(n) * logan c b = O(n log n)

Therefore, the complexity of the external sort is O(n logn).

23.8.1 Describe how external sort works. What is the complexity of the external sort algorithm?

23.8.2 Ten numbers {2, 3, 4, 0, 5, 6, 7, 9, 8, 1} are stored in the external file largedata.dat. Trace the SortLargeFile program by hand with MAX_ARRAY_SIZE 2.

Point Check

Key terms bubble sort 906 bucket sort 923 complete binary tree 917 external sort 925 heap 917

heap sort 916 height of a heap 922 merge sort 922 quick sort 912 radix sort 924

ChapTer summary 1. The worst-case complexity for a selection sort, insertion sort, bubble sort, and quick

sort is O(n2).

2. The average- and worst-case complexity for a merge sort is O(n logn). The average time for a quick sort is also O(n logn).

3. Heaps are a useful data structure for designing efficient algorithms such as sorting. You learned how to define and implement a heap class, and how to insert and delete elements to/from a heap.

4. The time complexity for a heap sort is O(n logn).

5. Bucket and radix sorts are specialized sorting algorithms for integer keys. These algo- rithms sort keys using buckets rather than by comparing keys. They are more efficient than general sorting algorithms.

M23_LIAN1878_11_GE_C23.indd 932 12/29/17 4:40 PM

Programming Exercises 933

6. A variation of the merge sort—called an external sort—can be applied to sort large amounts of data from external files.

Quiz Answer the quiz for this chapter online at the book Companion Website.

programming exerCiSeS

Sections 23.3–23.5 23.1 (Generic bubble sort) Write the following two generic methods using bubble

sort. The first method sorts the elements using the Comparable interface, and the second uses the Comparator interface.

public static <E extends Comparable<E>> void bubbleSort(E[] list) public static <E> void bubbleSort(E[] list, Comparator<? super E> comparator)

23.2 (Generic merge sort) Write the following two generic methods using merge sort. The first method sorts the elements using the Comparable interface and the second uses the Comparator interface.

public static <E extends Comparable<E>> void mergeSort(E[] list) public static <E> void mergeSort(E[] list, Comparator<? super E> comparator)

23.3 (Generic quick sort) Write the following two generic methods using quick sort. The first method sorts the elements using the Comparable interface, and the second uses the Comparator interface.

public static <E extends Comparable<E>> void quickSort(E[] list) public static <E> void quickSort(E[] list, Comparator<? super E> comparator)

23.4 (Improve quick sort) The quick-sort algorithm presented in the book selects the first element in the list as the pivot. Revise it by selecting the median among the first, middle, and the last elements in the list.

*23.5 (Generic Heap using Comparator) Revise Heap in Listing 23.9, using a generic parameter and a Comparator for comparing objects. Define the class as follows:

class HeapWithComparator<E> { private Comparator<? super E> comparator; // For comparing elements

public HeapWithComparator() { // Implement no−arg constructor by creating a comparator for natural order }

public HeapWithComparator(Comparator<? super E> comparator) { this.comparator = comparator; } // Implement all add, remove, and getSize method }

M23_LIAN1878_11_GE_C23.indd 933 12/29/17 4:40 PM

934 Chapter 23 Sorting

23.6 (Generic insertion sort) Write the following two generic methods using insertion sort. The first method sorts the elements using the Comparable interface and the second uses the Comparator interface.

public static <E extends Comparable<E>> voidinsertionSort (E[] list) public static <E void insertionSort(E[] list, Comparator<?super E> comparator)

Section 23.6 23.7 (Min-heap) The heap presented in the text is also known as a max-heap, in which

each node is greater than or equal to any of its children. A min-heap is a heap in which each node is less than or equal to any of its children. Min-heaps are often used to implement priority queues. Revise the Heap class in Listing 23.9 to implement a min-heap.

23.8 (Generic insertion sort) Write the following two generic methods using insertion sort. The first method sorts the elements using the Comparable interface, and the second uses the Comparator interface.

public static <E extends Comparable<E>> void insertionSort(E[] list) public static <E> void insertionSort(E[] list, Comparator<? super E> comparator)

*23.9 (Generic heap sort) Write the following two generic methods using heap sort. The first method sorts the elements using the Comparable interface, and the sec- ond uses the Comparator interface. (Hint: Use the Heap class in Programming Exercise 23.5.)

public static <E extends Comparable<E>> void heapSort(E[] list) public static <E> void heapSort(E[] list, Comparator<? super E> comparator)

**23.10 (Heap visualization) Write a program that displays a heap graphically, as shown in Figure 23.10. The program lets you insert and delete an element from the heap.

23.11 (Heap clone and equals) Implement the clone and equals method in the Heap class.

Section 23.7 *23.12 (Radix sort) Write a program that randomly generates 1,200,000 integers and

sorts them using radix sort. *23.13 (Execution time for sorting) Write a program that obtains the execution time of

selection sort, bubble sort, merge sort, quick sort, heap sort, and radix sort for

max-heap min-heap

M23_LIAN1878_11_GE_C23.indd 934 12/29/17 4:40 PM

Programming Exercises 935

input size 60,000, 120,000, 180,000, 240,000, 300,000, and 360,000. Your pro- gram should create data randomly and print a table like this:

Array size

Selection Sort

Bubble Sort

Merge Sort

Quick Sort

Heap Sort

Radix Sort

60,000 120,000 180,000 240,000 300,000 360,000

(Hint: You can use the following code template to obtain the execution time.)

long startTime = System.currentTimeMillis(); perform the task; long endTime = System.currentTimeMillis(); long executionTime = endTime − startTime;

The text gives a recursive quick sort. Write a nonrecursive version in this exercise.

Section 23.8 *23.14 (Execution time for external sorting) Write a program that obtains the execution

time of external sorts for integers of size 6,000,000, 12,000,000, 18,000,000, 24,000,000, 30,000,000, and 36,000,000. Your program should print a table like this:

File size 6,000,000 12,000,000 18,000,000 24,000,000 30,000,000 36,000,000 Time

Comprehensive *23.15 (Selection-sort animation) Write a program that animates the selection sort algo-

rithm. Create an array that consists of 25 distinct numbers from 1 to 25 in a random order. The array elements are displayed in a histogram, as shown in Figure 23.20a. Clicking the Step button causes the program to perform an iteration of the outer loop in the algorithm and repaints the histogram for the new array. Color the last bar in the sorted subarray. When the algorithm is finished, display a message to inform the user. Clicking the Reset button creates a new random array for a new start. (You can easily modify the program to animate the insertion algorithm.)

Figure 23.20 (a) The program animates selection sort. (b) The program animates bubble sort.

(b)(a)

M23_LIAN1878_11_GE_C23.indd 935 12/29/17 4:40 PM

936 Chapter 23 Sorting

*23.16 (Bubble-sort animation) Write a program that animates the bubble sort algo- rithm. Create an array that consists of 25 distinct numbers from 1 to 25 in a random order. The array elements are displayed in a histogram, as shown in Figure 23.20b. Clicking the Step button causes the program to perform one com- parison in the algorithm and repaints the histogram for the new array. Color the bar that represents the number being considered in the swap. When the algo- rithm is finished, display a message to inform the user. Clicking the Reset button creates a new random array for a new start.

*23.17 (Radix-sort animation) Write a program that animates the radix sort algorithm. Cre- ate an array that consists of 20 random numbers from 0 to 800. The array elements are displayed, as shown in Figure 23.21. Clicking the Step button causes the program to place a number in a bucket. The number that has just been placed is displayed in red. Once all the numbers are placed in the buckets, clicking the Step button col- lects all the numbers from the buckets and moves them back to the array. When the algorithm is finished, clicking the Step button displays a message to inform the user. Clicking the Reset button creates a new random array for a new start.

Figure 23.21 The program animates radix sort.

Figure 23.22 The program animates a merge of two sorted lists. (b) The program animates a partition for quick sort.

(b)(a)

*23.18 (Merge animation) Write a program that animates the merge of two sorted lists. Create two arrays, list1 and list2, each of which consists of 8 random numbers from 1 to 999. The array elements are displayed, as shown in Figure 23.22a. Clicking the Step button causes the program to move an element from list1 or list2 to temp. Clicking the Reset button creates two new random arrays for a new start. When the algorithm is finished, clicking the Step button displays a message to inform the user.

M23_LIAN1878_11_GE_C23.indd 936 12/29/17 4:40 PM

Programming Exercises 937

*23.19 (Quick-sort partition animation) Write a program that animates the partition for a quick sort. The program creates a list that consists of 25 random numbers from 1 to 999. The list is displayed, as shown in Figure 23.22b. Clicking the Step but- ton causes the program to move low to the right or high to the left, or swap the elements at low and high. Clicking the Reset button creates a new list of random numbers for a new start. When the algorithm is finished, clicking the Step button displays a message to inform the user.

*23.20 (Modify merge sort) Rewrite the mergeSort method to recursively sort the first half of the array and the second half of the array without creating new temporary arrays, then merge the two into a temporary array and copy its contents to the original array, as shown in Figure 23.6b.

M23_LIAN1878_11_GE_C23.indd 937 12/29/17 4:40 PM

A01_GIPE3898_09_SE_FM.indd 3 12/20/17 5:56 PM

This page intentionally left blank

Objectives ■■ To design common operations of lists in an interface and make the

interface a subtype of Collection (§24.2).

■■ To design and implement an array list using an array (§24.3).

■■ To design and implement a linked list using a linked structure (§24.4).

■■ To design and implement a stack class using an array list and a queue class using a linked list (§24.5).

■■ To design and implement a priority queue using a heap (§24.6).

Implementing Lists, Stacks, Queues, and Priority Queues

Chapter

M24_LIAN1878_11_GE_C24.indd 939 12/29/17 4:41 PM

940 Chapter 24 Implementing Lists, Stacks, Queues, and Priority Queues

24.1 Introduction This chapter focuses on implementing data structures.

Lists, stacks, queues, and priority queues are classic data structures typically covered in a data structures course. They are supported in the Java API, and their uses were presented in Chapter 20, Lists, Stacks, Queues, and Priority Queues. This chapter will examine how these data structures are implemented under the hood. Implementation of sets and maps will be covered in Chapter 27. Through these implementations, you will gain valuable insight on data structures and learn how to design and implement custom data structures.

24.2 Common Operations for Lists Common operations of lists are defined in the List interface.

A list is a popular data structure for storing data in sequential order—for example, a list of students, a list of available rooms, a list of cities, and a list of books. You can perform the following operations on a list:

■■ Retrieve an element from the list.

■■ Insert a new element into the list.

■■ Delete an element from the list.

■■ Find out how many elements are in the list.

■■ Determine whether an element is in the list.

■■ Check whether the list is empty.

There are two ways to implement a list. One is to use an array to store the elements. Array size is fixed. If the capacity of the array is exceeded, you need to create a new, larger array and copy all the elements from the current array to the new array. The other approach is to use a linked structure. A linked structure consists of nodes. Each node is dynamically created to hold an element. All the nodes are linked together to form a list. Thus, you can define two classes for lists. For convenience, let’s name these two classes MyArrayList and MyLinkedList. These two classes have common operations but dif- ferent implementations.

Design Guide Prior to Java 8, a popular design strategy for Java data structures is to define common operations in interfaces and provide convenient abstract classes for partially implement- ing the interfaces. So, the concrete classes can simply extend the convenient abstract classes without implementing the full interfaces. Java 8 enables you to define default methods. You can provide default implementation for some of the methods in the interfaces rather than in convenient abstract classes. Using default methods eliminate the need for convenient abstract classes.

Pedagogical Note For an interactive demo on how array lists and linked lists work, go to liveexample .pearsoncmg.com/dsanimation/ArrayListeBook.html and liveexample.pearsoncmg.com/ dsanimation/LinkedListeBook.html, as shown in Figure 24.1.

Let us name the interface MyList and define it as a subtype of Collection so the common operations in the Collection interface are also available in MyList. Figure 24.2

Point Key

default methods in interfaces

list animation on Companion Website

Point Key

M24_LIAN1878_11_GE_C24.indd 940 12/29/17 4:41 PM

24.2 Common Operations for Lists 941

Figure 24.1 The animation tool enables you to see how array lists and linked lists work.

(b) LinkedList animation

Figure 24.2 MyList defines a common interface for MyArrayList and MyLinkedList.

java.util.Collection MyList

MyArrayList

MyLinkedList

java.util.Iterable

shows the relationship of Collection, MyList, MyArrayList, and MyLinkedList. The methods in MyList are shown in Figure 24.3. Listing 24.1 gives the source code for MyList.

M24_LIAN1878_11_GE_C24.indd 941 12/29/17 4:41 PM

942 Chapter 24 Implementing Lists, Stacks, Queues, and Priority Queues

Listing 24.1 MyList.java 1 import java.util.Collection; 2 3 public interface MyList<E> extends Collection<E> { 4 /** Add a new element at the specified index in this list */ 5 public void add(int index, E e); 6 7 /** Return the element from this list at the specified index */ 8 public E get(int index); 9 10 /** Return the index of the first matching element in this list. 11 * Return −1 if no match. */ 12 public int indexOf(Object e); 13 14 /** Return the index of the last matching element in this list 15 * Return −1 if no match. */ 16 public int lastIndexOf(E e); 17 18 /** Remove the element at the specified position in this list 19 * Shift any subsequent elements to the left. 20 * Return the element that was removed from the list. */ 21 public E remove(int index); 22 23 /** Replace the element at the specified position in this list 24 * with the specified element and returns the new set. */ 25 public E set(int index, E e); 26 27 @Override /** Add a new element at the end of this list */ 28 public default boolean add(E e) { 29 add(size(), e); 30 return true;

remove(e)

set(index, e)

lastIndexOf(e)

indexOf(e)

get(index)

default add(e)

add(index, e)

Figure 24.3 MyList defines the methods for manipulating a list and partially implements some of the methods defined in the Collection interface.

«interface» java.util.Collection<E>

«interface» MyList<E>

«interface» java.util.Iterable<E>

+add(index: int, e: E): void +get(index: int): E +indexOf(e: Object): int +lastIndexOf(e: E): int +remove(index: int): E +set(index: int, e: E): E

Inserts a new element at the speci�ed index in this list. Returns the element from this list at the speci�ed index. Returns the index of the �rst matching element in this list. Returns the index of the last matching element in this list. Removes the element at the speci�ed index and returns the removed element. Sets the element at the speci�ed index and returns the element being replaced.

Override the add, isEmpty, remove, containsAll, addAll, removeAll, retainAll, toArray(), and toArray(T[]) methods de�ned in Collection using default methods.

M24_LIAN1878_11_GE_C24.indd 942 12/29/17 4:41 PM

24.2 Common Operations for Lists 943

31 } 32 33 @Override /** Return true if this list contains no elements */ 34 public default boolean isEmpty() { 35 return size() == 0; 36 } 37 38 @Override /** Remove the first occurrence of the element e 39 * from this list. Shift any subsequent elements to the left. 40 * Return true if the element is removed. */ 41 public default boolean remove(Object e) { 42 if (indexOf(e) >= 0) { 43 remove(indexOf(e)); 44 return true; 45 } 46 else 47 return false; 48 } 49 50 @Override 51 public default boolean containsAll(Collection<?> c) { 52 // Left as an exercise 53 return true; 54 } 55 56 @Override 57 public default boolean addAll(Collection<? extends E> c) { 58 // Left as an exercise 59 return true; 60 } 61 62 @Override 63 public default boolean removeAll(Collection<?> c) { 64 // Left as an exercise 65 return true; 66 } 67 68 @Override 69 public default boolean retainAll(Collection<?> c) { 70 // Left as an exercise 71 return true; 72 } 73 74 @Override 75 public default Object[] toArray() { 76 // Left as an exercise 77 return null; 78 } 79 80 @Override 81 public default <T> T[] toArray(T[] array) { 82 // Left as an exercise 83 return null; 84 } 85 }

The methods isEmpty(), add(E), remove(E), containsAll, addAll, removeAll, retainAll, toArray(), and toArray(T[]) are defined in the Collection interface. Since these methods are implementable in MyList, they are overridden in the MyList interface as default methods. The implementation for isEmpty(), add(E), and remove(E) are

implement retainAll

implement toArray(T[])

implement toArray()

implement removeAll

implement addAll

implement containsAll

implement remove(E e)

default isEmpty()

M24_LIAN1878_11_GE_C24.indd 943 12/29/17 4:41 PM

944 Chapter 24 Implementing Lists, Stacks, Queues, and Priority Queues

provided and the implementation for other default methods are left as exercises in Programming Exercise 24.1.

The following sections give the implementation for MyArrayList and MyLinkedList, respectively.

24.2.1 Suppose list is an instance of MyList, can you get an iterator for list using list .iterator()?

24.2.2 Can you create a list using new MyList()? 24.2.3 What methods in Collection are overridden as default methods in MyList? 24.2.4 What are the benefits of overriding the methods in Collection as default methods

in MyList?

24.3 Array Lists An array list is implemented using an array.

An array is a fixed-size data structure. Once an array is created, its size cannot be changed. Nevertheless, you can still use arrays to implement dynamic data structures. The trick is to create a larger new array to replace the current array, if the current array cannot hold new elements in the list.

Initially, an array, say data of E[] type, is created with a default size. When inserting a new element into the array, first make sure there is enough room in the array. If not, create a new array twice as large as the current one. Copy the elements from the current array to the new array. The new array now becomes the current array. Before inserting a new element at a specified index, shift all the elements after the index to the right and increase the list size by 1, as shown in Figure 24.4.

Point Check

Point Key

Figure 24.4 Inserting a new element into the array requires that all the elements after the insertion point be shifted one position to the right, so the new element can be inserted at the insertion point.

0 1 … i i + 1 k – 1 k + 1Before inserting e at insertion point i

…

data.length – 1 Insertion point

0 1 … i i + 1After inserting e at insertion point i, list size is incremented by 1

…

…shift…

data.length – 1e inserted here

k + 1ki + 2

e1 … ei ei+1 … ek–1 ekei–1

e0 e1 … e ei+1ei … ek–1 ekei–1

Note The data array is of type E[]. Each cell in the array actually stores the reference of an object.

To remove an element at a specified index, shift all the elements after the index to the left by one position and decrease the list size by 1, as shown in Figure 24.5.

M24_LIAN1878_11_GE_C24.indd 944 12/29/17 4:41 PM

24.3 Array Lists 945

MyArrayList uses an array to implement MyList, as shown in Figure 24.6. Its implemen- tation is given in Listing 24.2.

Figure 24.5 Deleting an element from the array requires that all the elements after the deletion point be shifted one position to the left.

Before deleting the element at index i e0

0 1 … i i + 1 k – 1

e1 … ei ei+1

…

… ek–1

data.length – 1Delete this element ...shift...

0 1 … iAfter deleting the element, list size is decremented by 1

e1 …

…

… ek

data.length – 1

ei–1

k – 1

ei+1

k – 2

ek–1

Figure 24.6 MyArrayList implements a list using an array.

–data: E[] –size: int

+MyArrayList() +MyArrayList(objects: E[]) +trimToSize(): void

–ensureCapacity(): void –checkIndex(index: int): void

Creates a default array list. Creates an array list from an array of objects. Trims the capacity of this array list to the list’s current size. Doubles the current array size if needed. Throws an exception if the index is out of bounds in the list.

Array for storing elements in this array list. The number of elements in the array list.

MyArrayList<E>

MyList<E> «interface»

Listing 24.2 MyArrayList.java 1 public class MyArrayList<E> implements MyList<E> { 2 public static final int INITIAL_CAPACITY = 16; 3 private E[] data = (E[])new Object[INITIAL_CAPACITY]; 4 private int size = 0; // Number of elements in the list 5 6 /** Create an empty list */ 7 public MyArrayList() { 8 } 9 10 /** Create a list from an array of objects */

no-arg constructor

number of elements create an array initial capacity

M24_LIAN1878_11_GE_C24.indd 945 12/29/17 4:41 PM

946 Chapter 24 Implementing Lists, Stacks, Queues, and Priority Queues

11 public MyArrayList(E[] objects) { 12 for (int i = 0; i < objects.length; i++) 13 add(objects[i]); // Warning: don't use super(objects)! 14 } 15 16 @Override /** Add a new element at the specified index */ 17 public void add(int index, E e) { 18 // Ensure the index is in the right range 19 if (index < 0 || index > size) 20 throw new IndexOutOfBoundsException 21 ("Index: " + index + ", Size: " + size); 22 23 ensureCapacity(); 24 25 // Move the elements to the right after the specified index 26 for (int i = size − 1; i >= index; i−−) 27 data[i + 1] = data[i]; 28 29 // Insert new element to data[index] 30 data[index] = e; 31 32 // Increase size by 1 33 size++; 34 } 35 36 /** Create a new larger array, double the current size + 1 */ 37 private void ensureCapacity() { 38 if (size >= data.length) { 39 E[] newData = (E[])(new Object[size * 2 + 1]); 40 System.arraycopy(data, 0, newData, 0, size); 41 data = newData; 42 } 43 } 44 45 @Override /** Clear the list */ 46 public void clear() { 47 data = (E[])new Object[INITIAL_CAPACITY]; 48 size = 0; 49 } 50 51 @Override /** Return true if this list contains the element */ 52 public boolean contains(Object e) { 53 for (int i = 0; i < size; i++) 54 if (e.equals(data[i])) return true; 55 56 return false; 57 } 58 59 @Override /** Return the element at the specified index */ 60 public E get(int index) { 61 checkIndex(index); 62 return data[index]; 63 } 64 65 private void checkIndex(int index) { 66 if (index < 0 || index >= size) 67 throw new IndexOutOfBoundsException 68 ("Index: " + index + ", Size: " + size); 69 } 70

constructor

add

ensureCapacity

double capacity + 1

clear

contains

get

checkIndex

M24_LIAN1878_11_GE_C24.indd 946 12/29/17 4:41 PM

24.3 Array Lists 947

71 @Override /** Return the index of the first matching element 72 * in this list. Return −1 if no match. */ 73 public int indexOf(Object e) { 74 for (int i = 0; i < size; i++) 75 if (e.equals(data[i])) return i; 76 77 return −1; 78 } 79 80 @Override /** Return the index of the last matching element 81 * in this list. Return −1 if no match. */ 82 public int lastIndexOf(E e) { 83 for (int i = size − 1; i >= 0; i−−) 84 if (e.equals(data[i])) return i; 85 86 return −1; 87 } 88 89 @Override /** Remove the element at the specified position 90 * in this list. Shift any subsequent elements to the left. 91 * Return the element that was removed from the list. */ 92 public E remove(int index) { 93 checkIndex(index); 94 95 E e = data[index]; 96 97 // Shift data to the left 98 for (int j = index; j < size − 1; j++) 99 data[j] = data[j + 1]; 100 101 data[size − 1] = null; // This element is now null 102 103 // Decrement size 104 size−−; 105 106 return e; 107 } 108 109 @Override /** Replace the element at the specified position 110 * in this list with the specified element. */ 111 public E set(int index, E e) { 112 checkIndex(index); 113 E old = data[index]; 114 data[index] = e; 115 return old; 116 } 117 118 @Override 119 public String toString() { 120 StringBuilder result = new StringBuilder("["); 121 122 for (int i = 0; i < size; i++) { 123 result.append(data[i]); 124 if (i < size − 1) result.append(", "); 125 } 126 127 return result.toString() + "]"; 128 } 129 130 /** Trims the capacity to current size */

indexOf

lastIndexOf

remove

set

toString

M24_LIAN1878_11_GE_C24.indd 947 12/29/17 4:41 PM

948 Chapter 24 Implementing Lists, Stacks, Queues, and Priority Queues

131 public void trimToSize() { 132 if (size != data.length) { 133 E[] newData = (E[])(new Object[size]); 134 System.arraycopy(data, 0, newData, 0, size); 135 data = newData; 136 } // If size == capacity, no need to trim 137 } 138 139 @Override /** Override iterator() defined in Iterable */ 140 public java.util.Iterator<E> iterator() { 141 return new ArrayListIterator(); 142 } 143 144 private class ArrayListIterator 145 implements java.util.Iterator<E> { 146 private int current = 0; // Current index 147 148 @Override 149 public boolean hasNext() { 150 return current < size; 151 } 152 153 @Override 154 public E next() { 155 return data[current++]; 156 } 157 158 @Override // Remove the element returned by the last next() 159 public void remove() { 160 if (current == 0) // next() has not been called yet 161 throw new IllegalStateException(); 162 MyArrayList.this.remove(−−current); 163 } 164 } 165 166 @Override /** Return the number of elements in this list */ 167 public int size() { 168 return size; 169 } 170 }

The constant INITIAL_CAPACITY (line 2) is used to create an initial array data (line 3). Owing to generics type erasure (see Restriction 2 in Section 19.8), you cannot create a generic array using the syntax new e[INITIAL_CAPACITY]. To circumvent this limitation, an array of the Object type is created in line 3 and cast into E[]. The size data field tracks the number of elements in the list (line 4).

The add(int index, E e) method (lines 17–34) inserts the element e at the specified index in the array. This method first invokes ensureCapacity() (line 23), which ensures that there is a space in the array for the new element. It then shifts all the elements after the index one position to the right before inserting the element (lines 26 and 27). After the element is added, size is incremented by 1 (line 33).

The ensureCapacity() method (lines 37–43) checks whether the array is full. If so, the program creates a new array that doubles the current array size + 1, copies the current array to the new array using the System.arraycopy method, and sets the new array as the current array. Note the current size might be 0 after invoking the trimToSize() method. new Object[2 * size + 1] (line 39) ensures that the new size is not 0.

The clear() method (lines 46–49) creates a new array using the size as INITIAL_ CAPACITY and resets the variable size to 0. The class will work if line 47 is deleted. However, the class

size

iterator

trimToSize

add

ensureCapacity

clear

M24_LIAN1878_11_GE_C24.indd 948 12/29/17 4:41 PM

24.3 Array Lists 949

will have a memory leak because the elements are still in the array, although they are no longer needed. By creating a new array and assigning it to data, the old array and the elements stored in the old array become garbage, which will be automatically collected by the JVM.

The contains(Object e) method (lines 52–57) checks whether element e is contained in the array by comparing e with each element in the array using the equals method.

The get(int index) method (lines 60–63) checks if index is within the range and returns data[index] if index is in the range.

The checkIndex(int index) method (lines 65–69) checks if index is within the range. If not, the method throws an IndexOutOfBoundsException (line 67).

The indexOf(Object e) method (lines 73–78) compares element e with the elements in the array, starting from the first one. If a match is found, the index of the element is returned; otherwise, −1 is returned.

The lastIndexOf(Object e) method (lines 82–87) compares element e with the ele- ments in the array, starting from the last one. If a match is found, the index of the element is returned; otherwise, −1 is returned.

The remove(int index) method (lines 92–107) shifts all the elements after the index one position to the left (lines 98 and 99) and decrements size by 1 (line 104). The last element is not used anymore and is set to null (line 101).

The set(int index, E e) method (lines 111–116) simply assigns e to data[index] to replace the element at the specified index with element e.

The toString() method (lines 119–128) overrides the toString method in the Object class to return a string representing all the elements in the list.

The trimToSize() method (lines 131–137) creates a new array whose size matches the current array-list size (line 133), copies the current array to the new array using the System. arraycopy method (line 134), and sets the new array as the current array (line 135). Note if size == capacity, there is no need to trim the size of the array.

The iterator() method defined in the java.lang.Iterable interface is implemented to return an instance on java.util.Iterator (lines 140–142). The ArrayListIterator class implements Iterator with concrete methods for hasNext, next, and remove (lines 144–164). It uses current to denote the current position of the element being traversed (line 146).

The size() method simply returns the number of elements in the array list (lines 167–169).

Listing 24.3 gives an example that creates a list using MyArrayList. It uses the add method to add strings to the list, and the remove method to remove strings. Since M yArrayList implements Iterable, the elements can be traversed using a foreach loop (lines 35 and 36).

Listing 24.3 TestMyArrayList.java 1 public class TestMyArrayList { 2 public static void main(String[] args) { 3 // Create a list 4 MyList<String> list = new MyArrayList<>(); 5 6 // Add elements to the list 7 list.add("America"); // Add it to the list 8 System.out.println("(1) " + list); 9 10 list.add(0, "Canada"); // Add it to the beginning of the list 11 System.out.println("(2) " + list); 12 13 list.add("Russia"); // Add it to the end of the list 14 System.out.println("(3) " + list); 15 16 list.add("France"); // Add it to the end of the list 17 System.out.println("(4) " + list);

contains

checkIndex

indexOf

lastIndexOf

remove

set

toString

trimToSize

iterator

size()

add to list

create a list

M24_LIAN1878_11_GE_C24.indd 949 12/29/17 4:41 PM

950 Chapter 24 Implementing Lists, Stacks, Queues, and Priority Queues

18 19 list.add(2, "Germany"); // Add it to the list at index 2 20 System.out.println("(5) " + list); 21 22 list.add(5, "Norway"); // Add it to the list at index 5 23 System.out.println("(6) " + list); 24 25 // Remove elements from the list 26 list.remove("Canada"); // Same as list.remove(0) in this case 27 System.out.println("(7) " + list); 28 29 list.remove(2); // Remove the element at index 2 30 System.out.println("(8) " + list); 31 32 list.remove(list.size() − 1); // Remove the last element 33 System.out.print("(9) " + list + "\n(10) "); 34 35 for (String s: list) 36 System.out.print(s.toUpperCase() + " "); 37 } 38 }

use iterator

remove from list

(1) [America] (2) [Canada, America] (3) [Canada, America, Russia] (4) [Canada, America, Russia, France] (5) [Canada, America, Germany, Russia, France] (6) [Canada, America, Germany, Russia, France, Norway] (7) [America, Germany, Russia, France, Norway] (8) [America, Germany, France, Norway] (9) [America, Germany, France] (10) AMERICA GERMANY FRANCE

24.3.1 What are the limitations of the array data type? 24.3.2 MyArrayList is implemented using an array, and an array is a fixed-size data struc-

ture. Why is MyArrayList considered a dynamic data structure?

24.3.3 Show the length of the array in MyArrayList after each of the following statements is executed:

1 MyArrayList<Double> list = new MyArrayList<>(); 2 list.add(1.5); 3 list.trimToSize(); 4 list.add(3.4); 5 list.add(7.4); 6 list.add(17.4);

24.3.4 What is wrong if lines 11 and 12 in Listing 24.2, MyArrayList.java,

for (int i = 0; i < objects.length; i++) add(objects[i]);

are replaced by

data = objects; size = objects.length;

Point Check

M24_LIAN1878_11_GE_C24.indd 950 12/29/17 4:41 PM

24.4 Linked Lists 951

24.3.5 If you change the code in line 33 in Listing 24.2, MyArrayList.java, from

E[] newData = (E[])(new Object[size * 2 + 1]);

E[] newData = (E[])(new Object[size * 2]);

the program is incorrect. Can you find the reason?

24.3.6 Will the MyArrayList class have memory leak if the following code in line 41 is deleted?

data = (E[])new Object[INITIAL_CAPACITY];

24.3.7 The get(index) method invokes the checkIndex(index) method (lines 59–63 in Listing 24.2) to throw an IndexOutOfBoundsException if the index is out of bounds. Suppose the add(index, e) method is implemented as follows:

public void add(int index, E e) { checkIndex(index);

// Same as lines 23−33 in Listing 24.2 MyArrayList.java }

What will happen if you run the following code?

MyArrayList<String> list = new MyArrayList<>(); list.add("New York");

24.4 Linked Lists A linked list is implemented using a linked structure.

Since MyArrayList is implemented using an array, the methods get(int index) and set(int index, E e) for accessing and modifying an element through an index and the add(E e) method for adding an element at the end of the list are efficient. However, the methods add(int index, E e) and remove(int index) are inefficient because they require shifting a potentially large number of elements. You can use a linked structure to implement a list to improve efficiency for adding and removing an element at the beginning of a list.

24.4.1 Nodes In a linked list, each element is contained in an object, called the node. When a new element is added to the list, a node is created to contain it. Each node is linked to its next neighbor, as shown in Figure 24.7.

A node can be created from a class defined as follows:

class Node<E> { E element; Node<E> next;

public Node(E e) { element = e; } }

Point Key

M24_LIAN1878_11_GE_C24.indd 951 12/29/17 4:41 PM

952 Chapter 24 Implementing Lists, Stacks, Queues, and Priority Queues

We use the variable head to refer to the first node in the list and the variable tail to the last node. If the list is empty, both head and tail are null. Here is an example that creates a linked list to hold three nodes. Each node stores a string element.

Step 1: Declare head and tail.

Node<String> head = null; The list is empty now Node<String> tail = null;

head and tail are both null. The list is empty.

Step 2: Create the first node and append it to the list, as shown in Figure 24.8. After the first node is inserted in the list, head and tail point to this node.

Figure 24.8 Append the first node to the list.

After the �rst node is inserted

head

head = new Node<>("Chicago"); tail = head;

tail "Chicago" next: null

Figure 24.9 Append the second node to the list.

head "Chicago" next

"Denver" next: null

tail

head "Chicago" next

"Denver" next: null

tail

tail.next = new Node<>("Denver");

tail = tail.next;

(a)

(b)

Figure 24.7 A linked list consists of any number of nodes chained together.

element 1head next

Node 1 element 2 next

Node 2 … element n

null

Node n tail

Step 3: Create the second node and append it into the list, as shown in Figure 24.9a. To append the second node to the list, link the first node with the new node. The new node is now the tail node, so you should move tail to point to this new node, as shown in Figure 24.9b.

Step 4: Create the third node and append it to the list, as shown in Figure 24.10a. To append the new node to the list, link the last node in the list with the new node. The new node is now the tail node, so you should move tail to point to this new node, as shown in Figure 24.10b.

M24_LIAN1878_11_GE_C24.indd 952 12/29/17 4:41 PM

24.4 Linked Lists 953

Each node contains the element and a data field named next that points to the next element. If the node is the last in the list, its pointer data field next contains the value null. You can use this property to detect the last node. For example, you can write the following loop to traverse all the nodes in the list:

1 Node<E> current = head; 2 while (current != null) { 3 System.out.println(current.element); 4 current = current.next; 5 }

The variable current points initially to the first node in the list (line 1). In the loop, the element of the current node is retrieved (line 3) then current points to the next node (line 4). The loop continues until the current node is null.

24.4.2 The MyLinkedList Class The MyLinkedList class uses a linked structure to implement a dynamic list. It imple- ments MyList. In addition, it provides the methods addFirst, addLast, removeFirst, removeLast, getFirst, and getLast, as shown in Figure 24.11.

next node

check last node current pointer

Figure 24.10 Append the third node to the list.

head "Chicago" next

"Denver" next

"Dallas" next: null

tail

head "Chicago" next

"Denver" next

"Dallas" next: null

tail

tail.next = new Node<>("Dallas");

tail = tail.next;

(a)

(b)

Figure 24.11 MyLinkedList implements a list using a linked list of nodes.

–head: Node<E> –tail: Node<E> –size: int

+MyLinkedList() +MyLinkedList(elements: E[]) +addFirst(e: E): void +addLast(e: E): void +getFirst(): E +getLast(): E +removeFirst(): E +removeLast(): E

Link

«interface» MyList<E>

Creates a default linked list. Creates a linked list from an array of elements. Adds an element to the head of the list. Adds an element to the tail of the list. Returns the �rst element in the list. Returns the last element in the list. Removes the �rst element from the list. Removes the last element from the list.

MyLinkedList<E>

element: E next: Node<E>

Node<E>

The head of the list. The tail of the list. The number of elements in the list.

M24_LIAN1878_11_GE_C24.indd 953 12/29/17 4:41 PM

954 Chapter 24 Implementing Lists, Stacks, Queues, and Priority Queues

Assuming the class has been implemented, Listing 24.4 gives a test program that uses the class.

Listing 24.4 TestMyLinkedList.java 1 public class TestMyLinkedList { 2 /** Main method */ 3 public static void main(String[] args) { 4 // Create a list for strings 5 MyLinkedList<String> list = new MyLinkedList<>(); 6 7 // Add elements to the list 8 list.add("America"); // Add it to the list 9 System.out.println("(1) " + list); 10 11 list.add(0, "Canada"); // Add it to the beginning of the list 12 System.out.println("(2) " + list); 13 14 list.add("Russia"); // Add it to the end of the list 15 System.out.println("(3) " + list); 16 17 list.addLast("France"); // Add it to the end of the list 18 System.out.println("(4) " + list); 19 20 list.add(2, "Germany"); // Add it to the list at index 2 21 System.out.println("(5) " + list); 22 23 list.add(5, "Norway"); // Add it to the list at index 5 24 System.out.println("(6) " + list); 25 26 list.add(0, "Poland"); // Same as list.addFirst("Poland") 27 System.out.println("(7) " + list); 28 29 // Remove elements from the list 30 list.remove(0); // Same as list.remove("Poland") in this case 31 System.out.println("(8) " + list); 32 33 list.remove(2); // Remove the element at index 2 34 System.out.println("(9) " + list); 35 36 list.remove(list.size() − 1); // Remove the last element 37 System.out.print("(10) " + list + "\n(11) "); 38 39 for (String s: list) 40 System.out.print(s.toUpperCase() + " "); 41 42 list.clear(); 43 System.out.println("\nAfter clearing the list, the list size is " 44 + list.size()); 45 } 46 }

append element

insert element

create list

print list append element

traverse using iterator

remove element

insert element

(1) [America] (2) [Canada, America] (3) [Canada, America, Russia] (4) [Canada, America, Russia, France] (5) [Canada, America, Germany, Russia, France] (6) [Canada, America, Germany, Russia, France, Norway] (7) [Poland, Canada, America, Germany, Russia, France, Norway]

M24_LIAN1878_11_GE_C24.indd 954 12/29/17 4:41 PM

24.4 Linked Lists 955

24.4.3 Implementing MyLinkedList Now let us turn our attention to implementing the MyLinkedList class. We will discuss how to implement the methods addFirst, addLast, add(index, e), removeFirst, removeLast, and remove(index) and leave the other methods in the MyLinkedList class as exercises.

24.4.3.1 Implementing addFirst(e)

The addFirst(e) method creates a new node for holding element e. The new node becomes the first node in the list. It can be implemented as follows:

1 public void addFirst(E e) { 2 Node<E> newNode = new Node<>(e); // Create a new node 3 newNode.next = head; // link the new node with the head 4 head = newNode; // head points to the new node 5 size++; // Increase list size 6 7 if (tail == null) // The new node is the only node in list 8 tail = head; 9 }

The addFirst(e) method creates a new node to store the element (line 2) and inserts the node at the beginning of the list (line 3), as shown in Figure 24.12a. After the insertion, head should point to this new element node (line 4), as shown in Figure 24.12b.

was empty?

increase size head to new node link with head create a node

Figure 24.12 A new element is added to the beginning of the list.

(a) Before a new node is inserted.

This is the new node

(b) After a new node is inserted.

head

…

A new node to be inserted here

tail

…e0 next

e next

ei next

ei+1 next

ek null

… …e0 next

e next

ei next

ei+1 next

ek null

tail

(8) [Canada, America, Germany, Russia, France, Norway] (9) [Canada, America, Russia, France, Norway] (10) [Canada, America, Russia, France] (11) CANADA AMERICA RUSSIA FRANCE After clearing the list, the list size is 0

If the list is empty (line 7), both head and tail will point to this new node (line 8). After the node is created, size should be increased by 1 (line 5).

M24_LIAN1878_11_GE_C24.indd 955 12/29/17 4:41 PM

956 Chapter 24 Implementing Lists, Stacks, Queues, and Priority Queues

24.4.3.2 Implementing addLast(e)

The addLast(e) method creates a node to hold the element and appends the node at the end of the list. It can be implemented as follows:

1 public void addLast(E e) { 2 Node<E> newNode = new Node<>(e); // Create a new node for e 3 4 if (tail == null) { 5 head = tail = newNode; // The only node in list 6 } 7 else { 8 tail.next = newNode; // Link the new node with the last node 9 tail = newNode; // tail now points to the last node 10 } 11 12 size++; // Increase size 13 }

The addLast(e) method creates a new node to store the element (line 2) and appends it to the end of the list. Consider two cases:

1. If the list is empty (line 4), both head and tail will point to this new node (line 5);

2. Otherwise, link the node with the last node in the list (line 8). tail should now point to this new node (line 9). Figures 24.13a and 13b show the new node for element e before and after the insertion.

In any case, after the node is created, the size should be increased by 1 (line 12).

increase size

create a node

Figure 24.13 A new element is added at the end of the list.

(a) Before a new node is inserted.

A new node is appended in the list

(b) After a new node is inserted.

head

… A new node to be inserted here

tail

…e0 next

e null

ei next

ei+1 next

ek null

head

… tail…e0 next

e null

ei next

ei+1 next

ek next

24.4.3.3 Implementing add(index, e)

The add(index, e) method inserts an element into the list at the specified index. It can be implemented as follows:

1 public void add(int index, E e) { 2 if (index == 0) addFirst(e); // Insert first 3 else if (index >= size) addLast(e); // Insert last 4 else { // Insert in the middle 5 Node<E> current = head;

insert last insert first

M24_LIAN1878_11_GE_C24.indd 956 12/29/17 4:41 PM

24.4 Linked Lists 957

6 for (int i = 1; i < index; i++) 7 current = current.next; 8 Node<E> temp = current.next; 9 current.next = new Node<>(e); 10 (current.next).next = temp; 11 size++; 12 } 13 }

There are three cases when inserting an element into the list:

1. If index is 0, invoke addFirst(e) (line 2) to insert the element at the beginning of the list.

2. If index is greater than or equal to size, invoke addLast(e) (line 3) to insert the element at the end of the list.

3. Otherwise, create a new node to store the new element and locate where to insert it. As shown in Figure 24.14a, the new node is to be inserted between the nodes current and temp. The method assigns the new node to current.next and assigns temp to the new node’s next, as shown in Figure 24.14b. The size is now increased by 1 (line 11).

increase size

create a node

Figure 24.14 A new element is inserted in the middle of the list.

(a) Before a new node is inserted.

(b) After a new node is inserted.

head

…

A new node to be inserted here

tail

…e0 next

ei next

ei+1 next

ek null

current temp

e null

A new node is inserted in the list

current temphead

…

tail

…e0 next

e next

ei next

ei+1 next

ek null

24.4.3.4 Implementing removeFirst()

The removeFirst() method removes the first element from the list. It can be implemented as follows:

1 public E removeFirst() { 2 if (size == 0) return null; // Nothing to delete 3 else { 4 Node<E> temp = head; // Keep the first node temporarily 5 head = head.next; // Move head to point to next node 6 size−−; // Reduce size by 1 7 if (head == null) tail = null; // List becomes empty destroy the node

decrease size new head keep old head

nothing to remove

M24_LIAN1878_11_GE_C24.indd 957 12/29/17 4:41 PM

958 Chapter 24 Implementing Lists, Stacks, Queues, and Priority Queues

8 return temp.element; // Return the deleted element 9 } 10 }

Consider two cases:

1. If the list is empty, there is nothing to delete, so return null (line 2).

2. Otherwise, remove the first node from the list by pointing head to the second node. Figures 24.15a and 15b show the linked list before and after the deletion. The size is reduced by 1 after the deletion (line 6). If the list becomes empty, after removing the element, tail should be set to null (line 7).

Figure 24.15 The first node is deleted from the list.

(a) Before the node is deleted.

(b) After the node is deleted.

… e1 next

ei next

… ei+1 next

ek null

e0 next

head

…

Delete this node

tail

e1 next

ei next

… ei+1 next

ek null

e0 next

head

This node is deleted

tail

24.4.3.5 Implementing removeLast()

The removeLast() method removes the last element from the list. It can be implemented as follows:

1 public E removeLast() { 2 if (size == 0) return null; // Nothing to remove 3 else if (size == 1) { // Only one element in the list 4 Node<E> temp = head; 5 head = tail = null; // list becomes empty 6 size = 0; 7 return temp.element; 8 } 9 else { 10 Node<E> current = head; 11 12 for (int i = 0; i < size − 2; i++) 13 current = current.next; 14 15 Node<E> temp = tail; 16 tail = current; 17 tail.next = null; 18 size−−; 19 return temp.element; 20 } 21 }

return element reduce size

move tail

size 7 1

return element size is 0 head and tail null

size 1? empty?

M24_LIAN1878_11_GE_C24.indd 958 12/29/17 4:41 PM

24.4 Linked Lists 959

Consider three cases:

1. If the list is empty, return null (line 2).

2. If the list contains only one node, this node is destroyed; head and tail both become null (line 5). The size becomes 0 after the deletion (line 6) and the element value of the deleted node is returned (line 7).

3. Otherwise, the last node is destroyed (line 17) and the tail is repositioned to point to the second-to-last node. Figures 24.16a and 16b show the last node before and after it is deleted. The size is reduced by 1 after the deletion (line 18) and the element value of the deleted node is returned (line 19).

Figure 24.16 The last node is deleted from the list.

(a) Before the node is deleted.

(b) After the node is deleted.

head

…

Delete this node

tail

e1 next

ek–2 next

ek–1 next

ek null

e0 next

current

head

…

This node is deleted

tail

e1 next

ek–2 next

ek–1 null

ek null

e0 next

24.4.3.6 Implementing remove(index)

The remove(index) method finds the node at the specified index then removes it. It can be implemented as follows:

1 public E remove(int index) { 2 if (index < 0 || index >= size) return null; // Out of range 3 else if (index == 0) return removeFirst(); // Remove first 4 else if (index == size − 1) return removeLast(); // Remove last 5 else { 6 Node<E> previous = head; 7 8 for (int i = 1; i < index; i++) { 9 previous = previous.next; 10 } 11 12 Node<E> current = previous.next; 13 previous.next = current.next; 14 size−−; 15 return current.element; 16 } 17 }

Consider four cases:

1. If index is beyond the range of the list (i.e., index < 0 || index >= size), return null (line 2).

return element reduce size remove from list locate current

locate previous

remove last remove first out of range

M24_LIAN1878_11_GE_C24.indd 959 12/29/17 4:41 PM

960 Chapter 24 Implementing Lists, Stacks, Queues, and Priority Queues

2. If index is 0, invoke removeFirst() to remove the first node (line 3).

3. If index is size − 1, invoke removeLast() to remove the last node (line 4).

4. Otherwise, locate the node at the specified index. Let current denote this node and previous denote the node before this node, as shown in Figure 24.17a. Assign current.next to previous.next to eliminate the current node, as shown in Figure 24.17b.

Figure 24.17 A node is deleted from the list.

(a) Before the node is deleted.

(b) After the node is deleted.

head

…

Delete this node

e0 next

Ek–1 next

ek next

previous current tail

…ek–1 next

ek null

current.next

previous current.nexthead

…

tail

…e0 next

ek–1 next

ek null

Listing 24.5 gives the implementation of MyLinkedList. The implementation of get(index), indexOf(e), lastIndexOf(e), contains(e), and set(index, e) is omitted and left as an exercise. The iterator() method defined in the java.lang . Iterable interface is implemented to return an instance on java.util.Iterator (lines 128–130). The LinkedListIterator class implements Iterator with concrete methods for hasNext, next, and remove (lines 132–152). This implementation uses current to point to the current position of the element being traversed (line 134). Initially, current points to the head of the list.

Listing 24.5 MyLinkedList.java 1 public class MyLinkedList<E> implements MyList<E> { 2 private Node<E> head, tail; 3 private int size = 0; // Number of elements in the list 4 5 /** Create an empty list */ 6 public MyLinkedList() { 7 } 8 9 /** Create a list from an array of objects */ 10 public MyLinkedList(E[] objects) { 11 for (int i = 0; i < objects.length; i++) 12 add(objects[i]); 13 } 14 15 /** Return the head element in the list */ 16 public E getFirst() { 17 if (size == 0) { 18 return null; 19 } 20 else {

iterator

getFirst

constructor

no-arg constructor

number of elements head, tail

M24_LIAN1878_11_GE_C24.indd 960 12/29/17 4:41 PM

24.4 Linked Lists 961

21 return head.element; 22 } 23 } 24 25 /** Return the last element in the list */ 26 public E getLast() { 27 if (size == 0) { 28 return null; 29 } 30 else { 31 return tail.element; 32 } 33 } 34 35 /** Add an element to the beginning of the list */ 36 public void addFirst(E e) { 37 // Implemented in Section 24.4.3.1, so omitted here 38 } 39 40 /** Add an element to the end of the list */ 41 public void addLast(E e) { 42 // Implemented in Section 24.4.3.2, so omitted here 43 } 44 45 @Override /** Add a new element at the specified index 46 * in this list. The index of the head element is 0 */ 47 public void add(int index, E e) { 48 // Implemented in Section 24.4.3.3, so omitted here 49 } 50 51 /** Remove the head node and 52 * return the object that is contained in the removed node. */ 53 public E removeFirst() { 54 // Implemented in Section 24.4.3.4, so omitted here 55 } 56 57 /** Remove the last node and 58 * return the object that is contained in the removed node. */ 59 public E removeLast() { 60 // Implemented in Section 24.4.3.5, so omitted here 61 } 62 63 @Override /** Remove the element at the specified position in this 64 * list. Return the element that was removed from the list. */ 65 public E remove(int index) { 66 // Implemented earlier in Section 24.4.3.6, so omitted 67 } 68 69 @Override /** Override toString() to return elements in the list */ 70 public String toString() { 71 StringBuilder result = new StringBuilder("["); 72 73 Node<E> current = head; 74 for (int i = 0; i < size; i++) { 75 result.append(current.element); 76 current = current.next; 77 if (current != null) { 78 result.append(", "); // Separate two elements with a comma 79 } 80 else { 81 result.append("]"); // Insert the closing ] in the string

getLast

addFirst

addLast

add

removeFirst

removeLast

remove

toString

M24_LIAN1878_11_GE_C24.indd 961 12/29/17 4:41 PM

962 Chapter 24 Implementing Lists, Stacks, Queues, and Priority Queues

82 } 83 } 84 85 return result.toString(); 86 } 87 88 @Override /** Clear the list */ 89 public void clear() { 90 size = 0; 91 head = tail = null; 92 } 93 94 @Override /** Return true if this list contains the element e */ 95 public boolean contains(Object e) { 96 // Left as an exercise 97 return true; 98 } 99 100 @Override /** Return the element at the specified index */ 101 public E get(int index) { 102 // Left as an exercise 103 return null; 104 } 105 106 @Override /** Return the index of the head matching element in 107 * this list. Return −1 if no match. */ 108 public int indexOf(Object e) { 109 // Left as an exercise 110 return 0; 111 } 112 113 @Override /** Return the index of the last matching element in 114 * this list. Return −1 if no match. */ 115 public int lastIndexOf(E e) { 116 // Left as an exercise 117 return 0; 118 } 119 120 @Override /** Replace the element at the specified position 121 * in this list with the specified element. */ 122 public E set(int index, E e) { 123 // Left as an exercise 124 return null; 125 } 126 127 @Override /** Override iterator() defined in Iterable */ 128 public java.util.Iterator<E> iterator() { 129 return new LinkedListIterator(); 130 } 131 132 private class LinkedListIterator 133 implements java.util.Iterator<E> { 134 private Node<E> current = head; // Current index 135 136 @Override 137 public boolean hasNext() { 138 return (current != null); 139 } 140 141 @Override

clear

contains

get

indexOf

lastIndexOf

set

iterator

LinkedListIterator class

M24_LIAN1878_11_GE_C24.indd 962 12/29/17 4:41 PM

24.4 Linked Lists 963

142 public E next() { 143 E e = current.element; 144 current = current.next; 145 return e; 146 } 147 148 @Override 149 public void remove() { 150 // Left as an exercise 151 } 152 } 153 154 private static class Node<E> { 155 E element; 156 Node<E> next; 157 158 public Node(E element) { 159 this.element = element; 160 } 161 } 162 163 @Override /** Return the number of elements in this list */ 164 public int size() { 165 return size; 166 } 167 }

24.6.4 MyArrayList vs. MyLinkedList Both MyArrayList and MyLinkedList can be used to store a list. MyArrayList is imple- mented using an array, and MyLinkedList is implemented using a linked list. The overhead of MyArrayList is smaller than that of MyLinkedList. However, MyLinkedList is more efficient if you need to insert elements into and delete elements from the beginning of the list. Table 24.1 summarizes the complexity of the methods in MyArrayList and MyLinkedList. Note MyArrayList is the same as java.util.ArrayList, and MyLinkedList is the same as java.util.LinkedList.

Node inner class

Methods MyArrayList/ArrayList MyLinkedList/LinkedList

add(e: E) O(1) O(1)

add(index: int, e: E) O(n) O(n)

clear() O(1) O(1)

contains(e: E) O(n) O(n)

get(index: int) O(1) O(n)

indexOf(e: E) O(n) O(n)

isEmpty() O(1) O(1)

lastIndexOf(e: E) O(n) O(n)

remove(e: E) O(n) O(n)

size() O(1) O(1)

remove(index: int) O(n) O(n)

set(index: int, e: E) O(n) O(n)

addFirst(e: E) O(n) O(1)

removeFirst() O(n) O(1)

tabLe 24.1 Time Complexities for Methods in MyArrayList and MyLinkedList

M24_LIAN1878_11_GE_C24.indd 963 12/29/17 4:41 PM

964 Chapter 24 Implementing Lists, Stacks, Queues, and Priority Queues

24.4.5 Variations of Linked Lists The linked list introduced in the preceding sections is known as a singly linked list. It contains a pointer to the list’s first node, and each node contains a pointer to the next node sequentially. Several variations of the linked list are useful in certain applications.

A circular, singly linked list is like a singly linked list, except that the pointer of the last node points back to the first node, as shown in Figure 24.18a. Note tail is not needed for circular linked lists. head points to the current node in the list. Insertion and deletion take place at the current node. A good application of a circular linked list is in the operating system that serves multiple users in a timesharing fashion. The system picks a user from a circular list and grants a small amount of CPU time then moves on to the next user in the list.

Figure 24.18 Linked lists may appear in various forms.

(a) Circular linked list

... Node 1 Node 2 Node n

element1 next

element2 next

head

(b) Doubly linked list

...

Node 1

element1 next null

head tail Node 2

element2 next

Node n

element2 null

Node 1

element1 next

head Node 2

element2 next

Node n

element2 next

A doubly linked list contains nodes with two pointers. One points to the next node and the other to the previous node, as shown in Figure 24.18b. These two pointers are conveni- ently called a forward pointer and a backward pointer. Thus, a doubly linked list can be traversed forward and backward. The java.util.LinkedList class is implemented using a doubly linked list, and it supports traversing of the list forward and backward using the ListIterator.

A circular, doubly linked list is like a doubly linked list, except that the forward pointer of the last node points to the first node, and the backward pointer of the first pointer points to the last node, as shown in Figure 24.18c.

The implementations of these linked lists are left as exercises.

24.4.1 If a linked list does not contain any nodes, what are the values in head and tail? 24.4.2 If a linked list has only one node, is head == tail true? List all cases in which

head == tail is true.

24.4.3 When a new node is inserted to the head of a linked list, will the head and the tail be changed?

24.4.4 When a new node is appended to the end of a linked list, will the head and the tail be changed?

Point Check

M24_LIAN1878_11_GE_C24.indd 964 12/29/17 4:41 PM

24.5 Stacks and Queues 965

Pedagogical Note For an interactive demo on how stacks and queues work, go to liveexample.pearsoncmg .com/dsanimation/StackeBook.html, and liveexample.pearsoncmg.com/dsanimation/ QueueeBook.html, as shown in Figure 24.20.

stack and queue animation on Companion Website

Figure 24.19 A queue holds objects in a first-in, first-out fashion.

Data1 Data2

Data1 Data1 Data2 Data3

Data2 Data3

Data1 Data2

Data3

Data1

Data3

24.4.5 Both MyArrayList and MyLinkedList are used to store a list of objects. Why do we need both types of lists?

24.4.6 Draw a diagram to show the linked list after each of the following statements is executed:

MyLinkedList<Double> list = new MyLinkedList<>(); list.add(1.5); list.add(6.2); list.add(3.4); list.add(7.4); list.remove(1.5); list.remove(2);

24.4.7 What is the time complexity of the addFirst(e) and removeFirst() methods in MyLinkedList?

24.4.8 Suppose you need to store a list of elements. If the number of elements in the program is fixed, what data structure should you use? If the number of elements in the program changes, what data structure should you use?

24.4.9 If you have to add or delete the elements at the beginning of a list, should you use MyArrayList or MyLinkedList? If most of the operations on a list involve retrieving an element at a given index, should you use MyArrayList or MyLinkedList?

24.4.10 Simplify the code in lines 77–82 in Listing 24.5 using a conditional expression. 24.4.11 Simplify the code for the removeLast() method by invoking the removeFirst()

method when the size is less than or equal to 1. Is the new code more efficient in execution time?

24.5 Stacks and Queues Stacks can be implemented using array lists and queues can be implemented using linked lists.

A stack can be viewed as a special type of list whose elements are accessed, inserted, and deleted only from the end (top), as shown in Figure 10.11. A queue represents a waiting list. It can be viewed as a special type of list whose elements are inserted into the end (tail) of the queue and are accessed and deleted from the beginning (head), as shown in Figure 24.19.

Point Key

M24_LIAN1878_11_GE_C24.indd 965 12/29/17 4:41 PM

966 Chapter 24 Implementing Lists, Stacks, Queues, and Priority Queues

(a) Stack animation

(b) Queue animation

Figure 24.21 GenericStack and GenericQueue may be implemented using inheritance or composition.

(a) Using inheritance

(b) Using composition

ArrayList GenericStack LinkedList GenericQueue

GenericStack ArrayList GenericQueue LinkedList

Since the insertion and deletion operations on a stack are made only at the end of the stack, it is more efficient to implement a stack with an array list than a linked list. Since deletions are made at the beginning of the list, it is more efficient to implement a queue using a linked list than an array list. This section implements a stack class using an array list and a queue class using a linked list.

There are two ways to design the stack and queue classes:

■■ Using inheritance: You can define a stack class by extending ArrayList, and a queue class by extending LinkedList, as shown in Figure 24.21a.

■■ Using composition: You can define an array list as a data field in the stack class and a linked list as a data field in the queue class, as shown in Figure 24.21b.

inheritance

composition

M24_LIAN1878_11_GE_C24.indd 966 12/29/17 4:41 PM

24.5 Stacks and Queues 967

Both designs are fine, but using composition is better because it enables you to define a completely new stack class and queue class without inheriting the unnecessary and inappropri- ate methods from the array list and linked list. The implementation of the stack class using the composition approach was given in Listing 19.1, GenericStack.java. Listing 24.6 implements the GenericQueue class using the composition approach. Figure 24.22 shows the UML of the class.

Figure 24.22 GenericQueue uses a linked list to provide a first-in, first-out data structure.

GenericQueue<E>

–list: java.util.LinkedList<E>

+enqueue(e: E): void +dequeue(): E +getSize(): int

Adds an element to this queue.

Returns the number of elements in this queue. Removes an element from this queue.

Listing 24.6 GenericQueue.java 1 public class GenericQueue<E> { 2 private java.util.LinkedList<E> list 3 = new java.util.LinkedList<>(); 4 5 public void enqueue(E e) { 6 list.addLast(e); 7 } 8 9 public E dequeue() { 10 return list.removeFirst(); 11 } 12 13 public int getSize() { 14 return list.size(); 15 } 16 17 @Override 18 public String toString() { 19 return "Queue: " + list.toString(); 20 } 21 }

A linked list is created to store the elements in a queue (lines 2 and 3). The enqueue(e) method (lines 5–7) adds element e into the tail of the queue. The dequeue() method (lines 9–11) removes an element from the head of the queue and returns the removed element. The getSize() method (lines 13–15) returns the number of elements in the queue.

Listing 24.7 gives an example that creates a stack using GenericStack and a queue using GenericQueue. It uses the push (enqueue) method to add strings to the stack (queue) and the pop (dequeue) method to remove strings from the stack (queue).

Listing 24.7 TestStackQueue.java 1 public class TestStackQueue { 2 public static void main(String[] args) { 3 // Create a stack 4 GenericStack<String> stack = new GenericStack<>(); 5

toString

getSize

dequeue

enqueue

linked list

M24_LIAN1878_11_GE_C24.indd 967 12/29/17 4:41 PM

968 Chapter 24 Implementing Lists, Stacks, Queues, and Priority Queues

(1) stack: [Tom] (2) stack: [Tom, Susan] (3) stack: [Tom, Susan, Kim, Michael] (4) Michael (5) Kim (6) stack: [Tom, Susan] (7) Queue: [Tom] (8) Queue: [Tom, Susan] (9) Queue: [Tom, Susan, Kim, Michael] (10) Tom (11) Susan (12) Queue: [Kim, Michael]

6 // Add elements to the stack 7 stack.push("Tom"); // Push it to the stack 8 System.out.println("(1) " + stack); 9 10 stack.push("Susan"); // Push it to the the stack 11 System.out.println("(2) " + stack); 12 13 stack.push("Kim"); // Push it to the stack 14 stack.push("Michael"); // Push it to the stack 15 System.out.println("(3) " + stack); 16 17 // Remove elements from the stack 18 System.out.println("(4) " + stack.pop()); 19 System.out.println("(5) " + stack.pop()); 20 System.out.println("(6) " + stack); 21 22 // Create a queue 23 GenericQueue<String> queue = new GenericQueue<>(); 24 25 // Add elements to the queue 26 queue.enqueue("Tom"); // Add it to the queue 27 System.out.println("(7) " + queue); 28 29 queue.enqueue("Susan"); // Add it to the queue 30 System.out.println("(8) " + queue); 31 32 queue.enqueue("Kim"); // Add it to the queue 33 queue.enqueue("Michael"); // Add it to the queue 34 System.out.println("(9) " + queue); 35 36 // Remove elements from the queue 37 System.out.println("(10) " + queue.dequeue()); 38 System.out.println("(11) " + queue.dequeue()); 39 System.out.println("(12) " + queue); 40 } 41 }

For a stack, the push(e) method adds an element to the top of the stack, and the pop() method removes the top element from the stack and returns the removed element. It is easy to see that the time complexity for the push and pop methods is O(1).

For a queue, the enqueue(e) method adds an element to the tail of the queue, and the dequeue() method removes the element from the head of the queue. It is easy to see that the time complexity for the enqueue and dequeue methods is O(1).

stack time complexity

queue time complexity

M24_LIAN1878_11_GE_C24.indd 968 12/29/17 4:41 PM

24.6 Priority Queues 969

24.5.1 You can use inheritance or composition to design the data structures for stacks and queues. Discuss the pros and cons of these two approaches.

24.5.2 If LinkedList is replaced by ArrayList in lines 2 and 3 in Listing 24.6, Gener- icQueue.java, what will be the time complexity for the enqueue and dequeue methods.

24.5.3 Which lines of the following code are wrong?

1 List<String> list = new ArrayList<>(); 2 list.add("Tom"); 3 list = new LinkedList<>(); 4 list.add("Tom"); 5 list = new GenericStack<>(); 6 list.add("Tom");

24.6 Priority Queues Priority queues can be implemented using heaps.

An ordinary queue is a first-in, first-out data structure. Elements are appended to the end of the queue and removed from the beginning. In a priority queue, elements are assigned with priorities. When accessing elements, the element with the highest priority is removed first. For example, the emergency room in a hospital assigns priority numbers to patients; the patient with the highest priority is treated first.

A priority queue can be implemented using a heap, in which the root is the object with the highest priority in the queue. Heaps were introduced in Section 23.6, Heap Sort. The class dia- gram for the priority queue is shown in Figure 24.23. Its implementation is given in Listing 24.8.

Point Check

Point Key

Figure 24.23 MyPriorityQueue uses a heap to provide a largest-in, first-out data structure.

MyPriorityQueue <E extends Comparable<E>>

–heap: Heap<E>

+enqueue(element: E): void +dequeue(): E +getSize(): int

Adds an element to this queue. Removes an element from this queue. Returns the number of elements in this queue.

Listing 24.8 MyPriorityQueue.java 1 public class MyPriorityQueue<E extends Comparable<E>> { 2 private Heap<E> heap = new Heap<>(); 3 4 public void enqueue(E newObject) { 5 heap.add(newObject); 6 } 7 8 public E dequeue() { 9 return heap.remove(); 10 } 11 12 public int getSize() { 13 return heap.getSize(); 14 } 15 }

getsize

dequeue

enqueue

heap for priority queue

M24_LIAN1878_11_GE_C24.indd 969 12/29/17 4:41 PM

970 Chapter 24 Implementing Lists, Stacks, Queues, and Priority Queues

Cindy(priority:7) Tim(priority:5) John(priority:2) Jim(priority:1)

Listing 24.9 gives an example of using a priority queue for patients. The Patient class is defined in lines 19–37. Four patients are created with associated priority values in lines 3–6. Line 8 creates a priority queue. The patients are enqueued in lines 10–13. Line 16 dequeues a patient from the queue.

Listing 24.9 TestPriorityQueue.java 1 public class TestPriorityQueue { 2 public static void main(String[] args) { 3 Patient patient1 = new Patient("John", 2); 4 Patient patient2 = new Patient("Jim", 1); 5 Patient patient3 = new Patient("Tim", 5); 6 Patient patient4 = new Patient("Cindy", 7); 7 8 MyPriorityQueue<Patient> priorityQueue 9 = new MyPriorityQueue<>(); 10 priorityQueue.enqueue(patient1); 11 priorityQueue.enqueue(patient2); 12 priorityQueue.enqueue(patient3); 13 priorityQueue.enqueue(patient4); 14 15 while (priorityQueue.getSize() > 0) 16 System.out.print(priorityQueue.dequeue() + " "); 17 } 18 19 static class Patient implements Comparable<Patient> { 20 private String name; 21 private int priority; 22 23 public Patient(String name, int priority) { 24 this.name = name; 25 this.priority = priority; 26 } 27 28 @Override 29 public String toString() { 30 return name + "(priority:" + priority + ")"; 31 } 32 33 @Override 34 public int compareTo(Patient patient) { 35 return this.priority − patient.priority; 36 } 37 } 38 }

remove from queue

add to queue

create a priority queue

inner class Patient

create a patient

compareTo

24.6.1 What is a priority queue? 24.6.2 What are the time complexity of the enqueue, dequeue, and getSize methods in

MyPriorityQueue?

24.6.3 Which of the following statements are wrong?

1 MyPriorityQueue<Object> q1 = new MyPriorityQueue<>(); 2 MyPriorityQueue<Number> q2 = new MyPriorityQueue<>(); 3 MyPriorityQueue<Integer> q3 = new MyPriorityQueue<>(); 4 MyPriorityQueue<Date> q4 = new MyPriorityQueue<>(); 5 MyPriorityQueue<String> q5 = new MyPriorityQueue<>();

Point Check

M24_LIAN1878_11_GE_C24.indd 970 12/29/17 4:41 PM

Programming Exercises 971

ChapTer summary 1. You learned how to implement array lists, linked lists, stacks, and queues.

2. To define a data structure is essentially to define a class. The class for a data structure should use data fields to store data and provide methods to support operations such as insertion and deletion.

3. To create a data structure is to create an instance from the class. You can then apply the methods on the instance to manipulate the data structure, such as inserting an element into the data structure or deleting an element from the data structure.

4. You learned how to implement a priority queue using a heap.

Quiz Answer the quiz for this chapter online at the book Companion Website.

programming exerCises 24.1 (Implement set operations in MyList) The implementations of the meth-

ods addAll, removeAll, retainAll, toArray(), and toArray(T[]) are omitted in the MyList interface. Implement these methods. Test your new MyList class using the code at liveexample.pearsoncmg.com/test/ Exercise24_01Test.txt.

*24.2 (Implement MyLinkedList) The implementations of the methods contains(E e), get(int index), indexOf(E e), lastIndexOf(E e), and set(int index, E e) are omitted in the MyLinkedList class. Implement these methods.

*24.3 (Implement a doubly linked list) The MyLinkedList class used in Listing 24.5 is a one-way directional linked list that enables one-way traversal of the list. Modify the Node class to add the new data field name previous to refer to the previous node in the list, as follows:

public class Node<E> { E element; Node<E> next; Node<E> previous;

public Node(E e) { element = e; } }

Implement a new class named TwoWayLinkedList that uses a doubly linked list to store elements. Define TwoWayLinkedList to implements MyList. You need to implement all the methods defined in MyLinkedList as well as the methods listIterator() and listIterator(int index). Both return an instance of java.util.ListIterator<E> (see Figure 20.4). The former sets the cursor to the head of the list and the latter to the element at the specified index.

24.4 (Use the GenericStack class) Write a program that displays the first 100 prime numbers in descending order. Use a stack to store the prime numbers.

M24_LIAN1878_11_GE_C24.indd 971 12/29/17 4:41 PM

972 Chapter 24 Implementing Lists, Stacks, Queues, and Priority Queues

24.5 (Implement GenericQueue using inheritance) In Section 24.5, Stacks and Queues, GenericQueue is implemented using composition. Define a new queue class that extends java.util.LinkedList.

*24.6 (Generic PriorityQueue using Comparator) Revise MyPriorityQueue in Listing 24.8, using a generic parameter for comparing objects. Define a new constructor with a Comparator as its argument as follows:

PriorityQueue(Comparator<? super E> comparator)

**24.7 (Animation: linked list) Write a program to animate search, insertion, and dele- tion in a linked list, as shown in Figure 24.1b. The Search button searches the specified value in the list. The Delete button deletes the specified value from the list. The Insert button appends the value into the list if the index is not specified; otherwise, it inserts the value into the specified index in the list.

*24.8 (Animation: array list) Write a program to animate search, insertion, and deletion in an array list, as shown in Figure 24.1a. The Search button searches the speci- fied value in the list. The Delete button deletes the specified value from the list. The Insert button appends the value into the list if the index is not specified; otherwise, it inserts the value into the specified index in the list.

*24.9 (Animation: array list in slow motion) Improve the animation in the preceding programming exercise by showing the insertion and deletion operations in a slow motion.

*24.10 (Animation: stack) Write a program to animate push and pop in a stack, as shown in Figure 24.20a.

*24.11 (Animation: doubly linked list) Write a program to animate search, insertion, and deletion in a doubly linked list, as shown in Figure 24.24. The Search button searches the specified value in the list. The Delete button deletes the specified value from the list. The Insert button appends the value into the list if the index is not specified; otherwise, it inserts the value into the specified index in the list. Also add two buttons named Forward Traversal and Backward Traversal for displaying the elements in a forward and backward order, respectively, using iterators, as shown in Figure 24.24. The elements are displayed in a label.

*24.12 (Animation: queue) Write a program to animate the enqueue and dequeue operations on a queue, as shown in Figure 24.20b.

*24.13 (Fibonacci number iterator) Define an iterator class named Fibonacci Iterator for iterating Fibonacci numbers. The constructor takes an argument that specifies the limit of the maximum Fibonacci number. For example, new

M24_LIAN1878_11_GE_C24.indd 972 12/29/17 4:41 PM

FibonacciIterator(23302) creates an iterator that iterates Fibonacci num- bers less than or equal to 23302. Write a test program that uses this iterator to display all Fibonacci numbers less than or equal to 120000.

*24.14 (Prime number iterator) Define an iterator class named PrimeIterator for iterating prime numbers. The constructor takes an argument that specifies the limit of the maximum prime number. For example, new PrimeIterator(23302) creates an iterator that iterates prime numbers less than or equal to 23302. Write a test program that uses this iterator to display all prime numbers less than or equal to 120000.

**24.15 (Test MyArrayList) Design and write a complete test program to test if the MyArrayList class in Listing 24.2 meets all requirements.

**24.16 (Test MyLinkedList) Design and write a complete test program to test if the MyLinkedList class in Listing 24.5 meets all requirements.

**24.17 (Revise MyPriorityQueue) Listing 24.8 uses a heap to implement the priority queue. Revise the implementation using a sorted array list to store the elements and name the new class PriorityUsingSortedArrayList. The elements in the array list are sorted in increasing order of their priority with the last element having the highest priority. Write a test program that generates 5 million integers and enqueues them to the priority and dequeues from the queue. Use the same numbers for MyPriorityQueue and PriorityUsingSortedArrayList and display their execution times.

Programming Exercises 973

M24_LIAN1878_11_GE_C24.indd 973 12/29/17 4:41 PM

A01_GIPE3898_09_SE_FM.indd 3 12/20/17 5:56 PM

This page intentionally left blank

Objectives ■■ To design and implement a binary search tree (§25.2).

■■ To represent binary trees using linked data structures (§25.2.1).

■■ To search an element in a binary search tree (§25.2.2).

■■ To insert an element into a binary search tree (§25.2.3).

■■ To traverse elements in a binary tree (§25.2.4).

■■ To design and implement the Tree interface and the BST class (§25.2.5).

■■ To delete elements from a binary search tree (§25.3).

■■ To display a binary tree graphically (§25.4).

■■ To create iterators for traversing a binary tree (§25.5).

■■ To implement Huffman coding for compressing data using a binary tree (§25.6).

Binary Search Trees

Chapter

M25_LIAN1878_11_GE_C25.indd 975 12/29/17 4:43 PM

976 Chapter 25 Binary Search Trees

25.1 Introduction A binary search tree is more efficient than a list for search, insertion, and deletion operations.

The preceding chapter gives the implementation for array lists and linked lists. The time complexity of search, insertion, and deletion operations in these data structures is O(n). This chapter presents a new data structure called binary search tree, which takes O(logn) average time for search, insertion, and deletion of elements.

25.2 Binary Search Trees A binary search tree can be implemented using a linked structure.

Recall that lists, stacks, and queues are linear structures that consist of a sequence of elements. A binary tree is a hierarchical structure. It either is empty or consists of an element, called the root, and two distinct binary trees, called the left subtree and right subtree, either or both of which may be empty, as shown in Figure 25.1a. Examples of binary trees are shown in Figures 25.1a and b.

Point Key

binary tree

root left subtree right subtree

Figure 25.1 Each node in a binary tree has zero, one, or two subtrees.

67 107

5745

M T

(b) (c)

100

Root

(a)

Left subtree

Right subtree

The length of a path is the number of the edges in the path. The depth of a node is the length of the path from the root to the node. The set of all nodes at a given depth is sometimes called a level of the tree. Siblings are nodes that share the same parent node. The root of a left (right) subtree of a node is called a left (right) child of the node. A node without children is called a leaf. The height of a nonempty tree is the length of the path from the root node to its furthest leaf. The height of a tree that contains a single node is 0. Conventionally, the height of an empty tree is -1. Consider the tree in Figure 25.1b. The length of the path from node 60 to 45 is 2. The depth of node 60 is 0, the depth of node 55 is 1, and the depth of node 45 is 2. The height of the tree is 2. Nodes 45 and 57 are siblings. Nodes 45, 57, 67, and 107 are at the same level.

A special type of binary tree called a binary search tree (BST) is often useful. A BST (with no duplicate elements) has the property that for every node in the tree, the value of any node in its left subtree is less than the value of the node, and the value of any node in its right subtree is greater than the value of the node. The binary trees in Figures 25.1b-c are all BSTs.

Pedagogical Note For an interactive GUI demo to see how a BST works, go to liveexample.pearsoncmg. com/dsanimation/BSTeBook.html, as shown in Figure 25.2.

length

depth

level sibling

leaf height

binary search tree

BST animation on Companion Website

Point Key

M25_LIAN1878_11_GE_C25.indd 976 12/29/17 4:43 PM

25.2 Binary Search Trees 977

25.2.1 Representing Binary Search Trees A binary tree can be represented using a set of linked nodes. Each node contains a value and two links named left and right that reference the left child and right child, respectively, as shown in Figure 25.3.

Figure 25.3 A binary tree can be represented using a set of linked nodes.

45 57

root

100

67 107

A node can be defined as a class, as follows:

class TreeNode<E> { protected E element; protected TreeNode<E> left; protected TreeNode<E> right;

public TreeNode(E e) { element = e; } }

M25_LIAN1878_11_GE_C25.indd 977 12/29/17 4:43 PM

978 Chapter 25 Binary Search Trees

We use the variable root to refer to the root node of the tree. If the tree is empty, root is null. The following code creates the first three nodes of the tree in Figure 25.3:

// Create the root node TreeNode<Integer> root = new TreeNode<>(60);

// Create the left child node root.left = new TreeNode<>(55);

// Create the right child node root.right = new TreeNode<>(100);

25.2.2 Searching for an Element To search for an element in the BST, you start from the root and scan down from it until a match is found or you arrive at an empty subtree. The algorithm is described in Listing 25.1. Let current point to the root (line 2). Repeat the following steps until current is null (line 4) or the element matches current.element (line 12):

■■ If element is less than current.element, assign current.left to current (line 6).

■■ If element is greater than current.element, assign current.right to current (line 9).

■■ If element is equal to current.element, return true (line 12).

If current is null, the subtree is empty and the element is not in the tree (line 14).

Listing 25.1 Searching for an Element in a BST 1 public boolean search(E element) { 2 TreeNode<E> current = root; // Start from the root 3 4 while (current != null) 5 if (element < current.element) { 6 current = current.left; // Go left 7 } 8 else if (element > current.element) { 9 current = current.right; // Go right 10 } 11 else // Element matches current.element 12 return true; // Element is found 13 14 return false; // Element is not in the tree 15 }

25.2.3 Inserting an Element into a BST To insert an element into a BST, you need to locate where to insert it in the tree. The key idea is to locate the parent for the new node. Listing 25.2 gives the algorithm.

Listing 25.2 Inserting an Element into a BST 1 boolean insert(E e) { 2 if (tree is empty) 3 // Create the node for e as the root; 4 else { 5 // Locate the parent node 6 parent = current = root; 7 while (current != null) 8 if (e < the value in current.element) {

start from root

left subtree

right subtree

found

not found

create a new node

locate parent

M25_LIAN1878_11_GE_C25.indd 978 12/29/17 4:43 PM

25.2 Binary Search Trees 979

9 parent = current; // Keep the parent 10 current = current.left; // Go left 11 } 12 else if (e > the value in current.element) { 13 parent = current; // Keep the parent 14 current = current.right; // Go right 15 } 16 else 17 return false; // Duplicate node not inserted 18 19 // Create a new node for e and attach it to parent 20 21 return true; // Element inserted 22 } 23 }

If the tree is empty, create a root node with the new element (lines 2 and 3). Otherwise, locate the parent node for the new element node (lines 6–17). Create a new node for the element and link this node to its parent node. If the new element is less than the parent element, the node for the new element will be the left child of the parent. If the new element is greater than the parent element, the node for the new element will be the right child of the parent.

For example, to insert 101 into the tree in Figure 25.3, after the while loop finishes in the algorithm, parent points to the node for 107, as shown in Figure 25.4a. The new node for 101 becomes the left child of the parent. To insert 59 into the tree, after the while loop finishes in the algorithm, the parent points to the node for 57, as shown in Figure 25.4b. The new node for 59 becomes the right child of the parent.

left child

right child

Figure 25.4 Two new elements are inserted into the tree.

(a) Inserting 101

parent

45 57

root

100

67 107

101

(b) Inserting 59

parent

45 57

root

100

67 107

10159

25.2.4 Tree Traversal Tree traversal is the process of visiting each node in the tree exactly once. There are several ways to traverse a tree. This section presents inorder, postorder, preorder, depth-first, and breadth-first traversals.

With inorder traversal, the left subtree of the current node is visited first recursively, then the current node, and finally the right subtree of the current node recursively. The inorder traversal displays all the nodes in a BST in increasing order.

With postorder traversal, the left subtree of the current node is visited recursively first, then recursively the right subtree of the current node, and finally the current node itself.

With preorder traversal, the current node is visited first, then recursively the left subtree of the current node, and finally the right subtree of the current node recursively.

tree traversal

inorder traversal

postorder traversal

preorder traversal

M25_LIAN1878_11_GE_C25.indd 979 12/29/17 4:43 PM

980 Chapter 25 Binary Search Trees

Note You can reconstruct a binary search tree by inserting the elements in their preorder. The reconstructed tree preserves the parent and child relationship for the nodes in the original binary search tree.

Depth-first traversal is to visit the root then recursively visit its left subtree and right subtree in an arbitrary order. The preorder traversal can be viewed as a special case of depth-first traversal, which recursively visits its left subtree then its right subtree.

With breadth-first traversal, the nodes are visited level by level. First the root is visited, then all the children of the root from left to right, then the grandchildren of the root from left to right, and so on.

For example, in the tree in Figure 25.4b, the inorder is

45 55 57 59 60 67 100 101 107

The postorder is

45 59 57 55 67 101 107 100 60

The preorder is

60 55 45 57 59 100 67 107 101

The breadth-first traversal is

60 55 100 45 57 67 107 59 101

You can use the following simple tree to help remember inorder, postorder, and preorder.

reconstruct a tree

depth-first traversal

breadth-first traversal

1 2

The inorder is 1 + 2, the postorder is 1 2 +, and the preorder is + 1 2.

25.2.5 The BST Class Following the design pattern for Java Collections Framework and utilizing the default methods in Java 8, we use an interface named Tree to define all common operations for trees and define Tree to be a subtype of Collection so we can use common operations in Collection for trees, as shown in Figure 25.5. A concrete BST class can be defined to implement Tree, as shown in Figure 25.6.

Listing 25.3 gives the implementation for Tree. It provides default implementations for the add, isEmpty, remove, containsAll, addAll, removeAll, retainAll, toArray(), and toArray(T[]) methods inherited from the Collection interface as well as the inorder(), preorder(), and postorder() defined in the Tree interface.

Listing 25.3 Tree.java 1 import java.util.Collection; 2 3 public interface Tree<E> extends Collection<E> { 4 /** Return true if the element is in the tree */ 5 public boolean search(E e); 6 7 /** Insert element e into the binary tree 8 * Return true if the element is inserted successfully */ 9 public boolean insert(E e); 10

interface

insert

M25_LIAN1878_11_GE_C25.indd 980 12/29/17 4:43 PM

25.2 Binary Search Trees 981

11 /** Delete the specified element from the tree 12 * Return true if the element is deleted successfully */ 13 public boolean delete(E e); 14 15 /** Get the number of elements in the tree */ 16 public int getSize(); 17 18 /** Inorder traversal from the root*/ 19 public default void inorder() { 20 } 21 22 /** Postorder traversal from the root */ 23 public default void postorder() { 24 }

Figure 25.5 The Tree interface defines common operations for trees, and partially implements Collection.

Returns true if the speci�ed element is in the tree.

Returns true if the element is added successfully.

Returns true if the element is removed from the tree successfully.

Prints the nodes in inorder traversal.

Prints the nodes in preorder traversal.

Prints the nodes in postorder traversal.

Returns the number of elements in the tree.

Returns true if the tree is empty.

Removes all elements from the tree.

+search(e: E): boolean

«interface» Tree<E>

+clear(): void

+isEmpty(): boolean

+getSize(): int

+postorder(): void

+preorder(): void

+inorder(): void

+delete(e: E): boolean

+insert(e: E): boolean

«interface» java.util.Collection<E>

Override the add, isEmpty, remove, containsAll, addAll, removeAll, retainAll, toArray(), and toArray(T[]) methods de�ned in Collection using default methods.

Figure 25.6 The BST class defines a concrete BST.

BST<E extends Comparable<E>>TreeNode<E>

#root: TreeNode<E>

#size: int

+BST()

+BST(objects: E[])

+path(e: E): java.util.List<TreeNode<E>>

#element: E

#left: TreeNode<E>

#right: TreeNode<E>

Link

The root of the tree.

The number of nodes in the tree.

Creates a default BST.

Creates a BST from an array of elements.

Returns the path of nodes from the root leading to the node for the speci�ed element. The element may not be in the tree.

«interface» Tree<E>

delete

getSize

inorder

postorder

M25_LIAN1878_11_GE_C25.indd 981 12/29/17 4:43 PM

982 Chapter 25 Binary Search Trees

25 26 /** Preorder traversal from the root */ 27 public default void preorder() { 28 } 29 30 @Override /** Return true if the tree is empty */ 31 public default boolean isEmpty() { 32 return size() == 0; 33 } 34 35 @Override 36 public default boolean contains(Object e) { 37 return search((E)e); 38 } 39 40 @Override 41 public default boolean add(E e) { 42 return insert(e); 43 } 44 45 @Override 46 public default boolean remove(Object e) { 47 return delete((E)e); 48 } 49 50 @Override 51 public default int size() { 52 return getSize(); 53 } 54 55 @Override 56 public default boolean containsAll(Collection<?> c) { 57 // Left as an exercise 58 return false; 59 } 60 61 @Override 62 public default boolean addAll(Collection<? extends E> c) { 63 // Left as an exercise 64 return false; 65 } 66 67 @Override 68 public default boolean removeAll(Collection<?> c) { 69 // Left as an exercise 70 return false; 71 } 72 73 @Override 74 public default boolean retainAll(Collection<?> c) { 75 // Left as an exercise 76 return false; 77 } 78 79 @Override 80 public default Object[] toArray() { 81 // Left as an exercise 82 return null; 83 } 84

preorder

default isEmpty

default contains

default add

default size

default containsAll

default addAll

default removeAll

default retainAll

default toArray()

default remove

M25_LIAN1878_11_GE_C25.indd 982 12/29/17 4:43 PM

25.2 Binary Search Trees 983

85 @Override 86 public default <T> T[] toArray(T[] array) { 87 // Left as an exercise 88 return null; 89 } 90 }

Listing 25.4 gives the implementations for the BST class.

Listing 25.4 BST.java 1 public class BST<E extends Comparable<E>> implements Tree<E> { 2 protected TreeNode<E> root; 3 protected int size = 0 4 5 /** Create an empty binary tree */ 6 public BST() { 7 } 8 9 /** Create a binary tree from an array of objects */ 10 public BST(E[] objects) { 11 for (int i = 0; i < objects.length; i++) 12 add(objects[i]); 13 } 14 15 @Override /** Returns true if the element is in the tree */ 16 public boolean search(E e) { 17 TreeNode<E> current = root; // Start from the root 18 19 while (current != null) { 20 if (e.compareTo(current.element) < 0) { 21 current = current.left; 22 } 23 else if (e.compareTo(current.element) > 0) { 24 current = current.right; 25 } 26 else // element matches current.element 27 return true; // Element is found 28 } 29 30 return false; 31 } 32 33 @Override /** Insert element e into the binary tree 34 * Return true if the element is inserted successfully */ 35 public boolean insert(E e) { 36 if (root == null) 37 root = createNewNode(e); // Create a new root 38 else { 39 // Locate the parent node 40 TreeNode<E> parent = null; 41 TreeNode<E> current = root; 42 while (current != null) 43 if (e.compareTo(current.element) < 0) { 44 parent = current; 45 current = current.left; 46 } 47 else if (e.compareTo(current.element) > 0) { 48 parent = current; 49 current = current.right;

default toArray(T[])

BST class root size

no-arg constructor

constructor

compare objects

insert

new root

compare objects

M25_LIAN1878_11_GE_C25.indd 983 12/29/17 4:43 PM

984 Chapter 25 Binary Search Trees

50 } 51 else 52 return false; // Duplicate node not inserted 53 54 // Create the new node and attach it to the parent node 55 if (e.compareTo(parent.element) < 0) 56 parent.left = createNewNode(e); 57 else 58 parent.right = createNewNode(e); 59 } 60 61 size++; 62 return true; // Element inserted successfully 63 } 64 65 protected TreeNode<E> createNewNode(E e) { 66 return new TreeNode<>(e); 67 } 68 69 @Override /** Inorder traversal from the root */ 70 public void inorder() { 71 inorder(root); 72 } 73 74 /** Inorder traversal from a subtree */ 75 protected void inorder(TreeNode<E> root) { 76 if (root == null) return; 77 inorder(root.left); 78 System.out.print(root.element + " "); 79 inorder(root.right); 80 } 81 82 @Override /** Postorder traversal from the root */ 83 public void postorder() { 84 postorder(root); 85 } 86 87 /** Postorder traversal from a subtree */ 88 protected void postorder(TreeNode<E> root) { 89 if (root == null) return; 90 postorder(root.left); 91 postorder(root.right); 92 System.out.print(root.element + " "); 93 } 94 95 @Override /** Preorder traversal from the root */ 96 public void preorder() { 97 preorder(root); 98 } 99 100 /** Preorder traversal from a subtree */ 101 protected void preorder(TreeNode<E> root) { 102 if (root == null) return; 103 System.out.print(root.element + " "); 104 preorder(root.left); 105 preorder(root.right); 106 } 107 108 /** This inner class is static, because it does not access 109 any instance members defined in its outer class */

link to parent

increase size

create new node

inorder

recursive helper method

postorder

recursive helper method

preorder

recursive helper method

M25_LIAN1878_11_GE_C25.indd 984 12/29/17 4:43 PM

25.2 Binary Search Trees 985

110 public static class TreeNode<E> { 111 protected E element; 112 protected TreeNode<E> left; 113 protected TreeNode<E> right; 114 115 public TreeNode(E e) { 116 element = e; 117 } 118 } 119 120 @Override /** Get the number of nodes in the tree */ 121 public int getSize() { 122 return size; 123 } 124 125 /** Returns the root of the tree */ 126 public TreeNode<E> getRoot() { 127 return root; 128 } 129 130 /** Returns a path from the root leading to the specified element */ 131 public java.util.ArrayList<TreeNode<E>> path(E e) { 132 java.util.ArrayList<TreeNode<E>> list = 133 new java.util.ArrayList<>(); 134 TreeNode<E> current = root; // Start from the root 135 136 while (current != null) { 137 list.add(current); // Add the node to the list 138 if (e.compareTo(current.element) < 0) { 139 current = current.left; 140 } 141 else if (e.compareTo(current.element) > 0) { 142 current = current.right; 143 } 144 else 145 break; 146 } 147 148 return list; // Return an array list of nodes 149 } 150 151 @Override /** Delete an element from the binary tree. 152 * Return true if the element is deleted successfully 153 * Return false if the element is not in the tree */ 154 public boolean delete(E e) { 155 // Locate the node to be deleted and also locate its parent node 156 TreeNode<E> parent = null; 157 TreeNode<E> current = root; 158 while (current != null) { 159 if (e.compareTo(current.element) < 0) { 160 parent = current; 161 current = current.left; 162 } 163 else if (e.compareTo(current.element) > 0) { 164 parent = current; 165 current = current.right; 166 } 167 else 168 break; // Element is in the tree pointed at by current 169 }

inner class

getSize

getRoot

path

delete

locate parent locate current

current found

M25_LIAN1878_11_GE_C25.indd 985 12/29/17 4:43 PM

986 Chapter 25 Binary Search Trees

170 171 if (current == null) 172 return false; // Element is not in the tree 173 174 // Case 1: current has no left child 175 if (current.left == null) { 176 // Connect the parent with the right child of the current node 177 if (parent == null) { 178 root = current.right; 179 } 180 else { 181 if (e.compareTo(parent.element) < 0) 182 parent.left = current.right; 183 else 184 parent.right = current.right; 185 } 186 } 187 else { 188 // Case 2: The current node has a left child 189 // Locate the rightmost node in the left subtree of 190 // the current node and also its parent 191 TreeNode<E> parentOfRightMost = current; 192 TreeNode<E> rightMost = current.left; 193 194 while (rightMost.right != null) { 195 parentOfRightMost = rightMost; 196 rightMost = rightMost.right; // Keep going to the right 197 } 198 199 // Replace the element in current by the element in rightMost 200 current.element = rightMost.element; 201 202 // Eliminate rightmost node 203 if (parentOfRightMost.right == rightMost) 204 parentOfRightMost.right = rightMost.left; 205 else 206 // Special case: parentOfRightMost == current 207 parentOfRightMost.left = rightMost.left; 208 } 209 210 size−−; 211 return true; // Element deleted successfully 212 } 213 214 @Override /** Obtain an iterator. Use inorder. */ 215 public java.util.Iterator<E> iterator() { 216 return new InorderIterator(); 217 } 218 219 // Inner class InorderIterator 220 private class InorderIterator implements java.util.Iterator<E> { 221 // Store the elements in a list 222 private java.util.ArrayList<E> list = 223 new java.util.ArrayList<>(); 224 private int current = 0; // Point to the current element in list 225 226 public InorderIterator() { 227 inorder(); // Traverse binary tree and store elements in list 228 } 229

not found

Case 1

reconnect parent

Case 2

locate parentOfRightMost locate rightMost

replace current

reconnect parentOfRightMost

reduce size successful deletion

iterator

iterator class

internal list

current position

M25_LIAN1878_11_GE_C25.indd 986 12/29/17 4:43 PM

25.2 Binary Search Trees 987

230 /** Inorder traversal from the root*/ 231 private void inorder() { 232 inorder(root); 233 } 234 235 /** Inorder traversal from a subtree */ 236 private void inorder(TreeNode<E> root) { 237 if (root == null)return; 238 inorder(root.left); 239 list.add(root.element); 240 inorder(root.right); 241 } 242 243 @Override /** More elements for traversing? */ 244 public boolean hasNext() { 245 if (current < list.size()) 246 return true; 247 248 return false; 249 } 250 251 @Override /** Get the current element and move to the next */ 252 public E next() { 253 return list.get(current++); 254 } 255 256 @Override /** Remove the current element */ 257 public void remove() { 258 if (current == 0) // next() has not been called yet 259 throw new IllegalStateException(); 260 261 delete(list.get(––current)); 262 list.clear(); // Clear the list 263 inorder(); // Rebuild the list 264 } 265 } 266 267 @Override /** Remove all elements from the tree */ 268 public void clear() { 269 root = null; 270 size = 0; 271 } 272 }

The insert(E e) method (lines 35–63) creates a node for element e and inserts it into the tree. If the tree is empty, the node becomes the root. Otherwise, the method finds an appropriate parent for the node to maintain the order of the tree. If the element is already in the tree, the method returns false; otherwise it returns true.

The inorder() method (lines 70–80) invokes inorder(root) to traverse the entire tree. The method inorder(TreeNode root) traverses the tree with the specified root. This is a recursive method. It recursively traverses the left subtree, then the root, and finally the right subtree. The traversal ends when the tree is empty.

The postorder() method (lines 83–93) and the preorder() method (lines 96–106) are implemented similarly using recursion.

The path(E e) method (lines 131–149) returns a path of the nodes as an array list. The path starts from the root leading to the element. The element may not be in the tree. For exam- ple, in Figure 25.4a, path(45) contains the nodes for elements 60, 55, and 45, and path(58) contains the nodes for elements 60, 55, and 57.

obtain inorder list

hasNext in iterator?

get next element

remove the current

refresh list clear list

tree clear method

M25_LIAN1878_11_GE_C25.indd 987 12/29/17 4:43 PM

988 Chapter 25 Binary Search Trees

The implementation of delete() and iterator() (lines 154–265) will be discussed in Sections 25.3 and 25.5.

Listing 25.5 gives an example that creates a binary search tree using BST (line 4). The pro- gram adds strings into the tree (lines 5–11), traverses the tree in inorder, postorder, and preorder (lines 14–20), searches for an element (line 24), and obtains a path from the node containing Peter to the root (lines 28–31).

Listing 25.5 TestBST.java 1 public class TestBST { 2 public static void main(String[] args) { 3 // Create a BST 4 BST<String> tree = new BST<>(); 5 tree.insert("George"); 6 tree.insert("Michael"); 7 tree.insert("Tom"); 8 tree.insert("Adam"); 9 tree.insert("Jones"); 10 tree.insert("Peter"); 11 tree.insert("Daniel"); 12 13 // Traverse tree 14 System.out.print("Inorder (sorted): "); 15 tree.inorder(); 16 System.out.print("\nPostorder: "); 17 tree.postorder(); 18 System.out.print("\nPreorder: "); 19 tree.preorder(); 20 System.out.print("\nThe number of nodes is " + tree.getSize()); 21 22 // Search for an element 23 System.out.print("\nIs Peter in the tree? " + 24 tree.search("Peter")); 25 26 // Get a path from the root to Peter 27 System.out.print("\nA path from the root to Peter is: "); 28 java.util.ArrayList<BST.TreeNode<String>> path 29 = tree.path("Peter"); 30 for (int i = 0; path != null && i < path.size(); i++) 31 System.out.print(path.get(i).element + " "); 32 33 Integer[] numbers = {2, 4, 3, 1, 8, 5, 6, 7}; 34 BST<Integer> intTree = new BST<>(numbers); 35 System.out.print("\nInorder (sorted): "); 36 intTree.inorder(); 37 } 38 }

create tree insert

inorder

postorder

preorder getSize

Inorder (sorted): Adam Daniel George Jones Michael Peter Tom Postorder: Daniel Adam Jones Peter Tom Michael George Preorder: George Adam Daniel Michael Jones Tom Peter The number of nodes is 7 Is Peter in the tree? true A path from the root to Peter is: George Michael Tom Peter Inorder (sorted): 1 2 3 4 5 6 7 8

The program checks path != null in line 30 to ensure that the path is not null before invoking path.get(i). This is an example of defensive programming to avoid potential runtime errors.

M25_LIAN1878_11_GE_C25.indd 988 12/29/17 4:43 PM

25.3 Deleting Elements from a BST 989

The program creates another tree for storing int values (line 34). After all the elements are inserted in the trees, the trees should appear as shown in Figure 25.7.

If the elements are inserted in a different order (e.g., Daniel, Adam, Jones, Peter, Tom, Michael, and George), the tree will look different. However, the inorder traversal prints elements in the same order as long as the set of elements is the same. The inorder traversal displays a sorted list.

25.2.1 Show the result of inserting 44 into Figure 25.4b. 25.2.2 Show the inorder, preorder, and postorder of traversing the elements in the binary

tree shown in Figure 25.1c.

25.2.3 If a set of elements is inserted into a BST in two different orders, will the two cor- responding BSTs look the same? Will the inorder traversal be the same? Will the postorder traversal be the same? Will the preorder traversal be the same?

25.2.4 What is the time complexity of inserting an element into a BST? 25.2.5 Implement the search(element) method using recursion.

25.3 Deleting Elements from a BST To delete an element from a BST, first locate it in the tree then consider two cases— whether or not the node has a left child—before deleting the element and reconnecting the tree.

The insert(element) method is presented in Section 25.2.3. Often, you need to delete an element from a binary search tree. Doing so is far more complex than adding an element into a binary search tree.

To delete an element from a binary search tree, you need to first locate the node that contains the element and also its parent node. Let current point to the node that contains the element in the binary search tree and parent point to the parent of the current node. The current node may be a left child or a right child of the parent node. There are two cases to consider.

Point Check

Point Key

locating element

Figure 25.7 The BSTs in Listing 25.5 are pictured here after they are created.

George

MichaelAdam

Daniel Jones Tom

Peter

root

(b)(a)

root

3 8

M25_LIAN1878_11_GE_C25.indd 989 12/29/17 4:43 PM

990 Chapter 25 Binary Search Trees

Case 1: The current node does not have a left child, as shown in Figure 25.8a. In this case, simply connect the parent with the right child of the current node, as shown in Figure 25.8b.

For example, to delete node 10 in Figure 25.9a, you would connect the parent of node 10 with the right child of node 10, as shown in Figure 25.9b.

Figure 25.9 Case 1: Deleting node 10 from (a) results in (b).

(a)

root

30 80

5027

(b)

root

30 80

5027

Figure 25.8 Case 1: The current node has no left child.

parentparent

Subtree

(a) (b)

Subtree

current may be a left or a right child of parent

Subtree may be a left or a right subtree of parent

current points the node to be deleted

current

No left child

Note If the current node is a leaf, it falls into Case 1. For example, to delete element 16 in Figure 25.9a, connect its right child (in this case, it is null) to the parent of node 16.

Case 2: The current node has a left child. Let rightMost point to the node that contains the largest element in the left subtree of the current node and parentOfRightMost point to the parent node of the rightMost node, as shown in Figure 25.10a. Note the rightMost node cannot have a right child but may have a left child. Replace the element value in the current node with the one in the rightMost node, connect the parentOfRightMost node with the left child of the rightMost node, and delete the rightMost node, as shown in Figure 25.10b.

For example, consider deleting node 20 in Figure 25.11a. The rightMost node has the element value 16. Replace the element value 20 with 16 in the current node and make node 10 the parent for node 14, as shown in Figure 25.11b.

delete a leaf

M25_LIAN1878_11_GE_C25.indd 990 12/29/17 4:43 PM

25.3 Deleting Elements from a BST 991

Note If the left child of current does not have a right child, current.left points to the largest element in the left subtree of current. In this case, rightMost is current. left and parentOfRightMost is current. You have to take care of this special case to reconnect the left child of rightMost with parentOfRightMost. special case

Figure 25.10 Case 2: The current node has a left child.

parent

Right subtree

(a) (b)

current may be a left or a right child of parent

current points to the node to be deleted

current

parentOfRightMost

rightMost

leftChildOfRightMost

parent

Right subtree

current

parentOfRightMost

Content copied to current and the node is deleted

leftChildOfRightMost

The content of the current node is replaced by the content of the rightMost node. The rightMost node is deleted.

null

Figure 25.11 Case 2: Deleting node 20 from (a) results in (b).

(a) (b)

root

rightMost

30 80

502714

root

30 80

502714

M25_LIAN1878_11_GE_C25.indd 991 12/29/17 4:43 PM

992 Chapter 25 Binary Search Trees

The algorithm for deleting an element from a binary search tree can be described in Listing 25.6.

Listing 25.6 Deleting an Element from a BST 1 boolean delete(E e) { 2 Locate element e in the tree; 3 if element e is not found 4 return false; 5 6 Let current be the node that contains e and parent be 7 the parent of current; 8 9 if (current has no left child) // Case 1 10 Connect the right child of current with parent; 11 Now current is not referenced, so it is eliminated; 12 else // Case 2 13 Locate the rightmost node in the left subtree of current. 14 Copy the element value in the rightmost node to current. 15 Connect the parent of the rightmost node to the left child 16 of rightmost node; 17 18 return true; // Element deleted 19 }

The complete implementation of the delete method is given in lines 154–212 in Listing 25.4. The method locates the node (named current) to be deleted and also locates its parent (named parent) in lines 156–169. If current is null, the element is not in the tree. Therefore, the method returns false (line 172). Please note that if current is root, parent is null. If the tree is empty, both current and parent are null.

Case 1 of the algorithm is covered in lines 175–186. In this case, the current node has no left child (i.e., current.left == null). If parent is null, assign current.right to root (lines 177–179). Otherwise, assign current.right to either parent.left or par- ent.right, depending on whether current is a left or a right child of parent (181–184).

Case 2 of the algorithm is covered in lines 187–208. In this case, current has a left child. The algorithm locates the rightmost node (named rightMost) in the left subtree of the current node and also its parent (named parentOfRightMost) (lines 194–197). Replace the element in current by the element in rightMost (line 200); assign rightMost.left to either parentOfRightMost.right or parentOfRightMost.left (lines 203–207), depending on whether rightMost is a right or a left child of parentOfRightMost.

Listing 25.7 gives a test program that deletes the elements from the binary search tree.

Listing 25.7 TestBSTDelete.java 1 public class TestBSTDelete { 2 public static void main(String[] args) { 3 BST<String> tree = new BST<>(); 4 tree.insert("George"); 5 tree.insert("Michael"); 6 tree.insert("Tom"); 7 tree.insert("Adam"); 8 tree.insert("Jones"); 9 tree.insert("Peter"); 10 tree.insert("Daniel"); 11 printTree(tree); 12 13 System.out.println("\nAfter delete George:");

delete method

not in the tree

locate current locate parent

Case 1

Case 2

M25_LIAN1878_11_GE_C25.indd 992 12/29/17 4:43 PM

25.3 Deleting Elements from a BST 993

14 tree.delete("George"); 15 printTree(tree); 16 17 System.out.println("\nAfter delete Adam:"); 18 tree.delete("Adam"); 19 printTree(tree); 20 21 System.out.println("\nAfter delete Michael:"); 22 tree.delete("Michael"); 23 printTree(tree); 24 } 25 26 public static void printTree(BST tree) { 27 // Traverse tree 28 System.out.print("Inorder (sorted): "); 29 tree.inorder(); 30 System.out.print("\nPostorder: "); 31 tree.postorder(); 32 System.out.print("\nPreorder: "); 33 tree.preorder(); 34 System.out.print("\nThe number of nodes is " + tree.getSize()); 35 System.out.println(); 36 } 37 }

delete an element

Inorder (sorted): Adam Daniel George Jones Michael Peter Tom Postorder: Daniel Adam Jones Peter Tom Michael George Preorder: George Adam Daniel Michael Jones Tom Peter The number of nodes is 7

After delete George: Inorder (sorted): Adam Daniel Jones Michael Peter Tom Postorder: Adam Jones Peter Tom Michael Daniel Preorder: Daniel Adam Michael Jones Tom Peter The number of nodes is 6

After delete Adam: Inorder (sorted): Daniel Jones Michael Peter Tom Postorder: Jones Peter Tom Michael Daniel Preorder: Daniel Michael Jones Tom Peter The number of nodes is 5

After delete Michael: Inorder (sorted): Daniel Jones Peter Tom Postorder: Peter Tom Jones Daniel Preorder: Daniel Jones Tom Peter The number of nodes is 4

Figures 25.12–25.14 show how the tree evolves as the elements are deleted from it.

Note It is obvious that the time complexity for the inorder, preorder, and postorder is O(n), since each node is traversed only once. The time complexity for search, insertion, and deletion is the height of the tree. In the worst case, the height of the tree is O(n). On average, the height of the tree is O(logn). So, the average time for search, insertion, dele- tion in a BST is O(logn).

BST time complexity

M25_LIAN1878_11_GE_C25.indd 993 12/29/17 4:43 PM

994 Chapter 25 Binary Search Trees

Figure 25.13 Deleting Adam falls into Case 1.

Daniel

MichaelAdam

Jones Tom

Peter

(a) Deleting Adam (b) After Adam is deleted

Delete this node

Daniel

Michael

Jones Tom

Peter

Figure 25.14 Deleting Michael falls into Case 2.

Daniel

Michael

Jones Tom

Peter

(a) Deleting Michael (b) After Michael is deleted

Daniel

Jones

Tom

Peter

Delete this node

Figure 25.12 Deleting George falls into Case 2.

George

MichaelAdam

Daniel Jones Tom

Peter

Delete this node

(a) Deleting George (b) After George is deleted

Daniel

MichaelAdam

Jones Tom

Peter

M25_LIAN1878_11_GE_C25.indd 994 12/29/17 4:43 PM

25.4 Tree Visualization and MVC 995

25.3.1 Show the result of deleting 55 from the tree in Figure 25.4b. 25.3.2 Show the result of deleting 60 from the tree in Figure 25.4b. 25.3.3 What is the time complexity of deleting an element from a BST? 25.3.4 Is the algorithm correct if lines 203–207 in Listing 25.4 in Case 2 of the

delete() method are replaced by the following code?

parentOfRightMost.right = rightMost.left;

25.4 Tree Visualization and MVC You can use recursion to display a binary tree.

Pedagogical Note One challenge facing the data-structure course is to motivate students. Displaying a binary tree graphically will not only help you understand the working of a binary tree but perhaps also stimulate your interest in programming. This section introduces the techniques to visualize binary trees. You can also apply visualization techniques to other projects.

How do you display a binary tree? It is a recursive structure, so you can display a binary tree using recursion. You can simply display the root, then display the two subtrees recursively. The techniques for displaying the Sierpinski triangle in Listing 18.9 can be applied to displaying a binary tree. For simplicity, we assume that the keys are positive integers less than 100. Listings 25.8 and 25.9 give the program, and Figure 25.15 shows some sample runs of the program.

Point Check

Point Key

Listing 25.8 BSTAnimation.java 1 import javafx.application.Application; 2 import javafx.geometry.Pos; 3 import javafx.stage.Stage; 4 import javafx.scene.Scene; 5 import javafx.scene.control.Button; 6 import javafx.scene.control.Label; 7 import javafx.scene.control.TextField; 8 import javafx.scene.layout.BorderPane; 9 import javafx.scene.layout.HBox; 10 11 public class BSTAnimation extends Application { 12 @Override // Override the start method in the Application class 13 public void start(Stage primaryStage) { 14 BST<Integer> tree = new BST<>(); // Create a tree 15 16 BorderPane pane = new BorderPane(); 17 BTView view = new BTView(tree); // Create a View

create a tree

view for tree

M25_LIAN1878_11_GE_C25.indd 995 12/29/17 4:43 PM

996 Chapter 25 Binary Search Trees

18 pane.setCenter(view); 19 20 TextField tfKey = new TextField(); 21 tfKey.setPrefColumnCount(3); 22 tfKey.setAlignment(Pos.BASELINE_RIGHT); 23 Button btInsert = new Button("Insert"); 24 Button btDelete = new Button("Delete"); 25 HBox hBox = new HBox(5); 26 hBox.getChildren().addAll(new Label("Enter a key: "), 27 tfKey, btInsert, btDelete); 28 hBox.setAlignment(Pos.CENTER); 29 pane.setBottom(hBox); 30 31 btInsert.setOnAction(e -> { 32 int key = Integer.parseInt(tfKey.getText()); 33 if (tree.search(key)) { // key is in the tree already 34 view.displayTree(); 35 view.setStatus(key + " is already in the tree"); 36 } 37 else { 38 tree.insert(key); // Insert a new key 39 view.displayTree(); 40 view.setStatus(key + " is inserted in the tree"); 41 } 42 }); 43 44 btDelete.setOnAction(e -> { 45 int key = Integer.parseInt(tfKey.getText()); 46 if (!tree.search(key)) { // key is not in the tree 47 view.displayTree(); 48 view.setStatus(key + " is not in the tree"); 49 } 50 else { 51 tree.delete(key); // Delete a key 52 view.displayTree(); 53 view.setStatus(key + " is deleted from the tree"); 54 } 55 }); 56 57 // Create a scene and place the pane in the stage 58 Scene scene = new Scene(pane, 450, 250); 59 primaryStage.setTitle("BSTAnimation"); // Set the stage title 60 primaryStage.setScene(scene); // Place the scene in the stage 61 primaryStage.show(); // Display the stage 62 } 63 }

Listing 25.9 BTView.java 1 import javafx.scene.layout.Pane; 2 import javafx.scene.paint.Color; 3 import javafx.scene.shape.Circle; 4 import javafx.scene.shape.Line; 5 import javafx.scene.text.Text; 6 7 public class BTView extends Pane { 8 private BST<Integer> tree = new BST<>(); 9 private double radius = 15; // Tree node radius 10 private double vGap = 50; // Gap between two levels in a tree 11

place tree view

place hBox

handle insertion

insert key display the tree

handle deletion

delete key display the tree

tree to display

M25_LIAN1878_11_GE_C25.indd 996 12/29/17 4:43 PM

25.4 Tree Visualization and MVC 997

12 BTView(BST<Integer> tree) { 13 this.tree = tree; 14 setStatus("Tree is empty"); 15 } 16 17 public void setStatus(String msg) { 18 getChildren().add(new Text(20, 20, msg)); 19 } 20 21 public void displayTree() { 22 this.getChildren().clear(); // Clear the pane 23 if (tree.getRoot() != null) { 24 // Display tree recursively 25 displayTree(tree.getRoot(), getWidth() / 2, vGap, 26 getWidth() / 4); 27 } 28 } 29 30 /** Display a subtree rooted at position (x, y) */ 31 private void displayTree(BST.TreeNode<Integer> root, 32 double x, double y, double hGap) { 33 if (root.left != null) { 34 // Draw a line to the left node 35 getChildren().add(new Line(x - hGap, y + vGap, x, y)); 36 // Draw the left subtree recursively 37 displayTree(root.left, x − hGap, y + vGap, hGap / 2); 38 } 39 40 if (root.right != null) { 41 // Draw a line to the right node 42 getChildren().add(new Line(x + hGap, y + vGap, x, y)); 43 // Draw the right subtree recursively 44 displayTree(root.right, x + hGap, y + vGap, hGap / 2); 45 } 46 47 // Display a node 48 Circle circle = new Circle(x, y, radius); 49 circle.setFill(Color.WHITE); 50 circle.setStroke(Color.BLACK); 51 getChildren().addAll(circle, 52 new Text(x - 4, y + 4, root.element + "")); 53 } 54 }

In Listing 25.8, BSTAnimation.java, a tree is created (line 14) and a tree view is placed in the pane (line 18). After a new key is inserted into the tree (line 38), the tree is repainted (line 39) to reflect the change. After a key is deleted (line 51), the tree is repainted (line 52) to reflect the change.

In Listing 25.9, BTView.java, the node is displayed as a circle with radius 15 (line 48). The distance between two levels in the tree is defined in vGap 50 (line 25). hGap (line 32) defines the distance between two nodes horizontally. This value is reduced by half (hGap / 2) in the next level when the displayTree method is called recursively (lines 37 and 44). Note that vGap is not changed in the tree.

The method displayTree is recursively invoked to display a left subtree (lines 33–38) and a right subtree (lines 40–45) if a subtree is not empty. A line is added to the pane to con- nect two nodes (lines 35 and 42). Note the method first adds the lines to the pane then adds the circle into the pane (line 52) so the circles will be painted on top of the lines to achieve desired visual effects.

set a tree

clear the display

display tree recursively

connect two nodes

draw left subtree

connect two nodes

draw right subtree

display a node

M25_LIAN1878_11_GE_C25.indd 997 12/29/17 4:43 PM

998 Chapter 25 Binary Search Trees

The program assumes the keys are integers. You can easily modify the program with a generic type to display keys of characters or short strings.

Tree visualization is an example of the model-view-controller (MVC) software architec- ture. This is an important architecture for software development. The model is for storing and handling data. The view is for visually presenting the data. The controller handles the user interaction with the model and controls the view, as shown in Figure 25.16.

The MVC architecture separates data storage and handling from the visual representation of the data. It has two major benefits:

■■ It makes multiple views possible so data can be shared through the same model. For example, you can create a new view that displays the tree with the root on the left and the tree grows horizontally to the right (see Programming Exercise 25.11).

■■ It simplifies the task of writing complex applications and makes the components scalable and easy to maintain. Changes can be made to the view without affecting the model, and vice versa.

25.4.1 How many times will the displayTree method be invoked if the tree is empty? How many times will the displayTree method be invoked if the tree has 100 nodes?

25.4.2 In what order are the nodes in the tree visited by the displayTree method: inor- der, preorder, or postorder?

25.4.3 What would happen if the code in lines 47–52 in Listing 25.9, BTView.java is moved to line 33?

25.4.4 What is MVC? What are the benefits of the MVC? 25.4.5 Write one statement that displays the maximum and minimum element in a BST

object named tree.

25.5 Iterators BST is iterable because it is defined as a subtype of the java.lang.Iterable interface.

The methods inorder(), preorder(), and postorder() display the elements in inorder, preorder, and postorder in a binary tree. These methods are limited to displaying the ele- ments in a tree. If you wish to process the elements in a binary tree rather than display them, these methods cannot be used. Recall that an iterator is provided for traversing the elements in a set or list. You can apply the same approach in a binary tree to provide a uniform way of traversing the elements in a binary tree.

The java.lang.Iterable interface defines the iterator method, which returns an instance of the java.util.Iterator interface. The java.util.Iterator interface (see Figure 25.17) defines the common features of iterators.

Point Check

Point Key

iterator

Figure 25.16 The controller obtains data and stores it in a model. The view displays the data stored in the model.

ModelView

BSTAnimation

Controller

BSTBTView

M25_LIAN1878_11_GE_C25.indd 998 12/29/17 4:43 PM

25.5 Iterators 999

The Tree interface extends java.util.Collection. Since Collection extends java. lang.Iterable, BST is also a subclass of Iterable. The Iterable interface contains the iterator() method that returns an instance of java.util.Iterator.

You can traverse a binary tree in inorder, preorder, or postorder. Since inorder is used fre- quently, we will use inorder for traversing the elements in a binary tree. We define an iterator class named InorderIterator to implement the java.util.Iterator interface in Listing 25.4 (lines 220–265). The iterator method simply returns an instance of InorderItera- tor (line 216).

The InorderIterator constructor invokes the inorder method (line 227). The inorder(root) method (lines 236–241) stores all the elements from the tree in list. The elements are traversed in inorder.

Once an Iterator object is created, its current value is initialized to 0 (line 224), which points to the first element in the list. Invoking the next() method returns the current element and moves current to point to the next element in the list (line 252).

The hasNext() method checks whether current is still in the range of list (line 245). The remove() method removes the element returned by the last next() (line 258). After-

ward, a new list is created (lines 261–262). Note that current does not need to be changed. Listing 25.10 gives a test program that stores the strings in a BST and displays all strings

in uppercase.

Listing 25.10 TestBSTWithIterator.java 1 public class TestBSTWithIterator { 2 public static void main(String[] args) { 3 BST<String> tree = new BST<>(); 4 tree.insert("George"); 5 tree.insert("Michael"); 6 tree.insert("Tom"); 7 tree.insert("Adam"); 8 tree.insert("Jones"); 9 tree.insert("Peter"); 10 tree.insert("Daniel"); 11 12 for (String s: tree) 13 System.out.print(s.toUpperCase() + " "); 14 } 15 }

The foreach loop (lines 12 and 13) uses an iterator to traverse all elements in the tree.

how to create an iterator

use an iterator get uppercase letters

Figure 25.17 The Iterator interface defines a uniform way of traversing the elements in a container.

«interface» java.util.Iterator<E>

+hasNext(): boolean

+next(): E

+remove(): void

Returns true if the iterator has more elements.

Returns the next element in the iterator.

Removes from the underlying container the last element returned by the iterator (optional operation).

ADAM DANIEL GEORGE JONES MICHAEL PETER TOM

M25_LIAN1878_11_GE_C25.indd 999 12/29/17 4:43 PM

1000 Chapter 25 Binary Search Trees

Design Guide Iterator is an important software design pattern. It provides a uniform way of traversing the elements in a container, while hiding the container’s structural details. By implement- ing the same interface java.util.Iterator, you can write a program that traverses the elements of all containers in the same way.

Note java.util.Iterator defines a forward iterator, which traverses the elements in the iterator in a forward direction, and each element can be traversed only once. The Java API also provides the java.util.ListIterator, which supports traversing in both forward and backward directions. If your data structure warrants flexible traversing, you may define iterator classes as a subtype of java.util.ListIterator.

The implementation of the iterator is not efficient. Every time you remove an element through the iterator, the whole list is rebuilt (line 263 in Listing 25.4, BST.java). The client should always use the delete method in the BST class to remove an element. To prevent the user from using the remove method in the iterator, implement the iterator as follows:

public void remove() { throw new UnsupportedOperationException ("Removing an element from the iterator is not supported"); }

After making the remove method unsupported by the iterator class, you can implement the iterator more efficiently without having to maintain a list for the elements in the tree. You can use a stack to store the nodes, and the node on the top of the stack contains the element that is to be returned from the next() method. If the tree is well balanced, the maximum stack size will be O(logn).

25.5.1 What is an iterator? 25.5.2 What method is defined in the java.lang.Iterable<E> interface? 25.5.3 Suppose you delete implements Collection<E> from line 3 in Listing 25.3,

Tree.java. Will Listing 25.10 still compile?

25.5.4 What is the benefit of being a subtype of Iterable<E>? 25.5.5 Write one statement that displays the maximum and minimum element in a BST

object named tree.

25.6 Case Study: Data Compression Huffman coding compresses data by using fewer bits to encode characters that occur more frequently. The codes for the characters are constructed based on the occurrence of the characters in the text using a binary tree, called the Huffman coding tree.

Compressing data is a common task. There are many utilities available for compressing files. This section introduces Huffman coding, invented by David Huffman in 1952.

In ASCII, every character is encoded in 8 bits. If a text consists of 100 characters, it will take 800 bits to represent the text. The idea of Huffman coding is to use a fewer bits to encode frequently used characters in the text and more bits to encode less frequently used characters to reduce the overall size of the file. In Huffman coding, the characters’ codes are constructed based on the characters’ occurrence in the text using a binary tree, called the Huffman coding tree. Suppose the text is Mississippi. Its Huffman tree is as shown in Figure 25.18a. The left and right edges of a node are assigned the values 0 and 1, respectively. Each character is a leaf in the tree. The code for the character consists of the edge values in the path from the

iterator pattern advantages of iterators

variations of iterators

Point Check

Point Key

Huffman coding

M25_LIAN1878_11_GE_C25.indd 1000 12/29/17 4:43 PM

25.6 Case Study: Data Compression 1001

root to the leaf, as shown in Figure 25.18b. Since i and s appear more than M and p in the text, they are assigned shorter codes.

The coding tree is also used for decoding a sequence of bits into characters. To do so, start with the first bit in the sequence and determine whether to go to the left or right branch of the tree’s root based on the bit value. Consider the next bit and continue to go down to the left or right branch based on the bit value. When you reach a leaf, you have found a character. The next bit in the stream is the first bit of the next character. For example, the stream 011001 is decoded to sip, with 01 matching s, 1 matching i, and 001 matching p.

Based on the coding scheme in Figure 25.18,

is encoded to is decoded to Mississippi

is encoded to = = = = = = = = =

000101011010110010011 is decoded to

= = = = = = = = = =

Mississippi

decoding

Figure 25.18 The codes for characters are constructed based on the occurrence of characters in the text using a coding tree.

Character M p s i

Code 000 001 01 1

Frequency 1 2 4 4

M p

10 s

0 i

(a) Huffman coding tree (b) Character code table

To construct a Huffman coding tree, use the following algorithm:

1. Begin with a forest of trees. Each tree contains a node for a character. The weight of the node is the frequency of the character in the text.

2. Repeat the following action to combine trees until there is only one tree: Choose two trees with the smallest weight and create a new node as their parent. The weight of the new tree is the sum of the weight of the subtrees.

3. For each interior node, assign its left edge a value 0 and right edge a value 1. All leaf nodes represent characters in the text.

Here is an example of building a coding tree for the text Mississippi. The frequency table for the characters is shown in Figure 25.18b. Initially, the forest contains single-node trees, as shown in Figure 25.19a. The trees are repeatedly combined to form large trees until only one tree is left, as shown in Figures 25.19b–d.

It is worth noting that no code is a prefix of another code. This property ensures that the streams can be decoded unambiguously.

The algorithm used here is an example of a greedy algorithm. A greedy algorithm is often used in solving optimization problems. The algorithm makes the choice that is optimal locally in the hope that this choice will lead to a globally optimal solution. In this case, the algorithm always chooses two trees with the smallest weight and creates a new node as their parent. This intuitive optimal local solution indeed leads to a final optimal solution for constructing a Huffman tree. As another example, consider changing money into the fewest possible coins. A greedy algorithm would take the largest possible coin first. For example, for 98¢, you would use three quarters to make 75¢, additional two dimes to make 95¢, and additional three pennies to make the 98¢. The greedy algorithm finds an optimal solution for this problem. However,

construct coding tree

prefix property

greedy algorithm

M25_LIAN1878_11_GE_C25.indd 1001 12/29/17 4:43 PM

1002 Chapter 25 Binary Search Trees

a greedy algorithm is not always going to find the optimal result; see the bin packing problem in Programming Exercise 11.19.

Listing 25.11 gives a program that prompts the user to enter a string, displays the frequency table of the characters in the string, and displays the Huffman code for each character.

Listing 25.11 HuffmanCode.java 1 import java.util.Scanner; 2 3 public class HuffmanCode { 4 public static void main(String[] args) { 5 Scanner input = new Scanner(System.in); 6 System.out.print("Enter text: "); 7 String text = input.nextLine(); 8 9 int[] counts = getCharacterFrequency(text); // Count frequency 10 11 System.out.printf("%-15s%-15s%-15s%-15s\n", 12 "ASCII Code", "Character", "Frequency", "Code"); 13 14 Tree tree = getHuffmanTree(counts); // Create a Huffman tree 15 String[] codes = getCode(tree.root); // Get codes 16 17 for (int i = 0; i < codes.length; i++) 18 if (counts[i] != 0) // (char)i is not in text if counts[i] is 0 19 System.out.printf("%−15d%−15s%−15d%−15s\n", 20 i, (char)i + "", counts[i], codes[i]); 21 } 22 23 /** Get Huffman codes for the characters 24 * This method is called once after a Huffman tree is built

count frequency

get Huffman tree code for each character

Figure 25.19 The coding tree is built by repeatedly combining the two smallest-weighted trees.

(b)(a)

weight: 1 ‘M’

weight: 4 ‘s’

weight: 4 ‘i’

weight: 2 ‘p’

weight: 4 ‘s’

weight: 4 ‘i’

weight: 3

weight: 1 ‘M’

weight: 2 ‘p’

(c)

weight: 1 ‘M’

weight: 2 ‘p’

weight: 4 ‘i’

weight: 7

weight: 3 weight: 4 ‘s’

(d)

weight: 1 ‘M’

weight: 2 ‘p’

weight: 4 ‘i’

weight: 7

0 1

weight: 11

weight: 3 weight: 4 ‘s’

M25_LIAN1878_11_GE_C25.indd 1002 12/29/17 4:43 PM

25.6 Case Study: Data Compression 1003

25 */ 26 public static String[] getCode(Tree.Node root) { 27 if (root == null) return null; 28 String[] codes = new String[2 * 128]; 29 assignCode(root, codes); 30 return codes; 31 } 32 33 /* Recursively get codes to the leaf node */ 34 private static void assignCode(Tree.Node root, String[] codes) { 35 if (root.left != null) { 36 root.left.code = root.code + "0"; 37 assignCode(root.left, codes); 38 39 root.right.code = root.code + "1"; 40 assignCode(root.right, codes); 41 } 42 else { 43 codes[(int)root.element] = root.code; 44 } 45 } 46 47 /** Get a Huffman tree from the codes */ 48 public static Tree getHuffmanTree(int[] counts) { 49 // Create a heap to hold trees 50 Heap<Tree> heap = new Heap<>(); // Defined in Listing 23.9 51 for (int i = 0; i < counts.length; i++) { 52 if (counts[i] > 0) 53 heap.add(new Tree(counts[i], (char)i)); // A leaf node tree 54 } 55 56 while (heap.getSize() > 1) { 57 Tree t1 = heap.remove(); // Remove the smallest−weight tree 58 Tree t2 = heap.remove(); // Remove the next smallest 59 heap.add(new Tree(t1, t2)); // Combine two trees 60 } 61 62 return heap.remove(); // The final tree 63 } 64 65 /** Get the frequency of the characters */ 66 public static int[] getCharacterFrequency(String text) { 67 int[] counts = new int[256]; // 256 ASCII characters 68 69 for (int i = 0; i < text.length(); i++) 70 counts[(int)text.charAt(i)]++; // Count the characters in text 71 72 return counts; 73 } 74 75 /** Define a Huffman coding tree */ 76 public static class Tree implements Comparable<Tree> { 77 Node root; // The root of the tree 78 79 /** Create a tree with two subtrees */ 80 public Tree(Tree t1, Tree t2) { 81 root = new Node(); 82 root.left = t1.root; 83 root.right = t2.root; 84 root.weight = t1.root.weight + t2.root.weight;

getCode

assignCode

getHuffmanTree

getCharacterFrequency

Huffman tree

M25_LIAN1878_11_GE_C25.indd 1003 12/29/17 4:43 PM

1004 Chapter 25 Binary Search Trees

85 } 86 87 /** Create a tree containing a leaf node */ 88 public Tree(int weight, char element) { 89 root = new Node(weight, element); 90 } 91 92 @Override /** Compare trees based on their weights */ 93 public int compareTo(Tree t) { 94 if (root.weight < t.root.weight) // Purposely reverse the order 95 return 1; 96 else if (root.weight == t.root.weight) 97 return 0; 98 else 99 return -1; 100 } 101 102 public class Node { 103 char element; // Stores the character for a leaf node 104 int weight; // weight of the subtree rooted at this node 105 Node left; // Reference to the left subtree 106 Node right; // Reference to the right subtree 107 String code = ""; // The code of this node from the root 108 109 /** Create an empty node */ 110 public Node() { 111 } 112 113 /** Create a node with the specified weight and character */ 114 public Node(int weight, char element) { 115 this.weight = weight; 116 this.element = element; 117 } 118 } 119 } 120 }

tree node

Enter text: Welcome

ASCII Code Character Frequency Code 87 W 1 110

99 c 1 111

101 e 2 10

108 l 1 011

109 m 1 010

111 o 1 00

The program prompts the user to enter a text string (lines 5–7) and counts the frequency of the characters in the text (line 9). The getCharacterFrequency method (lines 66–73) cre- ates an array counts to count the occurrences of each of the 256 ASCII characters in the text. If a character appears in the text, its corresponding count is increased by 1 (line 70).

The program obtains a Huffman coding tree based on counts (line 14). The tree consists of linked nodes. The Node class is defined in lines 102–118. Each node consists of properties

getCharacterFrequency

M25_LIAN1878_11_GE_C25.indd 1004 12/29/17 4:43 PM

Quiz 1005

element (storing character), weight (storing weight of the subtree under this node), left (linking to the left subtree), right (linking to the right subtree), and code (storing the Huffman code for the character). The Tree class (lines 76–119) contains the root property. From the root, you can access all the nodes in the tree. The Tree class implements Comparable. The trees are comparable based on their weights. The compare order is purposely reversed (lines 93–100) so the smallest-weight tree is removed first from the heap of trees.

The getHuffmanTree method returns a Huffman coding tree. Initially, the single-node trees are created and added to the heap (lines 50–54). In each iteration of the while loop (lines 56–60), two smallest-weight trees are removed from the heap and are combined to form a big tree, then the new tree is added to the heap. This process continues until the heap contains just one tree, which is our final Huffman tree for the text.

The assignCode method assigns the code for each node in the tree (lines 34–45). The getCode method gets the code for each character in the leaf node (lines 26–31). The element codes[i] contains the code for character (char)i, where i is from 0 to 255. Note codes[i] is null if (char)i is not in the text.

25.6.1 Every internal node in a Huffman tree has two children. Is it true? 25.6.2 What is a greedy algorithm? Give an example. 25.6.3 If the Heap class in line 50 in Listing 25.9 is replaced by java.util.

PriorityQueue, will the program still work?

25.6.4 How do you replace lines 94–99 in Listing 25.11 using one line?

Node class

Tree class

getHuffmanTree

assignCode

getCode

Point Check

Key Terms binary search tree 976 binary tree 976 breadth-first traversal 980 depth 976 depth-first traversal 980 greedy algorithm 1001 height 976 Huffman coding 1000

inorder traversal 979 leaf 976 length 976 level 976 postorder traversal 979 preorder traversal 979 sibling 976 tree traversal 979

ChapTer summary 1. A binary search tree (BST) is a hierarchical data structure. You learned how to define and

implement a BST class, how to insert and delete elements into/from a BST, and how to traverse a BST using inorder, postorder, preorder, depth-first, and breadth-first searches.

2. An iterator is an object that provides a uniform way of traversing the elements in a con- tainer, such as a set, a list, or a binary tree. You learned how to define and implement iterator classes for traversing the elements in a binary tree.

3. Huffman coding is a scheme for compressing data by using fewer bits to encode characters that occur more frequently. The codes for characters are constructed based on the occur- rence of characters in the text using a binary tree, called the Huffman coding tree.

Quiz Answer the quiz for this chapter online at the book Companion Website.

M25_LIAN1878_11_GE_C25.indd 1005 12/29/17 4:43 PM

1006 Chapter 25 Binary Search Trees

programming exerCises

Sections 25.2–25.6 *25.1 (Add new methods in BST) Add the following new methods in BST.

/** Display the nodes in a breadth-first traversal */ public void breadthFirstTraversal()

/** Return the height of this binary tree */ public int height()

*25.2 (Test perfect binary tree) A perfect binary tree is a complete binary tree with all levels fully filled. Add a method in the BST class to return true if the tree is a perfect binary tree. (Hint: The number of nodes in a nonempty perfect binary tree is 2height - 1)

/** Returns true if the tree is a perfect binary tree */ boolean isPerfectBST()

**25.3 (Implement inorder traversal without using recursion) Implement the inorder method in BST using a stack instead of recursion. Write a test program that prompts the user to enter 15 integers, stores them in a BST, and invokes the inorder method to display the elements.

**25.4 (Implement preorder traversal without using recursion) Implement the preorder method in BST using a stack instead of recursion. Write a test pro- gram that prompts the user to enter 15 integers, stores them in a BST, and invokes the preorder method to display the elements.

**25.5 (Implement postorder traversal without using recursion) Implement the postorder method in BST using a stack instead of recursion. Write a test program that prompts the user to enter 15 integers, stores them in a BST, and invokes the postorder method to display the elements.

**25.6 (Find the leaves) Add a method in the BST class to return the number of the leaves as follows:

/** Return the number of leaf nodes */ public int getNumberOfLeaves()

**25.7 (Find the nonleaves) Add a method in the BST class to return the number of the nonleaves as follows:

/** Return the number of nonleaf nodes */ public int getNumberofNonLeaves()

***25.8 (Implement bidirectional iterator) The java.util.Iterator interface defines a forward iterator. The Java API also provides the java.util.ListIterator interface that defines a bidirectional iterator. Study ListIterator and define a bidirectional iterator for the BST class.

**25.9 (Tree clone and equals) Implement the clone and equals methods in the BST class. Two BST trees are equal if they contain the same elements. The clone method returns an identical copy of a BST.

25.10 (Preorder iterator) Add the following method in the BST class that returns an iterator for traversing the elements in a BST in preorder.

/** Return an iterator for traversing the elements in preorder */ java.util.Iterator<E> preorderIterator()

M25_LIAN1878_11_GE_C25.indd 1006 12/29/17 4:43 PM

Programming Exercises 1007

25.11 (Inorder iterator) Add the following method in the BST class that returns an iterator for traversing the elements in a BST in inorder.

/** Returns an iterator for traversing the elements in inorder */ java.util.Iterator<E> inorderIterator()

**25.12 (Test BST) Design and write a complete test program to test if the BST class in Listing 25.4 meets all requirements.

**25.13 (Add new buttons in BSTAnimation) Modify Listing 25.8, BSTAnimation.java, to add three new buttons—Show Inorder, Show Preorder, and Show Postorder— to display the result in a label, as shown in Figure 25.20. You need also to modify Listing 25.4, BST.java to implement the inorderList(), preorderList(), and postorderList() methods so each of these methods returns a List of the node elements in inorder, preorder, and postorder, as follows:

public java.util.List<E> inorderList(); public java.util.List<E> preorderList(); public java.util.List<E> postorderList();

Figure 25.20 When you click the Show Inorder, Show Preorder, or Show Postorder button, the elements are displayed in an inorder, preorder, or postorder in a label. Source: Copyright © 1995–2016 Oracle and/or its affiliates. All rights reserved. Used with permission.

*25.14 (Modify BST using Comparator) Revise BST in Listing 25.4 using a Com- parator for comparing objects. Define the new class as BST<E> with two constructors:

BST(); // Compare elements using their natural order BST(Comparator<? super E> comparator)

M25_LIAN1878_11_GE_C25.indd 1007 12/29/17 4:43 PM

1008 Chapter 25 Binary Search Trees

Hint: You need to add a data field for Comparator in the BST class as follows:

protected Comparator<E> c = (e1, e2) -> ((Comparable<E>)e1).compareTo(e2);

The lambda expression gives the default comparator using a natural order. You need to use comparator c to replace e.compareTo(anotherElement) with c.compare(e, anotherElement) in Listing 25.4.

*25.15 (Parent reference for BST) Redefine TreeNode by adding a reference to a node’s parent, as shown below:

BST.TreeNode<E>

#element: E

#left: TreeNode<E>

#right: TreeNode<E>

#parent: TreeNode<E>

Enter 10 integers: 45 54 67 56 50 45 23 59 23 67

[50, 54, 23] [59, 56, 67, 54, 23]

Reimplement the insert and delete methods in the BST class to update the parent for each node in the tree. Add the following new method in BST:

/** Return the node for the specified element. * Return null if the element is not in the tree. */ private TreeNode<E> getNode(E element)

/** Return true if the node for the element is a leaf */ private boolean isLeaf(E element)

/** Return the path of elements from the specified element * to the root in an array list. */ public ArrayList<E> getPath(E e)

Write a test program that prompts the user to enter 10 integers, adds them to the tree, deletes the first integer from the tree, and displays the paths for all leaf nodes. Here is a sample run:

***25.16 (Data compression: Huffman coding) Write a program that prompts the user to enter a file name, then displays the frequency table of the characters in the file and the Huffman code for each character.

***25.17 (Data compression: Huffman coding animation) Write a program that enables the user to enter text and displays the Huffman coding tree based on the text, as shown in Figure 25.21a. Display the weight of the subtree inside the subtree’s root circle. Display each leaf node’s character. Display the encoded bits for the text in a label. When the user clicks the Decode Text button, a bit string is decoded into text displayed in the label, as shown in Figure 25.21b.

M25_LIAN1878_11_GE_C25.indd 1008 12/29/17 4:43 PM

Programming Exercises 1009

***25.18 (Compress a file) Write a program that compresses a source file into a target file using the Huffman coding method. First, use ObjectOutputStream to output the Huffman codes into the target file, then use BitOutputStream in Programming Exercise 17.17 to output the encoded binary contents to the target file. Pass the files from the command line using the following command:

java Exercise25_18 sourcefile targetfile

***25.19 (Decompress a file) The preceding exercise compresses a file. The compressed file contains the Huffman codes and the compressed contents. Write a program that decompresses a source file into a target file using the following command:

java Exercise25_19 sourcefile targetfile

Figure 25.21 (a) The animation shows the coding tree for a given text string and the encoded bits for the text are displayed in the label; (b) You can enter a bit string to display its text in the label. Source: Copyright © 1995–2016 Oracle and/or its affiliates. All rights reserved. Used with permission.

(a)

(b)

M25_LIAN1878_11_GE_C25.indd 1009 12/29/17 4:43 PM

A01_GIPE3898_09_SE_FM.indd 3 12/20/17 5:56 PM

This page intentionally left blank

Objectives ■■ To know what an AVL tree is (§26.1).

■■ To understand how to rebalance a tree using the LL rotation, LR rota- tion, RR rotation, and RL rotation (§26.2).

■■ To design the AVLTree class by extending the BST class (§26.3).

■■ To insert elements into an AVL tree (§26.4).

■■ To implement tree rebalancing (§26.5).

■■ To delete elements from an AVL tree (§26.6).

■■ To implement the AVLTree class (§26.7).

■■ To test the AVLTree class (§26.8).

■■ To analyze the complexity of search, insertion, and deletion operations in AVL trees (§26.9).

AVL Trees

Chapter

M26_LIAN1878_11_GE_C26.indd 1011 12/29/17 4:44 PM

1012 Chapter 26 AVL Trees

26.1 Introduction AVL Tree is a balanced binary search tree.

Chapter 25 introduced binary search trees. The search, insertion, and deletion times for a binary tree depend on the height of the tree. In the worst case, the height is O(n). If a tree is perfectly balanced—that is, a complete binary tree—its height is log n. Can we maintain a perfectly balanced tree? Yes, but doing so will be costly. The compromise is to maintain a well-balanced tree—that is, the heights of every node’s two subtrees are about the same. This chapter introduces AVL trees. Web Chapters 40 and 41 will introduce 2–4 trees and red–black trees.

AVL trees are well balanced. AVL trees were invented in 1962 by two Russian computer scientists, G. M. Adelson-Velsky and E. M. Landis (hence the name AVL). In an AVL tree, the difference between the heights of every node’s two subtrees is 0 or 1. It can be shown that the maximum height of an AVL tree is O(log n).

The process for inserting or deleting an element in an AVL tree is the same as in a binary search tree, except that you may have to rebalance the tree after an insertion or deletion operation. The balance factor of a node is the height of its right subtree minus the height of its left subtree. For example, the balance factor for the node 87 in Figure 26.1a is 0, for the node 67 is 1, and for the node 55 is −1. A node is said to be balanced if its balance factor is −1, 0, or 1. A node is considered left-heavy if its balance factor is −1 or less, and right-heavy if its balance factor is +1 or greater.

Point Key

perfectly balanced tree

well-balanced tree

AVL tree

O(log n)

balance factor

balanced left-heavy right-heavy

Figure 26.1 A balance factor determines whether a node is balanced.

(a) (b)

100

67 107

100

67 107

87 187105

Pedagogical Note For an interactive GUI demo to see how an AVL tree works, go to liveexample . pearsoncmg.com/dsanimation/AVLTreeeBook.html, as shown in Figure 26.2.

26.2 Rebalancing Trees After inserting or deleting an element from an AVL tree, if the tree becomes unbalanced, perform a rotation operation to rebalance the tree.

If a node is not balanced after an insertion or deletion operation, you need to rebalance it. The process of rebalancing a node is called rotation. There are four possible rotations: LL, RR, LR, and RL.

LL rotation: An LL imbalance occurs at a node A, such that A has a balance factor of −2 and a left child B with a balance factor of −1 or 0, as shown in Figure 26.3a. This type of imbalance can be fixed by performing a single right rotation at A, as shown in Figure 26.3b.

RR rotation: An RR imbalance occurs at a node A, such that A has a balance factor of +2 and a right child B with a balance factor of +1 or 0, as shown in Figure 26.4a. This type of imbalance can be fixed by performing a single left rotation at A, as shown in Figure 26.4b.

AVL tree animation on Companion Website

Point Key

rotation

LL rotation

LL imbalance

RR rotation

RR imbalance

M26_LIAN1878_11_GE_C26.indd 1012 12/29/17 4:45 PM

26.2 Rebalancing Trees 1013

LR rotation: An LR imbalance occurs at a node A, such that A has a balance factor of −2 and a left child B with a balance factor of +1, as shown in Figure 26.5a. Assume B’s right child is C. This type of imbalance can be fixed by performing a double rotation (first a single left rotation at B, then a single right rotation at A), as shown in Figure 26.5b.

RL rotation: An RL imbalance occurs at a node A, such that A has a balance factor of +2 and a right child B with a balance factor of −1, as shown in Figure 26.6a. Assume B’s left child is C. This type of imbalance can be fixed by performing a double rotation (first a single right rotation at B, then a single left rotation at A), as shown in Figure 26.6b.

26.2.1 What is an AVL tree? Describe the following terms: balance factor, left-heavy, and right-heavy.

26.2.2 Show the balance factor of each node in the trees shown in Figure 26.1. 26.2.3 Describe LL rotation, RR rotation, LR rotation, and RL rotation for an AVL tree.

LR rotation

LR imbalance

RL rotation RL imbalance

Point Check

Figure 26.3 An LL rotation fixes an LL imbalance.

A 22

B21 or 0

T1h 1 1

T2’s height is h or h 1 1

h 1 1

A 0 or 21

B0 or 1

T2 T3

h h

(a) (b)

M26_LIAN1878_11_GE_C26.indd 1013 12/29/17 4:45 PM

1014 Chapter 26 AVL Trees

Figure 26.4 An RR rotation fixes an RR imbalance.

A 12

B 11 or 0

T1 h 1 1

T2’s height is h or h 1 1

h 1 1

A0 or 11

B 0 or 21

T2T3

(a) (b)

Figure 26.5 An LR rotation fixes an LR imbalance.

B0 or 21 0 or 1

C 0

T2T1 h h

T4T3 hh

(a) (b)

A 22

B11

21, 0, or 1C

T3T2 hh

T2 and T3 may have different heights, but at least one has a height of h.

Figure 26.6 An RL rotation fixes an RL imbalance.

A0 or 21 0 or 1

C 0

T2T1 h h

T4T3 hh

(a) (b)

A 12

B21

0, 21, or 1 C

T3T2 hh T2 and T3 may have different heights, but at least one has a height of h.

M26_LIAN1878_11_GE_C26.indd 1014 12/29/17 4:45 PM

26.3 Designing Classes for AVL Trees 1015

26.3 Designing Classes for AVL Trees Since an AVL tree is a binary search tree, AVLTree is designed as a subclass of BST.

An AVL tree is a binary tree, so you can define the AVLTree class to extend the BST class, as shown in Figure 26.7. The BST and TreeNode classes were defined in Section 25.2.5. Point

Key

Figure 26.7 The AVLTree class extends BST with new implementations for the insert and delete methods.

AVLTree<E extends Comparable<E>>

+AVLTree()

+AVLTree(objects: E[])

#createNewNode(): AVLTreeNode<E>

+insert(e: E): boolean

+delete(e: E): boolean

–updateHeight(node: AVLTreeNode<E>): void

Creates an empty AVL tree.

Creates an AVL tree from an array of objects.

Overrides this method to create an AVLTreeNode.

Returns true if the element is added successfully.

Balances the nodes in the path from the node for the element to the root if needed.

BST<E extends Comparable<E>>

Link

–balancePath(e: E): void

Returns the balance factor of the node.–balanceFactor(node: AVLTreeNode<E>): int

Performs LL balance.–balanceLL(A: TreeNode, parentOfA: TreeNode<E>): void

Performs LR balance.–balanceLR(A: TreeNode<E>, parentOfA: TreeNode<E>): void

Performs RR balance.–balanceRR(A: TreeNode<E>, parentOfA: TreeNode<E>): void

Performs RL balance.–balanceRL(A: TreeNode<E>, parentOfA: TreeNode<E>): void

AVLTreeNode<E>

height: int

TreeNode<E>

Returns true if the element is removed from the tree successfully.

Updates the height of the speci�ed node.

In order to balance the tree, you need to know each node’s height. For convenience, store the height of each node in AVLTreeNode and define AVLTreeNode to be a subclass of BST.TreeNode. Note that TreeNode is defined as a static inner class in BST. AVLTreeNode will be defined as a static inner class in AVLTree. TreeNode contains the data fields element, left, and right, which are inherited by AVLTreeNode. Thus, AVLTreeNode contains four data fields, as shown in Figure 26.8.

AVLTreeNode

Figure 26.8 An AVLTreeNode contains the protected data fields element, height, left, and right.

#element: E

#height: int

#left: TreeNode<E>

#right: TreeNode<E>

node: AVLTreeNode<E>

The element stored in this node.

The height of this node.

The left child of this node.

The right child of this node.

M26_LIAN1878_11_GE_C26.indd 1015 12/29/17 4:45 PM

1016 Chapter 26 AVL Trees

In the BST class, the createNewNode() method creates a TreeNode object. This method is overridden in the AVLTree class to create an AVLTreeNode. Note the return type of the createNewNode() method in the BST class is TreeNode, but the return type of the createNewNode() method in the AVLTree class is AVLTreeNode. This is fine, since AVLTreeNode is a subclass of TreeNode.

Searching for an element in an AVLTree is the same as searching in a binary search tree, so the search method defined in the BST class also works for AVLTree.

The insert and delete methods are overridden to insert and delete an element and perform rebalancing operations if necessary to ensure that the tree is balanced.

26.3.1 What are the data fields in the AVLTreeNode class? 26.3.2 True or false: AVLTreeNode is a subclass of TreeNode. 26.3.3 True or false: AVLTree is a subclass of BST.

26.4 Overriding the insert Method Inserting an element into an AVL tree is the same as inserting it to a BST, except that the tree may need to be rebalanced.

A new element is always inserted as a leaf node. As a result of adding a new node, the heights of the new leaf node’s ancestors may increase. After inserting a new node, check the nodes along the path from the new leaf node up to the root. If an unbalanced node is found, perform an appropriate rotation using the algorithm in Listing 26.1.

Listing 26.1 Balancing Nodes on a Path 1 balancePath(E e) { 2 Get the path from the node that contains element e to the root, 3 as illustrated in Figure 26.9; 4 for each node A in the path leading to the root { 5 Update the height of A; 6 Let parentOfA denote the parent of A, 7 which is the next node in the path, or null if A is the root; 8 9 switch (balanceFactor(A)) { 10 case −2: if balanceFactor(A.left) == −1 or 0 11 Perform LL rotation; // See Figure 26.3 12 else 13 Perform LR rotation; // See Figure 26.5 14 break; 15 case +2: if balanceFactor(A.right) == +1 or 0 16 Perform RR rotation; // See Figure 26.4 17 else 18 Perform RL rotation; // See Figure 26.6 19 } // End of switch 20 } // End of for 21 } // End of method

The algorithm considers each node in the path from the new leaf node to the root. Update the height of the node on the path. If a node is balanced, no action is needed. If a node is not balanced, perform an appropriate rotation.

26.4.1 For the AVL tree in Figure 26.1a, show the new AVL tree after adding element 40. What rotation do you perform in order to rebalance the tree? Which node was unbalanced?

createNewNode()

Point Check

Point Key

get the path

update node height get parent node

is balanced?

LL rotation

LR rotation

RR rotation

RL rotation

Point Check

M26_LIAN1878_11_GE_C26.indd 1016 12/29/17 4:45 PM

26.5 Implementing Rotations 1017

26.4.2 For the AVL tree in Figure 26.1a, show the new AVL tree after adding element 50. What rotation do you perform in order to rebalance the tree? Which node was unbalanced?

26.4.3 For the AVL tree in Figure 26.1a, show the new AVL tree after adding element 80. What rotation do you perform in order to rebalance the tree? Which node was unbalanced?

26.4.4 For the AVL tree in Figure 26.1a, show the new AVL tree after adding element 89. What rotation do you perform in order to rebalance the tree? Which node was unbalanced?

26.5 Implementing Rotations An unbalanced tree becomes balanced by performing an appropriate rotation operation.

Section 26.2, Rebalancing Trees, illustrated how to perform rotations at a node. Listing 26.2 gives the algorithm for the LL rotation, as illustrated in Figure 26.3.

Listing 26.2 LL Rotation Algorithm 1 balanceLL(TreeNode A, TreeNode parentOfA) { 2 Let B be the left child of A. 3 4 if (A is the root) 5 Let B be the new root 6 else { 7 if (A is a left child of parentOfA) 8 Let B be a left child of parentOfA; 9 else 10 Let B be a right child of parentOfA; 11 } 12 13 Make T2 the left subtree of A by assigning B.right to A.left; 14 Make A the right child of B by assigning A to B.right;

15 Update the height of node A and node B; 16 } // End of method

Note the height of nodes A and B can be changed, but the heights of other nodes in the tree are not changed. You can implement the RR, LR, and RL rotations in a similar manner.

26.5.1 Use Listing 26.2 as a template to describe the algorithms for implementing the RR, LR, and RL rotations.

Point Key

left child of A

reconnect B’s parent

move subtrees

adjust height

Point Check

Figure 26.9 The nodes along the path from the new leaf node may become unbalanced.

New node contains element e

root

parentOfA

M26_LIAN1878_11_GE_C26.indd 1017 12/29/17 4:45 PM

1018 Chapter 26 AVL Trees

26.6 Implementing the delete Method Deleting an element from an AVL tree is the same as deleting it from a BST, except that the tree may need to be rebalanced.

As discussed in Section 25.3, Deleting Elements from a BST, to delete an element from a binary tree, the algorithm first locates the node that contains the element. Let current point to the node that contains the element in the binary tree and parent point to the parent of the current node. The current node may be a left child or a right child of the parent node. Two cases arise when deleting an element.

Case 1: The current node does not have a left child, as shown in Figure 25.10a. To delete the current node, simply connect the parent node with the right child of the current node, as shown in Figure 25.10b.

The height of the nodes along the path from the parent node up to the root may have decreased. To ensure that the tree is balanced, invoke

balancePath(parent.element); // Defined in Listing 26.1

Case 2: The current node has a left child. Let rightMost point to the node that contains the largest element in the left subtree of the current node and parentOfRightMost point to the parent node of the rightMost node, as shown in Figure 25.12a. The rightMost node cannot have a right child, but may have a left child. Replace the element value in the current node with the one in the rightMost node, connect the parentOfRightMost node with the left child of the rightMost node, and delete the rightMost node, as shown in Figure 25.12b.

The height of the nodes along the path from parentOfRightMost up to the root may have decreased. To ensure the tree is balanced, invoke

balancePath(parentOfRightMost); // Defined in Listing 26.1

26.6.1 For the AVL tree in Figure 26.1a, show the new AVL tree after deleting element 107. What rotation do you perform in order to rebalance the tree? Which node was unbalanced?

26.6.2 For the AVL tree in Figure 26.1a, show the new AVL tree after deleting element 60. What rotation do you perform in order to rebalance the tree? Which node was unbalanced?

26.6.3 For the AVL tree in Figure 26.1a, show the new AVL tree after deleting element 55. What rotation did you perform in order to rebalance the tree? Which node was unbalanced?

26.6.4 For the AVL tree in Figure 26.1b, show the new AVL tree after deleting elements 67 and 87. What rotation did you perform in order to rebalance the tree? Which node was unbalanced?

26.7 The AVLTree Class The AVLTree class extends the BST class to override the insert and delete meth- ods to rebalance the tree if necessary.

Listing 26.3 gives the complete source code for the AVLTree class.

Listing 26.3 AVLTree.java 1 public class AVLTree<E extends Comparable<E>> extends BST<E> { 2 /** Create an empty AVL tree */ 3 public AVLTree() {

Point Key

Point Check

Point Key

no-arg constructor

M26_LIAN1878_11_GE_C26.indd 1018 12/29/17 4:45 PM

26.7 The AVLTree Class 1019

4 } 5 6 /** Create an AVL tree from an array of objects */ 7 public AVLTree(E[] objects) { 8 super(objects); 9 } 10 11 @Override /** Override createNewNode to create an AVLTreeNode */ 12 protected AVLTreeNode<E> createNewNode(E e) { 13 return new AVLTreeNode<E>(e); 14 } 15 16 @Override /** Insert an element and rebalance if necessary */ 17 public boolean insert(E e) { 18 boolean successful = super.insert(e); 19 if (!successful) 20 return false; // e is already in the tree 21 else {

22 balancePath(e); // Balance from e to the root if necessary 23 } 24 25 return true; // e is inserted 26 } 27 28 /** Update the height of a specified node */ 29 private void updateHeight(AVLTreeNode<E> node) { 30 if (node.left == null && node.right == null) // node is a leaf 31 node.height = 0; 32 else if (node.left == null) // node has no left subtree 33 node.height = 1 + ((AVLTreeNode<E>)(node.right)).height; 34 else if (node.right == null) // node has no right subtree 35 node.height = 1 + ((AVLTreeNode<E>)(node.left)).height; 36 else 37 node.height = 1 + 38 Math.max(((AVLTreeNode<E>)(node.right)).height, 39 ((AVLTreeNode<E>)(node.left)).height); 40 } 41 42 /** Balance the nodes in the path from the specified 43 * node to the root if necessary 44 */ 45 private void balancePath(E e) { 46 java.util.ArrayList<TreeNode<E>> path = path(e); 47 for (int i = path.size() − 1; i >= 0; i—–) { 48 AVLTreeNode<E> A = (AVLTreeNode<E>)(path.get(i)); 49 updateHeight(A); 50 AVLTreeNode<E> parentOfA = (A == root) ? null : 51 (AVLTreeNode<E>)(path.get(i − 1)); 52 53 switch (balanceFactor(A)) { 54 case −2: 55 if (balanceFactor((AVLTreeNode<E>)A.left) <= 0) { 56 balanceLL(A, parentOfA); // Perform LL rotation 57 } 58 else { 59 balanceLR(A, parentOfA); // Perform LR rotation 60 } 61 break;

constructor

create AVL tree node

override insert

balance tree

update node height

balance nodes get path

consider a node update height get height

left-heavy

LL rotation

LR rotation

M26_LIAN1878_11_GE_C26.indd 1019 12/29/17 4:45 PM

1020 Chapter 26 AVL Trees

62 case +2: 63 if (balanceFactor((AVLTreeNode<E>)A.right) >= 0) { 64 balanceRR(A, parentOfA); // Perform RR rotation 65 } 66 else { 67 balanceRL(A, parentOfA); // Perform RL rotation 68 } 69 } 70 } 71 } 72 73 /** Return the balance factor of the node */ 74 private int balanceFactor(AVLTreeNode<E> node) { 75 if (node.right == null) // node has no right subtree 76 return −node.height; 77 else if (node.left == null) // node has no left subtree 78 return +node.height; 79 else 80 return ((AVLTreeNode<E>)node.right).height − 81 ((AVLTreeNode<E>)node.left).height; 82 } 83 84 /** Balance LL (see Figure 26.3) */ 85 private void balanceLL(TreeNode<E> A, TreeNode<E> parentOfA) { 86 TreeNode<E> B = A.left; // A is left-heavy and B is left−heavy 87 88 if (A == root) { 89 root = B; 90 } 91 else { 92 if (parentOfA.left == A) { 93 parentOfA.left = B; 94 } 95 else { 96 parentOfA.right = B; 97 } 98 } 99 100 A.left = B.right; // Make T2 the left subtree of A 101 B.right = A; // Make A the left child of B 102 updateHeight((AVLTreeNode<E>)A); 103 updateHeight((AVLTreeNode<E>)B); 104 } 105 106 /** Balance LR (see Figure 26.5) */ 107 private void balanceLR(TreeNode<E> A, TreeNode<E> parentOfA) { 108 TreeNode<E> B = A.left; // A is left−heavy 109 TreeNode<E> C = B.right; // B is right−heavy 110 111 if (A == root) { 112 root = C; 113 } 114 else { 115 if (parentOfA.left == A) { 116 parentOfA.left = C; 117 } 118 else { 119 parentOfA.right = C;

right-heavy

RR rotation

RL rotation

get balance factor

LL rotation

update height

LR rotation

M26_LIAN1878_11_GE_C26.indd 1020 12/29/17 4:45 PM

120 } 121 } 122 123 A.left = C.right; // Make T3 the left subtree of A 124 B.right = C.left; // Make T2 the right subtree of B 125 C.left = B; 126 C.right = A; 127 128 // Adjust heights 129 updateHeight((AVLTreeNode<E>)A); 130 updateHeight((AVLTreeNode<E>)B); 131 updateHeight((AVLTreeNode<E>)C); 132 } 133 134 /** Balance RR (see Figure 26.4) */ 135 private void balanceRR(TreeNode<E> A, TreeNode<E> parentOfA) { 136 TreeNode<E> B = A.right; // A is right-heavy and B is right-heavy 137 138 if (A == root) { 139 root = B; 140 } 141 else { 142 if (parentOfA.left == A) { 143 parentOfA.left = B; 144 } 145 else { 146 parentOfA.right = B; 147 } 148 } 149 150 A.right = B.left; // Make T2 the right subtree of A 151 B.left = A; 152 updateHeight((AVLTreeNode<E>)A); 153 updateHeight((AVLTreeNode<E>)B); 154 } 155 156 /** Balance RL (see Figure 26.6) */ 157 private void balanceRL(TreeNode<E> A, TreeNode<E> parentOfA) { 158 TreeNode<E> B = A.right; // A is right-heavy 159 TreeNode<E> C = B.left; // B is left-heavy 160 161 if (A == root) { 162 root = C; 163 } 164 else { 165 if (parentOfA.left == A) { 166 parentOfA.left = C; 167 } 168 else { 169 parentOfA.right = C; 170 } 171 } 172 173 A.right = C.left; // Make T2 the right subtree of A 174 B.left = C.right; // Make T3 the left subtree of B 175 C.left = A; 176 C.right = B; 177

26.7 The AVLTree Class 1021

update height

RR rotation

update height

RL rotation

M26_LIAN1878_11_GE_C26.indd 1021 12/29/17 4:45 PM

1022 Chapter 26 AVL Trees

178 // Adjust heights 179 updateHeight((AVLTreeNode<E>)A); 180 updateHeight((AVLTreeNode<E>)B); 181 updateHeight((AVLTreeNode<E>)C); 182 } 183 184 @Override /** Delete an element from the AVL tree. 185 * Return true if the element is deleted successfully 186 * Return false if the element is not in the tree */ 187 public boolean delete(E element) { 188 if (root == null) 189 return false; // Element is not in the tree 190 191 // Locate the node to be deleted and also locate its parent node 192 TreeNode<E> parent = null; 193 TreeNode<E> current = root; 194 while (current != null) { 195 if (element.compareTo(current.element) < 0) { 196 parent = current; 197 current = current.left; 198 } 199 else if (element.compareTo(current.element) > 0) { 200 parent = current; 201 current = current.right; 202 } 203 else 204 break; // Element is in the tree pointed by current 205 } 206 207 if (current == null) 208 return false; // Element is not in the tree 209 210 // Case 1: current has no left children (see Figure 25.10) 211 if (current.left == null) { 212 // Connect the parent with the right child of the current node 213 if (parent == null) { 214 root = current.right; 215 } 216 else { 217 if (element.compareTo(parent.element) < 0) 218 parent.left = current.right; 219 else 220 parent.right = current.right; 221 222 // Balance the tree if necessary 223 balancePath(parent.element); 224 } 225 } 226 else { 227 // Case 2: The current node has a left child 228 // Locate the rightmost node in the left subtree of 229 // the current node and also its parent 230 TreeNode<E> parentOfRightMost = current; 231 TreeNode<E> rightMost = current.left; 232 233 while (rightMost.right != null) { 234 parentOfRightMost = rightMost; 235 rightMost = rightMost.right; // Keep going to the right 236 }

update height

override delete

balance nodes

M26_LIAN1878_11_GE_C26.indd 1022 12/29/17 4:45 PM

237 238 // Replace the element in current by the element in rightMost 239 current.element = rightMost.element; 240 241 // Eliminate rightmost node 242 if (parentOfRightMost.right == rightMost) 243 parentOfRightMost.right = rightMost.left; 244 else 245 // Special case: parentOfRightMost is current 246 parentOfRightMost.left = rightMost.left; 247 248 // Balance the tree if necessary 249 balancePath(parentOfRightMost.element); 250 } 251 252 size—–; 253 return true; // Element inserted 254 } 255 256 /** AVLTreeNode is TreeNode plus height */ 257 protected static class AVLTreeNode<E> extends BST.TreeNode<E> { 258 protected int height = 0; // New data field 259 260 public AVLTreeNode(E e) { 261 super(e); 262 } 263 } 264 }

The AVLTree class extends BST. Like the BST class, the AVLTree class has a no-arg con- structor that constructs an empty AVLTree (lines 3 and 4) and a constructor that creates an initial AVLTree from an array of elements (lines 7–9).

The createNewNode() method defined in the BST class creates a TreeNode. This method is overridden to return an AVLTreeNode (lines 12–14).

The insert method in AVLTree is overridden in lines 17–26. The method first invokes the insert method in BST, then invokes balancePath(e) (line 22) to ensure that the tree is balanced.

The balancePath method first gets the nodes on the path from the node that contains element e to the root (line 46). For each node in the path, update its height (line 49), check its balance factor (line 53), and perform appropriate rotations if necessary (lines 53–69).

Four methods for performing rotations are defined in lines 85–182. Each method is invoked with two TreeNode arguments—A and parentOfA—to perform an appropriate rotation at node A. How each rotation is performed is illustrated in Figures 26.3–26.6. After the rotation, the heights of nodes A, B, and C are updated (lines 102, 129, 152, and 179).

The delete method in AVLTree is overridden in lines 187–254. The method is the same as the one implemented in the BST class, except that you have to rebalance the nodes after deletion in two cases (lines 223, 249).

26.7.1 Why is the createNewNode method defined protected? When is it invoked? 26.7.2 When is the updateHeight method invoked? When is the balanceFactor

method invoked? When is the balancePath method invoked? Will the program work if you replace the break in line 61 in the AVLTree class with a return and add a return at line 69?

26.7.3 What are data fields in the AVLTree class? 26.7.4 In the insert and delete methods, once you have performed a rotation to

balance a node in the tree, is it possible there are still unbalanced nodes?

balance nodes

inner AVLTreeNode class

node height

constructors

insert

balancePath

rotations

delete

Point Check

26.7 The AVLTree Class 1023

M26_LIAN1878_11_GE_C26.indd 1023 12/29/17 4:45 PM

1024 Chapter 26 AVL Trees

26.8 Testing the AVLTree Class This section gives an example of using the AVLTree class.

Listing 26.4 gives a test program. The program creates an AVLTree initialized with an array of the integers 25, 20, and 5 (lines 4 and 5), inserts elements in lines 9–18, and deletes elements in lines 22–28. Since AVLTree is a subclass of BST and the elements in a BST are iterable, the program uses a foreach loop to traverse all the elements in lines 33–35.

Listing 26.4 TestAVLTree.java 1 public class TestAVLTree { 2 public static void main(String[] args) { 3 // Create an AVL tree 4 AVLTree<Integer> tree = new AVLTree<Integer>(new Integer[]{25, 5 20, 5}); 6 System.out.print("After inserting 25, 20, 5:"); 7 printTree(tree); 8 9 tree.insert(34); 10 tree.insert(50); 11 System.out.print("\nAfter inserting 34, 50:"); 12 printTree(tree); 13 14 tree.insert(30); 15 System.out.print("\nAfter inserting 30"); 16 printTree(tree); 17 18 tree.insert(10); 19 System.out.print("\nAfter inserting 10"); 20 printTree(tree); 21 22 tree.delete(34); 23 tree.delete(30); 24 tree.delete(50); 25 System.out.print("\nAfter removing 34, 30, 50:"); 26 printTree(tree); 27 28 tree.delete(5); 29 System.out.print("\nAfter removing 5:"); 30 printTree(tree); 31 32 System.out.print("\nTraverse the elements in the tree: "); 33 for (int e: tree) { 34 System.out.print(e + " "); 35 } 36 } 37 38 public static void printTree(BST tree) { 39 // Traverse tree 40 System.out.print("\nInorder (sorted): "); 41 tree.inorder(); 42 System.out.print("\nPostorder: "); 43 tree.postorder(); 44 System.out.print("\nPreorder: "); 45 tree.preorder(); 46 System.out.print("\nThe number of nodes is " + tree.getSize());

Point Key

create an AVLTree

insert 34 insert 50

insert 30

insert 10

delete 34 delete 30 delete 50

delete 5

foreach loop

M26_LIAN1878_11_GE_C26.indd 1024 12/29/17 4:45 PM

26.8 Testing the AVLTree Class 1025

47 System.out.println(); 48 } 49 }

After inserting 25, 20, 5: Inorder (sorted): 5 20 25 Postorder: 5 25 20 Preorder: 20 5 25 The number of nodes is 3

After inserting 34, 50: Inorder (sorted): 5 20 25 34 50 Postorder: 5 25 50 34 20 Preorder: 20 5 34 25 50 The number of nodes is 5

After inserting 30 Inorder (sorted): 5 20 25 30 34 50 Postorder: 5 20 30 50 34 25 Preorder: 25 20 5 34 30 50 The number of nodes is 6

After inserting 10 Inorder (sorted): 5 10 20 25 30 34 50 Postorder: 5 20 10 30 50 34 25 Preorder: 25 10 5 20 34 30 50 The number of nodes is 7

After removing 34, 30, 50: Inorder (sorted): 5 10 20 25 Postorder: 5 20 25 10 Preorder: 10 5 25 20 The number of nodes is 4

After removing 5: Inorder (sorted): 10 20 25 Postorder: 10 25 20 Preorder: 20 10 25 The number of nodes is 3 Traverse the elements in the tree: 10 20 25

Figure 26.10 shows how the tree evolves as elements are added to the tree. After 25 and 20 are added, the tree is as shown in Figure 26.10a. 5 is inserted as a left child of 20, as shown in Figure 26.10b. The tree is not balanced. It is left-heavy at node 25. Perform an LL rotation to result in an AVL tree as shown in Figure 26.10c.

After inserting 34, the tree is as shown in Figure 26.10d. After inserting 50, the tree is as shown in Figure 26.10e. The tree is not balanced. It is right-heavy at node 25. Perform an RR rotation to result in an AVL tree as shown in Figure 26.10f.

After inserting 30, the tree is as shown in Figure 26.10g. The tree is not balanced. Perform an RL rotation to result in an AVL tree as shown in Figure 26.10h.

After inserting 10, the tree is as shown in Figure 26.10i. The tree is not balanced. Perform an LR rotation to result in an AVL tree as shown in Figure 26.10j.

M26_LIAN1878_11_GE_C26.indd 1025 12/29/17 4:45 PM

1026 Chapter 26 AVL Trees

Figure 26.10 The tree evolves as new elements are inserted.

Need LL rotation at node 25

Need RR rotation at node 25

RL rotation at node 20

LR rotation at node 20

20 20

25 5

25 50

30 50 5

30 5020

30 50

25 50

(a) Insert 25, 20

(e) Insert 50 (f) Balanced (g) Insert 30

(h) Balanced (i) Insert 10 (j) Balanced

(b) Insert 5 (c) Balanced (d) Insert 34

Figure 26.11a shows how the tree evolves as elements are deleted. After deleting 34, 30, and 50, the tree is as shown in Figure 26.11b. The tree is not balanced. Perform an LL rotation to result in an AVL tree as shown in Figure 26.11c.

After deleting 5, the tree is as shown in Figure 26.11d. The tree is not balanced. Perform an RL rotation to result in an AVL tree as shown in Figure 26.11e.

26.8.1 Show the change of an AVL tree when inserting 1, 2, 3, 4, 10, 9, 7, 5, 8, 6 into the tree, in this order.

26.8.2 For the tree built in the preceding question, show its change after 1, 2, 3, 4, 10, 9, 7, 5, 8, 6 are deleted from the tree in this order.

26.8.3 Can you traverse the elements in an AVL tree using a foreach loop?

Point Check

M26_LIAN1878_11_GE_C26.indd 1026 12/29/17 4:45 PM

26.9 AVL Tree Time Complexity Analysis 1027

26.9 AVL Tree Time Complexity Analysis Since the height of an AVL tree is O(log n), the time complexity of the search, insert, and delete methods in AVLTree is O(log n).

The time complexity of the search, insert, and delete methods in AVLTree depends on the height of the tree. We can prove that the height of the tree is O(log n).

Let G(h) denote the minimum number of nodes in an AVL tree with height h. Obviously, G(1) is 1 and G(2) is 2. The minimum number of nodes in an AVL tree with height h Ú 3 must have two minimum subtrees: one with height h - 1 and the other with height h - 2. Thus,

G(h) = G(h - 1) + G(h - 2) + 1

Recall that a Fibonacci number at index i can be described using the recurrence relation F(i) = F(i - 1) + F(i - 2). Therefore, the function G(h) is essentially the same as F(i). It can be proven that

h 6 1.4405 log(n + 2) - 1.3277

where n is the number of nodes in the tree. Hence, the height of an AVL tree is O(log n). The search, insert, and delete methods involve only the nodes along a path in the tree.

The updateHeight and balanceFactor methods are executed in a constant time for each node in the path. The balancePath method is executed in a constant time for a node in the path. Thus, the time complexity for the search, insert, and delete methods is O(log n).

26.9.1 What is the maximum/minimum height for an AVL tree of 3 nodes, 5 nodes, and 7 nodes?

26.9.2 If an AVL tree has a height of 3, what maximum number of nodes can the tree have? What minimum number of nodes can the tree have?

26.9.3 If an AVL tree has a height of 4, what maximum number of nodes can the tree have? What minimum number of nodes can the tree have?

Point Key

tree height

Point Check

Figure 26.11 The tree evolves as elements are deleted from the tree.

30 5020

205

(a) Delete 34, 30, 50 (b) After 34, 30, 50 are deleted

(d) After 5 is deleted (e) Balanced

LL rotation at node 25

RL rotation at 10 10

M26_LIAN1878_11_GE_C26.indd 1027 12/29/17 4:45 PM

1028 Chapter 26 AVL Trees

ChapTer summary 1. An AVL tree is a well-balanced binary tree. In an AVL tree, the difference between the

heights of two subtrees for every node is 0 or 1.

2. The process for inserting or deleting an element in an AVL tree is the same as in a binary search tree. The difference is that you may have to rebalance the tree after an insertion or deletion operation.

3. Imbalances in the tree caused by insertions and deletions are rebalanced through subtree rotations at the node of the imbalance.

4. The process of rebalancing a node is called a rotation. There are four possible rotations: LL rotation, LR rotation, RR rotation, and RL rotation.

5. The height of an AVL tree is O(log n). Therefore, the time complexities for the search, insert, and delete methods are O(log n).

Quiz Answer the quiz for this chapter online at the book Companion Website.

programming exerCises *26.1 (Display AVL tree graphically) Write a program that displays an AVL tree along

with its balance factor for each node.

26.2 (Compare performance) Write a test program that randomly generates 600,000 numbers and inserts them into a BST, reshuffles the 600,000 numbers and per- forms a search, and reshuffles the numbers again before deleting them from the tree. Write another test program that does the same thing for an AVLTree. Compare the execution times of these two programs.

***26.3 (AVL tree animation) Write a program that animates the AVL tree insert, delete, and search methods, as shown in Figure 26.2.

**26.4 (Parent reference for BST) Suppose the TreeNode class defined in BST con- tains a reference to the node’s parent, as shown in Programming Exercise 25.15. Implement the AVLTree class to support this change. Write a test program that adds numbers 1, 2, . . . , 100 to the tree and displays the paths for all leaf nodes.

**26.5 (The kth smallest element) You can find the kth smallest element in a BST in O(n) time from an inorder iterator. For an AVL tree, you can find it in O(log n) time. To achieve this, add a new data field named size in AVLTreeNode to store the number of nodes in the subtree rooted at this node. Note the size of a

Key terms AVL tree 1012 balance factor 1012 left-heavy 1012 LL rotation 1012 LR rotation 1013 perfectly balanced tree 1012

right-heavy 1012 RL rotation 1013 rotation 1012 RR rotation 1012 well-balanced tree 1012

M26_LIAN1878_11_GE_C26.indd 1028 12/29/17 4:45 PM

node v is one more than the sum of the sizes of its two children. Figure 26.12 shows an AVL tree and the size value for each node in the tree.

Figure 26.12 The size data field in AVLTreeNode stores the number of nodes in the subtree rooted at the node:

30 50

size: 6

size: 1

size: 2

size: 1 size: 1

size: 3

In the AVLTree class, add the following method to return the kth smallest element in the tree:

public E find(int k)

The method returns null if k < 1 or k > the size of the tree. This method can be implemented using the recursive method find(k, root), which returns the kth smallest element in the tree with the specified root. Let A and B be the left and right children of the root, respectively. Assuming the tree is not empty and k … root.size, find(k, root) can be recursively defined as follows:

find(k, root) = E root.element, if A is null and k is 1;B.element, if A is null and k is 2;find(k, A), if k 6 = A.size; root.element, if k = A.size + 1; find(k - A.size - 1, B), if k 7 A.size + 1;

Modify the insert and delete methods in AVLTree to set the correct value for the size property in each node. The insert and delete methods will still be in O(log n) time. The find(k) method can be implemented in O(log n) time. Therefore, you can find the kth smallest element in an AVL tree in O(log n) time.

Test your program using the code at

liveexample.pearsoncmg.com/test/Exercise26_05Test.txt.

**26.6 (Test AVLTree) Design and write a complete test program to test if the AVLTree class in Listing 26.3 meets all requirements.

Programming Exercises 1029

M26_LIAN1878_11_GE_C26.indd 1029 12/29/17 4:45 PM

A01_GIPE3898_09_SE_FM.indd 3 12/20/17 5:56 PM

This page intentionally left blank

Objectives ■■ To understand what hashing is and for what hashing is used (§27.2).

■■ To obtain the hash code for an object and design the hash function to map a key to an index (§27.3).

■■ To handle collisions using open addressing (§27.4).

■■ To know the differences among linear probing, quadratic probing, and double hashing (§27.4).

■■ To handle collisions using separate chaining (§27.5).

■■ To understand the load factor and the need for rehashing (§27.6).

■■ To implement MyHashMap using hashing (§27.7).

■■ To implement MyHashSet using hashing (§27.8).

Hashing

Chapter

M27_LIAN1878_11_GE_C27.indd 1031 12/29/17 4:47 PM

1032 Chapter 27 Hashing

27.1 Introduction Hashing is superefficient. It takes O(1) time to search, insert, and delete an element using hashing.

The preceding chapter introduced binary search trees. An element can be found in O(log n) time in a well-balanced search tree. Is there a more efficient way to search for an element in a container? This chapter introduces a technique called hashing. You can use hashing to imple- ment a map or a set to search, insert, and delete an element in O(1) time.

27.2 What Is Hashing? Hashing uses a hashing function to map a key to an index.

Before introducing hashing, let us review map, which is a data structure that is implemented using hashing. Recall that a map (introduced in Section 21.5) is a container object that stores entries. Each entry contains two parts: a key and a value. The key, also called a search key, is used to search for the corresponding value. For example, a dictionary can be stored in a map, in which the words are the keys and the definitions of the words are the values.

Note A map is also called a dictionary, a hash table, or an associative array.

The Java Collections Framework defines the java.util.Map interface for modeling maps. Three concrete implementations are java.util.HashMap, java.util.LinkedHashMap, and java.util.TreeMap. java.util.HashMap is implemented using hashing, java. util.LinkedHashMap using LinkedList, and java.util.TreeMap using red–black trees. (Bonus Chapter 41 will introduce red–black trees.) You will learn the concept of hashing and use it to implement a hash map in this chapter.

If you know the index of an element in the array, you can retrieve the element using the index in O(1) time. So does that mean we can store the values in an array and use the key as the index to find the value? The answer is yes—if you can map a key to an index. The array that stores the values is called a hash table. The function that maps a key to an index in the hash table is called a hash function. As shown in Figure 27.1, a hash function obtains an index from a key and uses the index to retrieve the value for the key. Hashing is a technique that retrieves the value using the index obtained from the key without performing a search.

Point Key

why hashing?

Point Key

map key value

dictionary hash table associative array

hash table hash function hashing

Figure 27.1 A hash function maps a key to an index in the hash table.

Hash function

i = hash(key)

N – 1

key value An entry

How do you design a hash function that produces an index from a key? Ideally, we would like to design a function that maps each search key to a different index in the hash table. Such a function is called a perfect hash function. However, it is difficult to find a perfect hash perfect hash function

M27_LIAN1878_11_GE_C27.indd 1032 12/29/17 4:47 PM

27.3 Hash Functions and Hash Codes 1033

function. When two or more keys are mapped to the same hash value, we say a collision has occurred. Although there are ways to deal with collisions, which will be are discussed later in this chapter, it is better to avoid collisions in the first place. Thus, you should design a fast and easy-to-compute hash function that minimizes collisions.

27.2.1 What is a hash function? What is a perfect hash function? What is a collision?

27.3 Hash Functions and Hash Codes A typical hash function first converts a search key to an integer value called a hash code, then compresses the hash code into an index to the hash table.

Java’s root class Object has the hashCode method, which returns an integer hash code. By default, the method returns the memory address for the object. The general contract for the hashCode method is as follows:

1. You should override the hashCode method whenever the equals method is overridden to ensure two equal objects return the same hash code.

2. During the execution of a program, invoking the hashCode method multiple times returns the same integer, provided that the object’s data are not changed.

3. Two unequal objects may have the same hash code, but you should implement the hashCode method to avoid too many such cases.

27.3.1 Hash Codes for Primitive Types For search keys of the type byte, short, int, and char, simply cast them to int. Therefore, two different search keys of any one of these types will have different hash codes.

For a search key of the type float, use Float.floatToIntBits(key) as the hash code. Note floatToIntBits(float f) returns an int value whose bit representation is the same as the bit representation for the floating number f. Thus, two different search keys of the float type will have different hash codes.

For a search key of the type long, simply casting it to int would not be a good choice, because all keys that differ in only the first 32 bits will have the same hash code. To take the first 32 bits into consideration, divide the 64 bits into two halves and perform the exclusive- or operation to combine the two halves. This process is called folding. The hash code for a long key is

int hashCode = (int)(key ^ (key >> 32));

Note >> is the right-shift operator that shifts the bits 32 positions to the right. For example, 1010110 >> 2 yields 0010101. The ^ is the bitwise exclusive-or operator. It operates on two corresponding bits of the binary operands. For example, 1010110 ^ 0110111 yields 1100001. For more on bitwise operations, see Appendix G, Bitwise Operations.

For a search key of the type double, first convert it to a long value using the Double. doubleToLongBits method, then perform a folding as follows:

long bits = Double.doubleToLongBits(key); int hashCode = (int)(bits ^ (bits >> 32));

27.3.2 Hash Codes for Strings Search keys are often strings, so it is important to design a good hash function for strings. An intuitive approach is to sum the Unicode of all characters as the hash code for the string. This approach may work if two search keys in an application don’t contain the same letters,

collision

Point Check

Point Key

hash code hashCode()

byte, short, int, char

float

long

folding

double

folding

M27_LIAN1878_11_GE_C27.indd 1033 12/29/17 4:47 PM

1034 Chapter 27 Hashing

but it will produce a lot of collisions if the search keys contain the same letters, such as tod and dot.

A better approach is to generate a hash code that takes the position of characters into con- sideration. Specifically, let the hash code be

s0 * b (n - 1) + s1 * b(n - 2) + g + sn - 1

where si is s.charAt(i). This expression is a polynomial for some positive b, so this is called a polynomial hash code. Using Horner’s rule for polynomial evaluation (see Section 6.7), the hash code can be calculated efficiently as follows:

( g ((s0 * b + s1) * b + s2) * b + g + sn - 2) * b + sn - 1 This computation can cause an overflow for long strings, but arithmetic overflow is ignored in Java. You should choose an appropriate value b to minimize collisions. Experiments show that good choices for b are 31, 33, 37, 39, and 41. In the String class, the hashCode is overridden using the polynomial hash code with b being 31.

27.3.3 Compressing Hash Codes The hash code for a key can be a large integer that is out of the range for the hash-table index, so you need to scale it down to fit in the index’s range. Assume the index for a hash table is between 0 and N-1. The most common way to scale an integer to between 0 and N−1 is to use

index = hashCode % N;

Ideally, you should choose a prime number for N to ensure the indices are spread evenly. However, it is time consuming to find a large prime number. In the Java API implementation for java.util.HashMap, N is set to an integer power of 2. There is a good reason for this choice. When N is an integer power of 2, you can use the & operator to compress a hash code to an index on the hash table as follows:

index = hashCode & (N − 1);

index will be between 0 and N − 1. The ampersand, &, is a bitwise AND operator (see Appendix G, Bitwise Operations). The AND of two corresponding bits yields a 1 if both bits are 1. For example, assume N = 4 and hashCode = 11. Thus, 11 & (4 − 1) = 1011 & 0011 = 0011.

To ensure the hashing is evenly distributed, a supplemental hash function is also used along with the primary hash function in the implementation of java.util.HashMap. This function is defined as:

private static int supplementalHash(int h) { h ^= (h >>> 20) ^ (h >>> 12); return h ^ (h >>> 7) ^ (h >>> 4); }

>>> is unsigned right-shift operations (also introduced in Appendix G). The bitwise operations are much faster than the multiplication, division, and remainder operations. You should replace these operations with the bitwise operations whenever possible.

The complete hash function is defined as:

h(hashCode) = supplementalHash(hashCode) & (N – 1)

The supplemental hash function helps avoid collisions for two numbers with the same lower bits. For example, both 11100101 & 00000111 and 11001101 & 00000111 yield 00000111. But supplementalHash(11100101) & 00000111 and supplemental- Hash(11001101) & 00000111 will be different. Using a supplemental function reduces this type of collision.

polynomial hash code

M27_LIAN1878_11_GE_C27.indd 1034 12/29/17 4:47 PM

27.4 Handling Collisions Using Open Addressing 1035

Note In Java, an int is a 32-bit signed integer. The hashCode() method returns an int and it may be negative. If a hash code is negative, hashCode % N would be negative. But hashCode & (N – 1) will be non-negative because the maximum hash-table size in Java is limited to 230, which is positive in a 32-bit int.anyInt & aNonNegativeInt will always be non-negative.

27.3.1 What is a hash code? What is the hash code for Byte, Short, Integer, and Character? 27.3.2 How is the hash code for a Float object computed? 27.3.3 How is the hash code for a Long object computed? 27.3.4 How is the hash code for a Double object computed? 27.3.5 How is the hash code for a String object computed? 27.3.6 How is a hash code compressed to an integer representing the index in a hash table? 27.3.7 If N is an integer power of the power of 2, is N / 2 same as N >> 1? 27.3.8 If N is an integer power of the power of 2, is m % N same as m & (N – 1) for a positive

integer m?

27.3.9 What is new Integer("–98").hashCode() and what is "ABCDEFGHIJK." hashCode()?

27.4 Handling Collisions Using Open Addressing A collision occurs when two keys are mapped to the same index in a hash table. Gener- ally, there are two ways for handling collisions: open addressing and separate chaining.

Open addressing is the process of finding an open location in the hash table in the event of a collision. Open addressing has several variations: linear probing, quadratic probing, and double hashing.

27.4.1 Linear Probing When a collision occurs during the insertion of an entry to a hash table, linear probing finds the next available location sequentially. For example, if a collision occurs at hashTable[k % N], check whether hashTable[(k+1) % N] is available. If not, check hashTable[(k+2) % N] and so on, until an available cell is found, as shown in Figure 27.2.

Note When probing reaches the end of the table, it goes back to the beginning of the table. Thus, the hash table is treated as if it were circular.

Point Check

Point Key

open addressing

add entry linear probing

circular hash table

Figure 27.2 Linear probing finds the next available location sequentially.

3 4

key: 44

key: 4

key: 16

key: 28

key: 21

For simplicity, only the keys are shown and the values are not shown. Here N is 11 and index = key % N.

New element with key 26 to be inserted

Probe 3 times before „nding an empty

cell

M27_LIAN1878_11_GE_C27.indd 1035 12/29/17 4:47 PM

1036 Chapter 27 Hashing

To search for an entry in the hash table, obtain the index, say k, from the hash function for the key. Check whether hashTable[k % N] contains the entry. If not, check whether hashTable[(k+1) % N] contains the entry, and so on, until it is found, or an empty cell is reached.

To remove an entry from the hash table, search the entry that matches the key. If the entry is found, place a special marker to denote that the entry is available. Each cell in the hash table has three possible states: occupied, marked, or empty. Note a marked cell is also available for insertion.

Linear probing tends to cause groups of consecutive cells in the hash table to be occupied. Each group is called a cluster. Each cluster is actually a probe sequence that you must search when retrieving, adding, or removing an entry. As clusters grow in size, they may merge into even larger clusters, further slowing down the search time. This is a big disadvantage of linear probing.

Pedagogical Note For an interactive GUI demo to see how linear probing works, go to http://liveexample. pearsoncmg.com/dsanimation/LinearProbingeBook.html, as shown in Figure 27.3.

27.4.2 Quadratic Probing Quadratic probing can avoid the clustering problem that can occur in linear probing. Linear probing looks at the consecutive cells beginning at index k. Quadratic probing, on the other hand, looks at the cells at indices (k + j2) , N, for j Ú 0, that is, k , N, (k + 1) , N, (k + 4) , n, (k + 9) , N, and so on, as shown in Figure 27.4.

search entry

remove entry

cluster

linear probing animation on Companion Website

quadratic probing

Figure 27.3 The animation tool shows how linear probing works.

M27_LIAN1878_11_GE_C27.indd 1036 12/29/17 4:47 PM

27.4 Handling Collisions Using Open Addressing 1037

Figure 27.4 Quadratic probing increases the next index in the sequence by j2 for j = 1, 2, 3, c.

3 4

key: 44

key: 4

key: 16

key: 28

key: 21

For simplicity, only the keys are shown and not the values. Here N is 11 and index = key % N.

New element with key 26 to be inserted

Quadratic probe 2 times before �nding

an empty cell

Quadratic probing works in the same way as linear probing except for a change in the search sequence. Quadratic probing avoids linear probing’s clustering problem, but it has its own clustering problem, called secondary clustering; that is, the entries that collide with an occu- pied entry use the same probe sequence.

Linear probing guarantees that an available cell can be found for insertion as long as the table is not full. However, there is no such guarantee for quadratic probing.

Pedagogical Note For an interactive GUI demo to see how quadratic probing works, go to http://liveexample .pearsoncmg.com/dsanimation/QuadraticProbingeBook.html, as shown in Figure 27.5.

27.4.3 Double Hashing Another open addressing scheme that avoids the clustering problem is known as double hashing. Starting from the initial index k, both linear probing and quadratic probing add an increment to k to define a search sequence. The increment is 1 for linear probing and j2 for quadratic probing. These increments are independent of the keys. Double hashing uses a secondary hash function h′(key) on the keys to determine the increments to avoid the clustering problem. Specifically, double hashing looks at the cells at indices (k + j * h′(key)) , N, for j Ú 0, that is, k , N, (k + h′(key)) , N, (k + 2 * h′(key)) , N, (k + 3 * h′(key)) , N, and so on.

For example, let the primary hash function h and secondary hash function h' on a hash table of size 11 be defined as follows:

h(key) = key % 11; h'(key) = 7 – key % 7;

For a search key of 12, we have

h(12) = 12 % 11 = 1; h'(12) = 7 – 12 % 7 = 2;

Suppose the elements with the keys 45, 58, 4, 28, and 21 are already placed in the hash table as shown in Figure 27.6. We now insert the element with key 12. The probe sequence for key 12 starts at index 1. Since the cell at index 1 is already occupied, search the next cell at index 3 (1 + 1 * 2). Since the cell at index 3 is already occupied, search the next cell at index 5 (1 + 2 * 2). Since the cell at index 5 is empty, the element for key 12 is now inserted at this cell.

The indices of the probe sequence are as follows: 1, 3, 5, 7, 9, 0, 2, 4, 6, 8, 10. This sequence reaches the entire table. You should design your functions to produce a probe sequence that

secondary clustering

quadratic probing animation on Companion Website

double hashing

M27_LIAN1878_11_GE_C27.indd 1037 12/29/17 4:47 PM

1038 Chapter 27 Hashing

reaches the entire table. Note the second function should never have a zero value, since zero is not an increment.

27.4.1 What is open addressing? What is linear probing? What is quadratic probing? What is double hashing?

27.4.2 Describe the clustering problem for linear probing. Point

Check

Figure 27.6 The secondary hash function in a double hashing determines the increment of the next index in the probe sequence.

key: 45

key: 4

key: 58

key: 28

key: 21

h(12)

key: 45

key: 4

key: 58

key: 28

key: 21

h(12) + h'(12)

key: 45

key: 4

key: 58

key: 28

key: 21

h(12) + 2*h'(12)

. . .

Figure 27.5 The animation tool shows how quadratic probing works.

M27_LIAN1878_11_GE_C27.indd 1038 12/29/17 4:47 PM

27.6 Load Factor and Rehashing 1039

27.4.3 What is secondary clustering? 27.4.4 Show the hash table of size 11 after inserting entries with keys 34, 29, 53, 44, 120,

39, 45, and 40, using linear probing.

27.4.5 Show the hash table of size 11 after inserting entries with keys 34, 29, 53, 44, 120, 39, 45, and 40, using quadratic probing.

27.4.6 Show the hash table of size 11 after inserting entries with keys 34, 29, 53, 44, 120, 39, 45, and 40, using double hashing with the following functions:

h(k) = k % 11; h'(k) = 7 – k % 7;

27.5 Handling Collisions Using Separate Chaining The separate chaining scheme places all entries with the same hash index in the same location, rather than finding new locations. Each location in the separate chaining scheme uses a bucket to hold multiple entries.

You can implement a bucket using an array, ArrayList, or LinkedList. We will use LinkedList for demonstration. You can view each cell in the hash table as the reference to the head of a linked list, and elements in the linked list are chained starting from the head, as shown in Figure 27.7.

Point Key

separate chaining

implementing bucket

Figure 27.7 Separate chaining scheme chains the entries with the same hash index in a bucket.

3 4

key: 44

key: 28

key: 21

key: 26key: 4

key: 16

New element with key 26 to be inserted

For simplicity, only the keys are shown, and not the values. Here N is 11 and index = key % N.

27.5.1 Show the hash table of size 11 after inserting entries with the keys 34, 29, 53, 44, 120, 39, 45, and 40, using separate chaining.

27.6 Load Factor and Rehashing The load factor measures how full a hash table is. If the load factor is exceeded, increase the hash-table size and reload the entries into a new larger hash table. This is called rehashing.

Load factor l (lambda) measures how full a hash table is. It is the ratio of the number of ele- ments to the size of the hash table, that is, l =

n N

, where n denotes the number of elements and N the number of locations in the hash table.

Note l is zero if the hash table is empty. For the open addressing scheme, l is between 0 and 1; l is 1 if the hash table is full. For the separate chaining scheme, l can be any value. As

Point Check

Point Key

rehashing

load factor

M27_LIAN1878_11_GE_C27.indd 1039 12/29/17 4:47 PM

1040 Chapter 27 Hashing

l increases, the probability of a collision also increases. Studies show you should maintain the load factor under 0.5 for the open addressing scheme, and under 0.9 for the separate chaining scheme.

Keeping the load factor under a certain threshold is important for the performance of hash- ing. In the implementation of the java.util.HashMap class in the Java API, the threshold 0.75 is used. Whenever the load factor exceeds the threshold, you need to increase the hash- table size and rehash all the entries in the map into a new larger hash table. Notice you need to change the hash functions, since the hash-table size has been changed. To reduce the likeli- hood of rehashing, since it is costly, you should at least double the hash-table size. Even with periodic rehashing, hashing is an efficient implementation for map.

Pedagogical Note For an interactive GUI demo to see how separate chaining works, go to http://liveexample .pearsoncmg.com/dsanimation/SeparateChainingeBook.html, as shown in Figure 27.8.

threshold

rehash

separate chaining animation on Companion Website

Figure 27.8 The animation tool shows how separate chaining works.

27.6.1 What is load factor? Assume the hash table has the initial size 4 and its load factor is 0.5; show the hash table after inserting entries with the keys 34, 29, 53, 44, 120, 39, 45, and 40, using linear probing.

27.6.2 Assume the hash table has the initial size 4 and its load factor is 0.5; show the hash table after inserting entries with the keys 34, 29, 53, 44, 120, 39, 45, and 40, using quadratic probing.

Point Check

M27_LIAN1878_11_GE_C27.indd 1040 12/29/17 4:47 PM

27.7 Implementing a Map Using Hashing 1041

27.6.3 Assume the hash table has the initial size 4 and its load factor is 0.5; show the hash table after inserting entries with the keys 34, 29, 53, 44, 120, 39, 45, and 40, using separate chaining.

27.7 Implementing a Map Using Hashing A map can be implemented using hashing.

Now you understand the concept of hashing. You know how to design a good hash function to map a key to an index in a hash table, how to measure performance using the load factor, and how to increase the table size and rehash to maintain the performance. This section dem- onstrates how to implement a map using separate chaining.

We design our custom Map interface to mirror java.util.Map and name the interface MyMap and a concrete class MyHashMap, as shown in Figure 27.9.

Point Key

Figure 27.9 MyHashMap implements the MyMap interface.

MyHashMap<K, V>

+clear(): void

+containsKey(key: K): boolean

+containsValue(value: V): boolean

+entrySet(): Set<Entry<K, V>>

+get(key: K): V

+isEmpty(): boolean

+keySet(): Set<K>

+put(key: K, value: V): V

+remove(key: K): void

+size(): int

+values(): Set<V>

Removes all entries from this map.

Returns true if this map contains an entry for the speci�ed key.

Returns true if this map maps one or more keys to the speci�ed value.

Returns a set consisting of the keys in this map.

Returns a set consisting of the entries in this map.

Returns a value for the speci�ed key in this map.

Returns true if this map contains no mappings.

Puts a mapping in this map.

Removes the entries for the speci�ed key.

Returns the number of mappings in this map.

Returns a set consisting of the values in this map.

Creates an empty map with default capacity 4 and default load–factor threshold 0.75f.

+MyHashMap()

Creates a map with a speci�ed capacity and default load–factor threshold 0.75f.

+MyHashMap(capacity: int)

Creates a map with a speci�ed capacity and load–factor threshold.

+MyHashMap(capacity: int, loadFactorThreshold: �oat)

Constructs an entry with the speci�ed key and value.

Returns the key in the entry.

Returns the value in the entry.

-key: K

-value: V

+Entry(key: K, value: V)

+getkey(): K

+getValue(): V

MyMap.Entry<K, V>

«interface» MyMap<K, V>

M27_LIAN1878_11_GE_C27.indd 1041 12/29/17 4:47 PM

1042 Chapter 27 Hashing

How do you implement MyHashMap? If you use an ArrayList and store a new entry at the end of the list, the search time will be O(n). If you implement MyHashMap using a binary tree, the search time will be O(log n) if the tree is well balanced. Nevertheless, you can implement MyHashMap using hashing to obtain an O(1) time search algorithm. Listing 27.1 shows the MyMap interface, and Listing 27.2 implements MyHashMap using separate chaining.

Listing 27.1 MyMap.java 1 public interface MyMap<K, V> { 2 /** Remove all of the entries from this map */ 3 public void clear(); 4 5 /** Return true if the specified key is in the map */ 6 public boolean containsKey(K key); 7 8 /** Return true if this map contains the specified value */ 9 public boolean containsValue(V value); 10 11 /** Return a set of entries in the map */ 12 public java.util.Set<Entry<K, V>> entrySet(); 13 14 /** Return the value that matches the specified key */ 15 public V get(K key); 16 17 /** Return true if this map doesn't contain any entries */ 18 public boolean isEmpty(); 19 20 /** Return a set consisting of the keys in this map */ 21 public java.util.Set<K> keySet(); 22 23 /** Add an entry (key, value) into the map */ 24 public V put(K key, V value); 25 26 /** Remove an entry for the specified key */ 27 public void remove(K key); 28 29 /** Return the number of mappings in this map */ 30 public int size(); 31 32 /** Return a set consisting of the values in this map */ 33 public java.util.Set<V> values(); 34 35 /** Define an inner class for Entry */ 36 public static class Entry<K, V> { 37 K key; 38 V value; 39 40 public Entry(K key, V value) { 41 this.key = key; 42 this.value = value; 43 } 44 45 public K getKey() { 46 return key; 47 } 48 49 public V getValue() { 50 return value; 51 } 52

interface MyMap

clear

containsKey

containsValue

entrySet

get

isEmpty

keySet

put

remove

size

values

Entry inner class

M27_LIAN1878_11_GE_C27.indd 1042 12/29/17 4:47 PM

27.7 Implementing a Map Using Hashing 1043

53 @Override 54 public String toString() { 55 return "[" + key + ", " + value + "]"; 56 } 57 } 58 }

Listing 27.2 MyHashMap.java 1 import java.util.LinkedList; 2 3 public class MyHashMap<K, V> implements MyMap<K, V> { 4 // Define the default hash-table size. Must be a power of 2 5 private static int DEFAULT_INITIAL_CAPACITY = 4; 6 7 // Define the maximum hash-table size. 1 << 30 is same as 2^30 8 private static int MAXIMUM_CAPACITY = 1 << 30; 9 10 // Current hash-table capacity. Capacity is a power of 2 11 private int capacity; 12 13 // Define default load factor 14 private static float DEFAULT_MAX_LOAD_FACTOR = 0.75f; 15 16 // Specify a load factor used in the hash table 17 private float loadFactorThreshold; 18 19 // The number of entries in the map 20 private int size = 0; 21 22 // Hash table is an array with each cell being a linked list 23 LinkedList<MyMap.Entry<K,V>>[] table; 24 25 /** Construct a map with the default capacity and load factor */ 26 public MyHashMap() { 27 this(DEFAULT_INITIAL_CAPACITY, DEFAULT_MAX_LOAD_FACTOR); 28 } 29 30 /** Construct a map with the specified initial capacity and 31 * default load factor */ 32 public MyHashMap(int initialCapacity) { 33 this(initialCapacity, DEFAULT_MAX_LOAD_FACTOR); 34 } 35 36 /** Construct a map with the specified initial capacity 37 * and load factor */ 38 public MyHashMap(int initialCapacity, float loadFactorThreshold) { 39 if (initialCapacity > MAXIMUM_CAPACITY) 40 this.capacity = MAXIMUM_CAPACITY; 41 else 42 this.capacity = trimToPowerOf2(initialCapacity); 43 44 this.loadFactorThreshold = loadFactorThreshold; 45 table = new LinkedList[capacity]; 46 } 47 48 @Override /** Remove all of the entries from this map */ 49 public void clear() { 50 size = 0; 51 removeEntries();

class MyHashMap

default initial capacity

maximum capacity

current capacity

default load factor

load-factor threshold

size

hash table

no-arg constructor

constructor

clear

M27_LIAN1878_11_GE_C27.indd 1043 12/29/17 4:47 PM

1044 Chapter 27 Hashing

52 } 53 54 @Override /** Return true if the specified key is in the map */ 55 public boolean containsKey(K key) { 56 if (get(key) != null) 57 return true; 58 else 59 return false; 60 } 61 62 @Override /** Return true if this map contains the value */ 63 public boolean containsValue(V value) { 64 for (int i = 0; i < capacity; i++) { 65 if (table[i] != null) { 66 LinkedList<Entry<K, V>> bucket = table[i]; 67 for (Entry<K, V> entry: bucket) 68 if (entry.getValue().equals(value)) 69 return true; 70 } 71 } 72 73 return false; 74 } 75 76 @Override /** Return a set of entries in the map */ 77 public java.util.Set<MyMap.Entry<K,V>> entrySet() { 78 java.util.Set<MyMap.Entry<K, V>> set = 79 new java.util.HashSet<>(); 80 81 for (int i = 0; i < capacity; i++) { 82 if (table[i] != null) { 83 LinkedList<Entry<K, V>> bucket = table[i]; 84 for (Entry<K, V> entry: bucket) 85 set.add(entry); 86 } 87 } 88 89 return set; 90 } 91 92 @Override /** Return the value that matches the specified key */ 93 public V get(K key) { 94 int bucketIndex = hash(key.hashCode()); 95 if (table[bucketIndex] != null) { 96 LinkedList<Entry<K, V>> bucket = table[bucketIndex]; 97 for (Entry<K, V> entry: bucket) 98 if (entry.getKey().equals(key)) 99 return entry.getValue(); 100 } 101 102 return null; 103 } 104 105 @Override /** Return true if this map contains no entries */ 106 public boolean isEmpty() { 107 return size == 0; 108 } 109 110 @Override /** Return a set consisting of the keys in this map */ 111 public java.util.Set<K> keySet() {

containsKey

containsValue

entrySet

get

isEmpty

keySet

M27_LIAN1878_11_GE_C27.indd 1044 12/29/17 4:47 PM

27.7 Implementing a Map Using Hashing 1045

112 java.util.Set<K> set = new java.util.HashSet<>(); 113 114 for (int i = 0; i < capacity; i++) { 115 if (table[i] != null) { 116 LinkedList<Entry<K, V>> bucket = table[i]; 117 for (Entry<K, V> entry: bucket) 118 set.add(entry.getKey()); 119 } 120 } 121 122 return set; 123 } 124 125 @Override /** Add an entry (key, value) into the map */ 126 public V put(K key, V value) { 127 if (get(key) != null) { // The key is already in the map 128 int bucketIndex = hash(key.hashCode()); 129 LinkedList<Entry<K, V>> bucket = table[bucketIndex]; 130 for (Entry<K, V> entry: bucket) 131 if (entry.getKey().equals(key)) { 132 V oldValue = entry.getValue(); 133 // Replace old value with new value 134 entry.value = value; 135 // Return the old value for the key 136 return oldValue; 137 } 138 } 139 140 // Check load factor 141 if (size >= capacity * loadFactorThreshold) { 142 if (capacity == MAXIMUM_CAPACITY) 143 throw new RuntimeException("Exceeding maximum capacity"); 144 145 rehash(); 146 } 147 148 int bucketIndex = hash(key.hashCode()); 149 150 // Create a linked list for the bucket if not already created 151 if (table[bucketIndex] == null) { 152 table[bucketIndex] = new LinkedList<Entry<K, V>>(); 153 } 154 155 // Add a new entry (key, value) to hashTable[index] 156 table[bucketIndex].add(new MyMap.Entry<K, V>(key, value)); 157 158 size++; // Increase size 159 160 return value; 161 } 162 163 @Override /** Remove the entries for the specified key */ 164 public void remove(K key) { 165 int bucketIndex = hash(key.hashCode()); 166 167 // Remove the first entry that matches the key from a bucket 168 if (table[bucketIndex] != null) { 169 LinkedList<Entry<K, V>> bucket = table[bucketIndex]; 170 for (Entry<K, V> entry: bucket) 171 if (entry.getKey().equals(key)) {

put

remove

M27_LIAN1878_11_GE_C27.indd 1045 12/29/17 4:47 PM

1046 Chapter 27 Hashing

172 bucket.remove(entry); 173 size—–; // Decrease size 174 break; // Remove just one entry that matches the key 175 } 176 } 177 } 178 179 @Override /** Return the number of entries in this map */ 180 public int size() { 181 return size; 182 } 183 184 @Override /** Return a set consisting of the values in this map */ 185 public java.util.Set<V> values() { 186 java.util.Set<V> set = new java.util.HashSet<>(); 187 188 for (int i = 0; i < capacity; i++) { 189 if (table[i] != null) { 190 LinkedList<Entry<K, V>> bucket = table[i]; 191 for (Entry<K, V> entry: bucket) 192 set.add(entry.getValue()); 193 } 194 } 195 196 return set; 197 } 198 199 /** Hash function */ 200 private int hash(int hashCode) { 201 return supplementalHash(hashCode) & (capacity − 1); 202 } 203 204 /** Ensure the hashing is evenly distributed */ 205 private static int supplementalHash(int h) { 206 h ^= (h >>> 20) ^ (h >>> 12); 207 return h ^ (h >>> 7) ^ (h >>> 4); 208 } 209 210 /** Return a power of 2 for initialCapacity */ 211 private int trimToPowerOf2(int initialCapacity) { 212 int capacity = 1; 213 while (capacity < initialCapacity) { 214 capacity <<= 1; // Same as capacity *= 2. <= is more efficient 215 } 216 217 return capacity; 218 } 219 220 /** Remove all entries from each bucket */ 221 private void removeEntries() { 222 for (int i = 0; i < capacity; i++) { 223 if (table[i] != null) { 224 table[i].clear(); 225 } 226 } 227 } 228 229 /** Rehash the map */ 230 private void rehash() { 231 java.util.Set<Entry<K, V>> set = entrySet(); // Get entries

size

values

hash

supplementalHash

trimToPowerOf2

removeEntries

rehash

M27_LIAN1878_11_GE_C27.indd 1046 12/29/17 4:47 PM

27.7 Implementing a Map Using Hashing 1047

232 capacity <<= 1; // Same as capacity *= 2. <= is more efficient 233 table = new LinkedList[capacity]; // Create a new hash table 234 size = 0; // Reset size to 0 235 236 for (Entry<K, V> entry: set) { 237 put(entry.getKey(), entry.getValue()); // Store to new table 238 } 239 } 240 241 @Override /** Return a string representation for this map */ 242 public String toString() { 243 StringBuilder builder = new StringBuilder("["); 244 245 for (int i = 0; i < capacity; i++) { 246 if (table[i] != null && table[i].size() > 0) 247 for (Entry<K, V> entry: table[i]) 248 builder.append(entry); 249 } 250 251 builder.append("]"); 252 return builder.toString(); 253 } 254 }

The MyHashMap class implements the MyMap interface using separate chaining. The param- eters that determine the hash-table size and load factors are defined in the class. The default initial capacity is 4 (line 5) and the maximum capacity is 230 (line 8). The current hash-table capacity is designed as a value of the power of 2 (line 11). The default load-factor threshold is 0.75f (line 14). You can specify a custom load-factor threshold when constructing a map. The custom load-factor threshold is stored in loadFactorThreshold (line 17). The data field size denotes the number of entries in the map (line 20). The hash table is an array. Each cell in the array is a linked list (line 23).

Three constructors are provided to construct a map. You can construct a default map with the default capacity and load-factor threshold using the no-arg constructor (lines 26–28), a map with the specified capacity and a default load-factor threshold (lines 32–34), and a map with the specified capacity and load-factor threshold (lines 38–46).

The clear method removes all entries from the map (lines 49–52). It invokes removeEntries(), which deletes all entries in the buckets (lines 221–227). The removeEntries() method takes O(capacity) time to clear all entries in the table.

The containsKey(key) method checks whether the specified key is in the map by invoking the get method (lines 55–60). Since the get method takes O(1) time, the containsKey(key) method takes O(1) time.

The containsValue(value) method checks whether the value is in the map (lines 63–74). This method takes O(capacity + size) time. It is actually O(capacity), since capacity 7 size.

The entrySet() method returns a set that contains all entries in the map (lines 77–90). This method takes O(capacity) time.

The get(key) method returns the value of the first entry with the specified key (lines 93–103). This method takes O(1) time.

The isEmpty() method simply returns true if the map is empty (lines 106–108). This method takes O(1) time.

The keySet() method returns all keys in the map as a set. The method finds the keys from each bucket and adds them to a set (lines 111–123). This method takes O(capacity) time.

The put(key, value) method adds a new entry into the map. The method first tests if the key is already in the map (line 127), if so, it locates the entry and replaces the old value with the new value in the entry for the key (line 134) and the old value is returned (line 136). If the key is

toString

hash-table parameters

three constructors

clear

containsKey

containsValue

entrySet

get

isEmpty

keySet

put

M27_LIAN1878_11_GE_C27.indd 1047 12/29/17 4:47 PM

1048 Chapter 27 Hashing

new in the map, the new entry is created in the map (line 156). Before inserting the new entry, the method checks whether the size exceeds the load-factor threshold (line 141). If so, the program invokes rehash() (line 145) to increase the capacity and store entries into a new larger hash table.

The rehash() method first copies all entries in a set (line 231), doubles the capacity (line 232), creates a new hash table (line 233), and resets the size to 0 (line 234). The method then copies the entries into the new hash table (lines 236–238). The rehash method takes O(capacity) time. If no rehash is performed, the put method takes O(1) time to add a new entry.

The remove(key) method removes the entry with the specified key in the map (lines 164–177). This method takes O(1) time.

The size() method simply returns the size of the map (lines 180–182). This method takes O(1) time.

The values() method returns all values in the map. The method examines each entry from all buckets and adds it to a set (lines 185–197). This method takes O(capacity) time.

The hash() method invokes the supplementalHash to ensure the hashing is evenly distrib- uted to produce an index for the hash table (lines 200–208). This method takes O(1) time.

Table 27.1 summarizes the time complexities of the methods in MyHashMap.

rehash

remove

size

values

hash

Methods Time

clear() O(capacity)

containsKey(key: Key) O(1)

containsValue(value: V) O(capacity)

entrySet() O(capacity)

get(key: K) O(1)

isEmpty() O(1)

keySet() O(capacity)

put(key: K, value: V) O(1)

remove(key: K) O(1)

size() O(1)

values() O(capacity)

rehash() O(capacity)

tabLe 27.1 Time Complexities for Methods in MyHashMap

Since rehashing does not happen very often, the time complexity for the put method is O(1). Note the complexities of the clear, entrySet, keySet, values, and rehash methods depend on capacity, so to avoid poor performance for these methods, you should choose an initial capacity carefully.

Listing 27.3 gives a test program that uses MyHashMap.

Listing 27.3 TestMyHashMap.java 1 public class TestMyHashMap { 2 public static void main(String[] args) { 3 // Create a map 4 MyMap<String, Integer> map = new MyHashMap<>(); 5 map.put("Smith", 30); 6 map.put("Anderson", 31); 7 map.put("Lewis", 29); 8 map.put("Cook", 29);

create a map put entries

M27_LIAN1878_11_GE_C27.indd 1048 12/29/17 4:47 PM

27.7 Implementing a Map Using Hashing 1049

9 map.put("Smith", 65); 10 11 System.out.println("Entries in map: " + map); 12 13 System.out.println("The age for Lewis is " + 14 map.get("Lewis")); 15 16 System.out.println("Is Smith in the map? " + 17 map.containsKey("Smith")); 18 System.out.println("Is age 33 in the map? " + 19 map.containsValue(33)); 20 21 map.remove("Smith"); 22 System.out.println("Entries in map: " + map); 23 24 map.clear(); 25 System.out.println("Entries in map: " + map); 26 } 27 }

display entries

get value

is key in map?

is value in map?

remove entry

Entries in map: [[Anderson, 31][Smith, 65][Lewis, 29][Cook, 29]] The age for Lewis is 29 Is Smith in the map? true Is age 33 in the map? false Entries in map: [[Anderson, 31][Lewis, 29][Cook, 29]] Entries in map: []

The program creates a map using MyHashMap (line 4) and adds five entries into the map (lines 5–9). Line 5 adds key Smith with value 30 and line 9 adds Smith with value 65. The latter value replaces the former value. The map actually has only four entries. The program displays the entries in the map (line 11), gets a value for a key (line 14), checks whether the map contains the key (line 17) and a value (line 19), removes an entry with the key Smith (line 21), and redisplays the entries in the map (line 22). Finally, the program clears the map (line 24) and displays an empty map (line 25).

27.7.1 What is 1 << 30 in line 8 in Listing 27.2? What are the integers resulted from 1 << 1, 1 << 2, and 1 << 3?

27.7.2 What are the integers resulted from 32 >> 1, 32 >> 2, 32 >> 3, and 32 >> 4? 27.7.3 In Listing 27.2, will the program work if LinkedList is replaced by ArrayList?

In Listing 27.2, how do you replace the code in lines 56–59 using one line of code?

27.7.4 Describe how the put(key, value) method is implemented in the MyHashMap class. 27.7.5 In Listing 27.2, the supplementalHash method is declared static. Can the hash

method be declared static?

27.7.6 Show the output of the following code:

MyMap<String, String> map = new MyHashMap<>(); map.put("Texas", "Dallas"); map.put("Oklahoma", "Norman"); map.put("Texas", "Austin"); map.put("Oklahoma", "Tulsa");

System.out.println(map.get("Texas")); System.out.println(map.size());

27.7.7 If x is a negative int value, will x & (N – 1) be negative?

Point Check

M27_LIAN1878_11_GE_C27.indd 1049 12/29/17 4:47 PM

1050 Chapter 27 Hashing

27.8 Implementing Set Using Hashing A hash set can be implemented using a hash map.

A set (introduced in Chapter 21) is a data structure that stores distinct values. The Java Collections Framework defines the java.util.Set interface for modeling sets. Three concrete implementations are java.util.HashSet, java.util.LinkedHashSet, and java.util.TreeSet. java.util.HashSet is implemented using hashing, java. util. LinkedHashSet using LinkedList, and java.util.TreeSet using binary search trees.

You can implement MyHashSet using the same approach as for implementing MyHashMap. The only difference is that key/value pairs are stored in the map, while elements are stored in the set.

Since all the methods in HashSet are inherited from Collection, we design our custom HashSet by implementing the Collection interface, as shown in Figure 27.10.

Point Key

set

hash map hash set

MyHashSet

Figure 27.10 MyHashSet implements the Collection interface.

MyHashSet<E>

«interface» java.lang.Collection<E>

+MyHashSet()

+MyHashMap(capacity: int)

+MyHashMap(capacity: int, loadFactorThreshold: float)

Creates an empty set with default capacity 4 and default load– factor threshold 0.75f.

Creates a set with a speci�ed capacity and default load– factor threshold 0.75f.

Creates a set with a speci�ed capacity and load–factor threshold.

Listing 27.4 implements MyHashSet using separate chaining.

Listing 27.4 MyHashSet.java 1 import java.util.*; 2 3 public class MyHashSet<E> implements Collection<E> { 4 // Define the default hash-table size. Must be a power of 2 5 private static int DEFAULT_INITIAL_CAPACITY = 4; 6 7 // Define the maximum hash-table size. 1 << 30 is same as 2^30 8 private static int MAXIMUM_CAPACITY = 1 << 30; 9 10 // Current hash-table capacity. Capacity is a power of 2 11 private int capacity; 12 13 // Define default load factor 14 private static float DEFAULT_MAX_LOAD_FACTOR = 0.75f; 15 16 // Specify a load-factor threshold used in the hash table 17 private float loadFactorThreshold; 18 19 // The number of elements in the set 20 private int size = 0;

class MyHashSet

default initial capacity

maximum capacity

current capacity

default max load factor

load-factor threshold

size

M27_LIAN1878_11_GE_C27.indd 1050 12/29/17 4:47 PM

27.8 Implementing Set Using Hashing 1051

21 22 // Hash table is an array with each cell being a linked list 23 private LinkedList<E>[] table; 24 25 /** Construct a set with the default capacity and load factor */ 26 public MyHashSet() { 27 this(DEFAULT_INITIAL_CAPACITY, DEFAULT_MAX_LOAD_FACTOR); 28 } 29 30 /** Construct a set with the specified initial capacity and 31 * default load factor */ 32 public MyHashSet(int initialCapacity) { 33 this(initialCapacity, DEFAULT_MAX_LOAD_FACTOR); 34 } 35 36 /** Construct a set with the specified initial capacity 37 * and load factor */ 38 public MyHashSet(int initialCapacity, float loadFactorThreshold) { 39 if (initialCapacity > MAXIMUM_CAPACITY) 40 this.capacity = MAXIMUM_CAPACITY; 41 else 42 this.capacity = trimToPowerOf2(initialCapacity); 43 44 this.loadFactorThreshold = loadFactorThreshold; 45 table = new LinkedList[capacity]; 46 } 47 48 @Override /** Remove all elements from this set */ 49 public void clear() { 50 size = 0; 51 removeElements(); 52 } 53 54 @Override /** Return true if the element is in the set */ 55 public boolean contains(E e) { 56 int bucketIndex = hash(e.hashCode()); 57 if (table[bucketIndex] != null) { 58 LinkedList<E> bucket = table[bucketIndex]; 59 return bucket.contains(e); 60 } 61 62 return false; 63 } 64 65 @Override /** Add an element to the set */ 66 public boolean add(E e) { 67 if (contains(e)) // Duplicate element not stored 68 return false; 69 70 if (size + 1 > capacity * loadFactorThreshold) { 71 if (capacity == MAXIMUM_CAPACITY) 72 throw new RuntimeException("Exceeding maximum capacity"); 73 74 rehash(); 75 } 76 77 int bucketIndex = hash(e.hashCode()); 78 79 // Create a linked list for the bucket if not already created 80 if (table[bucketIndex] == null) {

hash table

no-arg constructor

constructor

clear

contains

add

M27_LIAN1878_11_GE_C27.indd 1051 12/29/17 4:47 PM

1052 Chapter 27 Hashing

81 table[bucketIndex] = new LinkedList<E>(); 82 } 83 84 // Add e to hashTable[index] 85 table[bucketIndex].add(e); 86 87 size++; // Increase size 88 89 return true; 90 } 91 92 @Override /** Remove the element from the set */ 93 public boolean remove(E e) { 94 if (!contains(e)) 95 return false; 96 97 int bucketIndex = hash(e.hashCode()); 98 99 // Create a linked list for the bucket if not already created 100 if (table[bucketIndex] != null) { 101 LinkedList<E> bucket = table[bucketIndex]; 102 bucket.removed(e); 103 } 104 105 size——; // Decrease size 106 107 return true; 108 } 109 110 @Override /** Return true if the set contain no elements */ 111 public boolean isEmpty() { 112 return size == 0; 113 } 114 115 @Override /** Return the number of elements in the set */ 116 public int size() { 117 return size; 118 } 119 120 @Override /** Return an iterator for the elements in this set */ 121 public java.util.Iterator<E> iterator() { 122 return new MyHashSetIterator(this); 123 } 124 125 /** Inner class for iterator */ 126 private class MyHashSetIterator implements java.util.Iterator<E> { 127 // Store the elements in a list 128 private java.util.ArrayList<E> list; 129 private int current = 0; // Point to the current element in list 130 private MyHashSet<E> set; 131 132 /** Create a list from the set */ 133 public MyHashSetIterator(MyHashSet<E> set) { 134 this.set = set; 135 list = setToList(); 136 } 137 138 @Override /** Next element for traversing? */ 139 public boolean hasNext() { 140 return current < list.size(); 141 }

remove

isEmpty

size

iterator

inner class

M27_LIAN1878_11_GE_C27.indd 1052 12/29/17 4:47 PM

27.8 Implementing Set Using Hashing 1053

142 143 @Override /** Get current element and move cursor to the next */ 144 public E next() { 145 return list.get(current++); 146 } 147 148 /** Remove the current element returned by the last next() */ 149 public void remove() { 150 // Left as an exercise 151 // You need to remove the element from the set 152 // You also need to remove it from the list 153 } 154 } 155 156 /** Hash function */ 157 private int hash(int hashCode) { 158 return supplementalHash(hashCode) & (capacity − 1); 159 } 160 161 /** Ensure the hashing is evenly distributed */ 162 private static int supplementalHash(int h) { 163 h ^= (h >>> 20) ^ (h >>> 12); 164 return h ^ (h >>> 7) ^ (h >>> 4); 165 } 166 167 /** Return a power of 2 for initialCapacity */ 168 private int trimToPowerOf2(int initialCapacity) { 169 int capacity = 1; 170 while (capacity < initialCapacity) { 171 capacity <<= 1; // Same as capacity *= 2. <= is more efficient 172 } 173 174 return capacity; 175 } 176 177 /** Remove all e from each bucket */ 178 private void removeElements() { 179 for (int i = 0; i < capacity; i++) { 180 if (table[i] != null) { 181 table[i].clear(); 182 } 183 } 184 } 185 186 /** Rehash the set */ 187 private void rehash() { 188 java.util.ArrayList<E> list = setToList(); // Copy to a list 189 capacity <<= 1; // Same as capacity *= 2. <= is more efficient 190 table = new LinkedList[capacity]; // Create a new hash table 191 size = 0; 192 193 for (E element: list) { 194 add(element); // Add from the old table to the new table 195 } 196 } 197 198 /** Copy elements in the hash set to an array list */ 199 private java.util.ArrayList<E> setToList() { 200 java.util.ArrayList<E> list = new java.util.ArrayList<>(); 201 202 for (int i = 0; i < capacity; i++) {

hash

supplementalHash

trimToPowerOf2

rehash

setToList

M27_LIAN1878_11_GE_C27.indd 1053 12/29/17 4:47 PM

1054 Chapter 27 Hashing

203 if (table[i] != null) { 204 for (E e: table[i]) { 205 list.add(e); 206 } 207 } 208 } 209 210 return list; 211 } 212 213 @Override /** Return a string representation for this set */ 214 public String toString() { 215 java.util.ArrayList<E> list = setToList(); 216 StringBuilder builder = new StringBuilder("["); 217 218 // Add the elements except the last one to the string builder 219 for (int i = 0; i < list.size() − 1; i++) { 220 builder.append(list.get(i) + ", "); 221 } 222 223 // Add the last element in the list to the string builder 224 if (list.size() == 0) 225 builder.append("]"); 226 else 227 builder.append(list.get(list.size() − 1) + "]"); 228 229 return builder.toString(); 230 } 231 232 @Override 233 public boolean addAll(Collection<? extends E> arg0) { 234 // Left as an exercise 235 return false; 236 } 237 238 @Override 239 public boolean containsAll(Collection<?> arg0) { 240 // Left as an exercise 241 return false; 242 } 243 244 @Override 245 public boolean removeAll(Collection<?> arg0) { 246 // Left as an exercise 247 return false; 248 } 249 250 @Override 251 public boolean retainAll(Collection<?> arg0) { 252 // Left as an exercise 253 return false; 254 } 255 256 @Override 257 public Object[] toArray() { 258 // Left as an exercise 259 return null; 260 } 261 262 @Override

toString

override addAll

override containsAll

override removeAll

override retainAll

override toArray()

M27_LIAN1878_11_GE_C27.indd 1054 12/29/17 4:47 PM

27.8 Implementing Set Using Hashing 1055

263 public <T> T[] toArray(T[] arg0) { 264 // Left as an exercise 265 return null; 266 } 267 }

The MyHashSet class implements the MySet interface using separate chaining. Implement- ing MyHashSet is very similar to implementing MyHashMap except for the following differences:

1. The elements are stored in the hash table for MyHashSet, but the entries (key/value pairs) are stored in the hash table for MyHashMap.

2. MyHashSet implements Collection. Since Collection implements Iterable, the elements in MyHashSet are iterable.

Three constructors are provided to construct a set. You can construct a default set with the default capacity and load factor using the no-arg constructor (lines 26–28), a set with the speci- fied capacity and a default load factor (lines 32–34), and a set with the specified capacity and load factor (lines 38–46).

The clear method removes all elements from the set (lines 49–52). It invokes removeElements(), which clears all table cells (line 181). Each table cell is a linked list that stores the elements with the same hash table index. The removeElements() method takes O(capacity) time.

The contains(element) method checks whether the specified element is in the set by examining whether the designated bucket contains the element (line 59). This method takes O(1) time because the bucket size is considered very small.

The add(element) method adds a new element into the set. The method first checks if the element is already in the set (line 67). If so, the method returns false. The method then checks whether the size exceeds the load-factor threshold (line 70). If so, the program invokes rehash() (line 74) to increase the capacity and store elements into a new larger hash table.

The rehash() method first copies all elements to a list (line 188), doubles the capacity (line 189), creates a new hash table (line 190), and resets the size to 0 (line 191). The method then copies the elements into the new larger hash table (lines 193–195). The rehash method takes O(capacity) time. If no rehash is performed, the add method takes O(1) time to add a new element.

The remove(element) method removes the specified element in the set (lines 93–108). This method takes O(1) time.

The size() method simply returns the number of elements in the set (lines 116–118). This method takes O(1) time.

The iterator() method returns an instance of java.util.Iterator. The MyHash SetIterator class implements java.util.Iterator to create a forward iterator. When a MyHashSetIterator is constructed, it copies all the elements in the set to a list (line 135). The variable current points to the element in the list. Initially, current is 0 (line 129), which points to the first element in the list. MyHashSetIterator implements the methods has- Next(), next(), and remove() in java.util.Iterator. Invoking hasNext() returns true if current < list.size(). Invoking next() returns the current element and moves current to point to the next element (line 145). Invoking remove() removes the element called by the last next().

The hash() method invokes the supplementalHash to ensure the hashing is evenly distributed to produce an index for the hash table (lines 157–159). This method takes O(1) time.

The methods containsAll, addAll, removeAll, retainAll, toArray(), and toArray(T[]) defined in the Collection interface are overridden in MyHashSet. Their implementations are left as exercises in Programming Exercise 27.11.

override toArray(T[])

MyHashSet vs. MyHashMap

three constructors

clear

contains

add

rehash

remove

size

iterator

hash

M27_LIAN1878_11_GE_C27.indd 1055 12/29/17 4:47 PM

1056 Chapter 27 Hashing

Table 27.2 summarizes the time complexity of the methods in MyHashSet.

Methods Time

clear() O(capacity)

contains(e: E) O(1)

add(e: E) O(1)

remove(e: E) O(1)

isEmpty() O(1)

size() O(1)

iterator() O(capacity)

rehash() O(capacity)

tabLe 27.2 Time Complexities for Methods in MyHashSet

Elements in set: [Cook, Anderson, Smith, Lewis] Number of elements in set: 4 Is Smith in set? true Names in set in uppercase are COOK ANDERSON LEWIS Elements in set: []

Listing 27.5 gives a test program that uses MyHashSet.

Listing 27.5 TestMyHashSet.java 1 public class TestMyHashSet { 2 public static void main(String[] args) { 3 // Create a MyHashSet 4 java.util.Collection<String> set = new MyHashSet<>(); 5 set.add("Smith"); 6 set.add("Anderson"); 7 set.add("Lewis"); 8 set.add("Cook"); 9 set.add("Smith"); 10 11 System.out.println("Elements in set: " + set); 12 System.out.println("Number of elements in set: " + set.size()); 13 System.out.println("Is Smith in set? " + set.contains("Smith")); 14 15 set.remove("Smith"); 16 System.out.print("Names in set in uppercase are "); 17 for (String s: set) 18 System.out.print(s.toUpperCase() + " "); 19 20 set.clear(); 21 System.out.println("\nElements in set: " + set); 22 } 23 }

create a set add elements

display elements set size

remove element

foreach loop

clear set

The program creates a set using MyHashSet (line 4) and adds five elements to the set (lines 5–9). Line 5 adds Smith and line 9 adds Smith again. Since only nonduplicate elements are stored in the set, Smith appears in the set only once. The set actually has four elements. The program displays the elements (line 11), gets its size (line 12), checks whether the set contains

M27_LIAN1878_11_GE_C27.indd 1056 12/29/17 4:47 PM

Quiz 1057

a specified element (line 13), and removes an element (line 15). Since the elements in a set are iterable, a foreach loop is used to traverse all elements in the set (lines 17–18). Finally, the program clears the set (line 20) and displays an empty set (line 21).

27.8.1 Why can you use a foreach loop to traverse the elements in a set? 27.8.2 Describe how the add(e) method is implemented in the MyHashSet class. 27.8.3 Can lines 100–103 in Listing 27.4 be removed? 27.8.4 Implement the remove() method in lines 150–152?

Point Check

Key terms associative array 1032 cluster 1036 dictionary 1032 double hashing 1037 hash code 1033 hash function 1032 hash map 1050 hash set 1050 hash table 1051

linear probing 1035 load factor 1039 open addressing 1035 perfect hash function 1032 polynomial hash code 1034 quadratic probing 1036 rehashing 1039 separate chaining 1039

ChapTer summary 1. A map is a data structure that stores entries. Each entry contains two parts: a key and a

value. The key is also called a search key, which is used to search for the corresponding value. You can implement a map to obtain O(1) time complexity on searching, retrieval, insertion, and deletion using the hashing technique.

2. A set is a data structure that stores elements. You can use the hashing technique to implement a set to achieve O(1) time complexity on searching, insertion, and deletion for a set.

3. Hashing is a technique that retrieves the value using the index obtained from a key without performing a search. A typical hash function first converts a search key to an integer value called a hash code, then compresses the hash code into an index to the hash table.

4. A collision occurs when two keys are mapped to the same index in a hash table. Generally, there are two ways for handling collisions: open addressing and separate chaining.

5. Open addressing is the process of finding an open location in the hash table in the event of collision. Open addressing has several variations: linear probing, quadratic probing, and double hashing.

6. The separate chaining scheme places all entries with the same hash index into the same location, rather than finding new locations. Each location in the separate chaining scheme is called a bucket. A bucket is a container that holds multiple entries.

Quiz Answer the quiz for this chapter online at the book Companion Website.

M27_LIAN1878_11_GE_C27.indd 1057 12/29/17 4:47 PM

1058 Chapter 27 Hashing

programming exerCiSeS **27.1 (Implement MyMap using open addressing with linear probing) Create a new con-

crete class that implements MyMap using open addressing with linear probing. For simplicity, use f(key) = key % size as the hash function, where size is the hash-table size. Initially, the hash-table size is 6. The table size is doubled whenever the load factor exceeds the threshold (0.5).

**27.2 (Implement MyMap using open addressing with quadratic probing) Create a new concrete class that implements MyMap using open addressing with quadratic prob- ing. For simplicity, use f(key) = key % size as the hash function, where size is the hash-table size. Initially, the hash-table size is 6. The table size is doubled whenever the load factor exceeds the threshold (0.5).

**27.3 (Implement MyMap using open addressing with double hashing) Create a new concrete class that implements MyMap using open addressing with double hashing. For simplicity, use f(key) = key % size as the hash function, where size is the hash-table size. Initially, the hash-table size is 6. The table size is doubled whenever the load factor exceeds the threshold (0.5).

**27.4 (Modify MyHashMap with duplicate keys) Modify MyHashMap to allow dupli- cate keys for entries. You need to modify the implementation for the put(key, value) method. Also add a new method named getAll(key) that returns a set of values that match the key in the map.

**27.5 (Implement MyHashSet using MyHashMap) Implement MyHashSet using MyHashMap. Note you can create entries with (key, key), rather than (key, value).

**27.6 (Animate linear probing) Write a program that animates linear probing, as shown in Figure 27.3. You can change the initial size of the hash-table in the program. Assume that the load-factor threshold is 0.8.

**27.7 (Animate separate chaining) Write a program that animates MyHashMap, as shown in Figure 27.8. You can change the initial size of the table. Assume that the load- factor threshold is 0.8.

**27.8 (Animate quadratic probing) Write a program that animates quadratic probing, as shown in Figure 27.5. You can change the initial size of the hash-table in the program. Assume the load-factor threshold to be 0.8.

**27.9 (Implement hashCode for string) Write a method that returns a hash code for string using the approach described in Section 27.3.2 with b value 31. The function header is as follows:

public static int hashCodeForString(String s)

**27.10 (Compare MyHashSet and MyArrayList) MyArrayList is defined in Listing 24.2. Write a program that generates 1000000 random double values between 0 and 999999 and stores them in a MyArrayList and in a MyHashSet. Generate a list of 1000000 random double values between 0 and 1999999. For each number in the list, test if it is in the array list and in the hash set. Run your program to display the total test time for the array list and for the hash set.

**27.11 (Implement set operations in MyHashSet) The implementations of the methods addAll, removeAll, retainAll, toArray(), and toArray(T[]) are omitted in the MyHashSet class. Implement these methods. Also add a new constructor MyHashSet(E[] list) in the MyHashSet class. Test your new MyHashSet class using the code at liveexample.pearsoncmg.com/test/Exercise27_11Test.txt.

M27_LIAN1878_11_GE_C27.indd 1058 12/29/17 4:47 PM

Programming Exercises 1059

**27.12 (setToList) Write the following method that returns an ArrayList from a set:

public static <E> ArrayList<E> setToList(Set<E> s)

*27.13 (The Date class) Design a class named Date that meets the following requirements:

■■ Three data fields year, month, and day for representing a date ■■ A constructor that constructs a date with the specified year, month, and day ■■ Override the equals method ■■ Override the hashCode method. (For reference, see the implementation of the Date class in the Java API)

*27.14 (The Point class) Design a class named Point that meets the following requirements:

■■ Two data fields x and y for representing a point with getter methods ■■ A no-arg constructor that constructs a point for (0, 0) ■■ A constructor that constructs a point with the specified x and y values ■■ Override the equals method. Point p1 is said to be equal to point p2 if p1.x = = p2.x and p1.y = = p2.y.

■■ Override the hashCode method. (For reference, see the implementation of the Point2D class in the Java API.)

*27.15 (Modify Listing 27.4 MyHashSet.java) The book uses LinkedList for buck- ets. Replace LinkedList with AVLTree. Assume E is Comparable. Redefine MyHashSet as follows:

public class MyHashSet<E extends Comparable<E>> implements Collection { ... }

Test your program using the main method in Listing 27.5.

M27_LIAN1878_11_GE_C27.indd 1059 12/29/17 4:47 PM

A01_GIPE3898_09_SE_FM.indd 3 12/20/17 5:56 PM

This page intentionally left blank

Objectives ■■ To model real-world problems using graphs and explain the Seven

Bridges of Königsberg problem (§28.1).

■■ To describe the graph terminologies: vertices, edges, simple graphs, weighted/unweighted graphs, and directed/undirected graphs (§28.2).

■■ To represent vertices and edges using lists, edge arrays, edge objects, adjacency matrices, and adjacency lists (§28.3).

■■ To model graphs using the Graph interface and the UnweightedGraph class (§28.4).

■■ To display graphs visually (§28.5).

■■ To represent the traversal of a graph using the UnweightedGraph. SearchTree class (§28.6).

■■ To design and implement depth-first search (§28.7).

■■ To solve the connected-circle problem using depth-first search (§28.8).

■■ To design and implement breadth-first search (§28.9).

■■ To solve the nine tails problem using breadth-first search (§28.10).

Graphs and Applications

Chapter

M28_LIAN1878_11_GE_C28.indd 1061 12/29/17 4:49 PM

1062 Chapter 28 Graphs and Applications

28.1 Introduction Many real-world problems can be solved using graph algorithms.

Graphs are useful in modeling and solving real-world problems. For example, the problem to find the least number of flights between two cities can be modeled using a graph, where the ver- tices represent cities and the edges represent the flights between two adjacent cities, as shown in Figure 28.1. The problem of finding the minimal number of connecting flights between two cities is reduced to finding the shortest path between two vertices in a graph. At United Parcel Service, each driver makes an average 120 stops per day. There are many possible ways for ordering these stops. UPS spent hundreds of millions of dollars for 10 years to develop a system called Orion, which stands for On-Road Integrated Optimization and Navigation. The system uses graph algorithms to plan the most cost-efficient routes for each driver. This chapter studies the algorithms for unweighted graphs, and the next chapter studies those for weighted graphs.

Point Keyproblem

Figure 28.1 A graph can be used to model the flights between the cities.

Seattle (0)

San Francisco (1)

Los Angeles (2)

Dallas (10)

Houston (11)

Atlanta (8)

New York (7)

Boston (6) Chicago (5)

Denver (3)

Kansas City (4)

Miami (9)

Figure 28.2 Seven bridges connected islands and land.

(b) Graph model (a) Seven bridges sketch

Island 2 Island 1 C

The study of graph problems is known as graph theory. Graph theory was founded by Leonhard Euler in 1736, when he introduced graph terminology to solve the famous Seven Bridges of Königsberg problem. The city of Königsberg, Prussia (now Kaliningrad, Russia), was divided by the Pregel River. There were two islands on the river. The city and islands were connected by seven bridges, as shown in Figure 28.2a. The question is, can one take a walk, cross each bridge exactly once, and return to the starting point? Euler proved that it is not possible.

graph theory

Seven Bridges of Königsberg

To establish a proof, Euler first abstracted the Königsberg city map by eliminating all streets, producing the sketch shown in Figure 28.2a. Next, he replaced each land mass with

M28_LIAN1878_11_GE_C28.indd 1062 12/29/17 4:49 PM

28.2 Basic Graph Terminologies 1063

a dot, called a vertex or a node, and each bridge with a line, called an edge, as shown in Figure 28.2b. This structure with vertices and edges is called a graph.

Looking at the graph, we ask whether there is a path starting from any vertex, traversing all edges exactly once, and returning to the starting vertex. Euler proved that for such a path to exist, each vertex must have an even number of edges. Therefore, the Seven Bridges of Königsberg problem has no solution.

Graphs have many applications in various areas, such as in computer science, mathematics, biology, engineering, economics, genetics, and social sciences. This chapter presents the algo- rithms for depth-first search and breadth-first search, and their applications. The next chapter presents the algorithms for finding a minimum spanning tree and shortest paths in weighted graphs, and their applications.

28.2 Basic Graph Terminologies A graph consists of vertices, and edges that connect the vertices.

This chapter does not assume that you have any prior knowledge of graph theory or discrete mathematics. We use plain and simple terms to define graphs.

What is a graph? A graph is a mathematical structure that represents relationships among entities in the real world. For example, the graph in Figure 28.1 represents the flights among cities, and the graph in Figure 28.2b represents the bridges among land masses.

A graph consists of a nonempty set of vertices (also known as nodes or points), and a set of edges that connect the vertices. For convenience, we define a graph as G = (V, E), where V represents a set of vertices and E represents a set of edges. For example, V and E for the graph in Figure 28.1 are as follows:

V = {"Seattle", "San Francisco", "Los Angeles", "Denver", "Kansas City", "Chicago", "Boston", "New York", "Atlanta", "Miami", "Dallas", "Houston"};

E = {{"Seattle", "San Francisco"},{"Seattle", "Chicago"}, {"Seattle", "Denver"}, {"San Francisco", "Denver"}, ... };

A graph may be directed or undirected. In a directed graph, each edge has a direction, which indicates you can move from one vertex to the other through the edge. You can model parent/ child relationships using a directed graph, where an edge from vertex A to B indicates that A is a parent of B. Figure 28.3a shows a directed graph.

Point Key

what is a graph?

define a graph

directed vs. undirected graphs

Figure 28.3 Graphs may appear in many forms.

(a) A directed graph

Peter (0)

Cindy (3)

Wendy (4)

Jane (1)

Mark (2)

(b) A complete graph

In an undirected graph, you can move in both directions between vertices. The graph in Figure 28.1 is undirected.

M28_LIAN1878_11_GE_C28.indd 1063 12/29/17 4:49 PM

1064 Chapter 28 Graphs and Applications

Edges may be weighted or unweighted. For example, you can assign a weight for each edge in the graph in Figure 28.1 to indicate the flight time between the two cities.

Two vertices in a graph are said to be adjacent if they are connected by the same edge. Similarly, two edges are said to be adjacent if they are connected to the same vertex. An edge in a graph that joins two vertices is said to be incident to both vertices. The degree of a vertex is the number of edges incident to it.

Two vertices are called neighbors if they are adjacent. Similarly, two edges are called neighbors if they are adjacent.

A loop is an edge that links a vertex to itself. If two vertices are connected by two or more edges, these edges are called parallel edges. A simple graph is one that doesn’t have any loops or parallel edges. In a complete graph, every two pairs of vertices are connected, as shown in Figure 28.3b.

A graph is connected if there exists a path between any two vertices in the graph. A subgraph of a graph G is a graph whose vertex set is a subset of that of G and whose edge set is a subset of that of G. For example, the graph in Figure 28.3c is a subgraph of the graph in Figure 28.3b.

Assume the graph is connected and undirected. A cycle is a closed path that starts from a vertex and ends at the same vertex. A connected graph is a tree if it does not have cycles. A spanning tree of a graph G is a connected subgraph of G, and the subgraph is a tree that contains all vertices in G.

Pedagogical Note Before we introduce graph algorithms and applications, it is helpful to get acquainted with graphs using the interactive tool at liveexample.pearsoncmg.com/dsanimation /GraphLearningTooleBook.html, as shown in Figure 28.4. The tool allows you to add/ remove/move vertices and draw edges using mouse gestures. You can also find depth-first search (DFS) trees and breadth-first search (BFS) trees, and the shortest path between two vertices.

28.2.1 What is the famous Seven Bridges of Königsberg problem? 28.2.2 What is a graph? Explain the following terms: undirected graph, directed graph,

weighted graph, degree of a vertex, parallel edge, simple graph, complete graph, connected graph, cycle, subgraph, tree, and spanning tree.

28.2.3 How many edges are in a complete graph with 5 vertices? How many edges are in a tree of 5 vertices?

28.2.4 How many edges are in a complete graph with n vertices? How many edges are in a tree of n vertices?

28.3 Representing Graphs Representing a graph is to store its vertices and edges in a program. The data structure for storing a graph is arrays or lists.

To write a program that processes and manipulates graphs, you have to store or represent data for the graphs in the computer.

28.3.1 Representing Vertices The vertices can be stored in an array or a list. For example, you can store all the city names in the graph in Figure 28.1 using the following array:

String[] vertices = {"Seattle", "San Francisco", "Los Angeles", "Denver", "Kansas City", "Chicago", "Boston", "New York", "Atlanta", "Miami", "Dallas", "Houston"};

weighted vs. unweighted graphs

adjacent vertices incident edges degree

neighbor

loop

parallel edge simple graph

complete graph connected graph

subgraph

cycle tree

spanning tree

graph learning tool on Companion Website

Point Check

Point Key

M28_LIAN1878_11_GE_C28.indd 1064 12/29/17 4:49 PM

28.3 Representing Graphs 1065

Note The vertices can be objects of any type. For example, you can consider cities as objects that contain the information such as its name, population, and mayor. Thus, you may define vertices as follows:

City city0 = new City("Seattle", 608660, "Mike McGinn"); ... City city11 = new City("Houston", 2099451, "Annise Parker"); City[] vertices = {city0, city1, . . . , city11};

public class City { private String cityName; private int population; private String mayor;

public City(String cityName, int population, String mayor) { this.cityName = cityName; this.population = population; this.mayor = mayor; }

vertex type

Figure 28.4 You can use the tool to create a graph with mouse gestures and show DFS/BFS trees and shortest paths.

M28_LIAN1878_11_GE_C28.indd 1065 12/29/17 4:49 PM

1066 Chapter 28 Graphs and Applications

public String getCityName() { return cityName; }

public int getPopulation() { return population; }

public String getMayor() { return mayor; }

public void setMayor(String mayor) { this.mayor = mayor; }

public void setPopulation(int population) { this.population = population; } }

The vertices can be conveniently labeled using natural numbers 0, 1, 2, c , n - 1, for a graph of n vertices. Thus, vertices[0] represents "Seattle", vertices[1] represents "San Francisco", and so on, as shown in Figure 28.5.

Figure 28.5 An array stores the vertex names.

Seattle

San Francisco

Los Angeles

Dallas

Houston

Atlanta

New York

Boston

Chicago

Denver

Kansas City

Miami

vertices[0]

vertices[1]

vertices[2]

vertices[3]

vertices[4]

vertices[5]

vertices[6]

vertices[7]

vertices[8]

vertices[9]

vertices[10]

vertices[11]

Note You can reference a vertex by its name or its index, whichever is more convenient. Obviously, it is easy to access a vertex via its index in a program.

reference vertex

M28_LIAN1878_11_GE_C28.indd 1066 12/29/17 4:49 PM

28.3 Representing Graphs 1067

28.3.2 Representing Edges: Edge Array The edges can be represented using a two-dimensional array. For example, you can store all the edges in the graph in Figure 28.1 using the following array:

int[][] edges = { {0, 1}, {0, 3}, {0, 5}, {1, 0}, {1, 2}, {1, 3}, {2, 1}, {2, 3}, {2, 4}, {2, 10}, {3, 0}, {3, 1}, {3, 2}, {3, 4}, {3, 5}, {4, 2}, {4, 3}, {4, 5}, {4, 7}, {4, 8}, {4, 10}, {5, 0}, {5, 3}, {5, 4}, {5, 6}, {5, 7}, {6, 5}, {6, 7}, {7, 4}, {7, 5}, {7, 6}, {7, 8}, {8, 4}, {8, 7}, {8, 9}, {8, 10}, {8, 11}, {9, 8}, {9, 11}, {10, 2}, {10, 4}, {10, 8}, {10, 11}, {11, 8}, {11, 9}, {11, 10} };

This representation is known as the edge array. The vertices and edges in Figure 28.3a can be represented as follows:

String[] vertices = {"Peter", "Jane", "Mark", "Cindy", "Wendy"};

int[][] edges = {{0, 2}, {1, 2}, {2, 4}, {3, 4}};

28.3.3 Representing Edges: Edge Objects Another way to represent the edges is to define edges as objects and store the edges in a java.util.ArrayList. The Edge class can be defined as in Listing 28.1.

Listing 28.1 Edge.java public class Edge { int u; int v;

public Edge(int u, int v) { this.u = u; this.v = v; }

public boolean equals(Object o) { return u == ((Edge)o).u && v == ((Edge)o).v; } }

For example, you can store all the edges in the graph in Figure 28.1 using the following list:

java.util.ArrayList<Edge> list = new java.util.ArrayList<>(); list.add(new Edge(0, 1)); list.add(new Edge(0, 3)); list.add(new Edge(0, 5)); ...

Storing Edge objects in an ArrayList is useful if you don’t know the edges in advance. While representing edges using an edge array or Edge objects may be intuitive for input,

it’s not efficient for internal processing. The next two sections introduce the representation of graphs using adjacency matrices and adjacency lists. These two data structures are efficient for processing graphs.

edge array

M28_LIAN1878_11_GE_C28.indd 1067 12/29/17 4:49 PM

1068 Chapter 28 Graphs and Applications

28.3.4 Representing Edges: Adjacency Matrices Assume the graph has n vertices. You can use a two-dimensional n * n matrix, say adjacencyMatrix, to represent the edges. Each element in the array is 0 or 1. adjacencyMatrix[i][j] is 1 if there is an edge from vertex i to vertex j; otherwise, adjacencyMatrix[i][j] is 0. If the graph is undirected, the matrix is symmetric, because adjacencyMatrix[i][j] is the same as adjacencyMatrix[j][i]. For exam- ple, the edges in the graph in Figure 28.1 can be represented using an adjacency matrix as follows:

int[][] adjacencyMatrix = { {0, 1, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0}, // Seattle {1, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0}, // San Francisco {0, 1, 0, 1, 1, 1, 0, 0, 0, 0, 0, 0}, // Los Angeles {1, 1, 1, 0, 1, 1, 0, 0, 0, 0, 0, 0}, // Denver {0, 0, 1, 1, 0, 1, 0, 1, 1, 0, 1, 0}, // Kansas City {1, 0, 0, 1, 1, 0, 1, 1, 0, 0, 0, 0}, // Chicago {0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0}, // Boston {0, 0, 0, 0, 1, 1, 1, 0, 1, 0, 0, 0}, // New York {0, 0, 0, 1, 1, 0, 0, 1, 0, 1, 1, 1}, // Atlanta {0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1}, // Miami {0, 0, 1, 0, 1, 0, 0, 0, 1, 0, 0, 1}, // Dallas {0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 0} // Houston };

Note Since the matrix is symmetric for an undirected graph, to save storage you can use a ragged array.

The adjacency matrix for the directed graph in Figure 28.3a can be represented as follows:

int[][] a = {{0, 0, 1, 0, 0}, // Peter {0, 0, 1, 0, 0}, // Jane {0, 0, 0, 0, 1}, // Mark {0, 0, 0, 0, 1}, // Cindy {0, 0, 0, 0, 0} // Wendy };

28.3.5 Representing Edges: Adjacency Lists You can represent edges using adjacency vertex lists or adjacency edge lists. An adjacency vertex list for a vertex i contains the vertices that are adjacent to i and an adjacency edge list for a vertex i contains the edges that are adjacent to i. You may define an array of lists. The array has n entries, and each entry is a list. The adjacency vertex list for vertex i contains all the vertices j such that there is an edge from vertex i to vertex j. For example, to represent the edges in the graph in Figure 28.1, you can create an array of lists as follows:

java.util.List<Integer>[] neighbors = new java.util.List[12];

neighbors[0] contains all vertices adjacent to vertex 0 (i.e., Seattle), neighbors[1] contains all vertices adjacent to vertex 1 (i.e., San Francisco), and so on, as shown in Figure 28.6.

To represent the adjacency edge lists for the graph in Figure 28.1, you can create an array of lists as follows:

java.util.List<Edge>[] neighbors = new java.util.List[12];

neighbors[0] contains all edges adjacent to vertex 0 (i.e., Seattle), neighbors[1] contains all edges adjacent to vertex 1 (i.e., San Francisco), and so on, as shown in Figure 28.7.

adjacency matrix

ragged array

adjacency vertex lists adjacency edge lists

M28_LIAN1878_11_GE_C28.indd 1068 12/29/17 4:49 PM

28.3 Representing Graphs 1069

Note You can represent a graph using an adjacency matrix or adjacency lists. Which one is better? If the graph is dense (i.e., there are a lot of edges), using an adjacency matrix is preferred. If the graph is very sparse (i.e., very few edges), using adjacency lists is better, because using an adjacency matrix would waste a lot of space.

Both adjacency matrices and adjacency lists can be used in a program to make algorithms more efficient. For example, it takes O(1) constant time to check whether two vertices are connected using an adjacency matrix, and it takes linear time O(m) to print all edges in a graph using adjacency lists, where m is the number of edges.

adjacency matrices vs. adjacency lists

Figure 28.6 Edges in the graph in Figure 28.1 are represented using adjacency vertex lists.

Seattle

San Francisco

Los Angeles

Dallas

Houston

Atlanta

New York

Boston

Chicago

Denver

Kansas City

Miami

neighbors[0]

neighbors[1]

neighbors[2]

neighbors[3]

neighbors[4]

neighbors[5]

neighbors[6]

neighbors[7]

neighbors[8]

neighbors[9]

neighbors[10]

neighbors[11]

Figure 28.7 Edges in the graph in Figure 28.1 are represented using adjacency edge lists.

Edge (0, 1) Edge (0, 3) Edge (0, 5)

Edge (1, 0) Edge (1, 2) Edge (1, 3)

Edge (2, 1) Edge (2, 3) Edge (2, 4) Edge (2, 10)

Edge (3, 0) Edge (3, 1) Edge (3, 2) Edge (3, 4) Edge (3, 5)

Edge (4, 2) Edge (4, 3) Edge (4, 5) Edge (4, 7) Edge (4, 8) Edge (4, 10)

Edge (5, 0) Edge (5, 3) Edge (5, 4) Edge (5, 6) Edge (5, 7)

Edge (6, 5) Edge (6, 7)

Edge (7, 4) Edge (7, 5) Edge (7, 6) Edge (7, 8)

Edge (8, 4) Edge (8, 7) Edge (8, 9) Edge (8, 10) Edge (8, 11)

Edge (9, 8) Edge (9, 11)

Edge (10, 2) Edge (10, 4) Edge (10, 8) Edge (10, 11)

Edge (11, 8) Edge (11, 9) Edge (11, 10)

neighbors[0]

neighbors[1]

neighbors[2]

neighbors[3]

neighbors[4]

neighbors[5]

neighbors[6]

neighbors[7]

neighbors[8]

neighbors[9]

neighbors[10]

neighbors[11]

Seattle

San Francisco

Los Angeles

Denver

Kansas City

Chicago

Boston

New York

Atlanta

Miami

Dallas

Houston

M28_LIAN1878_11_GE_C28.indd 1069 12/29/17 4:49 PM

1070 Chapter 28 Graphs and Applications

Note Adjacency vertex list is simpler for representing unweighted graphs. However, adjacency edge lists are more flexible for a wide range of graph applications. It is easy to add additional constraints on edges using adjacency edge lists. For this reason, this book will use adjacency edge lists to represent graphs.

You can use arrays, array lists, or linked lists to store adjacency lists. We will use lists instead of arrays, because the lists are easily expandable to enable you to add new vertices. Further we will use array lists instead of linked lists, because our algorithms only require searching for adjacent vertices in the list. Using array lists is more efficient for our algorithms. Using array lists, the adjacency edge list in Figure 28.6 can be built as follows:

List<ArrayList<Edge>> neighbors = new ArrayList<>(); neighbors.add(new ArrayList<Edge>()); neighbors.get(0).add(new Edge(0, 1)); neighbors.get(0).add(new Edge(0, 3)); neighbors.get(0).add(new Edge(0, 5)); neighbors.add(new ArrayList<Edge>()); neighbors.get(1).add(new Edge(1, 0)); neighbors.get(1).add(new Edge(1, 2)); neighbors.get(1).add(new Edge(1, 3)); ... ... neighbors.get(11).add(new Edge(11, 8)); neighbors.get(11).add(new Edge(11, 9)); neighbors.get(11).add(new Edge(11, 10));

28.3.1 How do you represent vertices in a graph? How do you represent edges using an edge array? How do you represent an edge using an edge object? How do you represent edges using an adjacency matrix? How do you represent edges using adjacency lists?

28.3.2 Represent the following graph using an edge array, a list of edge objects, an adjacency matrix, an adjacency vertex list, and an adjacency edge list, respectively.

adjacency vertex lists vs. adjacency edge lists

using ArrayList

Point Check

Figure 28.8 Common operations of graphs are defined in the interface and concrete classes define concrete graphs.

Graph WeightedGraphUnweightedGraph

28.4 Modeling Graphs The Graph interface defines the common operations for a graph.

The Java Collections Framework serves as a good example for designing complex data structures. The common features of data structures are defined in the interfaces (e.g., Collection, Set, List, Queue), as shown in Figure 20.1. This design pattern is useful for modeling graphs. We will define an interface named Graph that contains all the common operations of graphs. Many concrete graphs can be added to the design. For example, we will define such graphs named UnweightedGraph and WeightedGraph. The relationships of these interfaces and classes are illustrated in Figure 28.8.

Point Key

M28_LIAN1878_11_GE_C28.indd 1070 12/29/17 4:49 PM

28.4 Modeling Graphs 1071

What are the common operations for a graph? In general, you need to get the number of vertices in a graph, get all vertices in a graph, get the vertex object with a specified index, get the index of the vertex with a specified name, get the neighbors for a vertex, get the degree for a vertex, clear the graph, add a new vertex, add a new edge, perform a depth-first search, and perform a breadth-first search. Depth-first search and breadth-first search will be introduced in the next section. Figure 28.9 illustrates these methods in the UML diagram.

UnweightedGraph does not introduce any new methods. A list of vertices and an edge adjacency list are defined in the class. With these data fields, it is sufficient to implement all the methods defined in the Graph interface. For convenience, we assume the graph is a simple graph, that is, a vertex has no edge to itself and there are no parallel edges from vertex u to v.

Figure 28.9 The Graph interface defines the common operations for all types of graphs.

+getSize(): int

+getVertices(): List<V>

+getVertex(index: int): V

+getIndex(v: V): int

+getNeighbors(index: int): List<Integer>

+getDegree(index: int): int

+printEdges(): void

+clear(): void

+addVertex(v: V): boolean

+addEdge(u: int, v: int): boolean

+addEdge(e: Edge): boolean

+remove(v: V): boolean

+remove(u: int, v: int): boolean

+dfs(v: int): UnWeightedGraph<V>.SearchTree

+bfs(v: int): UnWeightedGraph<V>.SearchTree

Returns the number of vertices in the graph.

Returns the vertices in the graph.

Returns the vertex object for the speci�ed vertex index.

Returns the index for the speci�ed vertex and return –1 if v is not in the graph.

Returns the neighbors of vertex with the speci�ed index.

Returns the degree for a speci�ed vertex index.

Prints the edges.

Clears the graph.

Returns true if v is added to the graph. Returns false if v is already in the graph.

Adds an edge from u to v to the graph. Throws IllegalArgumentException if u or v is invalid. Returns true if the edge is added and false if (u, v) is already in the graph. Adds an edge into the adjacency edge list.

Removes a vertex from the graph.

Removes an edge from the graph. Obtains a depth-�rst search tree starting from v.

Obtains a breadth-�rst search tree starting from v.

«interface» Graph<V>

#vertices: List<V> #neighbors: List<List<Edge>>

UnweightedGraph<V>

Vertices in the graph. Neighbors for each vertex in the graph.

Constructs an empty graph. Constructs a graph with the speci�ed edges and vertices stored in arrays. Constructs a graph with the speci�ed edges and vertices stored in lists.

Constructs a graph with the speci�ed edges in an array and the integer vertices 1, 2, .... Constructs a graph with the speci�ed edges in a list and the integer vertices 1, 2, ….

The generic type V is the type for vertices.

+UnweightedGraph() +UnweightedGraph(vertices: V[], edges: int[][]) +UnweightedGraph(vertices: List<V>, edges: List<Edge>)

+UnweightedGraph(edges: int[][], numberOfVertices: int) +UnweightedGraph(edges: List<Edge>, numberOfVertices: int)

M28_LIAN1878_11_GE_C28.indd 1071 12/29/17 4:49 PM

1072 Chapter 28 Graphs and Applications

Note You can create a graph with any type of vertices. Each vertex is associated with an index, which is the same as the index of the vertex in the vertices list. If you create a graph without specifying the vertices, the vertices are the same as their indices.

Assume the Graph interface and the UnweightedGraph class are available. Listing 28.2 gives a test program that creates the graph in Figure 28.1 and another graph for the one in Figure 28.3a.

Listing 28.2 TestGraph.java 1 public class TestGraph { 2 public static void main(String[] args) { 3 String[] vertices = {"Seattle", "San Francisco", "Los Angeles", 4 "Denver", "Kansas City", "Chicago", "Boston", "New York", 5 "Atlanta", "Miami", "Dallas", "Houston"}; 6 7 // Edge array for graph in Figure 28.1 8 int[][] edges = { 9 {0, 1}, {0, 3}, {0, 5}, 10 {1, 0}, {1, 2}, {1, 3}, 11 {2, 1}, {2, 3}, {2, 4}, {2, 10}, 12 {3, 0}, {3, 1}, {3, 2}, {3, 4}, {3, 5}, 13 {4, 2}, {4, 3}, {4, 5}, {4, 7}, {4, 8}, {4, 10}, 14 {5, 0}, {5, 3}, {5, 4}, {5, 6}, {5, 7}, 15 {6, 5}, {6, 7}, 16 {7, 4}, {7, 5}, {7, 6}, {7, 8}, 17 {8, 4}, {8, 7}, {8, 9}, {8, 10}, {8, 11}, 18 {9, 8}, {9, 11}, 19 {10, 2}, {10, 4}, {10, 8}, {10, 11}, 20 {11, 8}, {11, 9}, {11, 10} 21 }; 22 23 Graph<String> graph1 = new UnweightedGraph<>(vertices, edges); 24 System.out.println("The number of vertices in graph1: " 25 + graph1.getSize()); 26 System.out.println("The vertex with index 1 is " 27 + graph1.getVertex(1)); 28 System.out.println("The index for Miami is " + 29 graph1.getIndex("Miami")); 30 System.out.println("The edges for graph1:"); 31 graph1.printEdges(); 32 33 // List of Edge objects for graph in Figure 28.3a 34 String[] names = {"Peter", "Jane", "Mark", "Cindy", "Wendy"}; 35 java.util.ArrayList<Edge> edgeList 36 = new java.util.ArrayList<>(); 37 edgeList.add(new Edge(0, 2)); 38 edgeList.add(new Edge(1, 2)); 39 edgeList.add(new Edge(2, 4)); 40 edgeList.add(new Edge(3, 4)); 41 // Create a graph with 5 vertices 42 Graph<String> graph2 = new UnweightedGraph<> 43 (java.util.Arrays.asList(names), edgeList); 44 System.out.println("\nThe number of vertices in graph2: " 45 + graph2.getSize()); 46 System.out.println("The edges for graph2:"); 47 graph2.printEdges(); 48 } 49 }

vertices and their indices

vertices

edges

create a graph

number of vertices

get vertex

get index

print edges

list of Edge objects

create a graph

print edges

M28_LIAN1878_11_GE_C28.indd 1072 12/29/17 4:49 PM

28.4 Modeling Graphs 1073

The program creates graph1 for the graph in Figure 28.1 in lines 3–23. The vertices for graph1 are defined in lines 3–5. The edges for graph1 are defined in 8–21. The edges are represented using a two-dimensional array. For each row i in the array, edges[i][0] and edges[i][1] indicate there is an edge from vertex edges[i][0] to vertex edges[i] [1]. For example, the first row, {0, 1}, represents the edge from vertex 0 (edges[0][0]) to vertex 1 (edges[0][1]). The row {0, 5} represents the edge from vertex 0 (edges[2] [0]) to vertex 5 (edges[2][1]). The graph is created in line 23. Line 31 invokes the printEdges() method on graph1 to display all edges in graph1.

The program creates graph2 for the graph in Figure 28.3a in lines 34–43. The edges for graph2 are defined in lines 37–40. graph2 is created using a list of Edge objects in line 43. Line 47 invokes the printEdges() method on graph2 to display all edges in graph2.

Note both graph1 and graph2 contain the vertices of strings. The vertices are associated with indices 0, 1, . . . , n−1. The index is the location of the vertex in vertices. For example, the index of vertex Miami is 9.

Now, we turn our attention to implementing the interface and classes. Listings 28.3 and 28.4 give the Graph interface and the UnweightedGraph class, respectively.

Listing 28.3 Graph.java 1 public interface Graph<V> { 2 /** Return the number of vertices in the graph */ 3 public int getSize(); 4 5 /** Return the vertices in the graph */ 6 public java.util.List<V> getVertices(); 7 8 /** Return the object for the specified vertex index */ 9 public V getVertex(int index); 10 11 /** Return the index for the specified vertex object */

getSize

getVertices

getVertex

The number of vertices in graph1: 12 The vertex with index 1 is San Francisco The index for Miami is 9 The edges for graph1: Seattle (0): (0, 1) (0, 3) (0, 5) San Francisco (1): (1, 0) (1, 2) (1, 3) Los Angeles (2): (2, 1) (2, 3) (2, 4) (2, 10) Denver (3): (3, 0) (3, 1) (3, 2) (3, 4) (3, 5) Kansas City (4): (4, 2) (4, 3) (4, 5) (4, 7) (4, 8) (4, 10) Chicago (5): (5, 0) (5, 3) (5, 4) (5, 6) (5, 7) Boston (6): (6, 5) (6, 7) New York (7): (7, 4) (7, 5) (7, 6) (7, 8) Atlanta (8): (8, 4) (8, 7) (8, 9) (8, 10) (8, 11) Miami (9): (9, 8) (9, 11) Dallas (10): (10, 2) (10, 4) (10, 8) (10, 11) Houston (11): (11, 8) (11, 9) (11, 10)

The number of vertices in graph2: 5 The edges for graph2: Peter (0): (0, 2) Jane (1): (1, 2) Mark (2): (2, 4) Cindy (3): (3, 4) Wendy (4):

M28_LIAN1878_11_GE_C28.indd 1073 12/29/17 4:49 PM

1074 Chapter 28 Graphs and Applications

12 public int getIndex(V v); 13 14 /** Return the neighbors of vertex with the specified index */ 15 public java.util.List<Integer> getNeighbors(int index); 16 17 /** Return the degree for a specified vertex */ 18 public int getDegree(int v); 19 20 /** Print the edges */ 21 public void printEdges(); 22 23 /** Clear the graph */ 24 public void clear(); 25 26 /** Add a vertex to the graph */ 27 public boolean addVertex(V vertex); 28 29 /** Add an edge to the graph */ 30 public boolean addEdge(int u, int v); 31 32 /** Add an edge to the graph */ 33 public boolean addEdge(Edge e); 34 35 /** Remove a vertex v from the graph, return true if successful */ 36 public boolean remove(V v); 37 38 /** Remove an edge (u, v) from the graph, return true if successful */ 39 public boolean remove(int u, int v); 40 41 /** Obtain a depth-first search tree */ 42 public UnweightedGraph<V>.SearchTree dfs(int v); 43 44 /** Obtain a breadth-first search tree */ 45 public UnweightedGraph<V>.SearchTree bfs(int v); 46 }

Listing 28.4 UnweightedGraph.java 1 import java.util.*; 2 3 public class UnweightedGraph<V> implements Graph<V> { 4 protected List<V> vertices = new ArrayList<>(); // Store vertices 5 protected List<List<Edge>> neighbors 6 = new ArrayList<>(); // Adjacency Edge lists 7 8 /** Construct an empty graph */ 9 protected UnweightedGraph() { 10 } 11 12 /** Construct a graph from vertices and edges stored in arrays */ 13 protected UnweightedGraph(V[] vertices, int[][] edges) { 14 for (int i = 0; i < vertices.length; i++) 15 addVertex(vertices[i]); 16 17 createAdjacencyLists(edges, vertices.length); 18 } 19 20 /** Construct a graph from vertices and edges stored in List */ 21 protected UnweightedGraph(List<V> vertices, List<Edge> edges) { 22 for (int i = 0; i < vertices.size(); i++)

getIndex

getNeighbors

getDegree

printEdges

clear

addVertex

addEdge

remove vertex

remove edge

dfs

bfs

no-arg constructor

constructor

M28_LIAN1878_11_GE_C28.indd 1074 12/29/17 4:49 PM

28.4 Modeling Graphs 1075

23 addVertex(vertices.get(i)); 24 25 createAdjacencyLists(edges, vertices.size()); 26 } 27 28 /** Construct a graph for integer vertices 0, 1, 2 and edge list */ 29 protected UnweightedGraph(List<Edge> edges, int numberOfVertices) { 30 for (int i = 0; i < numberOfVertices; i++) 31 addVertex((V)(new Integer(i))); // vertices is {0, 1, . . . } 32 33 createAdjacencyLists(edges, numberOfVertices); 34 } 35 36 /** Construct a graph from integer vertices 0, 1, and edge array */ 37 protected UnweightedGraph(int[][] edges, int numberOfVertices) { 38 for (int i = 0; i < numberOfVertices; i++) 39 addVertex((V)(new Integer(i))); // vertices is {0, 1, . . . } 40 41 createAdjacencyLists(edges, numberOfVertices); 42 } 43 44 /** Create adjacency lists for each vertex */ 45 private void createAdjacencyLists( 46 int[][] edges, int numberOfVertices) { 47 for (int i = 0; i < edges.length; i++) { 48 addEdge(edges[i][0], edges[i][1]); 49 } 50 } 51 52 /** Create adjacency lists for each vertex */ 53 private void createAdjacencyLists( 54 List<Edge> edges, int numberOfVertices) { 55 for (Edge edge: edges) { 56 addEdge(edge.u, edge.v); 57 } 58 } 59 60 @Override /** Return the number of vertices in the graph */ 61 public int getSize() { 62 return vertices.size(); 63 } 64 65 @Override /** Return the vertices in the graph */ 66 public List<V> getVertices() { 67 return vertices; 68 } 69 70 @Override /** Return the object for the specified vertex */ 71 public V getVertex(int index) { 72 return vertices.get(index); 73 } 74 75 @Override /** Return the index for the specified vertex object */ 76 public int getIndex(V v) { 77 return vertices.indexOf(v); 78 } 79 80 @Override /** Return the neighbors of the specified vertex */ 81 public List<Integer> getNeighbors(int index) { 82 List<Integer> result = new ArrayList<>();

constructor

getSize

getVertices

getVertex

getIndex

getNeighbors

M28_LIAN1878_11_GE_C28.indd 1075 12/29/17 4:49 PM

1076 Chapter 28 Graphs and Applications

83 for (Edge e: neighbors.get(index)) 84 result.add(e.v); 85 86 return result; 87 } 88 89 @Override /** Return the degree for a specified vertex */ 90 public int getDegree(int v) { 91 return neighbors.get(v).size(); 92 } 93 94 @Override /** Print the edges */ 95 public void printEdges() { 96 for (int u = 0; u < neighbors.size(); u++) { 97 System.out.print(getVertex(u) + " (" + u + "): "); 98 for (Edge e: neighbors.get(u)) { 99 System.out.print("(" + getVertex(e.u) + ", " + 100 getVertex(e.v) + ") "); 101 } 102 System.out.println(); 103 } 104 } 105 106 @Override /** Clear the graph */ 107 public void clear() { 108 vertices.clear(); 109 neighbors.clear(); 110 } 111 112 @Override /** Add a vertex to the graph */ 113 public boolean addVertex(V vertex) { 114 if (!vertices.contains(vertex)) { 115 vertices.add(vertex); 116 neighbors.add(new ArrayList<Edge>()); 117 return true; 118 } 119 else { 120 return false; 121 } 122 } 123 124 @Override /** Add an edge to the graph */ 125 public boolean addEdge(Edge e) { 126 if (e.u < 0 || e.u > getSize() − 1) 127 throw new IllegalArgumentException("No such index: " + e.u); 128 129 if (e.v < 0 || e.v > getSize() − 1) 130 throw new IllegalArgumentException("No such index: " + e.v); 131 132 if (!neighbors.get(e.u).contains(e)) { 133 neighbors.get(e.u).add(e); 134 return true; 135 } 136 else { 137 return false; 138 } 139 } 140 141 @Override /** Add an edge to the graph */ 142 public boolean addEdge(int u, int v) {

getDegree

printEdges

clear

addVertex

addEdge

addEdge overloaded

M28_LIAN1878_11_GE_C28.indd 1076 12/29/17 4:49 PM

28.4 Modeling Graphs 1077

143 return addEdge(new Edge(u, v)); 144 } 145 146 @Override /** Obtain a DFS tree starting from vertex v */ 147 /** To be discussed in Section 28.7 */ 148 public SearchTree dfs(int v) { 149 List<Integer> searchOrder = new ArrayList<>(); 150 int[] parent = new int[vertices.size()]; 151 for (int i = 0; i < parent.length; i++) 152 parent[i] = −1; // Initialize parent[i] to −1 153 154 // Mark visited vertices 155 boolean[] isVisited = new boolean[vertices.size()]; 156 157 // Recursively search 158 dfs(v, parent, searchOrder, isVisited); 159 160 // Return a search tree 161 return new SearchTree(v, parent, searchOrder); 162 } 163 164 /** Recursive method for DFS search */ 165 private void dfs(int v, int[] parent, List<Integer> searchOrder, 166 boolean[] isVisited) { 167 // Store the visited vertex 168 searchOrder.add(v); 169 isVisited[v] = true; // Vertex v visited 170 171 for (Edge e : neighbors.get(v)) {// e.u is v 172 if (!isVisited[e.v]) {// e.v is w in Listing 28.8 173 parent[e.v] = v; // The parent of vertex w is v 174 dfs(e.v, parent, searchOrder, isVisited); // Recursive search 175 } 176 } 177 } 178 179 @Override /** Starting bfs search from vertex v */ 180 /** To be discussed in Section 28.9 */ 181 public SearchTree bfs(int v) { 182 List<Integer> searchOrder = new ArrayList<>(); 183 int[] parent = new int[vertices.size()]; 184 for (int i = 0; i < parent.length; i++) 185 parent[i] = -1; // Initialize parent[i] to -1 186 187 java.util.LinkedList<Integer> queue = 188 new java.util.LinkedList<>(); // list used as a queue 189 boolean[] isVisited = new boolean[vertices.size()]; 190 queue.offer(v); // Enqueue v 191 isVisited[v] = true; // Mark it visited 192 193 while (!queue.isEmpty()) { 194 int u = queue.poll(); // Dequeue to u 195 searchOrder.add(u); // u searched 196 for (Edge e: neighbors.get(u)) {// Note that e.u is u 197 if (!isVisited[e.v]) {// e.v is w in Listing 28.11 198 queue.offer(e.v); // Enqueue w 199 parent[e.v] = u; // The parent of w is u 200 isVisited[e.v] = true; // Mark it visited 201 } 202 }

dfs method

bfs method

e.u e.v

e.v

e.u e.v

e.v

M28_LIAN1878_11_GE_C28.indd 1077 12/29/17 4:49 PM

1078 Chapter 28 Graphs and Applications

203 } 204 205 return new SearchTree(v, parent, searchOrder); 206 } 207 208 /** Tree inner class inside the UnweightedGraph class */ 209 /** To be discussed in Section 28.6 */ 210 public class SearchTree { 211 private int root; // The root of the tree 212 private int[] parent; // Store the parent of each vertex 213 private List<Integer> searchOrder; // Store the search order 214 215 /** Construct a tree with root, parent, and searchOrder */ 216 public SearchTree(int root, int[] parent, 217 List<Integer> searchOrder) { 218 this.root = root; 219 this.parent = parent; 220 this.searchOrder = searchOrder; 221 } 222 223 /** Return the root of the tree */ 224 public int getRoot() { 225 return root; 226 } 227 228 /** Return the parent of vertex v */ 229 public int getParent(int v) { 230 return parent[v]; 231 } 232 233 /** Return an array representing search order */ 234 public List<Integer> getSearchOrder() { 235 return searchOrder; 236 } 237 238 /** Return number of vertices found */ 239 public int getNumberOfVerticesFound() { 240 return searchOrder.size(); 241 } 242 243 /** Return the path of vertices from a vertex to the root */ 244 public List<V> getPath(int index) { 245 ArrayList<V> path = new ArrayList<>(); 246 247 do { 248 path.add(vertices.get(index)); 249 index = parent[index]; 250 } 251 while (index != −1); 252 253 return path; 254 } 255 256 /** Print a path from the root to vertex v */ 257 public void printPath(int index) { 258 List<V> path = getPath(index); 259 System.out.print("A path from " + vertices.get(root) + " to " + 260 vertices.get(index) + ": "); 261 for (int i = path.size() − 1; i >= 0; i−−) 262 System.out.print(path.get(i) + " "); 263 }

SearchTree inner class

M28_LIAN1878_11_GE_C28.indd 1078 12/29/17 4:49 PM

28.4 Modeling Graphs 1079

264 265 /** Print the whole tree */ 266 public void printTree() { 267 System.out.println("Root is: " + vertices.get(root)); 268 System.out.print("Edges: "); 269 for (int i = 0; i < parent.length; i++) { 270 if (parent[i] != −1) { 271 // Display an edge 272 System.out.print("(" + vertices.get(parent[i]) + ", " + 273 vertices.get(i) + ") "); 274 } 275 } 276 System.out.println(); 277 } 278 } 279 280 @Override /** Remove vertex v and return true if successful */ 281 public boolean remove(V v) { 282 return true; // Implementation left as an exercise 283 } 284 285 @Override /** Remove edge (u, v) and return true if successful */ 286 public boolean remove(int u, int v) { 287 return true; // Implementation left as an exercise 289 } 290 }

The code in the Graph interface in Listing 28.3 is straightforward. Let us digest the code in the UnweightedGraph class in Listing 28.4.

The UnweightedGraph class defines the data field vertices (line 4) to store verti- ces and neighbors (line 5) to store edges in adjacency edges lists. neighbors.get(i) stores all edges adjacent to vertex i. Four overloaded constructors are defined in lines 9–42 to create a default graph, or a graph from arrays or lists of edges and vertices. The createAdjacencyLists(int[][] edges, int numberOfVertices) method cre- ates adjacency lists from edges in an array (lines 45–50). The createAdjacencyLists (List<Edge> edges, int numberOfVertices) method creates adjacency lists from edges in a list (lines 53–58).

The getNeighbors(u) method (lines 81–87) returns a list of vertices adjacent to vertex u. The clear() method (lines 106–110) removes all vertices and edges from the graph. The addVertex(u) method (lines 112–122) adds a new vertex to vertices and returns true. It returns false if the vertex is already in the graph (line 120).

The addEdge(e) method (lines 124–139) adds a new edge to the adjacency edge list and returns true. It returns false if the edge is already in the graph. This method may throw IllegalArgumentExcepiton if the edge is invalid (lines 126–130).

The addEdge(u, v) method (lines 141–144) adds an edge (u, v) to the graph. If a graph is undirected, you should invoke addEdge(u, v) and addEdge(v, u) to add an edge between u and v.

The printEdges() method (lines 95–104) displays all vertices and edges adjacent to each vertex.

The code in lines 148–278 gives the methods for finding a depth-first search tree and a breadth-first search tree, which will be introduced in Sections 28.7 and 28.9, respectively.

28.4.1 Describe the methods in Graph and UnweightedGraph. 28.4.2 For the code in Listing 28.2, TestGraph.java, what is graph1.

getIndex("Seattle")? What is graph1.getDegree(5)? What is graph1. getVertex(4)?

Point Check

M28_LIAN1878_11_GE_C28.indd 1079 12/29/17 4:49 PM

1080 Chapter 28 Graphs and Applications

28.4.3 Show the output of the following code: public class Test { public static void main(String[] args) { Graph<Character> graph = new UnweightedGraph<>(); graph.addVertex('U'); graph.addVertex('V'); int indexForU = graph.getIndex('U'); int indexForV = graph.getIndex('V'); System.out.println("indexForU is " + indexForU); System.out.println("indexForV is " + indexForV); graph.addEdge(indexForU, indexForV); System.out.println("Degree of U is " + graph.getDegree(indexForU)); System.out.println("Degree of V is " + graph.getDegree(indexForV)); } }

28.4.4 What will getIndex(v) return if v is not in the graph? What happens to getVertex(index) if index is not in the graph? What happens to addVertex(v) if v is already in the graph? What happens to addEdge(u, v) if u or v is not in the graph?

28.5 Graph Visualization To display a graph visually, each vertex must be assigned a location.

The preceding section introduced the Graph interface and the UnweightedGraph class. This section discusses how to display graphs graphically. In order to display a graph, you need to know where each vertex is displayed and the name of each vertex. To ensure a graph can be displayed, we define an interface named Displayable that has the methods for obtaining the x- and y-coordinates and their names, and make vertices instances of Displayable, in Listing 28.5.

Listing 28.5 Displayable.java 1 public interface Displayable { 2 public double getX(); // Get x-coordinate of the vertex 3 public double getY(); // Get y-coordinate of the vertex 4 public String getName(); // Get display name of the vertex 5 }

A graph with Displayable vertices can now be displayed on a pane named GraphView, as shown in Listing 28.6.

Listing 28.6 GraphView.java 1 import javafx.scene.Group; 2 import javafx.scene.layout.BorderPane; 3 import javafx.scene.shape.Circle; 4 import javafx.scene.shape.Line; 5 import javafx.scene.text.Text; 6 7 public class GraphView extends BorderPane { 8 private Graph<? extends Displayable> graph; 9 private Group group = new Group(); 10 11 public GraphView(Graph<? extends Displayable> graph) { 12 this.graph = graph; 13 this.setCenter(group); // Center the group

Point Key

Displayable interface

extends BorderPane Displayable vertices

M28_LIAN1878_11_GE_C28.indd 1080 12/29/17 4:49 PM

28.5 Graph Visualization 1081

14 repaintGraph(); 15 } 16 17 private void repaintGraph() { 18 group.getChildren().clear(); // Clear group for a new display 19 20 // Draw vertices and text for vertices 21 java.util.List<? extends Displayable> vertices 22 = graph.getVertices(); 23 for (int i = 0; i < graph.getSize(); i++) { 24 double x = vertices.get(i).getX(); 25 double y = vertices.get(i).getY(); 26 String name = vertices.get(i).getName(); 27 28 group.getChildren().add(new Circle(x, y, 16)); 29 group.getChildren().add(new Text(x − 8, y − 18, name)); 30 } 31 32 // Draw edges for pairs of vertices 33 for (int i = 0; i < graph.getSize(); i++) { 34 java.util.List<Integer> neighbors = graph.getNeighbors(i); 35 double x1 = graph.getVertex(i).getX(); 36 double y1 = graph.getVertex(i).getY(); 37 for (int v: neighbors) { 38 double x2 = graph.getVertex(v).getX(); 39 double y2 = graph.getVertex(v).getY(); 40 41 // Draw an edge for (i, v) 42 group.getChildren().add(new Line(x1, y1, x2, y2)); 43 } 44 } 45 } 46 }

To display a graph on a pane, simply create an instance of GraphView by passing the graph as an argument in the constructor (line 11). The class for the graph’s vertex must implement the Displayable interface in order to display the vertices (lines 21–44). For each vertex index i, invoking graph.getNeighbors(i) returns its adjacency list (line 34). From this list, you can find all vertices that are adjacent to i and draw a line to connect i with its adjacent vertex (lines 35–42).

Listing 28.7 gives an example of displaying the graph in Figure 28.1, as shown in Figure 28.10.

Figure 28.10 The graph is displayed in the pane.

display a vertex display a text

draw an edge

M28_LIAN1878_11_GE_C28.indd 1081 12/29/17 4:49 PM

1082 Chapter 28 Graphs and Applications

Listing 28.7 DisplayUSMap.java 1 import javafx.application.Application; 2 import javafx.scene.Scene; 3 import javafx.stage.Stage; 4 5 public class DisplayUSMap extends Application { 6 @Override // Override the start method in the Application class 7 public void start(Stage primaryStage) { 8 City[] vertices = {new City("Seattle", 75, 50), 9 new City("San Francisco", 50, 210), 10 new City("Los Angeles", 75, 275), new City("Denver", 275, 175), 11 new City("Kansas City", 400, 245), 12 new City("Chicago", 450, 100), new City("Boston", 700, 80), 13 new City("New York", 675, 120), new City("Atlanta", 575, 295), 14 new City("Miami", 600, 400), new City("Dallas", 408, 325), 15 new City("Houston", 450, 360) }; 16 17 // Edge array for graph in Figure 28.1 18 int[][] edges = { 19 {0, 1}, {0, 3}, {0, 5}, {1, 0}, {1, 2}, {1, 3}, 20 {2, 1}, {2, 3}, {2, 4}, {2, 10}, 21 {3, 0}, {3, 1}, {3, 2}, {3, 4}, {3, 5}, 22 {4, 2}, {4, 3}, {4, 5}, {4, 7}, {4, 8}, {4, 10}, 23 {5, 0}, {5, 3}, {5, 4}, {5, 6}, {5, 7}, 24 {6, 5}, {6, 7}, {7, 4}, {7, 5}, {7, 6}, {7, 8}, 25 {8, 4}, {8, 7}, {8, 9}, {8, 10}, {8, 11}, 26 {9, 8}, {9, 11}, {10, 2}, {10, 4}, {10, 8}, {10, 11}, 27 {11, 8}, {11, 9}, {11, 10} 28 }; 29 30 Graph<City> graph = new UnweightedGraph<>(vertices, edges); 31 32 // Create a scene and place it in the stage 33 Scene scene = new Scene(new GraphView(graph), 750, 450); 34 primaryStage.setTitle("DisplayUSMap"); // Set the stage title 35 primaryStage.setScene(scene); // Place the scene in the stage 36 primaryStage.show(); // Display the stage 37 } 38 39 static class City implements Displayable { 40 private double x, y; 41 private String name; 42 43 City(String name, double x, double y) { 44 this.name = name; 45 this.x = x; 46 this.y = y; 47 } 48 49 @Override 50 public double getX() { 51 return x; 52 } 53 54 @Override 55 public double getY() { 56 return y; 57 } 58

create a graph

create a GraphView

City class

M28_LIAN1878_11_GE_C28.indd 1082 12/29/17 4:49 PM

28.6 Graph Traversals 1083

59 @Override 60 public String getName() { 61 return name; 62 } 63 } 64 }

The class City is defined to model the vertices with their coordinates and names (lines 39–63). The program creates a graph with the vertices of the City type (line 30). Since City implements Displayable, a GraphView object created for the graph displays the graph in the pane (line 33).

As an exercise to get acquainted with the graph classes and interfaces, add a city (e.g., Savannah) with appropriate edges into the graph.

28.5.1 Will Listing 28.7, DisplayUSMap.java work, if the code in lines 38–42 in Listing 28.6, GraphView.java is replaced by the following code?

if (i < v) { double x2 = graph.getVertex(v).getX(); double y2 = graph.getVertex(v).getY();

// Draw an edge for (i, v) getChildren().add(new Line(x1, y1, x2, y2)); }

28.5.2 For the graph1 object created in Listing 28.1, TestGraph.java, can you create a GraphView object as follows?

GraphView view = new GraphView(graph1);

28.6 Graph Traversals Depth-first and breadth-first are two common ways to traverse a graph.

Graph traversal is the process of visiting each vertex in the graph exactly once. There are two popular ways to traverse a graph: depth-first traversal (or depth-first search) and breadth-first traversal (or breadth-first search). Both traversals result in a spanning tree, which can be modeled using a class, as shown in Figure 28.11. Note SearchTree is an inner class defined in the UnweightedGraph class. UnweightedGraph<V>.SearchTree is different from the Tree interface defined in Section 25.2.5. UnweightedGraph<V>.SearchTree is a special- ized class designed for describing the parent–child relationship of the nodes, whereas the Tree interface defines common operations such as searching, inserting, and deleting in a tree. Since there is no need to perform these operations for a spanning tree, UnweightedGraph<V>. SearchTree is not defined as a subtype of Tree.

The SearchTree class is defined as an inner class in the UnweightedGraph class in lines 210–278 in Listing 28.4. The constructor creates a tree with the root, edges, and a search order.

The SearchTree class defines seven methods. The getRoot() method returns the root of the tree. You can get the order of the vertices searched by invoking the getSearchOrder() method. You can invoke getParent(v) to find the parent of vertex v in the search. Invoking getNumber- OfVerticesFound() returns the number of vertices searched. The method getPath(index) returns a list of vertices from the specified vertex index to the root. Invoking printPath(v) displays a path from the root to v. You can display all edges in the tree using the printTree() method.

Sections 28.7 and 28.9 will introduce depth-first search and breadth-first search, respec- tively. Both searches will result in an instance of the SearchTree class.

28.6.1 Does UnweightedGraph<V>.SearchTree implement the Tree interface defined in Listing 25.3, Tree.java?

28.6.2 What method do you use to find the parent of a vertex in the tree?

Point Check

Point Key

depth-first search breadth-first search

Point Check

M28_LIAN1878_11_GE_C28.indd 1083 12/29/17 4:49 PM

1084 Chapter 28 Graphs and Applications

28.7 Depth-First Search (DFS) The depth-first search of a graph starts from a vertex in the graph and visits all verti- ces in the graph as far as possible before backtracking.

The depth-first search of a graph is like the depth-first search of a tree discussed in Section 25.2.4, Tree Traversal. In the case of a tree, the search starts from the root. In a graph, the search can start from any vertex.

A depth-first search of a tree first visits the root, then recursively visits the subtrees of the root. Similarly, the depth-first search of a graph first visits a vertex, then it recursively visits all the vertices adjacent to that vertex. The difference is that the graph may contain cycles, which could lead to an infinite recursion. To avoid this problem, you need to track the vertices that have already been visited.

The search is called depth-first because it searches “deeper” in the graph as much as pos- sible. The search starts from some vertex v. After visiting v, it visits an unvisited neighbor of v. If v has no unvisited neighbor, the search backtracks to the vertex from which it reached v. We assume that the graph is connected and the search starting from any vertex can reach all the vertices. If this is not the case, see Programming Exercise 28.4 for finding connected components in a graph.

28.7.1 Depth-First Search Algorithm The algorithm for the depth-first search is described in Listing 28.8.

Listing 28.8 Depth-First Search Algorithm Input: G = (V, E) and a starting vertex v Output: a DFS tree rooted at v

1 SearchTree dfs(vertex v) { 2 visit v; 3 for each neighbor w of v 4 if (w has not been visited) { 5 set v as the parent for w in the tree; 6 dfs(w); 7 } 8 }

Point Key

Figure 28.11 The SearchTree class describes the nodes with parent–child relationships.

–root: int

–parent: int[]

–searchOrder: List<Integer>

+SearchTree(root: int, parent: int[],

searchOrder: List<Integer>)

+printPath(index: int): void +printTree(): void

Constructs a tree with the speci�ed root, parent, and searchOrder.

The root of the tree.

The parents of the vertices.

The orders for traversing the vertices.

Returns the root of the tree.

Returns the order of vertices searched.

Returns the parent for the speci�ed vertex index.

Returns the number of vertices searched.

Returns a list of vertices from the speci�ed vertex index to the root. Displays a path from the root to the speci�ed vertex. Displays tree with the root and all edges.

+getRoot(): int

+getSearchOrder(): List<Integer>

+getParent(index: int): int

+getNumberOfVerticesFound(): int

+getPath(index: int): List<V>

UnweightedGraph<V>.SearchTree

visit v

check a neighbor set a parent in the tree recursive search

M28_LIAN1878_11_GE_C28.indd 1084 12/29/17 4:49 PM

28.7 Depth-First Search (DFS) 1085

You can use an array named isVisited to denote whether a vertex has been visited. Initially, isVisited[i] is false for each vertex i. Once a vertex, say v, is visited, isVisited[v] is set to true.

Consider the graph in Figure 28.12a. Suppose we start the depth-first search from vertex 0. First visit 0, then any of its neighbors, say 1. Now 1 is visited, as shown in Figure 28.12b. Vertex 1 has three neighbors—0, 2, and 4. Since 0 has already been visited, you will visit either 2 or 4. Let us pick 2. Now 2 is visited, as shown in Figure 28.12c. Vertex 2 has three neighbors: 0, 1, and 3. Since 0 and 1 have already been visited, pick 3. 3 is now visited, as shown in Figure 28.12d. At this point, the vertices have been visited in this order:

0, 1, 2, 3

Since all the neighbors of 3 have been visited, backtrack to 2. Since all the vertices of 2 have been visited, backtrack to 1. 4 is adjacent to 1, but 4 has not been visited. Therefore, visit 4, as shown in Figure 28.12e. Since all the neighbors of 4 have been visited, backtrack to 1. Since all the neighbors of 1 have been visited, backtrack to 0. Since all the neighbors of 0 have been visited, the search ends.

Since each edge and each vertex is visited only once, the time complexity of the dfs method is O(|E| + |V|), where |E| denotes the number of edges and |V| the number of vertices.

DFS time complexity

Figure 28.12 Depth-first search visits a node and its neighbors recursively.

0 1

3 4

(a)

0 1

3 4

0 1

3 4

(b) (c)

0 1

3 4

0 1

3 4

(d) (e)

28.7.2 Implementation of Depth-First Search The algorithm for DFS in Listing 28.8 uses recursion. It is natural to use recursion to implement it. Alternatively, you can use a stack (see Programming Exercise 28.3).

The dfs(int v) method is implemented in lines 148–177 in Listing 28.4. It returns an instance of the SearchTree class with vertex v as the root. The method stores the vertices searched in the list searchOrder (line 149), the parent of each vertex in the array parent (line 150), and uses the isVisited array to indicate whether a vertex has been visited (line 155). It invokes the helper method dfs(v, parent, searchOrder, isVisited) to per- form a depth-first search (line 158).

In the recursive helper method, the search starts from vertex v. v is added to searchOrder in line 168 and is marked as visited (line 169). For each unvisited neighbor of v, the method is recursively invoked to perform a depth-first search. When a vertex e.v is visited, the parent

M28_LIAN1878_11_GE_C28.indd 1085 12/29/17 4:49 PM

1086 Chapter 28 Graphs and Applications

of e.v is stored in parent[e.v] (line 173). The method returns when all vertices are visited for a connected graph, or in a connected component.

Listing 28.9 gives a test program that displays a DFS for the graph in Figure 28.1 starting from Chicago. The graphical illustration of the DFS starting from Chicago is shown in Figure 28.13.

Listing 28.9 TestDFS.java 1 public class TestDFS { 2 public static void main(String[] args) { 3 String[] vertices = {"Seattle", "San Francisco", "Los Angeles", 4 "Denver", "Kansas City", "Chicago", "Boston", "New York", 5 "Atlanta", "Miami", "Dallas", "Houston"}; 6 7 int[][] edges = { 8 {0, 1}, {0, 3}, {0, 5}, 9 {1, 0}, {1, 2}, {1, 3}, 10 {2, 1}, {2, 3}, {2, 4}, {2, 10}, 11 {3, 0}, {3, 1}, {3, 2}, {3, 4}, {3, 5}, 12 {4, 2}, {4, 3}, {4, 5}, {4, 7}, {4, 8}, {4, 10}, 13 {5, 0}, {5, 3}, {5, 4}, {5, 6}, {5, 7}, 14 {6, 5}, {6, 7}, 15 {7, 4}, {7, 5}, {7, 6}, {7, 8}, 16 {8, 4}, {8, 7}, {8, 9}, {8, 10}, {8, 11}, 17 {9, 8}, {9, 11}, 18 {10, 2}, {10, 4}, {10, 8}, {10, 11}, 19 {11, 8}, {11, 9}, {11, 10} 20 }; 21 22 Graph<String> graph = new UnweightedGraph<>(vertices, edges); 23 UnweightedGraph<String>.SearchTree dfs = 24 graph.dfs(graph.getIndex("Chicago")); 25 26 java.util.List<Integer> searchOrders = dfs.getSearchOrder(); 27 System.out.println(dfs.getNumberOfVerticesFound() + 28 " vertices are searched in this DFS order:"); 29 for (int i = 0; i < searchOrders.size(); i++) 30 System.out.print(graph.getVertex(searchOrders.get(i)) + " "); 31 System.out.println(); 32 33 for (int i = 0; i < searchOrders.size(); i++) 34 if (dfs.getParent(i) != -1) 35 System.out.println("parent of " + graph.getVertex(i) + 36 " is " + graph.getVertex(dfs.getParent(i))); 37 } 38 }

U.S. Map Search

12 vertices are searched in this DFS order: Chicago Seattle San Francisco Los Angeles Denver Kansas City New York Boston Atlanta Miami Houston Dallas parent of Seattle is Chicago parent of San Francisco is Seattle parent of Los Angeles is San Francisco parent of Denver is Los Angeles parent of Kansas City is Denver parent of Boston is New York parent of New York is Kansas City parent of Atlanta is New York parent of Miami is Atlanta parent of Dallas is Houston parent of Houston is Miami

vertices

edges

create a graph

get DFS

get search order

M28_LIAN1878_11_GE_C28.indd 1086 12/29/17 4:49 PM

28.7 Depth-First Search (DFS) 1087

28.7.3 Applications of the DFS The depth-first search can be used to solve many problems, such as the following:

■■ Detecting whether a graph is connected. Search the graph starting from any vertex. If the number of vertices searched is the same as the number of vertices in the graph, the graph is connected. Otherwise, the graph is not connected. (See Programming Exercise 28.1.)

■■ Detecting whether there is a path between two vertices (see Programming Exercise 28.5).

■■ Finding a path between two vertices (see Programming Exercise 28.5).

■■ Finding all connected components. A connected component is a maximal connected subgraph in which every pair of vertices are connected by a path (see Programming Exercise 28.4).

■■ Detecting whether there is a cycle in the graph (see Programming Exercise 28.6).

■■ Finding a cycle in the graph (see Programming Exercise 28.7).

■■ Finding a Hamiltonian path/cycle. A Hamiltonian path in a graph is a path that visits each vertex in the graph exactly once. A Hamiltonian cycle visits each ver- tex in the graph exactly once and returns to the starting vertex (see Programming Exercise 28.17).

The first six problems can be easily solved using the dfs method in Listing 28.4. To find a Hamiltonian path/cycle, you have to explore all possible DFSs to find the one that leads to the longest path. The Hamiltonian path/cycle has many applications, including for solving the well-known Knight’s Tour problem, which is presented in Supplement VI.E on the Companion Website.

28.7.1 What is depth-first search? 28.7.2 Draw a DFS tree for the graph in Figure 28.3b starting from node A. 28.7.3 Draw a DFS tree for the graph in Figure 28.1 starting from vertex Atlanta. 28.7.4 What is the return type from invoking dfs(v)?

Point Check

Figure 28.13 A DFS search starts from Chicago. Source: © Mozilla Firefox.

M28_LIAN1878_11_GE_C28.indd 1087 12/29/17 4:49 PM

1088 Chapter 28 Graphs and Applications

28.7.5 The depth-first search algorithm described in Listing 28.8 uses recursion. Alterna- tively, you can use a stack to implement it, as shown below. Point out the error in this algorithm and give a correct algorithm.

// Wrong version SearchTree dfs(vertex v) { push v into the stack; mark v visited;

while (the stack is not empty) { pop a vertex, say u, from the stack visit u; for each neighbor w of u if (w has not been visited) push w into the stack; } }

28.8 Case Study: The Connected Circles Problem The connected circles problem is to determine whether all circles in a two- dimensional plane are connected. This problem can be solved using a depth-first traversal.

The DFS algorithm has many applications. This section applies the DFS algorithm to solve the connected circles problem.

In the connected circles problem, you determine whether all the circles in a two-dimensional plane are connected. If all the circles are connected, they are painted as filled circles, as shown in Figure 28.14a. Otherwise, they are not filled, as shown in Figure 28.14b.

Point Key

(a) Circles are connected (b) Circles are not connected

We will write a program that lets the user create a circle by clicking a mouse in a blank area that is not currently covered by a circle. As the circles are added, the circles are repainted filled if they are connected or unfilled otherwise.

We will create a graph to model the problem. Each circle is a vertex in the graph. Two circles are connected if they overlap. We apply the DFS in the graph, and if all vertices are found in the depth-first search, the graph is connected.

M28_LIAN1878_11_GE_C28.indd 1088 12/29/17 4:49 PM

28.8 Case Study: The Connected Circles Problem 1089

The program is given in Listing 28.10.

Listing 28.10 ConnectedCircles.java 1 import javafx.application.Application; 2 import javafx.geometry.Point2D; 3 import javafx.scene.Node; 4 import javafx.scene.Scene; 5 import javafx.scene.layout.Pane; 6 import javafx.scene.paint.Color; 7 import javafx.scene.shape.Circle; 8 import javafx.stage.Stage; 9 10 public class ConnectedCircles extends Application { 11 @Override // Override the start method in the Application class 12 public void start(Stage primaryStage) { 13 // Create a scene and place it in the stage 14 Scene scene = new Scene(new CirclePane(), 450, 350); 15 primaryStage.setTitle("ConnectedCircles"); // Set the stage title 16 primaryStage.setScene(scene); // Place the scene in the stage 17 primaryStage.show(); // Display the stage 18 } 19 20 /** Pane for displaying circles */ 21 class CirclePane extends Pane { 22 public CirclePane() { 23 this.setOnMouseClicked(e -> { 24 if (!isInsideACircle(new Point2D(e.getX(), e.getY()))) { 25 // Add a new circle 26 getChildren().add(new Circle(e.getX(), e.getY(), 20)); 27 colorIfConnected(); 28 } 29 }); 30 } 31 32 /** Returns true if the point is inside an existing circle */ 33 private boolean isInsideACircle(Point2D p) { 34 for (Node circle: this.getChildren()) 35 if (circle.contains(p)) 36 return true; 37 38 return false; 39 } 40 41 /** Color all circles if they are connected */ 42 private void colorIfConnected() { 43 if (getChildren().size() == 0) 44 return; // No circles in the pane 45 46 // Build the edges 47 java.util.List<Edge> edges 48 = new java.util.ArrayList<>(); 49 for (int i = 0; i < getChildren().size(); i++) 50 for (int j = i + 1; j < getChildren().size(); j++) 51 if (overlaps((Circle)(getChildren().get(i)), 52 (Circle)(getChildren().get(j)))) { 53 edges.add(new Edge(i, j)); 54 edges.add(new Edge(j, i)); 55 } 56 57 // Create a graph with circles as vertices

create a circle pane

pane for showing circles

handle mouse clicked is it inside another circle?

add a new circle color if all connected

contains the point?

create edges

M28_LIAN1878_11_GE_C28.indd 1089 12/29/17 4:49 PM

1090 Chapter 28 Graphs and Applications

58 Graph<Node> graph = new UnweightedGraph<> 59 ((java.util.List<Node>)getChildren(), edges); 60 UnweightedGraph<Node>.SearchTree tree = graph.dfs(0); 61 boolean isAllCirclesConnected = getChildren().size() == tree 62 .getNumberOfVerticesFound(); 63 64 for (Node circle: getChildren()) { 65 if (isAllCirclesConnected) { // All circles are connected 66 ((Circle)circle).setFill(Color.RED); 67 } 68 else { 69 ((Circle)circle).setStroke(Color.BLACK); 70 ((Circle)circle).setFill(Color.WHITE); 71 } 72 } 73 } 74 } 75 76 public static boolean overlaps(Circle circle1, Circle circle2) { 77 return new Point2D(circle1.getCenterX(), circle1.getCenterY()). 78 distance(circle2.getCenterX(), circle2.getCenterY()) 79 <= circle1.getRadius() + circle2.getRadius(); 80 } 81 }

The JavaFX Circle class contains the data fields x, y, and radius, which specify the circle’s center location and radius. It also defines the contains method for testing whether a point is inside the circle. The overlaps method (lines 76–80) checks whether two circles overlap.

When the user clicks the mouse outside of any existing circle, a new circle is created cen- tered at the mouse point and the circle is added to the list circles (line 26).

To detect whether the circles are connected, the program constructs a graph (lines 46–59). The circles are the vertices of the graph. The edges are constructed in lines 47–55. Two circle vertices are connected if they overlap (line 51). The DFS of the graph results in a tree (line 60). The tree’s getNumberOfVerticesFound() returns the number of vertices searched. If it is equal to the number of circles, all circles are connected (lines 61–62).

28.8.1 How is a graph created for the connected circles problem? 28.8.2 When you click the mouse inside a circle, does the program create a new circle? 28.8.3 How does the program know if all circles are connected?

28.9 Breadth-First Search (BFS) The breadth-first search of a graph visits the vertices level by level. The first level consists of the starting vertex. Each next level consists of the vertices adjacent to the vertices in the preceding level.

The breadth-first traversal of a graph is like the breadth-first traversal of a tree discussed in Section 25.2.4, Tree Traversal. With breadth-first traversal of a tree, the nodes are visited level by level. First the root is visited, then all the children of the root, then the grandchildren of the root, and so on. Similarly, the breadth-first search of a graph first visits a vertex, then all its adjacent vertices, then all the vertices adjacent to those vertices, and so on. To ensure each vertex is visited only once, it skips a vertex if it has already been visited.

28.9.1 Breadth-First Search Algorithm The algorithm for the breadth-first search starting from vertex v in a graph is described in Listing 28.11.

Point Check

Point Key

create a graph

get a search tree connected?

connected

not connected

two circles overlap?

M28_LIAN1878_11_GE_C28.indd 1090 12/29/17 4:49 PM

28.9 Breadth-First Search (BFS) 1091

Listing 28.11 Breadth-First Search Algorithm Input: G = (V, E) and a starting vertex v Output: a BFS tree rooted at v

1 SearchTree bfs(vertex v) { 2 create an empty queue for storing vertices to be visited; 3 add v into the queue; 4 mark v visited; 5 6 while (the queue is not empty) { 7 dequeue a vertex, say u, from the queue; 8 add u into a list of traversed vertices; 9 for each neighbor w of u 10 if w has not been visited { 11 add w into the queue; 12 set u as the parent for w in the tree; 13 mark w visited; 14 } 15 } 16 }

Consider the graph in Figure 28.15a. Suppose you start the breadth-first search from vertex 0. First visit 0, then visit all its neighbors, 1, 2, and 3, as shown in Figure 28.15b. Vertex 1 has three neighbors: 0, 2, and 4. Since 0 and 2 have already been visited, you will now visit just 4, as shown in Figure 28.15c. Vertex 2 has three neighbors, 0, 1, and 3, which have all been visited. Vertex 3 has three neighbors, 0, 2, and 4, which have all been visited. Vertex 4 has two neighbors, 1 and 3, which have all been visited. Hence, the search ends.

Since each edge and each vertex is visited only once, the time complexity of the bfs method is O(|E| + |V|), where |E| denotes the number of edges and |V| the number of vertices.

BFS time complexity

Figure 28.15 Breadth-first search visits a node, then its neighbors, then its neighbors’s neighbors, and so on.

0 1

3 4

(a)

0 1

3 4

0 1

3 4

(b) (c)

28.9.2 Implementation of Breadth-First Search The bfs(int v) method is defined in the Graph interface and implemented in the UnweightedGraph class in Listing 28.4 (lines 181–206). It returns an instance of the SearchTree class with vertex v as the root. The method stores the vertices searched in the list searchOrder (line 182), the parent of each vertex in the array parent (line 183), uses a linked list for a queue (lines 187–188), and uses the isVisited array to indicate whether a vertex has been visited (line 191). The search starts from vertex v. v is added to the queue in line 190 and is marked as visited (line 191). The method now examines each vertex u in the queue (line 193) and adds it to searchOrder (line 195). The method adds each unvisited neighbor e.v of u to the queue (line 198), sets its parent to u (line 199), and marks it as visited (line 200).

create a queue enqueue v

dequeue into u u traversed check a neighbor w is w visited? enqueue w

M28_LIAN1878_11_GE_C28.indd 1091 12/29/17 4:49 PM

1092 Chapter 28 Graphs and Applications

Listing 28.12 gives a test program that displays a BFS for the graph in Figure 28.1 start- ing from Chicago. The graphical illustration of the BFS starting from Chicago is shown in Figure 28.16.

Listing 28.12 TestBFS.java 1 public class TestBFS { 2 public static void main(String[] args) { 3 String[] vertices = {"Seattle", "San Francisco", "Los Angeles", 4 "Denver", "Kansas City", "Chicago", "Boston", "New York", 5 "Atlanta", "Miami", "Dallas", "Houston"}; 6 7 int[][] edges = { 8 {0, 1}, {0, 3}, {0, 5}, 9 {1, 0}, {1, 2}, {1, 3}, 10 {2, 1}, {2, 3}, {2, 4}, {2, 10}, 11 {3, 0}, {3, 1}, {3, 2}, {3, 4}, {3, 5}, 12 {4, 2}, {4, 3}, {4, 5}, {4, 7}, {4, 8}, {4, 10}, 13 {5, 0}, {5, 3}, {5, 4}, {5, 6}, {5, 7}, 14 {6, 5}, {6, 7}, 15 {7, 4}, {7, 5}, {7, 6}, {7, 8}, 16 {8, 4}, {8, 7}, {8, 9}, {8, 10}, {8, 11}, 17 {9, 8}, {9, 11}, 18 {10, 2}, {10, 4}, {10, 8}, {10, 11}, 19 {11, 8}, {11, 9}, {11, 10} 20 }; 21 22 Graph<String> graph = new UnweightedGraph<>(vertices, edges); 23 UnweightedGraph<String>.SearchTree bfs = 24 graph.bfs(graph.getIndex("Chicago")); 25 26 java.util.List<Integer> searchOrders = bfs.getSearchOrder(); 27 System.out.println(bfs.getNumberOfVerticesFound() + 28 " vertices are searched in this order:"); 29 for (int i = 0; i < searchOrders.size(); i++) 30 System.out.println(graph.getVertex(searchOrders.get(i))); 31 32 for (int i = 0; i < searchOrders.size(); i++) 33 if (bfs.getParent(i) != -1) 34 System.out.println("parent of " + graph.getVertex(i) + 35 " is " + graph.getVertex(bfs.getParent(i))); 36 } 37 }

12 vertices are searched in this order: Chicago Seattle Denver Kansas City Boston New York San Francisco Los Angeles Atlanta Dallas Miami Houston parent of Seattle is Chicago parent of San Francisco is Seattle parent of Los Angeles is Denver parent of Denver is Chicago parent of Kansas City is Chicago parent of Boston is Chicago parent of New York is Chicago parent of Atlanta is Kansas City parent of Miami is Atlanta parent of Dallas is Kansas City parent of Houston is Atlanta

vertices

edges

create a graph create a BFS tree

get search order

M28_LIAN1878_11_GE_C28.indd 1092 12/29/17 4:49 PM

28.10 Case Study: The Nine Tails Problem 1093

28.9.3 Applications of the BFS Many of the problems solved by the DFS can also be solved using the BFS. Specifically, the BFS can be used to solve the following problems:

■■ Detecting whether a graph is connected. A graph is connected if there is a path between any two vertices in the graph.

■■ Detecting whether there is a path between two vertices.

■■ Finding the shortest path between two vertices. You can prove that the path between the root and any node in the BFS tree is the shortest path between the root and the node. (See CheckPoint Question 28.9.5.)

■■ Finding all connected components. A connected component is a maximal connected subgraph in which every pair of vertices are connected by a path.

■■ Detecting whether there is a cycle in the graph (see Programming Exercise 28.6).

■■ Finding a cycle in the graph (see Programming Exercise 28.7).

■■ Testing whether a graph is bipartite. (A graph is bipartite if the vertices of the graph can be divided into two disjoint sets such that no edges exist between vertices in the same set.) (See Programming Exercise 28.8.)

28.9.1 What is the return type from invoking bfs(v)? 28.9.2 What is breadth-first search? 28.9.3 Draw a BFS tree for the graph in Figure 28.3b starting from node A. 28.9.4 Draw a BFS tree for the graph in Figure 28.1 starting from vertex Atlanta. 28.9.5 Prove the path between the root and any node in the BFS tree is the shortest path

between the root and the node.

28.10 Case Study: The Nine Tails Problem The nine tails problem can be reduced to the shortest path problem.

The nine tails problem is as follows. Nine coins are placed in a 3 * 3 matrix, with some face up and some face down. A legal move is to take any coin that is face up and reverse it, together with the coins adjacent to it (this does not include coins that are diagonally adjacent).

Point Check

Point Key

Figure 28.16 BFS search starts from Chicago. Source: © Mozilla Firefox.

M28_LIAN1878_11_GE_C28.indd 1093 12/29/17 4:49 PM

1094 Chapter 28 Graphs and Applications

Your task is to find the minimum number of moves that lead to all coins being face down. For example, start with the nine coins as shown in Figure 28.17a. After you flip the second coin in the last row, the nine coins are now as shown in Figure 28.17b. After you flip the second coin in the first row, the nine coins are all face down, as shown in Figure 28.17c. See liveexample .pearsoncmg.com/dsanimation/NineCoin.html for an interactive demo.

Figure 28.17 The problem is solved when all coins are face down.

(a)

(b)

(c)

Enter the initial nine coins Hs and Ts: HHHTTTHHH

The steps to flip the coins are HHH TTT HHH

HHH THT TTT

TTT TTT TTT

We will write a program that prompts the user to enter an initial state of the nine coins and displays the solution, as shown in the following sample run:

Each state of the nine coins represents a node in the graph. For example, the three states in Figure 28.17 correspond to three nodes in the graph. For convenience, we use a 3 * 3 matrix to represent all nodes and use 0 for heads and 1 for tails. Since there are nine cells and each cell is either 0 or 1, there are a total of 29 (512) nodes, labeled 0, 1, . . . , and 511, as shown in Figure 28.18.

Figure 28.18 There are total of 512 nodes labeled in this order: 0, 1, 2, . . . , 511.

. . . . .

511

We assign an edge from node v to u if there is a legal move from u to v. Figure 28.19 shows a partial graph. Note there is an edge from 511 to 47, since you can flip a cell in node 47 to become node 511.

The last node in Figure 28.18 represents the state of nine face-down coins. For convenience, we call this last node the target node. Thus, the target node is labeled 511. Suppose the initial state of the nine tails problem corresponds to the node s. The problem is reduced to finding the shortest path from node s to the target node, which is equivalent to finding the path from node s to the target node in a BFS tree rooted at the target node.

M28_LIAN1878_11_GE_C28.indd 1094 12/29/17 4:49 PM

28.10 Case Study: The Nine Tails Problem 1095

Now the task is to build a directed graph that consists of 512 nodes labeled 0, 1, 2, . . . , 511, and edges among the nodes. Once the graph is created, obtain a BFS tree rooted at node 511. From the BFS tree, you can find the shortest path from the root to any vertex. We will create a class named NineTailModel, which contains the method to get the shortest path from the target node to any other node. The class UML diagram is shown in Figure 28.19.

H H H H T T T H H HH

0 Position is 2 here in a node

A node is an array of nine characters

Visually, a node is represented in a 3 * 3 matrix with the letters H and T. In our program, we use a single-dimensional array of nine characters to represent a node. For example, the node for vertex 1 in Figure 28.18 is represented as {'H', 'H', 'H', 'H', 'H', 'H', 'H', 'H', 'T'} in the array.

The getEdges() method returns a list of Edge objects. The getNode(index) method returns the node for the specified index. For example,

getNode(0) returns the node that contains nine Hs. getNode(511) returns the node that contains nine Ts. The getIndex(node) method returns the index of the node.

Note the data field tree is defined as protected so it can be accessed from the WeightedNineTail subclass in the next chapter.

The getFlippedNode(char[] node, int position) method flips the node at the specified position and its adjacent positions. This method returns the index of the new node. The position is a value from 0 to 8, which points to a coin in the node, as shown in the fol- lowing figure.

Figure 28.19 The NineTailModel class models the nine tails problem using a graph.

NineTailModel

#tree: UnweightedGraph<Integer>.

SearchTree

+NineTailModel() +getShortestPath(nodeIndex: int): List<Integer>

–getEdges(): List<Edge>

+getNode(index: int): char[]

+getIndex(node: char[]): int

+getFlippedNode(node: char[], position: int): int +‚ipACell(node: char[], row: int, column: int): void +printNode(node: char[]): void

A tree rooted at node 511.

Constructs a model for the nine tails problem and obtains the tree. Returns a path from the speci�ed node to the root. The path returned consists of the node labels in a list.

Flips the node at the speci�ed row and column.

Displays the node on the console.

Returns a list of Edge objects for the graph.

Returns a node consisting of nine characters of Hs and Ts.

Returns the index of the speci�ed node. Flips the node at the speci�ed position and returns the index of the �ipped node.

For example, for node 56 in Figure 28.20, flip it at position 0, and you will get node 51. If you flip node 56 at position 1, you will get node 47.

The flipACell(char[] node, int row, int column) method flips a node at the specified row and column. For example, if you flip node 56 at row 0 and column 0, the new node is 408. If you flip node 56 at row 2 and column 0, the new node is 30.

M28_LIAN1878_11_GE_C28.indd 1095 12/29/17 4:49 PM

1096 Chapter 28 Graphs and Applications

Listing 28.13 shows the source code for NineTailModel.java.

Listing 28.13 NineTailModel.java 1 import java.util.*; 2 3 public class NineTailModel { 4 public final static int NUMBER_OF_NODES = 512; 5 protected UnweightedGraph<Integer>.SearchTree tree; 6 7 /** Construct a model */ 8 public NineTailModel() { 9 // Create edges 10 List<Edge> edges = getEdges(); 11 12 // Create a graph 13 UnweightedGraph<Integer> graph = new UnweightedGraph<>( 14 edges, NUMBER_OF_NODES); 15 16 // Obtain a BSF tree rooted at the target node 17 tree = graph.bfs(511); 18 } 19 20 /** Create all edges for the graph */ 21 private List<Edge> getEdges() { 22 List<Edge> edges = 23 new ArrayList<>(); // Store edges 24 25 for (int u = 0; u < NUMBER_OF_NODES; u++) { 26 for (int k = 0; k < 9; k++) { 27 char[] node = getNode(u); // Get the node for vertex u

Figure 28.20 If node u becomes node v after cells are flipped, assign an edge from v to u.

1 0 1 1

1 0

0 0

408

1 1 0 1

1 1

0 0

488

0 1 1 0

1 1

0 0

240

0 0 1 1

0 0

1 1

0 1 0 1

0 0

1 1

0 1 1 1

0 0

1 1 1 1

1 1

511

0 1 1 0

0 0

0 1

declare a tree

create edges

create graph

create tree

get edges

M28_LIAN1878_11_GE_C28.indd 1096 12/29/17 4:49 PM

28.10 Case Study: The Nine Tails Problem 1097

28 if (node[k] == 'H') { 29 int v = getFlippedNode(node, k); 30 // Add edge (v, u) for a legal move from node u to node v 31 edges.add(new Edge(v, u)); 32 } 33 } 34 } 35 36 return edges; 37 } 38 39 public static int getFlippedNode(char[] node, int position) { 40 int row = position / 3; 41 int column = position % 3; 42 43 flipACell(node, row, column); 44 flipACell(node, row − 1, column); 45 flipACell(node, row + 1, column); 46 flipACell(node, row, column − 1); 47 flipACell(node, row, column + 1); 48 49 return getIndex(node); 50 } 51 52 public static void flipACell(char[] node, int row, int column) { 53 if (row >= 0 && row <= 2 && column >= 0 && column <= 2) { 54 // Within the boundary 55 if (node[row * 3 + column] == 'H') 56 node[row * 3 + column] = 'T'; // Flip from H to T 57 else 58 node[row * 3 + column] = 'H'; // Flip from T to H 59 } 60 } 61 62 public static int getIndex(char[] node) { 63 int result = 0; 64 65 for (int i = 0; i < 9; i++) 66 if (node[i] == 'T') 67 result = result * 2 + 1; 68 else 69 result = result * 2 + 0; 70 71 return result; 72 } 73 74 public static char[] getNode(int index) { 75 char[] result = new char[9]; 76 77 for (int i = 0; i < 9; i++) { 78 int digit = index % 2; 79 if (digit == 0) 80 result[8 - i] = 'H'; 81 else 82 result[8 - i] = 'T'; 83 index = index / 2; 84 } 85 86 return result; 87 }

row

column

Flip

node: HHHHHHHTT

H H H H H H H T T

index: 3 For example:

For example:

T H H H H H H T T

node: THHHHHHTT index: 259

add an edge

flip cells

flip a cell

get index for a node

get node for an index

M28_LIAN1878_11_GE_C28.indd 1097 12/29/17 4:49 PM

1098 Chapter 28 Graphs and Applications

The constructor (lines 8–18) creates a graph with 512 nodes, and each edge corresponds to the move from one node to the other (line 10). From the graph, a BFS tree rooted at the target node 511 is obtained (line 17).

To create edges, the getEdges method (lines 21–37) checks each node u to see if it can be flipped to another node v. If so, add (v, u) to the Edge list (line 31). The getFlippedNode(node, position) method finds a flipped node by flipping an H cell and its neighbors in a node (lines 43–47). The flipACell(node, row, column) method actually flips an H cell and its neighbors in a node (lines 52–60).

The getIndex(node) method is implemented in the same way as converting a binary number to a decimal number (lines 62–72). The getNode(index) method returns a node consisting of the letters H and T (lines 74–87).

The getShortestpath(nodeIndex) method invokes the getPath(nodeIndex) method to get the vertices in a shortest path from the specified node to the target node (lines 89–91).

The printNode(node) method displays a node on the console (lines 93–101). Listing 28.14 gives a program that prompts the user to enter an initial node and displays

the steps to reach the target node.

Listing 28.14 NineTail.java 1 import java.util.Scanner; 2 3 public class NineTail { 4 public static void main(String[] args) { 5 // Prompt the user to enter nine coins' Hs and Ts 6 System.out.print("Enter the initial nine coins Hs and Ts: "); 7 Scanner input = new Scanner(System.in); 8 String s = input.nextLine(); 9 char[] initialNode = s.toCharArray(); 10 11 NineTailModel model = new NineTailModel(); 12 java.util.List<Integer> path = 13 model.getShortestPath(NineTailModel.getIndex(initialNode)); 14 15 System.out.println("The steps to flip the coins are "); 16 for (int i = 0; i < path.size(); i++) 17 NineTailModel.printNode( 18 NineTailModel.getNode(path.get(i).intValue())); 19 } 20 }

88 89 public List<Integer> getShortestPath(int nodeIndex) { 90 return tree.getPath(nodeIndex); 91 } 92 93 public static void printNode(char[] node) { 94 for (int i = 0; i < 9; i++) 95 if (i % 3 != 2) 96 System.out.print(node[i]); 97 else 98 System.out.println(node[i]); 99 100 System.out.println(); 101 } 102 }

For example:

T H H H H H H T T

node: THHHHHHTT Output:

shortest path

display a node

initial node

create model

get shortest path

M28_LIAN1878_11_GE_C28.indd 1098 12/29/17 4:49 PM

Programming Exercises 1099

The program prompts the user to enter an initial node with nine letters with a combination of Hs and Ts as a string in line 8, obtains an array of characters from the string (line 9), creates a graph model to get a BFS tree (line 11), obtains the shortest path from the initial node to the target node (lines 12–13), and displays the nodes in the path (lines 16–18).

28.10.1 How are the nodes created for the graph in NineTailModel? 28.10.2 How are the edges created for the graph in NineTailModel? 28.10.3 What is returned after invoking getIndex("HTHTTTHHH". toCharArray()) in

Listing 28.13? What is returned after invoking getNode(46) in Listing 28.13?

28.10.4 If lines 26 and 27 are swapped in Listing 28.13, NineTailModel.java, will the program work? Why not?

Point Check

Key terms adjacency list 1069 adjacency matrix 1069 adjacent vertices 1064 breadth-first search 1083 complete graph 1064 cycle 1064 degree 1064 depth-first search 1083 directed graph 1063 graph 1062

incident edges 1064 parallel edge 1064 Seven Bridges of Königsberg 1062 simple graph 1064 spanning tree 1064 tree 1064 undirected graph 1063 unweighted graph 1064 weighted graph 1064

ChapTer summary 1. A graph is a useful mathematical structure that represents relationships among entities

in the real world. You learned how to model graphs using classes and interfaces, how to represent vertices and edges using arrays and linked lists, and how to implement opera- tions for graphs.

2. Graph traversal is the process of visiting each vertex in the graph exactly once. You learned two popular ways for traversing a graph: the depth-first search (DFS) and breadth-first search (BFS).

3. DFS and BFS can be used to solve many problems such as detecting whether a graph is connected, detecting whether there is a cycle in the graph, and finding the shortest path between two vertices.

Quiz Answer the quiz for this chapter online at the book Companion Website.

programming exerCises

Sections 28.6–28.10 *28.1 (Test whether a graph is connected) Write a program that reads a graph from

a file and determines whether the graph is connected. The first line in the file contains a number that indicates the number of vertices (n). The vertices are labeled as 0, 1, . . . , n−1. Each subsequent line, with the format u v1 v2 ..., describes edges (u, v1), (u, v2), and so on. Figure 28.21 gives the examples of two files for their corresponding graphs.

M28_LIAN1878_11_GE_C28.indd 1099 12/29/17 4:49 PM

1100 Chapter 28 Graphs and Applications

Figure 28.21 The vertices and edges of a graph can be stored in a file.

(a) (b)

File 6 0 1 2 1 0 3 2 0 3 4 3 1 2 4 5 4 2 3 5 5 3 4

File 6 0 1 2 3 1 0 3 2 0 3 3 0 1 2 4 5 5 4

2 3

4 5

Enter a URL: https://liveexample.pearsoncmg.com/test/GraphSample1.txt

The number of vertices is 6 Vertex 0: (0, 1) (0, 2) Vertex 1: (1, 0) (1, 3) Vertex 2: (2, 0) (2, 3) (2, 4) Vertex 3: (3, 1) (3, 2) (3, 4) (3, 5) Vertex 4: (4, 2) (4, 3) (4, 5) Vertex 5: (5, 3) (5, 4) The graph is connected

Your program should prompt the user to enter a URL for the file, then it should read data from the file, create an instance g of UnweightedGraph, invoke g.printEdges() to display all edges, and invoke dfs() to obtain an instance tree of UnweightedGraph<V>.SearchTree . If tree. getNumberOfVerticesFound() is the same as the number of verti- ces in the graph, the graph is connected. Here is a sample run of the program:

(Hint: Use new UnweightedGraph(list, numberOfVertices) to create a graph, where list contains a list of Edge objects. Use new Edge(u, v) to create an edge. Read the first line to get the number of vertices. Read each subsequent line into a string s and use s.split("[\;\s+]") to extract the vertices from the string and create edges from the vertices.)

*28.2 (Create a file for a graph) Modify Listing 28.2, TestGraph.java to create a file representing graph1. The file format is described in Programming Exercise 28.1. Create the file from the array defined in lines 8–21 in Listing 28.2. The number of vertices for the graph is 12, which will be stored in the first line of the file. The contents of the file should be as follows:

12 0 1 3 5 1 0 2 3 2 1 3 4 10 3 0 1 2 4 5 4 2 3 5 7 8 10 5 0 3 4 6 7 6 5 7 7 4 5 6 8 8 4 7 9 10 11 9 8 11 10 2 4 8 11 11 8 9 10

M28_LIAN1878_11_GE_C28.indd 1100 12/29/17 4:49 PM

Programming Exercises 1101

*28.3 (Implement DFS using a stack) The depth-first search algorithm described in Listing 28.8, Depth-First Search Algorithm uses recursion. Design a new algorithm without using recursion. Describe it using pseudocode. Implement it by defining a new class named UnweightedGraphWithNonrecursiveDFS that extends UnweightedGraph and overriding the dfs method. Write a test program same as Listing 28.9, TestPFS.java except that UnweightedGraph is replaced by UnweightedGraphWithNonrecursiveDFS.

*28.4 (Find connected components) Create a new class named MyGraph as a subclass of UnweightedGraph that contains a method for finding all connected compo- nents in a graph with the following header:

public List<List<Integer>> getConnectedComponents();

The method returns a List<List<Integer>>. Each element in the list is another list that contains all the vertices in a connected component. For exam- ple, for the graph in Figure 28.21b, getConnectedComponents() returns [[0, 1, 2, 3], [4, 5]].

*28.5 (Find paths) Define a new class named UnweightedGraphWithGetPath that extends UnweightedGraph with a new method for finding a path between two vertices with the following header:

public List<Integer> getPath(int u, int v);

The method returns a List<Integer> that contains all the vertices in a path from u to v in this order. Using the BFS approach, you can obtain the shortest path from u to v. If there isn’t a path from u to v, the method returns null. Write a test program that creates a graph for Figure 28.1. The program prompts the user to enter two cities and displays their paths. Here is a sample run:

Enter a starting city: Seattle

Enter an ending city: Miami

The path is Seattle Denver Kansas City Atlanta Miami

*28.6 (Detect cycles) Define a new class named UnweightedGraphDetectCycle that extends UnweightedGraph with a new method for determining whether there is a cycle in the graph with the following header:

public boolean isCyclic();

Describe the algorithm in pseudocode and implement it. Note the graph may be a directed graph.

*28.7 (Find a cycle) Define a new class named UnweightedGraphFindCycle that extends UnweightedGraph with a new method for finding a cycle starting at vertex u with the following header:

public List<Integer> getACycle(int u);

The method returns a List that contains all the vertices in a cycle starting from u. If the graph doesn’t have any cycles, the method returns null. Describe the algorithm in pseudocode and implement it.

M28_LIAN1878_11_GE_C28.indd 1101 12/29/17 4:49 PM

1102 Chapter 28 Graphs and Applications

**28.8 (Test bipartite) Recall that a graph is bipartite if its vertices can be divided into two disjoint sets such that no edges exist between vertices in the same set. Define a new class named UnweightedGraphTestBipartite with the fol- lowing method to detect whether the graph is bipartite:

public boolean isBipartite();

**28.9 (Get bipartite sets) Add a new method in UnweightedGraph with the follow- ing header to return two bipartite sets if the graph is bipartite:

public List<List<Integer>> getBipartite();

The method returns a List that contains two sublists, each of which contains a set of vertices. If the graph is not bipartite, the method returns null.

*28.10 (Find the shortest path) Write a program that reads a connected graph from a file. The graph is stored in a file using the same format specified in Program- ming Exercise 28.1. Your program should prompt the user to enter the name of the file, then two vertices, and should display the shortest path between the two vertices. For example, for the graph in Figure 28.21a, the shortest path between 0 and 5 may be displayed as 0 1 3 5.

Here is a sample run of the program:

Enter a file name: c:\exercise\GraphSample1.txt

Enter two vertices (integer indexes): 0 5 The number of vertices is 6 Vertex 0: (0, 1) (0, 2) Vertex 1: (1, 0) (1, 3) Vertex 2: (2, 0) (2, 3) (2, 4) Vertex 3: (3, 1) (3, 2) (3, 4) (3, 5) Vertex 4: (4, 2) (4, 3) (4, 5) Vertex 5: (5, 3) (5, 4) The path is 0 1 3 5

**28.11 (Revise Listing 28.14, NineTail.java) The program in Listing 28.14 lets the user enter an input for the nine tails problem from the console and displays the result on the console. Write a program that lets the user set an initial state of the nine coins (see Figure 28.22a) and click the Solve button to display the solution, as shown in Figure 28.22b. Initially, the user can click the mouse button to flip a coin. Set a red color on the flipped cells.

(a) (b)

M28_LIAN1878_11_GE_C28.indd 1102 12/29/17 4:49 PM

Programming Exercises 1103

**28.12 (Variation of the nine tails problem) In the nine tails problem, when you flip a coin, the horizontal and vertical neighboring cells are also flipped. Rewrite the program, assuming all neighboring cells including the diagonal neighbors are also flipped.

**28.13 (4 * 4 16 tails problem) Listing 28.14, NineTail.java, presents a solution for the nine tails problem. Revise this program for the 4 * 4 16 tails problem. Note it is possible that a solution may not exist for a starting pattern. If so, report that no solution exists.

**28.14 (4 * 4 16 tails analysis) The nine tails problem in the text uses a 3 * 3 matrix. Assume you have 16 coins placed in a 4 * 4 matrix. Write a program to find out the number of the starting patterns that don’t have a solution.

*28.15 (4 * 4 16 tails GUI) Rewrite Programming Exercise 28.14 to enable the user to set an initial pattern of the 4 * 4 16 tails problem (see Figure 28.23a). The user can click the Solve button to display the solution, as shown in Figure 28.23b. Initially, the user can click the mouse button to flip a coin. If a solution does not exist, display a message dialog to report it.

(b)(a)

**28.16 (Induced subgraph) Given an undirected graph G = (V, E) and an integer k, find an induced subgraph H of G of maximum size such that all vertices of H have a degree 7 = k, or conclude that no such induced subgraph exists. Imple- ment the method with the following header:

public static <V> Graph<V> maxInducedSubgraph(Graph<V> g, int k)

The method returns an empty graph if such a subgraph does not exist.

(Hint: An intuitive approach is to remove vertices whose degree is less than k. As vertices are removed with their adjacent edges, the degrees of other vertices may be reduced. Continue the process until no vertices can be removed, or all the vertices are removed.)

*** 28.17 (Hamiltonian cycle) The Hamiltonian path algorithm is implemented in Supple- ment VI.E. Add the following getHamiltonianCycle method in the Graph interface and implement it in the UnweightedGraph class:

/** Return a Hamiltonian cycle * Return null if the graph doesn't contain a Hamiltonian cycle */

public List<Integer> getHamiltonianCycle()

M28_LIAN1878_11_GE_C28.indd 1103 12/29/17 4:49 PM

1104 Chapter 28 Graphs and Applications

***28.18 (Knight’s Tour cycle) Rewrite KnightTourApp.java in the case study in Sup- plement VI.E to find a knight’s tour that visits each square in a chessboard and returns to the starting square. Reduce the Knight’s Tour cycle problem to the problem of finding a Hamiltonian cycle.

**28.19 (Display a DFS/BFS tree in a graph) Modify GraphView in Listing 28.6 to add a new data field tree with a setter method. The edges in the tree are displayed in red. Write a program that displays the graph in Figure 28.1 and the DFS/BFS tree starting from a specified city, as shown in Figures 28.13 and 28.16. If a city not in the map is entered, the program displays an error message in the label.

*28.20 (Display a graph) Write a program that reads a graph from a file and displays it. The first line in the file contains a number that indicates the number of verti- ces (n). The vertices are labeled 0, 1, . . . , n−1. Each subsequent line, with the format u x y v1 v2 ..., describes the position of u at (x, y) and edges (u, v1), (u, v2), and so on. Figure 28.24a gives an example of the file for their cor- responding graph. Your program prompts the user to enter the name of the file, reads data from the file, and displays the graph on a pane using GraphView, as shown in Figure 28.24b.

(a) (b)

File 7 0 30 30 1 2 1 90 30 0 3 6 2 30 90 0 3 4 3 90 90 1 2 4 5 4 30 150 2 3 5 5 90 150 3 4 6 6 130 90 1 5

3 6

**28.21 (Display sets of connected circles) Modify Listing 28.10, ConnectedCircles. java to display sets of connected circles in different colors. That is, if two circles are connected, they are displayed using the same color; otherwise, they are not in same color, as shown in Figure 28.25. (Hint: See Programming Exercise 28.4.)

Figure 28.25 (a) Connected circles are displayed in the same color. (b) Rectangles are not filled with a color if they are not connected. (c) Rectangles are filled with a color if they are connected. Source: Copyright © 1995–2016 Oracle and/or its affiliates. All rights reserved. Used with permission.

(a) (b) (c)

M28_LIAN1878_11_GE_C28.indd 1104 12/29/17 4:49 PM

Programming Exercises 1105

*28.22 (Move a circle) Modify Listing 28.10, ConnectedCircles.java, to enable the user to drag and move a circle.

**28.23 (Connected rectangles) Listing 28.10, ConnectedCircles.java, allows the user to create circles and determine whether they are connected. Rewrite the pro- gram for rectangles. The program lets the user create a rectangle by clicking a mouse in a blank area that is not currently covered by a rectangle. As the rec- tangles are added, the rectangles are repainted as filled if they are connected or are unfilled otherwise, as shown in Figure 28.25b–c.

*28.24 (Remove a circle) Modify Listing 28.10, ConnectedCircles.java, to enable the user to remove a circle when the mouse is clicked inside the circle.

*28.25 (Implement remove(V v)) Modify Listing 28.4, UnweightedGraph.java, to override the remove(V v) method defined in the Graph interface.

*28.26 (Implement remove(int u, int v)) Modify Listing 28.4, UnweightedGraph.java, to override the remove(int u, int v) method defined in the Graph interface.

M28_LIAN1878_11_GE_C28.indd 1105 12/29/17 4:49 PM

A01_GIPE3898_09_SE_FM.indd 3 12/20/17 5:56 PM

This page intentionally left blank

Objectives ■■ To represent weighted edges using adjacency matrices and adjacency

lists (§29.2).

■■ To model weighted graphs using the WeightedGraph class that extends the UnweightedGraph class (§29.3).

■■ To design and implement the algorithm for finding a minimum spanning tree (§29.4).

■■ To define the MST class that extends the SearchTree class (§29.4).

■■ To design and implement the algorithm for finding single-source shortest paths (§29.5).

■■ To define the ShortestPathTree class that extends the SearchTree class (§29.5).

■■ To solve the weighted nine tails problem using the shortest-path algorithm (§29.6).

Weighted Graphs and Applications

Chapter

M29_LIAN1878_11_GE_C29.indd 1107 12/29/17 4:51 PM

1108 Chapter 29 Weighted Graphs and Applications

29.1 Introduction A graph is a weighted graph if each edge is assigned a weight. Weighted graphs have many practical applications.

Figure 28.1 assumes the graph represents the number of flights among cities. You can apply the Breadth-First Search (BFS) to find the fewest number of flights between two cities. Assume the edges represent the driving distances among the cities as shown in Figure 29.1. How do you find the minimal total distances for connecting all cities? How do you find the shortest path between two cities? This chapter will address these questions. The former is known as the minimum spanning tree (MST) problem, and the latter as the shortest path problem.

Point Key

problem

Figure 29.1 The graph models the distances among the cities.

Seattle (0)

San Francisco (1)

Los Angeles (2)

Dallas (10)

Houston (11)

Atlanta (8)

New York (7)

Boston (6)

Chicago (5)

Denver (3)

807

381

1267

2097

1331

1663

496

239

1187

810

661

781

864

888

214

1435

1003

533 599

1260

983

787

1015 Kansas City (4)

Miami (9)

The preceding chapter introduced the concept of graphs. You learned how to represent edges using edge arrays, edge lists, adjacency matrices, and adjacency lists, and how to model a graph using the Graph interface and the UnweightedGraph class. The preceding chapter also introduced two important techniques for traversing graphs: depth-first search and breadth-first search, and applied traversal to solve practical problems. This chapter will introduce weighted graphs. You will learn the algorithm for finding a minimum spanning tree in Section 29.4, and the algorithm for finding shortest paths in Section 29.5.

Pedagogical Note Before we introduce the algorithms and applications for weighted graphs, it is helpful to get acquainted with weighted graphs using the GUI interactive tool at liveexample.pearsoncmg.com/dsanimation/WeightedGraphLearningTooleBook .html, as shown in Figure 29.2. The tool allows you to enter vertices, specify edges and their weights, view the graph, and find an MST and all shortest paths from a single source.

weighted graph learning tool on Companion Website

M29_LIAN1878_11_GE_C29.indd 1108 12/29/17 4:51 PM

29.2 Representing Weighted Graphs 1109

29.2 Representing Weighted Graphs Weighted edges can be stored in adjacency lists.

There are two types of weighted graphs: vertex weighted and edge weighted. In a vertex-weighted graph, each vertex is assigned a weight. In an edge-weighted graph, each edge is assigned a weight. Of the two types, edge-weighted graphs have more applications. This chapter considers edge-weighted graphs.

Weighted graphs can be represented in the same way as unweighted graphs, except that you have to represent the weights on the edges. As with unweighted graphs, the vertices in weighted graphs can be stored in an array. This section introduces three representations for the edges in weighted graphs.

29.2.1 Representing Weighted Edges: Edge Array Weighted edges can be represented using a two-dimensional array. For example, you can store all the edges in the graph in Figure 29.3a using the array in Figure 29.3b.

Note Weights can be of any type: Integer, Double, BigDecimal, and so on. You can use a two-dimensional array of the Object type to represent weighted edges as follows:

Object[][] edges = { {new Integer(0), new Integer(1), new SomeTypeForWeight(2)}, {new Integer(0), new Integer(3), new SomeTypeForWeight(8)}, ... };

Point Key

vertex-weighted graph edge-weighted graph

Figure 29.2 You can use the tool to create a weighted graph with mouse gestures and show the MST and shortest paths. Source: © Mozilla Firefox.

M29_LIAN1878_11_GE_C29.indd 1109 12/29/17 4:51 PM

1110 Chapter 29 Weighted Graphs and Applications

29.2.2 Weighted Adjacency Matrices Assume the graph has n vertices. You can use a two-dimensional n * n matrix, say weights, to represent the weights on edges. weights[i][j] represents the weight on edge (i, j). If vertices i and j are not connected, weights[i][j] is null. For example, the weights in the graph in Figure 29.3a can be represented using an adjacency matrix as follows:

Integer[][] adjacencyMatrix = { {null, 2, null, 8, null}, {2, null, 7, 3, null}, {null, 7, null, 4, 5}, {8, 3, 4, null, 6}, {null, null, 5, 6, null} };

Figure 29.3 Each edge is assigned a weight in an edge-weighted graph.

(a) (b)

int[][] edges = {{0, 1, 2}, {0, 3, 8},

{1, 0, 2}, {1, 2, 7}, {1, 3, 3},

{2, 1, 7}, {2, 3, 4}, {2, 4, 5},

{3, 0, 8}, {3, 1, 3}, {3, 2, 4}, {3, 4, 6},

{4, 2, 5}, {4, 3, 6}

};

weightvertex

1 2

3 4

0 1 2 3 4

0 null 2 null 8 null

1 2 null 7 3 null

2 null 7 null 4 5

3 8 3 4 null 6

4 null null 5 6 null

29.2.3 Adjacency Lists Another way to represent the edges is to define edges as objects. The Edge class was defined to represent an unweighted edge in Listing 28.3. For weighted edges, we define the WeightedEdge class as shown in Listing 29.1.

Listing 29.1 WeightedEdge.java 1 public class WeightedEdge extends Edge 2 implements Comparable<WeightedEdge> { 3 public double weight; // The weight on edge (u, v) 4 5 /** Create a weighted edge on (u, v) */ 6 public WeightedEdge(int u, int v, double weight) { 7 super(u, v); 8 this.weight = weight; 9 } 10 11 @Override /** Compare two edges on weights */ 12 public int compareTo(WeightedEdge edge) { 13 if (weight > edge.weight) 14 return 1; 15 else if (weight == edge.weight) 16 return 0; 17 else 18 return −1; 19 } 20 }

edge weight

constructor

compare edges

M29_LIAN1878_11_GE_C29.indd 1110 12/29/17 4:51 PM

29.3 The WeightedGraph Class 1111

An Edge object represents an edge from vertex u to v. WeightedEdge extends Edge with a new property weight. To create a WeightedEdge object, use new WeightedEdge(i, j, w), where w is the weight on edge (i, j ). Often you need to compare the weights of the edges. For this reason, the WeightedEdge class implements the Comparable interface.

For unweighted graphs, we use adjacency lists to represent edges. For weighted graphs, we still use adjacency lists, the adjacency lists for the vertices in the graph in Figure 29.3a can be represented as follows:

java.util.List<WeightedEdge>[] list = new java.util.List[5];

list[i] stores all edges adjacent to vertex i. For flexibility, we will use an array list rather than a fixed-sized array to represent list

as follows:

List<List<WeightedEdge>> list = new java.util.ArrayList<>();

29.2.1 For the code WeightedEdge edge = new WeightedEdge(1, 2, 3.5), what is edge.u, edge.v, and edge.weight?

29.2.2 What is the output of the following code?

List<WeightedEdge> list = new ArrayList<>(); list.add(new WeightedEdge(1, 2, 3.5)); list.add(new WeightedEdge(2, 3, 4.5)); WeightedEdge e = java.util.Collections.max(list); System.out.println(e.u); System.out.println(e.v); System.out.println(e.weight);

29.3 The WeightedGraph Class The WeightedGraph class extends UnweightedGraph.

The preceding chapter designed the Graph interface and the UnweightedGraph class for modeling graphs. We now design WeightedGraph as a subclass of UnweightedGraph, as shown in Figure 29.4.

WeightedGraph simply extends UnweightedGraph with five constructors for creating concrete WeightedGraph instances. WeightedGraph inherits all methods from UnweightedGraph, overrides the clear and addVertex methods, implements a new addEdge method for adding a weighted edge, and also introduces new methods for obtaining minimum spanning trees and for finding all single-source shortest paths. Minimum spanning trees and shortest paths will be introduced in Sections 29.4 and 29.5, respectively.

Listing 29.2 implements WeightedGraph. Edge adjacency lists (lines 38–63) are used internally to store adjacent edges for a vertex. When a WeightedGraph is constructed, its edge

Point Check

Point Key

WeightedEdge (0, 1, 2) WeightedEdge (0, 3, 8)

WeightedEdge (1, 0, 2) WeightedEdge (1, 3, 3) WeightedEdge (1, 2, 7)

WeightedEdge (2, 3, 4) WeightedEdge (2, 4, 5) WeightedEdge (2, 1, 7)

WeightedEdge (3, 1, 3) WeightedEdge (3, 2, 4) WeightedEdge (3, 4, 6)

WeightedEdge (4, 2, 5) WeightedEdge (4, 3, 6)

WeightedEdge (3, 0, 8)

list[0]

list[1]

list[2]

list[3]

list[4]

M29_LIAN1878_11_GE_C29.indd 1111 12/29/17 4:51 PM

1112 Chapter 29 Weighted Graphs and Applications

adjacency lists are created (lines 47 and 57). The methods getMinimumSpanningTree() (lines 99–138) and getShortestPath() (lines 156–197) will be introduced in upcoming sections.

Listing 29.2 WeightedGraph.java 1 import java.util.*; 2 3 public class WeightedGraph<V>extends UnweightedGraph<V> { 4 /** Construct an empty */ 5 public WeightedGraph() { 6 } 7 8 /** Construct a WeightedGraph from vertices and edged in arrays */ 9 public WeightedGraph(V[] vertices, int[][] edges) { 10 createWeightedGraph(java.util.Arrays.asList(vertices), edges); 11 } 12 13 /** Construct a WeightedGraph from vertices and edges in list */ 14 public WeightedGraph(int[][] edges, int numberOfVertices) { 15 List<V> vertices = new ArrayList<>(); 16 for (int i = 0; i < numberOfVertices; i++) 17 vertices.add((V)(new Integer(i))); 18 19 createWeightedGraph(vertices, edges); 20 } 21

print edges

get edge weight

create list for vertices

create a weighted edge

create list for vertices

constructor

no-arg constructor

Figure 29.4 WeightedGraph extends UnweightedGraph.

UnweightedGraph<V>

WeightedGraph<V>

+WeightedGraph()

+WeightedGraph(vertices: V[], edges: int[][])

+WeightedGraph(vertices: List<V>, edges: List<WeightedEdge>)

+WeightedGraph(edges: int[][], numberOfVertices: int)

+WeightedGraph(edges: List<WeightedEdge>, numberOfVertices: int)

+printWeightedEdges(): void

+getWeight(int u, int v): double

+addEdges(u: int, v: int, weight: double): void

+getMinimumSpanningTree(): MST

+getMinimumSpanningTree(index: int): MST

+getShortestPath(index: int): ShortestPathTree

Constructs an empty graph.

Constructs a weighted graph with the speci�ed edges and the vertices in arrays.

Constructs a weighted graph with the speci�ed edges and the number of vertices.

Constructs a weighted graph with the speci�ed edges in an array and the number of vertices.

Constructs a weighted graph with the speci�ed edges in a list and the number of vertices.

Displays all edges and weights.

Returns the weight on the edge from u to v. Throw an exception if the edge does not exist.

Adds a weighted edge to the graph and throws an IllegalArgumentException if u, v, or w is invalid. If (u, v) is already in the graph, the new weight is set.

Returns a minimum spanning tree starting from vertex 0.

Returns a minimum spanning tree starting from vertex v.

Returns all single-source shortest paths.

De�ned in Figure 28.9.

M29_LIAN1878_11_GE_C29.indd 1112 12/29/17 4:51 PM

22 /** Construct a WeightedGraph for vertices 0, 1, 2 and edge list */ 23 public WeightedGraph(List<V> vertices, List<WeightedEdge> edges) { 24 createWeightedGraph(vertices, edges); 25 } 26 27 /** Construct a WeightedGraph from vertices 0, 1, and edge array */ 28 public WeightedGraph(List<WeightedEdge> edges, 29 int numberOfVertices) { 30 List<V> vertices = new ArrayList<>(); 31 for (int i = 0; i < numberOfVertices; i++) 32 vertices.add((V)(new Integer(i))); 33 34 createWeightedGraph(vertices, edges); 35 } 36 37 /** Create adjacency lists from edge arrays */ 38 private void createWeightedGraph(List<V> vertices, int[][] edges) { 39 this.vertices = vertices; 40 41 for (int i = 0; i < vertices.size(); i++) { 42 neighbors.add(new ArrayList<Edge>()); // Create a list for vertices 43 } 44 45 for (int i = 0; i < edges.length; i++) { 46 neighbors.get(edges[i][0]).add( 47 new WeightedEdge(edges[i][0], edges[i][1], edges[i][2])); 48 } 49 } 50 51 /** Create adjacency lists from edge lists */ 52 private void createWeightedGraph( 53 List<V> vertices, List<WeightedEdge> edges) { 54 this.vertices = vertices; 55 56 for (int i = 0; i < vertices.size(); i++) { 57 neighbors.add(new ArrayList<Edge>()); // Create a list for vertices 58 } 59 60 for (WeightedEdge edge: edges) { 61 neighbors.get(edge.u).add(edge); // Add an edge into the list 62 } 63 } 64 65 /** Return the weight on the edge (u, v) */ 66 public double getWeight(int u, int v) throws Exception { 67 for (Edge edge : neighbors.get(u)) { 68 if (edge.v == v) { 69 return ((WeightedEdge)edge).weight; 70 } 71 } 72 73 throw new Exception("Edge does not exit"); 74 } 75 76 /** Display edges with weights */ 77 public void printWeightedEdges() { 78 for (int i = 0; i < getSize(); i++) { 79 System.out.print(getVertex(i) + " (" + i + "): "); 80 for (Edge edge : neighbors.get(i)) { 81 System.out.print("(" + edge.u +

adjacency lists are created (lines 47 and 57). The methods getMinimumSpanningTree() (lines 99–138) and getShortestPath() (lines 156–197) will be introduced in upcoming sections.

print edges

get edge weight

create list for vertices

create a weighted edge

create list for vertices

constructor

no-arg constructor

29.3 The WeightedGraph Class 1113

M29_LIAN1878_11_GE_C29.indd 1113 12/29/17 4:51 PM

1114 Chapter 29 Weighted Graphs and Applications

82 ", " + edge.v + ", " + ((WeightedEdge)edge).weight + ") "); 83 } 84 System.out.println(); 85 } 86 } 87 88 /** Add edges to the weighted graph */ 89 public boolean addEdge(int u, int v, double weight) { 90 return addEdge(new WeightedEdge(u, v, weight)); 91 } 92 93 /** Get a minimum spanning tree rooted at vertex 0 */ 94 public MST getMinimumSpanningTree() { 95 return getMinimumSpanningTree(0); 96 } 97 98 /** Get a minimum spanning tree rooted at a specified vertex */ 99 public MST getMinimumSpanningTree(int startingVertex) { 100 // cost[v] stores the cost by adding v to the tree 101 double[] cost = new double[getSize()]; 102 for (int i = 0; i < cost.length; i++) { 103 cost[i] = Double.POSITIVE_INFINITY; // Initial cost 104 } 105 cost[startingVertex] = 0; // Cost of source is 0 106 107 int[] parent = new int[getSize()]; // Parent of a vertex 108 parent[startingVertex] = −1; // startingVertex is the root 109 double totalWeight = 0; // Total weight of the tree thus far 110 111 List<Integer> T = new ArrayList<>(); 112 113 // Expand T 114 while (T.size() < getSize()) { 115 // Find smallest cost u in V − T 116 int u = −1; // Vertex to be determined 117 double currentMinCost = Double.POSITIVE_INFINITY; 118 for (int i = 0; i < getSize(); i++) { 119 if (!T.contains(i) && cost[i] < currentMinCost) { 120 currentMinCost = cost[i]; 121 u = i; 122 } 123 } 124 125 if (u == −1) break; else T.add(u); // Add a new vertex to T 126 totalWeight += cost[u]; // Add cost[u] to the tree 127 128 // Adjust cost[v] for v that is adjacent to u and v in V − T 129 for (Edge e: neighbors.get(u)) { 130 if (!T.contains(e.v) && cost[e.v] > ((WeightedEdge)e).weight) { 131 cost[e.v] = ((WeightedEdge)e).weight; 132 parent[e.v] = u; 133 } 134 } 135 } // End of while 136 137 return new MST(startingVertex, parent, T, totalWeight); 138 } 139 140 /** MST is an inner class in WeightedGraph */ 141 public class MST extends SearchTree {

add edge

get an MST start from vertex 0

MST from a starting vertex

initialize cost

initialize parent

minimum spanning tree

expand MST update total cost

vertex with smallest cost

add to tree

adjust cost

create an MST

MST inner class

M29_LIAN1878_11_GE_C29.indd 1114 12/29/17 4:51 PM

142 private double totalWeight; // Total weight of all edges in the tree 143 144 public MST(int root, int[] parent, List<Integer> searchOrder, 145 double totalWeight) { 146 super(root, parent, searchOrder); 147 this.totalWeight = totalWeight; 148 } 149 150 public double getTotalWeight() { 151 return totalWeight; 152 } 153 } 154 155 /** Find single-source shortest paths */ 156 public ShortestPathTree getShortestPath(int sourceVertex) { 157 // cost[v] stores the cost of the path from v to the source 158 double[] cost = new double[getSize()]; 159 for (int i = 0; i < cost.length; i++) { 160 cost[i] = Double.POSITIVE_INFINITY; // Initial cost set to infinity 161 } 162 cost[sourceVertex] = 0; // Cost of source is 0 163 164 // parent[v] stores the previous vertex of v in the path 165 int[] parent = new int[getSize()]; 166 parent[sourceVertex] = −1; // The parent of source is set to −1 167 168 // T stores the vertices whose path found so far 169 List<Integer> T = new ArrayList<>(); 170 171 // Expand T 172 while (T.size() < getSize()) { 173 // Find smallest cost u in V − T 174 int u = −1; // Vertex to be determined 175 double currentMinCost = Double.POSITIVE_INFINITY; 176 for (int i = 0; i < getSize(); i++) { 177 if (!T.contains(i) && cost[i] < currentMinCost) { 178 currentMinCost = cost[i]; 179 u = i; 180 } 181 } 182 183 if (u == −1) break; else T.add(u); // Add a new vertex to T 184 185 // Adjust cost[v] for v that is adjacent to u and v in V − T 186 for (Edge e: neighbors.get(u)) { 187 if (!T.contains(e.v) 188 && cost[e.v] > cost[u] + ((WeightedEdge)e).weight) { 189 cost[e.v] = cost[u] + ((WeightedEdge)e).weight; 190 parent[e.v] = u; 191 } 192 } 193 } // End of while 194 195 // Create a ShortestPathTree 196 return new ShortestPathTree(sourceVertex, parent, T, cost); 197 } 198 199 /** ShortestPathTree is an inner class in WeightedGraph */ 200 public class ShortestPathTree extends SearchTree { 201 private double[] cost; // cost[v] is the cost from v to source

29.3 The WeightedGraph Class 1115

total weight in tree

getShortestPath

initialize cost

shortest-path tree

expand tree

vertex with smallest cost

add to T

adjust cost adjust parent

create a tree

shortest-path tree cost

M29_LIAN1878_11_GE_C29.indd 1115 12/29/17 4:51 PM

1116 Chapter 29 Weighted Graphs and Applications

202 203 /** Construct a path */ 204 public ShortestPathTree(int source, int[] parent, 205 List<Integer> searchOrder, double[] cost) { 206 super(source, parent, searchOrder); 207 this.cost = cost; 208 } 209 210 /** Return the cost for a path from the root to vertex v */ 211 public double getCost(int v) { 212 return cost[v]; 213 } 214 215 /** Print paths from all vertices to the source */ 216 public void printAllPaths() { 217 System.out.println("All shortest paths from " + 218 vertices.get(getRoot()) + " are:"); 219 for (int i = 0; i < cost.length; i++) { 220 printPath(i); // Print a path from i to the source 221 System.out.println("(cost: " + cost[i] + ")"); // Path cost 222 } 223 } 224 } 225 }

The WeightedGraph class extends the UnweightedGraph class (line 3). The proper- ties vertices and neighbors in UnweightedGraph are inherited in WeightedGraph. neighbors is a list. Each element is the list is another list that contains edges. For unweighted graph, each edge is an instance of Edge. For a weighted graph, each edge is an instance of WeightedEdge. WeightedEdge is a subtype of Edge. So you can add a weighted edge into neighbors.get(i) for a weighted graph (line 47).

The addEdge(u, v, weight) method (lines 88–91) adds an edge (u, v, weight) to the graph. If a graph is undirected, you should invoke addEdge(u, v, weight) and addEdge(v, u, weight) to add an edge between u and v.

Listing 29.3 gives a test program that creates a graph for the one in Figure 29.1, and another graph for the one in Figure 29.3a.

Listing 29.3 TestWeightedGraph.java 1 public class TestWeightedGraph { 2 public static void main(String[] args) { 3 String[] vertices = {"Seattle", "San Francisco", "Los Angeles", 4 "Denver", "Kansas City", "Chicago", "Boston", "New York", 5 "Atlanta", "Miami", "Dallas", "Houston"}; 6 7 int[][] edges = { 8 {0, 1, 807}, {0, 3, 1331}, {0, 5, 2097}, 9 {1, 0, 807}, {1, 2, 381}, {1, 3, 1267}, 10 {2, 1, 381}, {2, 3, 1015}, {2, 4, 1663}, {2, 10, 1435}, 11 {3, 0, 1331}, {3, 1, 1267}, {3, 2, 1015}, {3, 4, 599}, 12 {3, 5, 1003}, 13 {4, 2, 1663}, {4, 3, 599}, {4, 5, 533}, {4, 7, 1260}, 14 {4, 8, 864}, {4, 10, 496}, 15 {5, 0, 2097}, {5, 3, 1003}, {5, 4, 533}, 16 {5, 6, 983}, {5, 7, 787}, 17 {6, 5, 983}, {6, 7, 214}, 18 {7, 4, 1260}, {7, 5, 787}, {7, 6, 214}, {7, 8, 888}, 19 {8, 4, 864}, {8, 7, 888}, {8, 9, 661},

edges

vertices

constructor

get cost

print all paths

M29_LIAN1878_11_GE_C29.indd 1116 12/29/17 4:51 PM

20 {8, 10, 781}, {8, 11, 810}, 21 {9, 8, 661}, {9, 11, 1187}, 22 {10, 2, 1435}, {10, 4, 496}, {10, 8, 781}, {10, 11, 239}, 23 {11, 8, 810}, {11, 9, 1187}, {11, 10, 239} 24 }; 25 26 WeightedGraph<String> graph1 = 27 new WeightedGraph<>(vertices, edges); 28 System.out.println("The number of vertices in graph1: " 29 + graph1.getSize()); 30 System.out.println("The vertex with index 1 is " 31 + graph1.getVertex(1)); 32 System.out.println("The index for Miami is " + 33 graph1.getIndex("Miami")); 34 System.out.println("The edges for graph1:"); 35 graph1.printWeightedEdges(); 36 37 edges = new int[][] { 38 {0, 1, 2}, {0, 3, 8}, 39 {1, 0, 2}, {1, 2, 7}, {1, 3, 3}, 40 {2, 1, 7}, {2, 3, 4}, {2, 4, 5}, 41 {3, 0, 8}, {3, 1, 3}, {3, 2, 4}, {3, 4, 6}, 42 {4, 2, 5}, {4, 3, 6} 43 }; 44 WeightedGraph<Integer> graph2 = new WeightedGraph<>(edges, 5); 45 System.out.println("\nThe edges for graph2:"); 46 graph2.printWeightedEdges(); 47 } 48 }

create graph

print edges

edges

create graph

print edges

The number of vertices in graph1: 12 The vertex with index 1 is San Francisco The index for Miami is 9 The edges for graph1: Vertex 0: (0, 1, 807) (0, 3, 1331) (0, 5, 2097) Vertex 1: (1, 2, 381) (1, 0, 807) (1, 3, 1267) Vertex 2: (2, 1, 381) (2, 3, 1015) (2, 4, 1663) (2, 10, 1435) Vertex 3: (3, 4, 599) (3, 5, 1003) (3, 1, 1267) (3, 0, 1331) (3, 2, 1015) Vertex 4: (4, 10, 496) (4, 8, 864) (4, 5, 533) (4, 2, 1663) (4, 7, 1260) (4, 3, 599) Vertex 5: (5, 4, 533) (5, 7, 787) (5, 3, 1003) (5, 0, 2097) (5, 6, 983) Vertex 6: (6, 7, 214) (6, 5, 983) Vertex 7: (7, 6, 214) (7, 8, 888) (7, 5, 787) (7, 4, 1260) Vertex 8: (8, 9, 661) (8, 10, 781) (8, 4, 864) (8, 7, 888) (8, 11, 810) Vertex 9: (9, 8, 661) (9, 11, 1187) Vertex 10: (10, 11, 239) (10, 4, 496) (10, 8, 781) (10, 2, 1435) Vertex 11: (11, 10, 239) (11, 9, 1187) (11, 8, 810)

The edges for graph2: Vertex 0: (0, 1, 2) (0, 3, 8) Vertex 1: (1, 0, 2) (1, 2, 7) (1, 3, 3) Vertex 2: (2, 3, 4) (2, 1, 7) (2, 4, 5) Vertex 3: (3, 1, 3) (3, 4, 6) (3, 2, 4) (3, 0, 8) Vertex 4: (4, 2, 5) (4, 3, 6)

29.3 The WeightedGraph Class 1117

M29_LIAN1878_11_GE_C29.indd 1117 12/29/17 4:51 PM

1118 Chapter 29 Weighted Graphs and Applications

The program creates graph1 for the graph in Figure 29.1 in lines 3–27. The vertices for graph1 are defined in lines 3–5. The edges for graph1 are defined in lines 7–24. The edges are represented using a two-dimensional array. For each row i in the array, edges[i][0] and edges[i][1] indicate there is an edge from vertex edges[i][0] to vertex edges[i][1] and the weight for the edge is edges[i][2]. For example, {0, 1, 807} (line 8) represents the edge from vertex 0 (edges[0][0] ) to vertex 1 (edges[0] [1]) with weight 807 (edges[0][2]). {0, 5, 2097} (line 8) represents the edge from vertex 0 (edges[2][0]) to vertex 5 (edges[2][1]) with weight 2097 (edges[2][2]). Line 35 invokes the printWeightedEdges() method on graph1 to display all edges in graph1.

The program creates the edges for graph2 for the graph in Figure 29.3a in lines 37–44. Line 46 invokes the printWeightedEdges() method on graph2 to display all edges in graph2.

29.3.1 If a priority queue is used to store weighted edges, what is the output of the following code?

PriorityQueue<WeightedEdge> q = new PriorityQueue<>(); q.offer(new WeightedEdge(1, 2, 3.5)); q.offer(new WeightedEdge(1, 6, 6.5)); q.offer(new WeightedEdge(1, 7, 1.5)); System.out.println(q.poll().weight); System.out.println(q.poll().weight); System.out.println(q.poll().weight);

29.3.2 If a priority queue is used to store weighted edges, what is wrong in the following code? Fix it and show the output.

List<PriorityQueue<WeightedEdge>> queues = new ArrayList<>(); queues.get(0).offer(new WeightedEdge(0, 2, 3.5)); queues.get(0).offer(new WeightedEdge(0, 6, 6.5)); queues.get(0).offer(new WeightedEdge(0, 7, 1.5)); queues.get(1).offer(new WeightedEdge(1, 0, 3.5)); queues.get(1).offer(new WeightedEdge(1, 5, 8.5)); queues.get(1).offer(new WeightedEdge(1, 8, 19.5)); System.out.println(queues.get(0).peek() .compareTo(queues.get(1).peek()));

29.3.3 Show the output of the following code:

public class Test { public static void main(String[] args) { WeightedGraph<Character> graph = new WeightedGraph<>(); graph.addVertex('U'); graph.addVertex('V'); int indexForU = graph.getIndex('U'); int indexForV = graph.getIndex('V'); System.out.println("indexForU is " + indexForU); System.out.println("indexForV is " + indexForV); graph.addEdge(indexForU, indexForV, 2.5); System.out.println("Degree of U is " + graph.getDegree(indexForU)); System.out.println("Degree of V is " + graph.getDegree(indexForV)); System.out.println("Weight of UV is " + graph.getWeight(indexForU, indexOfV)); } }

Point Check

M29_LIAN1878_11_GE_C29.indd 1118 12/29/17 4:51 PM

29.4 Minimum Spanning Trees 1119

29.4 Minimum Spanning Trees A minimum spanning tree of a graph is a spanning tree with the minimum total weights.

A graph may have many spanning trees. Suppose the edges are weighted. A minimum spanning tree has the minimum total weights. For example, the trees in Figures 29.5b, 29.5c, 29.5d are spanning trees for the graph in Figure 29.5a. The trees in Figures 29.5c and 29.5d are minimum spanning trees.

Point Key

minimum spanning tree

Figure 29.5 The trees in (c) and (d) are minimum spanning trees of the graph in (a).

(d) Total weight is 38(c) Total weight is 38

(b) Total weight is 42(a)

7 7 8

6 7

7 8

The problem of finding a minimum spanning tree has many applications. Consider a com- pany with branches in many cities. The company wants to lease telephone lines to connect all the branches together. The phone company charges different rates to connect different pairs of cities. There are many ways to connect all branches together. The cheapest way is to find a spanning tree with the minimum total rates.

29.4.1 Minimum Spanning Tree Algorithms How do you find a minimum spanning tree? There are several well-known algorithms for doing so. This section introduces Prim’s algorithm. Prim’s algorithm starts with a spanning tree T that contains an arbitrary vertex. The algorithm expands the tree by repeatedly adding a vertex with the lowest-cost edge incident to a vertex already in the tree. Prim’s algorithm is a greedy algorithm, and it is described in Listing 29.4.

Listing 29.4 Prim's Minimum Spanning Tree Algorithm Input: A connected undirected weighted G = (V, E) with nonnegative weights Output: MST (a minimum spanning tree)

1 MST minimumSpanningTree() { 2 Let T be a set for the vertices in the spanning tree;

Prim’s algorithm

M29_LIAN1878_11_GE_C29.indd 1119 12/29/17 4:51 PM

1120 Chapter 29 Weighted Graphs and Applications

3 Initially, add the starting vertex, s, to T; 4 5 while (size of T < n) { 6 Find x in T and y in V – T with the smallest weight 7 on the edge (x, y), as shown in Figure 29.6; 8 Add y to T and set parent[y] = x; 9 } 10 }

add initial vertex

more vertices? find a vertex

add to tree

Figure 29.6 Find a vertex x in T that connects a vertex y in V – T with the smallest weight.

V – T Vertices already in the spanning tree

Vertices not currently in the spanning tree

The algorithm starts by adding the starting vertex into T. It then continuously adds a vertex (say y) from V – T into T. y is the vertex that is adjacent to a vertex in T with the smallest weight on the edge. For example, there are five edges connecting vertices in T and V – T as shown in Figure 29.6, and (x, y) is the one with the smallest weight. Consider the graph in Figure 29.7. The algorithm adds the vertices to T in this order:

1. Add vertex 0 to T.

2. Add vertex 5 to T, since WeightedEdge(5, 0, 5) has the smallest weight among all edges incident to a vertex in T, as shown in Figure 29.7a. The arrow line from 0 to 5 indicates that 0 is the parent of 5.

3. Add vertex 1 to T, since WeightedEdge(1, 0, 6) has the smallest weight among all edges incident to a vertex in T, as shown in Figure 29.7b.

4. Add vertex 6 to T, since WeightedEdge(6, 1, 7) has the smallest weight among all edges incident to a vertex in T, as shown in Figure 29.7c.

5. Add vertex 2 to T, since WeightedEdge(2, 6, 5) has the smallest weight among all edges incident to a vertex in T, as shown in Figure 29.7d.

6. Add vertex 4 to T, since WeightedEdge(4, 6, 7) has the smallest weight among all edges incident to a vertex in T, as shown in Figure 29.7e.

7. Add vertex 3 to T, since WeightedEdge(3, 2, 8) has the smallest weight among all edges incident to a vertex in T, as shown in Figure 29.7f.

Note A minimum spanning tree is not unique. For example, both (c) and (d) in Figure 29.5 are minimum spanning trees for the graph in Figure 29.5a. However, if the weights are distinct, the graph has a unique minimum spanning tree.

Note Assume the graph is connected and undirected. If a graph is not connected or directed, the algorithm will not work. You can modify the algorithm to find a spanning forest for any undirected graph. A spanning forest is a graph in which each connected component is a tree.

example

unique tree?

connected and undirected

M29_LIAN1878_11_GE_C29.indd 1120 12/29/17 4:51 PM

29.4 Minimum Spanning Trees 1121

29.4.2 Refining Prim’s MST Algorithm To make it easy to identify the next vertex to add into the tree, we use cost[v] to store the cost of adding a vertex v to the spanning tree T. Initially, cost[s] is 0 for a starting vertex and assign infinity to cost[v] for all other vertices. The algorithm repeatedly finds a vertex u in V–T with the smallest cost[u] and moves u to T. The refined version of the alogrithm is given in Listing 29.5.

Listing 29.5 Refined Version of Prim's Algorithm Input: A connected undirected weighted G = (V, E) with nonnegative weights Output: a minimum spanning tree with the starting vertex s as the root

Figure 29.7 The adjacent vertices with the smallest weight are added successively to T.

(b)

5 4

77 8

6 3

(c)

5 4

77 8

6 3

5 4

77 8

6 3

(e)

5 4

77 8

6 3

(f)

5 4

77 8

6 3

1 T

(a)

5 4

77 8

6 3

(d)

M29_LIAN1878_11_GE_C29.indd 1121 12/29/17 4:51 PM

1122 Chapter 29 Weighted Graphs and Applications

1 MST getMinimumSpanngingTree(s) { 2 Let T be a set that contains the vertices in the spanning tree; 3 Initially T is empty; 4 Set cost[s] = 0 and cost[v] = infinity for all other vertices in V; 5 6 while (size of T < n) { 7 Find u not in T with the smallest cost[u]; 8 Add u to T; 9 for (each v not in T and (u, v) in E) 10 if (cost[v] > w(u, v)) { // Adjust cost[v] 11 cost[v] = w(u, v); parent[v] = u; 12 } 13 } 14 }

For an interactive demo on how the refined Prim’s algorithm works, see liveexample .pearsoncmg.com/dsanimation/RefinedPrim.html.

29.4.3 Implementation of the MST Algorithm The getMinimumSpanningTree(int v) method is defined in the WeightedGraph class, as shown in Figure 29.4. It returns an instance of the MST class. The MST class is defined as an inner class in the WeightedGraph class, which extends the SearchTree class, as shown in Figure 29.8. The SearchTree class was shown in Figure 28.11. The MST class was implemented in lines 141–153 in Listing 29.2.

find next vertex add a vertex to T

adjust cost[v]

getMinimumSpanningTree()

Figure 29.8 The MST class extends the SearchTree class.

UnweightedGraph<V>.SearchTree

WeightedGraph<V>.MST

–totalWeight: double

+MST(root: int, parent: int[], searchOrder: List<Integer> totalWeight: double)

+getTotalWeight(): double

Total weight of the tree.

Constructs an MST with the speci�ed root, parent array, searchOrder, and total weight for the tree.

Returns the totalWeight of the tree.

The refined version of the Prim’s algorithm greatly simplifies the implementation. The getMinimumSpanningTree method was implemented using the refined version of the Prim’s algorithm in lines 99–138 in Listing 29.2. The getMinimumSpanningTree(int startingVertex) method sets cost[startingVertex] to 0 (line 105) and cost[v] to infinity for all other vertices (lines 102–104). The parent of startingVertex is set to −1 (line 108). T is a list that stores the vertices added into the spanning tree (line 111). We use a list for T rather than a set in order to record the order of the vertices added to T.

Initially, T is empty. To expand T, the method performs the following operations:

1. Find the vertex u with the smallest cost[u] (lines 118–123).

2. If u is found, add it to T (line 125). Note if u is not found (u == −1), the graph is not connected. The break statement exits the while loop in this case.

3. After adding u in T, update cost[v] and parent[v] for each v adjacent to u in V−T if cost[v] > w(u, v) (lines 129–134).

After a new vertex is added to T, totalWeight is updated (line 126). Once all vertices are added to T, an instance of MST is created (line 137). Note the method will not work if the graph is not connected. However, you can modify it to obtain a partial MST.

M29_LIAN1878_11_GE_C29.indd 1122 12/29/17 4:51 PM

29.4 Minimum Spanning Trees 1123

The MST class extends the SearchTree class (line 141). To create an instance of MST, pass root, parent, T, and totalWeight (lines 144–145). The data fields root, parent, and searchOrder are defined in the SearchTree class, which is an inner class defined in UnweightedGraph.

Note testing whether a vertex i is in T by invoking T.contains(i) takes O(n) time, since T is a list. Therefore, the overall time complexity for this implementation is O(n3). Interested readers may see Programming Exercise 29.20 for improving the implementation and reduce the complexity to O(n2).

Listing 29.6 gives a test program that displays minimum spanning trees for the graph in Figure 29.1 and the graph in Figure 29.3a, respectively.

Listing 29.6 TestMinimumSpanningTree.java 1 public class TestMinimumSpanningTree { 2 public static void main(String[] args) { 3 String[] vertices = {"Seattle", "San Francisco", "Los Angeles", 4 "Denver", "Kansas City", "Chicago", "Boston", "New York", 5 "Atlanta", "Miami", "Dallas", "Houston"}; 6 7 int[][] edges = { 8 {0, 1, 807}, {0, 3, 1331}, {0, 5, 2097}, 9 {1, 0, 807}, {1, 2, 381}, {1, 3, 1267}, 10 {2, 1, 381}, {2, 3, 1015}, {2, 4, 1663}, {2, 10, 1435}, 11 {3, 0, 1331}, {3, 1, 1267}, {3, 2, 1015}, {3, 4, 599}, 12 {3, 5, 1003}, 13 {4, 2, 1663}, {4, 3, 599}, {4, 5, 533}, {4, 7, 1260}, 14 {4, 8, 864}, {4, 10, 496}, 15 {5, 0, 2097}, {5, 3, 1003}, {5, 4, 533}, 16 {5, 6, 983}, {5, 7, 787}, 17 {6, 5, 983}, {6, 7, 214}, 18 {7, 4, 1260}, {7, 5, 787}, {7, 6, 214}, {7, 8, 888}, 19 {8, 4, 864}, {8, 7, 888}, {8, 9, 661}, 20 {8, 10, 781}, {8, 11, 810}, 21 {9, 8, 661}, {9, 11, 1187}, 22 {10, 2, 1435}, {10, 4, 496}, {10, 8, 781}, {10, 11, 239}, 23 {11, 8, 810}, {11, 9, 1187}, {11, 10, 239} 24 }; 25 26 WeightedGraph<String> graph1 = 27 new WeightedGraph<>(vertices, edges); 28 WeightedGraph<String>.MST tree1 = graph1.getMinimumSpanningTree(); 29 System.out.println("tree1: Total weight is " + 30 tree1.getTotalWeight()); 31 tree1.printTree(); 32 33 edges = new int[][] { 34 {0, 1, 2}, {0, 3, 8}, 35 {1, 0, 2}, {1, 2, 7}, {1, 3, 3}, 36 {2, 1, 7}, {2, 3, 4}, {2, 4, 5}, 37 {3, 0, 8}, {3, 1, 3}, {3, 2, 4}, {3, 4, 6}, 38 {4, 2, 5}, {4, 3, 6} 39 }; 40 41 WeightedGraph<Integer> graph2 = new WeightedGraph<>(edges, 5); 42 WeightedGraph<Integer>.MST tree2 = 43 graph2.getMinimumSpanningTree(1); 44 System.out.println("\ntree2: Total weight is " + 45 tree2.getTotalWeight()); 46 tree2.printTree();

time complexity

print tree total weight

MST for graph2

create graph2

create edges

print tree

total weight MST for graph1

create graph1

create edges

create vertices

M29_LIAN1878_11_GE_C29.indd 1123 12/29/17 4:51 PM

1124 Chapter 29 Weighted Graphs and Applications

47 48 System.out.println("\nShow the search order for tree1:"); 49 for (int i: tree1.getSearchOrder()) 50 System.out.print(graph1.getVertex(i) + " "); 51 } 52 }

display search order

Total weight is 6513.0

Root is: Seattle Edges: (Seattle, San Francisco) (San Francisco, Los Angeles) (Los Angeles, Denver) (Denver, Kansas City) (Kansas City, Chicago) (New York, Boston) (Chicago, New York) (Dallas, Atlanta) (Atlanta, Miami) (Kansas City, Dallas) (Dallas, Houston)

Total weight is 14.0 Root is: 1 Edges: (1, 0) (3, 2) (1, 3) (2, 4)

Show the search order for tree1: Seattle San Francisco Los Angeles Denver Kansas City Dallas

Houston Chicago Atlanta Miami New York Boston

Figure 29.9 The edges in a minimum spanning tree for the cities are highlighted.

Seattle

Los Angeles

Denver

Chicago

Houston

Boston

New York

Atlanta

Miami

661

888

1187

810 Dallas

1331

2097

1003 807

381

1015

1267

1663

1435

239

496

781

864

1260

983

787

214

533

599

San Francisco Kansas City

The program creates a weighted graph for Figure 29.1 in line 27. It then invokes getMinimumSpanningTree() (line 28) to return an MST that represents a minimum span- ning tree for the graph. Invoking printTree() (line 31) on the MST object displays the edges in the tree. Note that MST is a subclass of Tree. The printTree() method is defined in the SearchTree class.

The graphical illustration of the minimum spanning tree is shown in Figure 29.9. The vertices are added to the tree in this order: Seattle, San Francisco, Los Angeles, Denver, Kansas City, Dallas, Houston, Chicago, Atlanta, Miami, New York, and Boston.

graphical illustration

M29_LIAN1878_11_GE_C29.indd 1124 12/29/17 4:51 PM

29.5 Finding Shortest Paths 1125

29.4.1 Find a minimum spanning tree for the following graph: Point

Check

29.4.2 Is a minimum spanning tree unique if all edges have different weights? 29.4.3 If you use an adjacency matrix to represent weighted edges, what will be the time

complexity for Prim’s algorithm?

29.4.4 What happens to the getMinimumSpanningTree() method in WeightedGraph if the graph is not connected? Verify your answer by writing a test program that creates an unconnected graph and invokes the getMinimumSpanningTree() method. How do you fix the problem by obtaining a partial MST?

29.4.5 Show the output of the following code:

public class Test { public static void main(String[] args) { WeightedGraph<Character> graph = new WeightedGraph<>(); graph.addVertex('U'); graph.addVertex('V'); graph.addVertex('X'); int indexForU = graph.getIndex('U'); int indexForV = graph.getIndex('V'); int indexForX = graph.getIndex('X'); System.out.println("indexForU is " + indexForU); System.out.println("indexForV is " + indexForV); System.out.println("indexForX is " + indexForV); graph.addEdge(indexForU, indexForV, 3.5); graph.addEdge(indexForV, indexForU, 3.5); graph.addEdge(indexForU, indexForX, 2.1); graph.addEdge(indexForX, indexForU, 2.1); graph.addEdge(indexForV, indexForX, 3.1); graph.addEdge(indexForX, indexForV, 3.1); WeightedGraph<Character>.MST mst = graph.getMinimumSpanningTree(); graph.printWeightedEdges(); System.out.println(mst.getTotalWeight()); mst.printTree(); } }

29.5 Finding Shortest Paths The shortest path between two vertices is a path with the minimum total weights.

Given a graph with nonnegative weights on the edges, a well-known algorithm for finding a shortest path between two vertices was discovered by Edsger Dijkstra, a Dutch computer scientist. In order to find a shortest path from vertex s to vertex v, Dijkstra’s algorithm finds the shortest path from s to all vertices. So Dijkstra’s algorithm is known as a

Point Key

Dijkstra’s algorithm

5 4

77 8

6 3

M29_LIAN1878_11_GE_C29.indd 1125 12/29/17 4:51 PM

1126 Chapter 29 Weighted Graphs and Applications

single-source shortest-path algorithm. The algorithm uses cost[v] to store the cost of a shortest path from vertex v to the source vertex s. cost[s] is 0. Initially assign infinity to cost[v] for all other vertices. The algorithm repeatedly finds a vertex u in V–T with the smallest cost[u] and moves u to T.

The algorithm is described in Listing 29.7.

Listing 29.7 Dijkstra's Single-Source Shortest-Path Algorithm Input: a graph G = (V, E) with nonnegative weights Output: a shortest-path tree with the source vertex s as the root

1 ShortestPathTree getShortestPath(s) { 2 Let T be a set that contains the vertices whose 3 paths to s are known; Initially T is empty; 4 Set cost[s] = 0; and cost[v] = infinity for all other vertices in V; 5 6 while (size of T < n) { 7 Find u not in T with the smallest cost[u]; 8 Add u to T; 9 for (each v not in T and (u, v) in E) 10 if (cost[v] > cost[u] + w(u, v)) { 11 cost[v] = cost[u] + w(u, v); parent[v] = u; 12 } 13 } 14 }

This algorithm is very similar to Prim’s for finding a minimum spanning tree. Both algorithms divide the vertices into two sets: T and V − T. In the case of Prim’s algorithm, set T contains the vertices that are already added to the tree. In the case of Dijkstra’s, set T contains the ver- tices whose shortest paths to the source have been found. Both algorithms repeatedly find a vertex from V − T and add it to T. In the case of Prim’s algorithm, the vertex is adjacent to some vertex in the set with the minimum weight on the edge. In Dijkstra’s algorithm, the vertex is adjacent to some vertex in the set with the minimum total cost to the source.

The algorithm starts by setting cost[s] to 0 (line 4), sets cost[v] to infinity for all other vertices. It then continuously adds a vertex (say u) from V–T into T with smallest cost[u] (lines 7–8), as shown in Figure 29.10a. After adding u to T, the algorithm updates cost[v] and parent[v] for each v not in T if (u, v) is in T and cost[v] > cost[u] + w(u, v) (lines 10–12).

single-source shortest path shortest path

find next vertex add a vertex to T

adjust cost[v]

Figure 29.10 (a) Find a vertex u in V–T with the smallest cost[u]. (b) Update cost[v] for v in V–T and v is adjacent to u.

u T

V – T

T contains vertices whose shortest path to s are known

V – T contains vertices whose shortest path to s are not known yet.

V – T

T contains vertices whose shortest path to s are known

V – T contains vertices whose shortest path to s are not known yet.

(a) Before moving u to T (b) After moving u to T

M29_LIAN1878_11_GE_C29.indd 1126 12/29/17 4:51 PM

29.5 Finding Shortest Paths 1127

Let us illustrate Dijkstra’s algorithm using the graph in Figure 29.11a. Suppose the source vertex is 1. Therefore, cost[1] = 0 and the costs for all other vertices are initially ∞ , as shown in Figure 29.11b. We use the parent[i] to denote the parent of i in the path. For convenience, set the parent of the source node to −1.

Figure 29.11 The algorithm will find all shortest paths from source vertex 1.

0 5

72 5

6 4

0 1 2 3 4 5 6

parent

–1

0 1 2 3 4 5 6

cost

`` ` ` ` `

(b)(a)

Initially set T is empty. The algorithm selects the vertex with the smallest cost. In this case, the vertex is 1. The algorithm adds 1 to T, as shown in Figure 29.12a. Afterward, it adjusts the cost for each vertex adjacent to 1. The cost for vertices 2, 0, 6, and 3 and their parents are now updated, as shown in Figure 29.12b.

Figure 29.12 Now vertex 1 is in set T.

0 1 2 3 4 5 6

parent

–1 11 1 1

(b)(a)

0 5

72 5

6 4

1 T

0 1 2 3 4 5 6

cost

5 10 ` ` 9

Vertices 2, 0, 6, and 3 are adjacent to the source vertex, and vertex 2 is the one in V−T with the smallest cost, so add 2 to T, as shown in Figure 29.13 and update the cost and parent for vertices in V−T and adjacent to 2. cost[0] is now updated to 6 and its parent is set to 2. The arrow line from 1 to 2 indicates 1 is the parent of 2 after 2 is added into T.

M29_LIAN1878_11_GE_C29.indd 1127 12/29/17 4:51 PM

1128 Chapter 29 Weighted Graphs and Applications

Now T contains {1, 2}. Vertex 0 is the one in V−T with the smallest cost, so add 0 to T, as shown in Figure 29.14 and update the cost and parent for vertices in V−T and adjacent to 0 if applicable. cost[5] is now updated to 10 and its parent is set to 0 and cost[6] is now updated to 8 and its parent is set to 0.

Figure 29.13 Now vertices 1 and 2 are in set T.

0 1 2 3 4 5 6

parent

2 1 1–1

(b)(a)

0 5

72 5

6 4

T 10

0 1 2 3 4 5 6

cost

56 10 ` ` 9

Figure 29.14 Now vertices {1, 2, 0} are in set T.

0 1 2 3 4 5 6

parent

2 1 1–1

(b)(a)

0 5

72 5

6 4

T 10

0 1 2 3 4 5 6

cost

56 10 ` 10 8

Figure 29.15 Now vertices {1, 2, 0, 6} are in set T.

0 1 2 3 4 5 6

parent

2 1–1

(b)(a)

0 5

72 5

6 4

0 1 2 3 4 5 6

cost

56 10 ` 10 8

01 0

Now T contains {1, 2, 0}. Vertex 6 is the one in V−T with the smallest cost, so add 6 to T, as shown in Figure 29.15 and update the cost and parent for vertices in V−T and adjacent to 6 if applicable.

M29_LIAN1878_11_GE_C29.indd 1128 12/29/17 4:51 PM

29.5 Finding Shortest Paths 1129

Now T contains {1, 2, 0, 6}. Vertex 3 or 5 is the one in V−T with the smallest cost. You may add either 3 or 5 into T. Let us add 3 to T, as shown in Figure 29.16 and update the cost and parent for vertices in V−T and adjacent to 3 if applicable. cost[4] is now updated to 18 and its parent is set to 3.

Figure 29.16 Now vertices {1, 2, 0, 6, 3} are in set T.

0 1 2 3 4 5 6

parent

2 1–1

(b)(a)

0 5

72 5

6 4

0 1 2 3 4 5 6

cost

56 10 18 10 8

01 03

Now T contains {1, 2, 0, 6, 3}. Vertex 5 is the one in V−T with the smallest cost, so add 5 to T, as shown in Figure 29.17 and update the cost and parent for vertices in V−T and adjacent to 5 if applicable. cost[4] is now updated to 15, and its parent is set to 5.

0 1 2 3 4 5 6

parent

2 1–1

(b)(a)

0 5

72 5

6 4

0 1 2 3 4 5 6

cost

56 10 15 10 8

01 05

Now T contains {1, 2, 0, 6, 3, 5}. Vertex 4 is the one in V−T with the smallest cost, so add 4 to T, as shown in Figure 29.18.

As you can see, the algorithm essentially finds all shortest paths from a source vertex, which produces a tree rooted at the source vertex. We call this tree a single-source all-shortest-path tree (or simply a shortest-path tree). To model this tree, define a class named ShortestPathTree that extends the SearchTree class, as shown in Figure 29.19. ShortestPathTree is defined as an inner class in WeightedGraph in lines 200–224 in Listing 29.2.

The getShortestPath(int sourceVertex) method was implemented in lines 156–197 in Listing 29.2. The method sets cost[sourceVertex] to 0 (line 162) and cost[v] to

shortest-path tree

M29_LIAN1878_11_GE_C29.indd 1129 12/29/17 4:51 PM

1130 Chapter 29 Weighted Graphs and Applications

infinity for all other vertices (lines 159–161). The parent of sourceVertex is set to −1 (line 166). T is a list that stores the vertices added into the shortest-path tree (line 169). We use a list for T rather than a set in order to record the order of the vertices added to T.

Initially, T is empty. To expand T, the method performs the following operations:

1. Find the vertex u with the smallest cost[u] (lines 175–181).

2. If u is found, add it to T (line 183). Note that if u is not found (u == −1), the graph is not connected. The break statement exits the while loop in this case.

3. After adding u in T, update cost[v] and parent[v] for each v adjacent to u in V−T if cost[v] > cost[u] + w(u, v) (lines 186–192).

Once all vertices from s are added to T, an instance of ShortestPathTree is created (line 196). The ShortestPathTree class extends the SearchTree class (line 200). To create an

instance of ShortestPathTree, pass sourceVertex, parent, T, and cost (lines 204–205). sourceVertex becomes the root in the tree. The data fields root, parent, and searchOrder are defined in the SearchTree class, which is an inner class defined in UnweightedGraph.

Note testing whether a vertex i is in T by invoking T.conatins(i) takes O(n) time, since T is a list. Therefore, the overall time complexity for this implementation is O(n3). Interested readers may see Programming Exercise 29.20 for improving the implementation and reducing the complexity to O(n2).

Dijkstra’s algorithm is a combination of a greedy algorithm and dynamic programming. It is a greedy algorithm in the sense that it always adds a new vertex that has the shortest distance to the source. It stores the shortest distance of each known vertex to the source, and uses it later to avoid redundant computing, so Dijkstra’s algorithm also uses dynamic programming.

ShortestPathTree class

Dijkstra’s algorithm time complexity

greedy and dynamic programming

0 1 2 3 4 5 6

parent

2 1–1

(b)(a)

0 5

72 5

6 4

0 1 2 3 4 5 6

cost

56 10 15 10 8

01 05

UnweightedGraph<V>.SearchTree

WeightedGraph<V>.ShortestPathTree

–cost:double[]

+ShortestPathTree(source: int, parent: int[], searchOrder: List<Integer>, cost: double[])

+getCost(v: int): double +printAllPaths(): void

cost[v] stores the cost for the path from the source to v.

Constructs a shortest–path tree with the speci�ed source, parent array, searchOrder, and cost array.

Returns the cost for the path from the source to vertex v. Displays all paths from the source.

M29_LIAN1878_11_GE_C29.indd 1130 12/29/17 4:51 PM

29.5 Finding Shortest Paths 1131

Listing 29.8 gives a test program that displays the shortest paths from Chicago to all other cities in Figure 29.1, and the shortest paths from vertex 3 to all vertices for the graph in Figure 29.3a, respectively.

Listing 29.8 TestShortestPath.java 1 public class TestShortestPath { 2 public static void main(String[] args) { 3 String[] vertices = {"Seattle", "San Francisco", "Los Angeles", 4 "Denver", "Kansas City", "Chicago", "Boston", "New York", 5 "Atlanta", "Miami", "Dallas", "Houston"}; 6 7 int[][] edges = { 8 {0, 1, 807}, {0, 3, 1331}, {0, 5, 2097}, 9 {1, 0, 807}, {1, 2, 381}, {1, 3, 1267}, 10 {2, 1, 381}, {2, 3, 1015}, {2, 4, 1663}, {2, 10, 1435}, 11 {3, 0, 1331}, {3, 1, 1267}, {3, 2, 1015}, {3, 4, 599}, 12 {3, 5, 1003}, 13 {4, 2, 1663}, {4, 3, 599}, {4, 5, 533}, {4, 7, 1260}, 14 {4, 8, 864}, {4, 10, 496}, 15 {5, 0, 2097}, {5, 3, 1003}, {5, 4, 533}, 16 {5, 6, 983}, {5, 7, 787}, 17 {6, 5, 983}, {6, 7, 214}, 18 {7, 4, 1260}, {7, 5, 787}, {7, 6, 214}, {7, 8, 888}, 19 {8, 4, 864}, {8, 7, 888}, {8, 9, 661}, 20 {8, 10, 781}, {8, 11, 810}, 21 {9, 8, 661}, {9, 11, 1187}, 22 {10, 2, 1435}, {10, 4, 496}, {10, 8, 781}, {10, 11, 239}, 23 {11, 8, 810}, {11, 9, 1187}, {11, 10, 239} 24 }; 25 26 WeightedGraph<String> graph1 = 27 new WeightedGraph<>(vertices, edges); 28 WeightedGraph<String>.ShortestPathTree tree1 = 29 graph1.getShortestPath(graph1.getIndex("Chicago")); 30 tree1.printAllPaths(); 31 32 // Display shortest paths from Houston to Chicago 33 System.out.print("Shortest path from Houston to Chicago: "); 34 java.util.List<String> path 35 = tree1.getPath(graph1.getIndex("Houston")); 36 for (String s: path) { 37 System.out.print(s + " "); 38 } 39 40 edges = new int[][] { 41 {0, 1, 2}, {0, 3, 8}, 42 {1, 0, 2}, {1, 2, 7}, {1, 3, 3}, 43 {2, 1, 7}, {2, 3, 4}, {2, 4, 5}, 44 {3, 0, 8}, {3, 1, 3}, {3, 2, 4}, {3, 4, 6}, 45 {4, 2, 5}, {4, 3, 6} 46 }; 47 WeightedGraph<Integer> graph2 = new WeightedGraph<>(edges, 5); 48 WeightedGraph<Integer>.ShortestPathTree tree2 = 49 graph2.getShortestPath(3); 50 System.out.println("\n"); 51 tree2.printAllPaths(); 52 } 53 }

vertices

edges

create graph1

shortest path

create edges

create graph2

print paths

M29_LIAN1878_11_GE_C29.indd 1131 12/29/17 4:51 PM

1132 Chapter 29 Weighted Graphs and Applications

All shortest paths from Chicago are: A path from Chicago to Seattle: Chicago Seattle (cost: 2097.0) A path from Chicago to San Francisco: Chicago Denver San Francisco (cost: 2270.0) A path from Chicago to Los Angeles: Chicago Denver Los Angeles (cost: 2018.0) A path from Chicago to Denver: Chicago Denver (cost: 1003.0) A path from Chicago to Kansas City: Chicago Kansas City (cost: 533.0) A path from Chicago to Chicago: Chicago (cost: 0.0) A path from Chicago to Boston: Chicago Boston (cost: 983.0) A path from Chicago to New York: Chicago New York (cost: 787.0) A path from Chicago to Atlanta: Chicago Kansas City Atlanta (cost: 1397.0) A path from Chicago to Miami: Chicago Kansas City Atlanta Miami (cost: 2058.0) A path from Chicago to Dallas: Chicago Kansas City Dallas (cost: 1029.0) A path from Chicago to Houston: Chicago Kansas City Dallas Houston (cost: 1268.0) Shortest path from Houston to Chicago: Houston Dallas Kansas City Chicago

All shortest paths from 3 are: A path from 3 to 0: 3 1 0 (cost: 5.0) A path from 3 to 1: 3 1 (cost: 3.0) A path from 3 to 2: 3 2 (cost: 4.0) A path from 3 to 3: 3 (cost: 0.0) A path from 3 to 4: 3 4 (cost: 6.0)

The program creates a weighted graph for Figure 29.1 in line 27. It then invokes the getShortestPath(graph1.getIndex("Chicago")) method to return a Path object that contains all shortest paths from Chicago. Invoking printAllPaths() on the ShortestPathTree object displays all the paths (line 30).

The graphical illustration of all shortest paths from Chicago is shown in Figure 29.20. The shortest paths from Chicago to the cities are found in this order: Kansas City, New York, Boston, Denver, Dallas, Houston, Atlanta, Los Angeles, Miami, Seattle, and San Francisco.

29.5.1 Trace Dijkstra’s algorithm for finding shortest paths from Boston to all other cities in Figure 29.1.

29.5.2 Is a shortest path between two vertices unique if all edges have different weights? 29.5.3 If you use an adjacency matrix to represent weighted edges, what would be the time

complexity for Dijkstra’s algorithm?

29.5.4 What happens to the getShortestPath() method in WeightedGraph if the source vertex cannot reach all vertices in the graph? Verify your answer by writing a test program that creates an unconnected graph and invoke the getShortestPath() method. How do you fix the problem by obtaining a partial shortest-path tree?

Point Check

M29_LIAN1878_11_GE_C29.indd 1132 12/29/17 4:51 PM

29.5 Finding Shortest Paths 1133

29.5.5 If there is no path from vertex v to the source vertex, what will be cost[v]?

29.5.6 Assume the graph is connected; will the getShortestPath method find the short- est paths correctly if lines 159–161 in WeightedGraph are deleted?

29.5.7 Show the output of the following code:

Figure 29.20 The shortest paths from Chicago to all other cities are highlighted.

Seattle

San Francisco

Los Angeles

Denver

Chicago

Kansas City

Houston

Boston

New York

Atlanta

Miami

661

888

1187

810 Dallas

1331

2097

1003 807

381

1015

1267

1663

1435

239

496

781

864

1260

983

787

214

533

599

M29_LIAN1878_11_GE_C29.indd 1133 12/29/17 4:51 PM

1134 Chapter 29 Weighted Graphs and Applications

graph.printWeightedEdges(); tree.printTree(); } }

29.6 Case Study: The Weighted Nine Tails Problem The weighted nine tails problem can be reduced to the weighted shortest path problem.

Section 28.10 presented the nine tails problem and solved it using the BFS algorithm. This section presents a variation of the nine tails problem and solves it using the shortest-path algorithm.

The nine tails problem is to find the minimum number of the moves that lead to all coins facing down. Each move flips a head coin and its neighbors. The weighted nine tails problem assigns the number of flips as a weight on each move. For example, you can move from the coins in Figure 29.21a to those in Figure 29.21b by flipping the first coin in the first row and its two neighbors. Thus, the weight for this move is 3. You can move from the coins in Figure 29.21c to Figure 29.21d by flipping the five coins. So the weight for this move is 5.

Point Key

Figure 29.21 The weight for each move is the number of flips for the move.

T T T

H H

H H H

(a)

H T T

T H

H H H

(b)

H H T

T H

H H H

(c)

T T H

H H

H T H

(d)

The weighted nine tails problem can be reduced to finding a shortest path from a starting node to the target node in an edge-weighted graph. The graph has 512 nodes. Create an edge from node v to u if there is a move from node u to node v. Assign the number of flips to be the weight of the edge.

Recall in Section 28.10, we defined a class NineTailModel for modeling the nine tails problem. We now define a new class named WeightedNineTailModel that extends NineTailModel, as shown in Figure 29.22.

The NineTailModel class creates a Graph and obtains a Tree rooted at the target node 511. WeightedNineTailModel is the same as NineTailModel except that it creates a WeightedGraph and obtains a ShortestPathTree rooted at the target node 511. The method getEdges() finds all edges in the graph. The getNumberOfFlips(int u, int v) method returns the number of flips from node u to node v. The getNumberOfFlips(int u) method returns the number of flips from node u to the target node.

Listing 29.9 implements the WeightedNineTailModel.

M29_LIAN1878_11_GE_C29.indd 1134 12/29/17 4:51 PM

29.6 Case Study: The Weighted Nine Tails Problem 1135

Listing 29.9 WeightedNineTailModel.java 1 import java.util.*; 2 3 public class WeightedNineTailModel extends NineTailModel { 4 /** Construct a model */ 5 public WeightedNineTailModel() { 6 // Create edges 7 List<WeightedEdge> edges = getEdges(); 8 9 // Create a graph 10 WeightedGraph<Integer> graph = new WeightedGraph<Integer>( 11 edges, NUMBER_OF_NODES); 12 13 // Obtain a shortest-path tree rooted at the target node 14 tree = graph.getShortestPath(511); 15 } 16 17 /** Create all edges for the graph */ 18 private List<WeightedEdge> getEdges() { 19 // Store edges 20 List<WeightedEdge> edges = new ArrayList<>();

extends NineTailModel

constructor

get edges

create a graph

get a tree

get weighted edges

Figure 29.22 The WeightedNineTailModel class extends NineTailModel.

WeightedNineTailModel

+WeightedNineTailModel() Constructs a model for the weighted nine tails problem and obtains a ShortestPathTree rooted from the target node.

NineTailModel

#tree: UnweightedGraph<Integer>.SeachTree A tree rooted at node 511.

+NineTailModel() Constructs a model for the nine tails problem and obtains the tree.

+getShortestPath(nodeIndex: int): List<Integer>

Returns a path from the speci�ed node to the root. The path returned consists of the node labels in a list. Returns a list of Edge objects for the graph.

Flips the node at the speci�ed position and returns the index of the �ipped node.

Returns a node consisting of nine characters of Hs and Ts. Returns the index of the speci�ed node.

Flips the node at the speci�ed row and column.

Displays the node to the console.

+getNumberOfFlips(u: int): int Returns the number of �ips from node u to the target node 511.

–getNumberOfFlips(u: int, v: int): int Returns the number of different cells between the two nodes.

–getEdges(): List<WeightedEdge> Gets the weighted edges for the weighted nine tail problem.

–getEdges(): List<AbstractGraph.Edge>

+getNode(index: int): char[]

+getIndex(node: char[]): int +getFlippedNode(node: char[], position: int): int

+f lipACell(node: char[], row: int, column: int): void

+printNode(node: char[]): void

M29_LIAN1878_11_GE_C29.indd 1135 12/29/17 4:51 PM

1136 Chapter 29 Weighted Graphs and Applications

21 22 for (int u = 0; u < NUMBER_OF_NODES; u++) { 23 for (int k = 0; k < 9; k++) { 24 char[] node = getNode(u); // Get the node for vertex u 25 if (node[k] == 'H') { 26 int v = getFlippedNode(node, k); 27 int numberOfFlips = getNumberOfFlips(u, v); 28 29 // Add edge (v, u) for a legal move from node u to node v 30 edges.add(new WeightedEdge(v, u, numberOfFlips)); 31 } 32 } 33 } 34 35 return edges; 36 } 37 38 private static int getNumberOfFlips(int u, int v) { 39 char[] node1 = getNode(u); 40 char[] node2 = getNode(v); 41 42 int count = 0; // Count the number of different cells 43 for (int i = 0; i < node1.length; i++) 44 if (node1[i] != node2[i]) count++; 45 46 return count; 47 } 48 49 public int getNumberOfFlips(int u) { 50 return (int)((WeightedGraph<Integer>.ShortestPathTree)tree) 51 .getCost(u); 52 } 53 }

WeightedNineTailModel extends NineTailModel to build a WeightedGraph to model the weighted nine tails problem (lines 10–11). For each node u, the getEdges() method finds a flipped node v and assigns the number of flips as the weight for edge (v, u ) (line 30). The getNumberOfFlips(int u, int v) method returns the number of flips from node u to node v (lines 38–47). The number of flips is the number of the different cells between the two nodes (line 44).

The WeightedNineTailModel obtains a ShortestPathTree rooted at the target node 511 (line 14). Note tree is a protected data field defined in NineTailModel and ShortestPathTree is a subclass of Tree. The methods defined in NineTailModel use the tree property.

The getNumberOfFlips(int u) method (lines 49–52) returns the number of flips from node u to the target node, which is the cost of the path from node u to the target node. This cost can be obtained by invoking the getCost(u) method defined in the ShortestPathTree class (line 51).

Listing 29.10 gives a program that prompts the user to enter an initial node and displays the minimum number of flips to reach the target node.

Listing 29.10 WeightedNineTail.java 1 import java.util.Scanner; 2 3 public class WeightedNineTail { 4 public static void main(String[] args) { 5 // Prompt the user to enter the nine coins' Hs and Ts

get adjacent node weight

add an edge

number of flips

total number of flips

M29_LIAN1878_11_GE_C29.indd 1136 12/29/17 4:51 PM

29.6 Case Study: The Weighted Nine Tails Problem 1137

6 System.out.print("Enter an initial nine coins' Hs and Ts: "); 7 Scanner input = new Scanner(System.in); 8 String s = input.nextLine(); 9 char[] initialNode = s.toCharArray(); 10 11 WeightedNineTailModel model = new WeightedNineTailModel(); 12 java.util.List<Integer> path = 13 model.getShortestPath(NineTailModel.getIndex(initialNode)); 14 15 System.out.println("The steps to flip the coins are "); 16 for (int i = 0; i < path.size(); i++) 17 NineTailModel.printNode(NineTailModel.getNode(path.get(i))); 18 19 System.out.println("The number of flips is " + 20 model.getNumberOfFlips(NineTailModel.getIndex(initialNode))); 21 } 22 }

initial node

create model

get shortest path

print node

number of flips

Enter an initial nine coins Hs and Ts: HHHTTTHHH The steps to flip the coins are HHH TTT HHH

HHH THT TTT

TTT TTT TTT

The number of flips is 8

The program prompts the user to enter an initial node with nine letters with a combination of Hs and Ts as a string in line 8, obtains an array of characters from the string (line 9), creates a model (line 11), obtains the shortest path from the initial node to the target node (lines 12–13), displays the nodes in the path (lines 16–17), and invokes getNumberOfFlips to get the number of flips needed to reach the target node (line 20).

29.6.1 Why is the tree data field in NineTailModel in Listing 28.13 defined protected?

29.6.2 How are the nodes created for the graph in WeightedNineTailModel? 29.6.3 How are the edges created for the graph in WeightedNineTailModel?

Point Check

Key Terms Dijkstra’s algorithm 1125 edge-weighted graph 1109 minimum spanning tree 1119 Prim’s algorithm 1119

shortest path 1126 single-source shortest path 1126 vertex-weighted graph 1109

M29_LIAN1878_11_GE_C29.indd 1137 12/29/17 4:51 PM

1138 Chapter 29 Weighted Graphs and Applications

ChapTer summary 1. You can use adjacency matrices or lists to store weighted edges in graphs.

2. A spanning tree of a graph is a subgraph that is a tree and connects all vertices in the graph.

3. Prim’s algorithm for finding a minimum spanning tree works as follows: the algorithm starts with a spanning tree T that contains an arbitrary vertex. The algorithm expands the tree by adding a vertex with the minimum-weight edge incident to a vertex already in the tree.

4. Dijkstra’s algorithm starts the search from the source vertex and keeps finding vertices that have the shortest path to the source until all vertices are found.

Quiz Answer the quiz for this chapter online at the book Companion Website.

programming exerCises *29.1 (Kruskal’s algorithm) The text introduced Prim’s algorithm for finding a mini-

mum spanning tree. Kruskal’s algorithm is another well-known algorithm for finding a minimum spanning tree. The algorithm repeatedly finds a minimum- weight edge and adds it to the tree if it does not cause a cycle. The process ends when all vertices are in the tree. Design and implement an algorithm for finding an MST using Kruskal’s algorithm.

*29.2 (Implement Prim’s algorithm using an adjacency matrix) The text implements Prim’s algorithm using lists for adjacent edges. Implement the algorithm using an adjacency matrix for weighted graphs.

*29.3 (Implement Dijkstra’s algorithm using an adjacency matrix) The text implements Dijkstra’s algorithm using lists for adjacent edges. Implement the algorithm using an adjacency matrix for weighted graphs.

*29.4 (Modify weight in the nine tails problem) In the text, we assign the number of the flips as the weight for each move. Assuming the weight is three times of the number of flips, revise the program.

*29.5 (Prove or disprove) The conjecture is that both NineTailModel and WeightedNineTailModel result in the same shortest path. Write a program to prove or disprove it. (Hint: Let tree1 and tree2 denote the trees rooted at node 511 obtained from NineTailModel and WeightedNineTailModel, respectively. If the depth of a node u is the same in tree1 and in tree2, the length of the path from u to the target is the same.)

**29.6 (Weighted 4 * 4 16 tails model) The weighted nine tails problem in the text uses a 3 * 3 matrix. Assume that you have 16 coins placed in a 4 * 4 matrix. Create a new model class named WeightedTailModel16. Create an instance of the model and save the object into a file named WeightedTailModel16.dat.

**29.7 (Weighted 4 * 4 16 tails) Revise Listing 29.9, WeightedNineTail.java, for the weighted 4 * 4 16 tails problem. Your program should read the model object created from the preceding exercise.

M29_LIAN1878_11_GE_C29.indd 1138 12/29/17 4:51 PM

**29.8 (Traveling salesperson problem) The traveling salesperson problem (TSP) is to find the shortest round-trip route that visits each city exactly once and then returns to the starting city. The problem is similar to finding a shortest Hamiltonian cycle in Programming Exercise 28.17. Add the following method in the WeightedGraph class:

// Return a shortest cycle // Return null if no such cycle exists public List<Integer> getShortestHamiltonianCycle()

*29.9 (Find a minimum spanning tree) Write a program that reads a connected graph from a file and displays its minimum spanning tree. The first line in the file contains a number that indicates the number of vertices (n). The vertices are labeled as 0, 1, ... , n−1. Each subsequent line describes the edges in the form of u1, v1, w1 | u2, v2, w2 | .... Each triplet in this form describes an edge and its weight. Figure 29.23 shows an example of the file for the cor- responding graph. Note that we assume the graph is undirected. If the graph has an edge (u, v), it also has an edge (v, u). Only one edge is represented in the file. When you construct a graph, both edges need to be added.

Figure 29.23 The vertices and edges of a weighted graph can be stored in a file.

(a) (b)

0 100

3 20

2 5

File 6 0, 1, 100 | 0, 2, 3 1, 3, 20 2, 3, 40 | 2, 4, 2 | 2, 5, 5 3, 5, 5 4, 5, 9

Programming Exercises 1139

Only one edge is represented in the file. When you construct a graph, both edges need to be added. Your program should prompt the user to enter the name of the file, read data from the file, create an instance g of WeightedGraph, invoke g. printWeightedEdges() to display all edges, invoke getMinimumSpanningTree() to obtain an instance tree of WeightedGraph. MST, invoke tree.getTotalWeight() to display the weight of the minimum spanning tree, and invoke tree. printTree() to display the tree. Here is a sample run of the program:

Enter a file name: c:\exercise\WeightedGraphSample2.txt The number of vertices is 6 Vertex 0: (0, 1, 100) (0, 2, 3) Vertex 1: (1, 0, 100) (1, 3, 20) Vertex 2: (2, 0, 3)(2, 3, 40) (2, 4, 2) (2, 5, 5) Vertex 3: (3, 1, 20) (3, 2, 40) (3, 5, 5) Vertex 4: (4, 2, 2) (4, 5, 9) Vertex 5: (5, 2, 5) (5, 3, 5) (5, 4, 9) Total weight in MST is 35 Root is: 0 Edges: (3, 1) (0, 2) (5, 3)(2, 4) (2, 5)

M29_LIAN1878_11_GE_C29.indd 1139 12/29/17 4:51 PM

1140 Chapter 29 Weighted Graphs and Applications

(Hint: Use new WeightedGraph(list, numberOfVertices) to cre- ate a graph, where list contains a list of WeightedEdge objects. Use new WeightedEdge(u, v, w) to create an edge. Read the first line to get the number of vertices. Read each subsequent line into a string s and use s.split("[\\|]") to extract the triplets. For each triplet, use triplet .split("[,]") to extract vertices and weight.)

*29.10 (Create a file for a graph) Modify Listing 29.3, TestWeightedGraph.java, to create a file for representing graph1. The file format is described in Program- ming Exercise 29.9. Create the file from the array defined in lines 7–24 in Listing 29.3. The number of vertices for the graph is 12, which will be stored in the first line of the file. An edge (u, v) is stored if u < v. The contents of the file should be as follows:

0, 1, 807 | 0, 3, 1331 | 0, 5, 2097 1, 2, 381 | 1, 3, 1267 2, 3, 1015 | 2, 4, 1663 | 2, 10, 1435 3, 4, 599 | 3, 5, 1003 4, 5, 533 | 4, 7, 1260 | 4, 8, 864 | 4, 10, 496 5, 6, 983 | 5, 7, 787 6, 7, 214 7, 8, 888 8, 9, 661 | 8, 10, 781 | 8, 11, 810 9, 11, 1187 10, 11, 239

Enter a file name: WeightedGraphSample2.txt

Enter two vertices (integer indexes): 0 1 The number of vertices is 6 Vertex 0: (0, 1, 100.0) (0, 2, 3.0) Vertex 1: (1, 0, 100.0) (1, 3, 20.0) Vertex 2: (2, 0, 3.0) (2, 3, 40.0) (2, 4, 2.0) (2, 5, 5.0) Vertex 3: (3, 1, 20.0) (3, 2, 40.0) (3, 5, 5.0) Vertex 4: (4, 2, 2.0) (4, 5, 9.0) Vertex 5: (5, 2, 5.0) (5, 3, 5.0) (5, 4, 9.0) A path from 0 to 1: 0 2 5 3 1

*29.11 (Find shortest paths) Write a program that reads a connected graph from a file. The graph is stored in a file using the same format specified in Program- ming Exercise 29.9. Your program should prompt the user to enter the name of the file then two vertices, and should display a shortest path between the two vertices. For example, for the graph in Figure 29.23, a shortest path between 0 and 1 can be displayed as 0 2 5 3 1. Here is a sample run of the program:

M29_LIAN1878_11_GE_C29.indd 1140 12/29/17 4:51 PM

*29.12 (Display weighted graphs) Revise GraphView in Listing 28.6 to display a weighted graph. Write a program that displays the graph in Figure 29.1 as shown in Figure 29.24. (Instructors may ask students to expand this program by adding new cities with appropriate edges into the graph).

Figure 29.24 Programming Exercise 29.12 displays a weighted graph.

Programming Exercises 1141

*29.13 (Display shortest paths) Revise GraphView in Listing 28.6 to display a weighted graph and a shortest path between the two specified cities, as shown in Figure 29.25. You need to add a data field path in GraphView. If a path is not null, the edges in the path are displayed in red. If a city not in the map is entered, the program displays a text to alert the user.

Figure 29.25 Programming Exercise 29.13 displays a shortest path.

M29_LIAN1878_11_GE_C29.indd 1141 12/29/17 4:51 PM

1142 Chapter 29 Weighted Graphs and Applications

*29.14 (Display a minimum spanning tree) Revise GraphView in Listing 28.6 to display a weighted graph and a minimum spanning tree for the graph in Figure 29.1, as shown in Figure 29.26. The edges in the MST are shown in red.

Figure 29.26 Programming Exercise 29.14 displays an MST.

Figure 29.27 The program can add vertices and edges and display a shortest path between two specified vertices. Source: Copyright © 1995–2016 Oracle and/or its affiliates. All rights reserved. Used with permission.

***29.15 (Dynamic graphs) Write a program that lets the users create a weighted graph dynamically. The user can create a vertex by entering its name and location, as shown in Figure 29.27. The user can also create an edge to connect two vertices. To simplify the program, assume vertex names are the same as vertex indices. You have to add the vertex indices 0, 1, . . . , and n, in this order. The user can specify two vertices and let the program display their shortest path in red.

***29.16 (Display a dynamic MST) Write a program that lets the user create a weighted graph dynamically. The user can create a vertex by entering its name and loca- tion, as shown in Figure 29.28. The user can also create an edge to connect two

M29_LIAN1878_11_GE_C29.indd 1142 12/29/17 4:51 PM

vertices. To simplify the program, assume vertex names are the same as those of vertex indices. You have to add the vertex indices 0, 1, . . . , and n, in this order. The edges in the MST are displayed in red. As new edges are added, the MST is redisplayed.

Programming Exercises 1143

***29.17 (Weighted graph visualization tool) Develop a GUI program as shown in Figure 29.2, with the following requirements: (1) The radius of each vertex is 20 pixels. (2) The user clicks the left mouse button to place a vertex centered at the mouse point, provided the mouse point is not inside or too close to an existing vertex. (3) The user clicks the right mouse button inside an existing vertex to remove the vertex. (4) The user presses a mouse button inside a vertex and drags to another vertex then releases the button to create an edge, and the distance between the two vertices is also displayed. (5) The user drags a vertex while pressing the CTRL key to move a vertex. (6) The vertices are numbers starting from 0. When a vertex is removed, the vertices are renumbered. (7) You can click the Show MST or Show All SP From the Source button to display an MST or SP tree from a starting vertex. (8) You can click the Show Shortest Path button to display the shortest path between the two specified vertices.

***29.18 (Alternative version of Dijkstra algorithm) An alternative version of the Dijkstra algorithm can be described as follows:

Input: a weighted graph G = (V, E) with nonnegative weights Output: A shortest-path tree from a source vertex s

1 ShortestPathTree getShortestPath(s) { 2 Let T be a set that contains the vertices whose 3 paths to s are known; 4 Initially T contains source vertex s with cost[s] = 0; 5 for (each u in V – T) 6 cost[u] = infinity; 7 8 while (size of T < n) { 9 Find v in V – T with the smallest cost[u] + w(u, v) value 10 among all u in T; 11 Add v to T and set cost[v] = cost[u] + w(u, v); 12 parent[v] = u; 13 } 14 }

add a vertex

find next vertex more vertex

add initial vertex

M29_LIAN1878_11_GE_C29.indd 1143 12/29/17 4:51 PM

1144 Chapter 29 Weighted Graphs and Applications

The algorithm uses cost[v] to store the cost of a shortest path from vertex v to the source vertex s. cost[s] is 0. Initially assign infinity to cost[v] to indicate that no path is found from v to s. Let V denote all vertices in the graph and T denote the set of the vertices whose costs are known. Initially, the source vertex s is in T. The algorithm repeatedly finds a vertex u in T and a vertex v in V–T such that cost[u] + w(u, v) is the smallest, and moves v to T. The shortest-path algorithm given in the text continuously updates the cost and par- ent for a vertex in V–T. This algorithm initializes the cost to infinity for each vertex and then changes the cost for a vertex only once when the vertex is added into T. Implement this algorithm and use Listing 29.7, TestShortestPath.java, to test your new algorithm.

***29.19 (Find u with smallest cost[u] efficiently) The getShortestPath method finds a u with the smallest cost[u] using a linear search, which takes O( � V � ). The search time can be reduced to O(log � V � ) using an AVL tree. Modify the method using an AVL tree to store the vertices in V–T. Use Listing 29.7, Test- ShortestPath.java, to test your new implementation.

***29.20 (Test if a vertex u is in T efficiently) Since T is implemented using a list in the getMinimumSpanningTree and getShortestPath methods in Listing 29.2 WeightedGraph.java, testing whether a vertex u is in T by invoking T. contains(u) takes O(n) time. Modify these two methods by introducing an array named isInT. Set isInT[u] to true when a vertex u is added to T. Testing whether a vertex u is in T can now be done in O(1) time. Write a test program using the following code, where graph1 is created from Figure 29.1:

WeightedGraph<String> graph1 = new WeightedGraph<>(edges, vertices);

WeightedGraph<String>.MST tree1 = graph1.getMinimumSpanningTree();

System.out.println("Total weight is " + tree1.getTotalWeight());

tree1.printTree();

WeightedGraph<String>.ShortestPathTree tree2 = graph1.getShortestPath(graph1.getIndex("Chicago"));

tree2.printAllPaths();

M29_LIAN1878_11_GE_C29.indd 1144 12/29/17 4:51 PM

Objectives ■■ To use aggregate operations on collection streams to simplify coding

and improve performance (§30.1).

■■ To create a stream pipeline, apply lazy intermediate methods (skip, limit, filter, distinct, sorted, map, and mapToInt), and terminal methods (count, sum, average, max, min, forEach, findFirst, firstAny, anyMatch, allMatch, noneMatch, and toArray) on a stream (§30.2).

■■ To process primitive data values using the IntStream, LongStream, and DoubleStream (§30.3).

■■ To create parallel streams for fast execution (§30.4).

■■ To reduce the elements in a stream into a single result using the reduce method (§30.5).

■■ To place the elements in a stream into a mutable collection using the collect method (§30.6).

■■ To group the elements in a stream and apply aggregate methods for the elements in the groups (§30.7).

■■ To use a variety of examples to demonstrate how to simplify coding using streams (§30.8).

Aggregate Operations for Collection Streams

Chapter

M30_LIAN1878_11_GE_C30.indd 1145 12/29/17 4:53 PM

1146 Chapter 30 Aggregate Operations for Collection Streams

30.1 Introduction Using aggregate operations on collection streams can greatly simplify coding and improve performance.

Often, you need to process data in an array or a collection. Suppose, for instance, that you need to count the number of elements in a set that is greater than 60. You may write the code using a foreach loop as follows:

Double[] numbers = {2.4, 55.6, 90.12, 26.6}; Set<Double> set = new HashSet<>(Arrays.asList(numbers)); int count = 0; for (double e: set) if (e > 60) count++; System.out.println("Count is " + count);

The code is fine. However, Java provides a better and simpler way for accomplishing the task. Using the aggregate operations, you can rewrite the code as follows:

System.out.println("Count is " + set.stream().filter(e −> e > 60).count());

Invoking the stream() method on a set returns a Stream for the elements from a set. The filter method specifies a condition for selecting the elements whose value is greater than 60. The count() method returns the number of elements in the stream that satisfy the condition.

A collection stream or simply stream is a sequence of elements. The operations on a stream is called aggregate operations (also known as stream operations) because they apply to all the data in the stream. The filter and count are the examples of aggregate operations. The code written using a foreach loop describes the process how to obtain the count, i.e., for each element, if it is greater than 60, increase the count. The code written using the aggregate operations tells the program to return the count for the elements greater than 60, but it does not specify how the count is obtained. Clearly, using the aggregate operations leaves the detailed implementation to the computer, therefore, makes the code concise and simpler. Moreover, the aggregate operations on a stream can be executed in parallel to take advantage of multiple processors. So, the code written using aggregate operations usually run faster than the ones using a foreach loop.

Java provides many aggregate operations and many different ways of using aggregate opera- tions. This chapter gives a comprehensive coverage on aggregate operations and streams.

30.1.1 What are the benefits of using aggregate operations on collection streams for processing data?

30.2 Stream Pipelines A stream pipeline consists of a stream created from a data source, zero or more inter- mediate methods, and a final terminal method.

An array or a collection is an object for storing data. A stream is a transient object for pro- cessing data. After data is processed, the stream is destroyed. Java 8 introduced a new default stream() method in the Collection interface to return a Stream object. The Stream interface extends the BaseStream interface and contains the aggregate methods and the utility methods as shown in Figure 30.1.

The methods in the Stream interface are divided into three groups: intermediate methods, terminal methods, and static methods. An intermediate method transforms the stream into another stream. A terminal method returns a result or performs actions. After a terminal method is executed, the stream is closed automatically. A static method creates a stream.

Point Key

stream aggregate operations

why using aggregate operations?

Point Check

Point Key

intermediate method terminal method

static method

M30_LIAN1878_11_GE_C30.indd 1146 12/29/17 4:53 PM

30.2 Stream Pipelines 1147

Figure 30.1 The Stream class defines the aggregate operations for the elements in a stream.

«interface» java.util.stream.Stream<T>

+distinct(): Stream<T>

+f ilter(p: Predicate<? super T): Stream<T>

+limit(n: long): Stream<T>

+skip(n: long): Stream<T>

+sorted(): Stream<T>

+sorted(comparator: Comparator<? super T>): Stream<T>

+map(mapper: Function<? super T, ? extends R>: Stream<R>

+mapToInt(mapper: ToIntFunction<? super T>): IntStream

+mapToLong(mapper: ToLongFunction<? super T>): LongStream

+mapToDouble(mapper: ToDoubleFunction<? super T>): DoubleStream

+count(): long

+max(c: Comparator<? super T>): Optional<T>

+min(c: Comparator<? super T>): Optional<T>

+f indFirst(): Optional<T>

+f indAny(): Optional<T>

+allMatch(p: Predicate<? super T): boolean

+anyMatch(p: Predicate<? super T): boolean

+noneMatch(p: Predicate<? super T): boolean

+forEach(action: Consumer<? super T>): void

+reduce(accumulator: BinaryOperator<T>): Optional<T>

+reduce(identity: T, accumulator: BinaryOperator<T>): T

+collect(collector: <? super <T, A, R>>): R

+toArray(): Object[]

+empty(): Stream<T>

+of(values: T...): Stream<T>

+of(values: T): Stream<T>

+concat(a1: Stream<? extends T>, a2: Stream<? extends T>): Stream<T>

Returns a stream consisting of distinct elements from this stream.

Returns a stream consisting of the elements matching the predicate.

Returns a stream consisting of the �rst n elements from this stream.

Returns a stream consisting of the remaining elements in this stream after discarding the �rst n elements.

Returns a stream consisting of the elements of this stream sorted in a natural order.

Returns a stream consisting of the elements of this stream sorted using the comparator.

Returns a stream consisting of the results of applying the function to the elements of this stream.

Returns an IntStream consisting of the results of applying the function to the elements of this stream.

Returns a LongStream consisting of the results of applying the function to the elements of this stream.

Returns a DoubleStream consisting of the results of applying the function to the elements of this stream.

Returns the number of elements in this stream.

Returns the maximum element in this stream based on the comparator.

Returns the minimum element in this stream based on the comparator.

Returns the �rst element from this stream.

Returns any element from this stream.

Returns true if all the elements in this stream match the predicate.

Returns true if one element in this stream matches the predicate.

Returns true if no element in this stream matches the predicate.

Performs an action for each element of this stream.

Reduces the elements in the stream to a value using the identity and an associative accumulation function. Return an Optional describing the reduced value..

Reduces the elements in the stream to a value using the identity and an associative accumulation function. Return the reduced value.

Performs a mutable reduction operation on the elements of this stream using a Collector.

Returns an array consisting of the elements in this stream.

Returns an empty sequential stream. (static method)

Returns a stream consisting of the speci�ed values. (static method)

Returns a stream consisting of a single value. (static method)

Returns a lazily concatenated stream consisting of the elements in a1 followed by the elements in a2. (static method)

«interface» java.util.stream.BaseStream<T, S extendsBaseStream<T, S>>

+close(): S

+parallel(): S

+sequential(): S

+isParallel(): boolean

Closes this stream.

Returns an equivalent stream that is executed in parallel.

Returns an equivalent stream that is executed in sequential.

Returns true if this stream is parallel.

Intermediate operations

Terminal operations

Static methods

The methods are invoked using a stream pipeline. A stream pipeline consists of a source (e.g., a list, a set, or an array), a method that creates a stream, zero or more intermediate methods, and a final terminal method. The following is a stream pipeline example:

Source Create a stream

Zero or more Intermediate methods

One terminal method

set.stream().limit(50).distinct().count()

stream pipeline

M30_LIAN1878_11_GE_C30.indd 1147 12/29/17 4:53 PM

1148 Chapter 30 Aggregate Operations for Collection Streams

Here, set is the source of the data, invoking stream() creates a stream for the data from the source, invoking limit(50) returns the first 50 elements from the stream, invoking dis- tinct() obtains a stream of distinct elements from the stream of the 50 elements, and invok- ing count() returns the number of elements in the final stream.

Streams are lazy, which means that the computation is performed only when the terminal operation is initiated. This allows the JVM to optimize computation.

Most of the arguments for stream methods are instances of functional interfaces. So the arguments can be created using lambda expressions or method references. Listing 30.1 gives an example that demonstrates creating a stream and applying methods on the streams.

Listing 30.1 StreamDemo.java 1 import java.util.stream.Stream; 2 3 public class StreamDemo { 4 public static void main(String[] args) { 5 String[] names = {"John", "Peter", "Susan", "Kim", "Jen", 6 "George", "Alan", "Stacy", "Michelle", "john"}; 7 8 // Display the first four names sorted 9 Stream.of(names).limit(4).sorted() 10 .forEach(e −> System.out.print(e + " ")); 11 12 // Skip four names and dispaly the rest sorted ignore case 13 System.out.println(); 14 Stream.of(names).skip(4) 15 .sorted((e1, e2) −> e1.compareToIgnoreCase(e2)) 16 .forEach(e −> System.out.print(e + " ")); 17 18 System.out.println(); 19 Stream.of(names).skip(4) 20 .sorted(String::compareToIgnoreCase) 21 .forEach(e −> System.out.print(e + " ")); 22 23 System.out.println("\nLargest string with length > 4: " 24 + Stream.of(names) 25 .filter(e −> e.length() > 4) 26 .max(String::compareTo).get()); 27 28 System.out.println("Smallest string alphabetically: " 29 + Stream.of(names).min(String::compareTo).get()); 30 31 System.out.println("Stacy is in names? " 32 + Stream.of(names).anyMatch(e −> e.equals("Stacy"))); 33 34 System.out.println("All names start with a capital letter? " 35 + Stream.of(names) 36 .allMatch(e −> Character.isUpperCase(e.charAt(0)))); 37 38 System.out.println("No name begins with Ko? " 39 + Stream.of(names).noneMatch(e −> e.startsWith("Ko"))); 40 41 System.out.println("Number of distinct case-insensitive strings: " 42 + Stream.of(names).map(e −> e.toUpperCase()) 43 .distinct().count()); 44 45 System.out.println("First element in this stream in lowercase: " 46 + Stream.of(names).map(String::toLowerCase).findFirst().get()); 47

lazy evaluation

create an array

Stream.of forEach

skip sorted

method reference

filter method reference

min

anyMatch

allMatch

distinct

findFirst

M30_LIAN1878_11_GE_C30.indd 1148 12/29/17 4:53 PM

30.2 Stream Pipelines 1149

48 System.out.println("Skip 4 and get any element in this stream:" 49 + Stream.of(names).skip(4).sorted().findAny().get()); 50 51 Object[] namesInLowerCase = 52 Stream.of(names).map(String::toLowerCase).toArray(); 53 System.out.println(java.util.Arrays.toString(namesInLowerCase)); 54 } 55 }

John Kim Peter Susan Alan George Jen john Michelle Stacy Alan George Jen john Michelle Stacy Largest string with length > 4: Susan Smallest string alphabetically: Alan Stacy is in names? true All names start with a capital letter? false No name begins with Ko? true Number of distinct case-insensitive strings: 9 First element in this stream in lowercase: john Skip 4 and get any element in this stream: Alan [john, peter, susan, kim, jen, george, alan, stacy, michelle, john]

We now introduce the stream methods through this example.

30.2.1 The Stream.of, limit, and forEach Methods The program creates an array of strings (lines 5 –6). In lines 9–10, invoking the static Stream. of(names) returns a Stream consisting of strings from the names array, invoking limit(4) returns a new Stream consisting the first four elements in the stream, invoking sorted() sorts the stream, and invoking the forEach method displays each element in the stream. The argument passed to the forEach method is a lambda expression. As introduced in Section 15.6, a lambda expression is a concise syntax to replace an anonymous inner class that imple- ments a functional interface. The argument passed to the forEach method is an instance of the functional interface Consumer<? super T> with an abstract function accept(T t). The statement in line 10 using a lambda expression in (a) is equivalent to using an anonymous inner class in (b) as shown below. The lambda expression not only simplifies the code, but also the concept of the method. You can now simply say that for each element in the stream, perform the action as specified in the expression.

forEach(e −> System.out.print(e + " ")) forEach( new java.util.function.Consumer<String>() { public void accept(String e) { System.out.print(e + " "); } } )

(a) Using a lambda expression (b) Using an anonymous inner class

30.2.2 The sorted Method The sorted method in line 15 sorts the strings in the stream using a Comparator. The Comparator is a functional interface. A lambda expression is used to implement the interface and specifies that two strings are compared ignoring cases. This lambda expression in (a) is

forEach method

lambda expression

findAny

toArray display array

M30_LIAN1878_11_GE_C30.indd 1149 12/29/17 4:53 PM

1150 Chapter 30 Aggregate Operations for Collection Streams

equivalent to the code using an anonymous inner class in (b). The lambda expression simply invokes a method in this case. So, it can be further simplified using a method reference in line 20 (also see in (c)). The method reference was introduced in Section 20.6.method reference

sorted((e1, e2) −> e1.compareToIgnoreCase(e2))

sorted( new java.util.Comparator<String>() { public int compare(String e1, String e2) { return e1.compareToIgnoreCase(e2); } } )

(a) Using a lambda expression

sorted(String::compareToIgnoreCase)

30.2.3 The filter Method The filter method takes an argument of the Predicate<? super T> type, which is a functional interface with an abstract method test(T t) that returns a Boolean value. The method selects the elements from the stream that satisfies the predicate. Line 25 uses a lambda expression to implement the Predicate interface as shown in (a), which is equivalent to the code using an anonymous inner class as shown in (b).

filter(e −> e.length() > 4) filter( new java.util.function.Predicate<String>() { public boolean test(String e) { return e.length() > 4; } } )

(a) Using a lambda expression (b) Using an anonymous inner class

30.2.4 The max and min Methods The max and min methods take an argument of the Comparator<? Super T> type. This argument specifies how the elements are compared in order to obtain the maximum and mini- mum elements. The program uses the method reference String::compareTo to simplify the code for creating a Comparator (lines 26 and 29). The max and min methods return an Optional<T> that describes the element. You need to invoke the get() method from the Optional class to return the element.

30.2.5 The anyMatch, allMatch, and noneMatch Methods The anyMatch, allMatch, and noneMatch methods take an argument of the Predicate<? super T> type to test if the stream contains an element, all elements, or no element that satisfies the predicate. The program tests whether the name Stacy is in the stream (line 32), whether all the names in the stream start with a capital letter (line 36), and whether any names starts with string Ko (line 39).

30.2.6 The map, distinct, and count Methods The map method returns a new stream by mapping each element in the stream into a new element. So, the map method in line 42 returns a new stream with all uppercase strings. The distinct() method obtains a new stream with all distinct elements. The count() method counts the number of the elements in the stream. So, the stream pipeline in line 43 counts the number of distinct strings in the array names.

Optional <T> get method

M30_LIAN1878_11_GE_C30.indd 1150 12/29/17 4:53 PM

30.2 Stream Pipelines 1151

The map method takes an argument of the Function<? super T, ? super R> type to return an instance of the Stream<R>. The Function is a functional interface with an abstract method apply(T t) that maps t into a value of the type R. Line 42 uses a lambda expression to implement the Function interface as shown in (a), which is equivalent to the code using an anonymous inner class as shown in (b). You can further simplify it using a method refer- ence as shown in (c).

map(e −> e.toUpperCase()) map( new java.util.function.Function<String, String>() { public String apply(String e) { return e.toUpperCase(); } } )

(a) Using a lambda expression

map(String::toUpperCase)

30.2.7 The findFirst, findAny, and toArray Methods The findFirst() method (line 46) returns the first element in the stream wrapped in an instance of Optional<T>. The actual element value is then returned by invoking the get() method in the Optional<T> class. The findAny() method (line 49) returns any element in the stream. Which element is selected depends on the internal state of the stream. The findAny() method is more efficient than the findFirst() method.

The toArray() method (line 52) returns an array of objects from the stream.

Note The BaseStream interface defines the close() method, which can be invoked to close a stream. You don’t need to use it because the stream is automatically closed after the terminal method is executed.

30.2.1 Show the output of the following code:

Character[] chars = {'D', 'B', 'A', 'C'}; System.out.println(Stream.of(chars).sorted().findFirst().get()); System.out.println(Stream.of(chars).sorted( java.util.Comparator.reverseOrder()).findFirst().get()); System.out.println(Stream.of(chars) .limit(2).sorted().findFirst().get()); System.out.println(Stream.of(chars).distinct() .skip(2).filter(e −> e > 'A').findFirst().get()); System.out.println(Stream.of(chars) .max(Character::compareTo).get()); System.out.println(Stream.of(chars) .max(java.util.Comparator.reverseOrder()).get()); System.out.println(Stream.of(chars) .filter(e −> e > 'A').findFirst().get()); System.out.println(Stream.of(chars) .allMatch(e −> e >= 'A')); System.out.println(Stream.of(chars) .anyMatch(e −> e > 'F')); System.out.println(Stream.of(chars) .noneMatch(e −> e > 'F')); Stream.of(chars).map(e −> e + "").map(e −> e.toLowerCase()) .forEach(System.out::println);

close method

Point Check

M30_LIAN1878_11_GE_C30.indd 1151 12/29/17 4:53 PM

1152 Chapter 30 Aggregate Operations for Collection Streams

Object[] temp = Stream.of(chars).map(e −> e + "Y") .map(e −> e.toLowerCase()).sorted().toArray(); System.out.println(java.util.Arrays.toString(temp));

30.2.2 What is wrong in the following code?

Character[] chars = {'D', 'B', 'A', 'C'}; Stream<Character> stream = Stream.of(chars).sorted(); System.out.println(stream.findFirst()); System.out.println(stream.skip(2).findFirst());

30.2.3 Rewrite (a) using a method reference and an anonymous inner class and (b) using lambda expression and an anonymous inner class:

(a) sorted((s1, s2) −> s1.compareToIgnoreCase(s2)) (b) forEach(System.out::println)

30.2.4 Given a map of the type Map<String, Double>, write an expression that returns the sum of all the values in map. For example, if the map contains {"john", 1.5} and {"Peter", 1.1}, the sum is 2.6.

30.3 IntStream, LongStream, and DoubleStream IntStream, LongStream, and DoubleStream are special type of streams for processing a sequence of primitive int, long, and double values.

Stream represents a sequence of objects. In addition to Stream, Java provides IntStream, LongStream, and DoubleStream for representing a sequence of int, long, and double values. These streams are also subinterfaces of BaseStream. You can use these streams in the same way like a Stream. Additionally, you can use the sum(), average(), and summary Statistics()methods for returning the sum, average, various statistics of the elements in the stream. You can use the mapToInt method to convert a Stream to an IntStream and use the map method to convert any stream including an IntStream to a Stream.

Listing 30.2 gives an example of using the IntStream.

Listing 30.2 IntStreamDemo.java 1 import java.util.IntSummaryStatistics; 2 import java.util.stream.IntStream; 3 import java.util.stream.Stream; 4 5 public class IntStreamDemo { 6 public static void main(String[] args) { 7 int[] values = {3, 4, 1, 5, 20, 1, 3, 3, 4, 6}; 8 9 System.out.println("The average of distinct even numbers > 3: " + 10 IntStream.of(values).distinct() 11 .filter(e −> e > 3 && e % 2 == 0).average().getAsDouble()); 12 13 System.out.println("The sum of the first 4 numbers is " + 14 IntStream.of(values).limit(4).sum()); 15 16 IntSummaryStatistics stats = 17 IntStream.of(values).summaryStatistics(); 18 19 System.out.printf("The summary of the stream is\n%−10s%10d\n" + 20 "%−10s%10d\n%−10s%10d\n%−10s%10d\n%−10s%10.2f\n",

Point Key

IntStream

LongStream

DoubleStream

create an int array

IntStream.of

average

sum

summaryStatistics

M30_LIAN1878_11_GE_C30.indd 1152 12/29/17 4:53 PM

21 " Count:", stats.getCount(), " Max:", stats.getMax(), 22 " Min:", stats.getMin(), " Sum:", stats.getSum(), 23 " Average:", stats.getAverage()); 24 25 String[] names = {"John", "Peter", "Susan", "Kim", "Jen", 26 "George", "Alan", "Stacy", "Michelle", "john"}; 27 28 System.out.println("Total character count for all names is " 29 + Stream.of(names).mapToInt(e −> e.length()).sum()); 30 31 System.out.println("The number of digits in array values is " + 32 Stream.of(values).map(e −> e + "") 33 .mapToInt(e −> e.length()).sum()); 34 } 35 }

The average of distinct even numbers > 3: 10.0

The sum of the first 4 numbers is 13 The summary of the stream is

Count: 10 Max: 20 Min: 1 Sum: 50 Average: 5.00

Total character count for all names is 47 The number of digits in array values is 11

The program creates an array of int values (line 7). The stream pipeline in line 10 applies the intermediate methods distinct and filter with a terminal method average. The average() method returns the average value in the stream as an OptionalDouble object (line 11). The actual average value is obtained by invoking the getAsDouble() method.

The stream pipeline in line 14 applies the intermediate method limit with a terminal method sum. The sum() method returns the sum of all values in the stream.

If you need to obtain multiple summary values from the stream, using the summaryStatistics() method is more efficient. This method (line 17) returns an instance of IntSummaryStatistics that contains summary values for count, min, max, sum, and average (lines 19–23). Note sum(), average(), and summaryStatistics() methods are only applicable to the IntStream, LongStream, and DoubleStream.

The mapToInt method returns an IntStream by mapping each element in the stream into an int value. The mapToInt method in the stream pipeline in line 29 maps each string into an int value that is the length of a string, and the sum method obtains the sum of all the int values in the IntStream. The stream pipeline in line 29 obtains the total count for all characters in the stream.

The mapToInt method takes an argument of the ToIntFunction<? super T> type to return an instance of the IntStream. The ToIntFunction is a functional interface with an abstract method applyAsInt(T t) that maps t into a value of the type int. Line 33 uses a lambda expression to implement the ToIntFunction interface as shown in (a), which is equivalent to the code using an anonymous inner class as shown in (b). You can also further simplify it using a method reference as shown in (c).

average method

sum method

OptionalDouble class

summaryStatistics method

mapToInt method

30.3 IntStream, LongStream, and DoubleStream 1153

mapToInt

map mapToInt

M30_LIAN1878_11_GE_C30.indd 1153 12/29/17 4:53 PM

1154 Chapter 30 Aggregate Operations for Collection Streams

The map method in line 32 returns a new stream of strings. Each string is converted from an integer in array values. The mapToInt method in line 33 returns a new stream of integers. Each integer represents the length of a string. The sum() method returns the sum of all int values in the final stream. So the stream pipeline in lines 32–33 obtains the total number of digits in the array values.

30.3.1 Show the output of the following code:

int[] numbers = {1, 4, 2, 3, 1}; System.out.println(IntStream.of(numbers) .sorted().findFirst().getAsInt()); System.out.println(IntStream.of(numbers) .limit(2).sorted().findFirst().getAsInt()); System.out.println(IntStream.of(numbers).distinct() .skip(1).filter(e −> e > 2).sum()); System.out.println(IntStream.of(numbers).distinct() .skip(1).filter(e −> e > 2).average().getAsDouble()); System.out.println(IntStream.of(numbers).max().getAsInt()); System.out.println(IntStream.of(numbers).max().getAsInt()); System.out.println(IntStream.of(numbers) .filter(e −> e > 1).findFirst().getAsInt()); System.out.println(IntStream.of(numbers) .allMatch(e −> e >= 1)); System.out.println(IntStream.of(numbers) .anyMatch(e −> e > 4)); System.out.println(IntStream.of(numbers).noneMatch(e −> e > 4)); IntStream.of(numbers).mapToObj(e −> (char)(e + 50)) .forEach(System.out::println);

Object[] temp = IntStream.of(numbers) .mapToObj(e −> (char)(e + 'A')).toArray(); System.out.println(java.util.Arrays.toString(temp));

30.3.2 What is wrong in the following code?

int[] numbers = {1, 4, 2, 3, 1}; DoubleSummaryStatistics stats = DoubleStream.of(numbers).summaryStatistics(); System.out.printf("The summary of the stream is\n%−10s%10d\n" + "%−10s%10.2f\n%−10s%10.2f\n%−10s%10.2f\n%−10s%10.2f\n", " Count:", stats.getCount(), " Max:", stats.getMax(), " Min:", stats.getMin(), " Sum:", stats.getSum(), " Average:", stats.getAverage());

30.3.3 Rewrite the following code that maps an int to a Character using an anony- mous inner class:

mapToObj(e −> (char)(e + 50))

Point Check

mapToInt(e −> e.length()) mapToInt( new java.util.function.ToIntFunction<String>() { public int applyAsInt(String e) { return e.length(); } } )

(a) Using a lambda expression

mapToInt(String::length)

M30_LIAN1878_11_GE_C30.indd 1154 12/29/17 4:53 PM

30.4 Parallel Streams 1155

30.3.4 Show the output of the following code:

int[][] m = {{1, 2}, {3, 4}, {5, 6}}; System.out.println(Stream.of(m) .mapToInt(e −> IntStream.of(e).sum()).sum());

30.3.5 Given an array names in Listing 30.1, write the code to display the total number of characters in names.

30.4 Parallel Streams Streams can be executed in parallel mode to improve performance.

The widespread use of multicore systems has created a revolution in software. In order to benefit from multiple processors, software needs to run in parallel. All stream operations can execute in parallel to utilize the multicore processors. The stream() method in the Collection interface returns a sequential stream. To execute operations in parallel, use the parallelStream() method in the Collection interface to obtain a parallel stream. Any stream can be turned to into a parallel stream by invoking the parallel() method defined in tbe BaseStream interface. Likewise, you can turn a parallel stream into a sequential stream by invoking the sequential() method.

Intermediate methods can be further divided into stateless and stateful methods. A stateless method such as filter and map can be executed independently from other elements in the stream. A stateful method such as distinct and sorted must be executed to take the entire stream into consideration. For example, to produce a result, the distinct method must con- sider all elements in the stream. Stateless methods are inherently parallelizable and can be executed in one pass in parallel. Stateful methods have to be executed in multiple passes in parallel.

Listing 30.3 gives an example to demonstrate the benefits of using parallel streams.

Listing 30.3 ParallelStreamDemo.java 1 import java.util.Arrays; 2 import java.util.Random; 3 import java.util.stream.IntStream; 4 5 public class ParallelStreamDemo { 6 public static void main(String[] args) { 7 Random random = new Random(); 8 int[] list = random.ints(200_000_000).toArray(); 9 10 System.out.println("Number of processors: " + 11 Runtime.getRuntime().availableProcessors()); 12 13 long startTime = System.currentTimeMillis(); 14 int[] list1 = IntStream.of(list).filter(e −> e > 0).sorted() 15 .limit(5).toArray(); 16 System.out.println(Arrays.toString(list1)); 17 long endTime = System.currentTimeMillis(); 18 System.out.println("Sequential execution time is " + 19 (endTime − startTime) + " milliseconds"); 20 21 startTime = System.currentTimeMillis(); 22 int[] list2 = IntStream.of(list).parallel().filter(e −> e > 0) 23 .sorted().limit(5).toArray(); 24 System.out.println(Arrays.toString(list2)); 25 endTime = System.currentTimeMillis();

Point Key

stateless methods stateful methods

create an array

available processors

sequential stream

parallel stream

M30_LIAN1878_11_GE_C30.indd 1155 12/29/17 4:53 PM

1156 Chapter 30 Aggregate Operations for Collection Streams

26 System.out.println("Parallel execution time is " + 27 (endTime − startTime) + " milliseconds"); 28 } 29 }

Number of processors: 8 [4, 9, 38, 42, 52] Sequential execution time is 12362 milliseconds [4, 9, 38, 42, 52] Parallel execution time is 3448 milliseconds

The Random class introduced in Section 9.6.2 can be used to generate random numbers. You can use its ints(n) method to generate an IntStream consisting of n number of random int values (line 8). You can also use ints(n, r1, r2) to generate an IntStream with n element in the range from r1 (inclusive) to r2 (exclusive), use doubles(n) and doubles(n, r1, r2)to generate a DoubleStream of random floating-point numbers. Invoking the toArray() method on an IntStream (line 8) returns an array of int values from the stream. Recall that you can use underscores in an integer 200_000_000 to improve readability (see Section 2.10.1).

Invoking Runtime.getRuntim() returns a Runtime object (line 11). Invoking Runtime object’s availableProcessors() returns the number of available processors for the JVM. In this case, the system has 8 processors.

An IntStream is created using IntStream.of(list) (line 14). The intermediate filter(e −> e > 0) method selects positive integers from the stream. The intermediate sorted() method sorts the filtered stream. The intermediate limit(5) method selects the first five integers in the sorted stream (line 15). Finally, the terminal method toArray() returns an array from the 5 integers in the stream. This is a sequential stream. To turn it into a parallel stream, simply invoke the parallel() method (line 22), which sets the stream for parallel execution. As you see from the sample run, the parallel execution is much faster than the sequential execution.

Several interesting questions arise.

1. The intermediate methods are lazy and are executed when the terminal method is initiated. This can be confirmed in the following code:

1 long startTime = System.currentTimeMillis(); 2 IntStream stream = IntStream.of(list).filter(e −> e > 0).sorted() 3 .limit(5); 4 System.out.println("The time for the preceding method is " + 5 (System.currentTimeMillis() − startTime) + " milliseconds"); 6 int[] list1 = stream.toArray(); 7 System.out.println("The execution time is " + 8 (System.currentTimeMillis() − startTime) + " milliseconds");

When you run the code, you will see that almost no time is spent on lines 2–3 because the intermediate methods are not executed yet. When the terminal method toArray() is invoked, all the methods for the stream pipeline are executed. So, the actual execution time for the stream pipeline is in line 6.

2. Does the order of the intermediate methods in a stream pipeline matter? Yes, it matters. For example, if the methods limit(5) and sorted() are swapped, the result will be different. It also matters to the performance even though the result is the same. For example, if the sorted() method is placed before filter(e −> e > 0), the result will be the same, but it would take more time to execute the stream because sorting a

lazy intermediate methods

order of methods

M30_LIAN1878_11_GE_C30.indd 1156 12/29/17 4:53 PM

30.5 Stream Reduction Using the reduce Method 1157

large number of elements takes more time to complete. Applying filter before sorted would eliminate roughly half of the elements for sorting.

3. Is a parallel stream always faster? Not necessarily. Parallel execution requires synchroniza- tion, which carries some overhead. If you replace IntStream.of(list) in lines 14 and 22 by random.ints(200_000_000), the parallel stream in lines 22–23 will takes longer time to execute than the sequential stream in lines 14–15. The reason is that the algorithm for generating a sequence of pseudo-random numbers is highly sequential. The overhead of a parallel stream is far greater than the time saving on parallel processing. So, you should test both sequential and parallel streams before choosing parallel streams for deployment.

4. When executing a stream method in parallel, the elements in the stream may be pro- cessed in any order. So, the following code may display the numbers in the stream in a random order:

IntStream.of(1, 2, 3, 4, 5).parallel() .forEach(e −> System.out.print(e + " "));

However, if it is executed sequentially, the numbers will be displayed as 1 2 3 4 5.

30.4.1 What is a stateless method? What is a stateful method? 30.4.2 How do you create a parallel stream? 30.4.3 Suppose names is a set of strings, which of the following two streams is better?

Object[] s = set.parallelStream().filter(e −> e.length() > 3) .sorted().toArray();

Object[] s = set.parallelStream().sorted() .filter(e −> e.length() > 3).toArray();

30.4.4 What will be the output of the following code?

int[] values = {3, 4, 1, 5, 20, 1, 3, 3, 4, 6}; System.out.print("The values are "); IntStream.of(values) .forEach(e −> System.out.print(e + " "));

30.4.5 What will be the output of the following code?

int[] values = {3, 4, 1, 5, 20, 1, 3, 3, 4, 6}; System.out.print("The values are "); IntStream.of(values).parallel() .forEach(e −> System.out.print(e + " "));

30.4.6 Write a statement to obtain an array of 1000 random double values between 0.0 and 1.0, excluding 1.0.

30.5 Stream Reduction Using the reduce Method You can use the reduce method to reduce the elements in a stream into a single value.

Often you need to process all the elements in a collection to produce a summary value such as the sum, the maximum, or the minimum. For example, the following code obtains the sum of all elements in set s:

int total = 0; for (int e: s) { total += e; }

parallel vs. sequential streams

order of parallel execution

Point Check

Point Key

M30_LIAN1878_11_GE_C30.indd 1157 12/29/17 4:53 PM

1158 Chapter 30 Aggregate Operations for Collection Streams

This is a simple code, but it specifies the exact steps on how to obtain the sum and it is highly sequential. The reduce method on a stream can be used to write the code in a high level for parallel execution.

A reduction takes the elements from a stream to produce a single value by repeated applica- tion of a binary operation such as addition, multiplication, or finding the maximum between two elements. Using reduction, you can write the code for finding the sum of all elements in a set as follows:

int sum = s.parallelStream().reduce(0, (e1, e2) −> e1 + e2);

Here, the reduce method takes two arguments. The first is an identity, i.e., the starting value. The second argument is an object of the functional interface IntBinaryOperator. This interface contains the abstract method applyAsInt(int e1, int e2) that returns an int value from applying a binary operation. The preceding lambda expression in (a) is equivalent to the code using an anonymous inner class in (b).

reduction

reduce(0, e −> (e1, e2) −> e1 + e2) reduce(0, new java.util.function.IntBinaryOperator() { public int applyAsInt(int e1, int e2) { return e1 + e2; } } )

(a) Using a lambda expression

(b) Using an anonymous inner class

The preceding reduce method is semantically equivalent to an imperative code as follows:

int total = identity (i.e., 0, in this case); for (int e: s) { total = applyAsInt(total, e); }

The reduce method makes the code concise. Moreover, the code can be parallelizable, because multiple processors can simultaneously invoke the applyAsInt method on two inte- gers repeatedly.

Using the reduce method, you can write the following code to return the maximum ele- ment in the set:

int result = s.parallelStream() .reduce(Integer.MIN_VALUE, (e1, e2) −> Math.max(e1, e2));

In fact, the sum, max, and min methods are implemented using the reduce method. Listing 30.4 gives an example of using the reduce method.

Listing 30.4 StreamReductionDemo.java 1 import java.util.stream.IntStream; 2 import java.util.stream.Stream; 3 4 public class StreamReductionDemo { 5 public static void main(String[] args) { 6 int[] values = {3, 4, 1, 5, 20, 1, 3, 3, 4, 6}; 7 8 System.out.print("The values are "); 9 IntStream.of(values).forEach(e −> System.out.print(e + " ")); 10

create an array

forEach

M30_LIAN1878_11_GE_C30.indd 1158 12/29/17 4:53 PM

11 System.out.println("\nThe result of multiplying all values is " + 12 IntStream.of(values).parallel().reduce(1, (e1, e2) −> e1 * e2)); 13 14 System.out.print("The values are " + 15 IntStream.of(values).mapToObj(e −> e + "") 16 .reduce((e1, e2) −> e1 + ", " + e2).get()); 17 18 String[] names = {"John", "Peter", "Susan", "Kim", "Jen", 19 "George", "Alan", "Stacy", "Michelle", "john"}; 20 System.out.print("\nThe names are: "); 21 System.out.println(Stream.of(names) 22 .reduce((x, y) −> x + ", " + y).get()); 23 24 System.out.print("Concat names: "); 25 System.out.println(Stream.of(names) 26 .reduce((x, y) −> x + y).get()); 27 28 System.out.print("Total number of characters: "); 29 System.out.println(Stream.of(names) 30 .reduce((x, y) −> x + y).get().length()); 31 } 32 }

The values are 3, 4, 1, 5, 20, 1, 3, 3, 4, 6 The result of multiplying all values is 259200 The values are 3, 4, 1, 5, 20, 1, 3, 3, 4, 6 The names are John, Peter, Susan, Kim, Jen, George, Alan, Stacy, Michelle, john Concat names: JohnPeterSusanKimJenGeorgeAlanStacyMichellejohn Total number of characters: 47

30.5 Stream Reduction Using the reduce Method 1159

The program creates an array of int values (line 6). The stream pipeline creates an IntStream from the int array and invokes the forEach method to display each integer in the stream (line 9).

The program creates a parallel stream pipeline for the int array and applies the reduce method to obtain the product of int values in the stream (line 12).

The mapToObj method returns a stream of string objects from the IntStream (line 15). The reduce method can be called without an identity. In this case, it returns an object of Optional<T>. The reduce method in line 16 reduces the strings in the stream into one com- posite string that consists of all strings in the stream separated by commas.

The reduce method in line 26 combines all strings in the stream together into one long string. Stream.of(names).reduce((x, y) −> x + y).get() returns a string that concatenates all strings in the stream into one string, and invoking the length() method on the string returns the number of characters in the string (line 30).

Note reduce((x, y) −> x + y) in line 30 can be simplified using a method reference as reduce(String::concat).

30.5.1 Show the output of the following code:

int[] values = {1, 2, 3, 4}; System.out.println(IntStream.of(values) .reduce(0, (e1, e2) −> e1 + e2)); System.out.println(IntStream.of(values) .reduce(1, (e1, e2) −> e1 * e2)); System.out.println(IntStream.of(values).map(e −> e * e) .reduce(0, (e1, e2) −> e1 + e2));

Point Check

reduce

mapToObj reduce

reduce

M30_LIAN1878_11_GE_C30.indd 1159 12/29/17 4:53 PM

1160 Chapter 30 Aggregate Operations for Collection Streams

System.out.println(IntStream.of(values).mapToObj(e −> "" + e) .reduce((e1, e2) −> e1 + " " + e2).get()); System.out.println(IntStream.of(values).mapToObj(e −> "" + e) .reduce((e1, e2) −> e1 + ", " + e2).get());

30.5.2 Show the output of the following code:

int[][] m = {{1, 2}, {3, 4}, {5, 6}}; System.out.println(Stream.of(m) .map(e −> IntStream.of(e).reduce(1, (e1, e2) −> e1 * e2)) .reduce(1, (e1, e2) −> e1 * e2));

30.5.3 Show the output of the following code:

int[][] m = {{1, 2}, {3, 4}, {5, 6}, {1, 3}}; Stream.of(m).map(e −> IntStream.of(e)) .reduce((e1, e2) −> IntStream.concat(e1, e2)) .get().distinct() .forEach(e −> System.out.print(e + " "));

30.5.4 Show the output of the following code:

int[][] m = {{1, 2}, {3, 4}, {5, 6}, {1, 3}}; System.out.println( Stream.of(m).map(e −> IntStream.of(e)) .reduce((e1, e2) −> IntStream.concat(e1, e2)) .get().distinct().mapToObj(e −> e + "") .reduce((e1, e2) −> e1 + ", " + e2).get());

30.6 Stream Reduction Using the collect Method You can use the collect method to reduce the elements in a stream into a mutable container.

In the preceding example, the String’s concat method is used in the reduce method for Stream.of(names).reduce((x, y) −> x + y). This operation causes a new string to be created when concatenating two strings, which is very inefficient. A better approach is to use a StringBuilder and accumulate the result into a StringBuilder. This can be accomplished using the collect method.

The collect method collects the elements in a stream into a mutable container such as a Collection object using the following syntax:

<R> R collect(Supplier<R> supplier, BiConsumer<R, ? super T> accumulator, BiConsumer<R, R> combiner)

The method takes three functional arguments: 1) a supplier function to construct a new instance of the result container, 2) an accumulator function to incorporate the elements from the stream to the result container, and 3) a combining function to merge the contents of one result con- tainer into another.

For example, to combine strings into a StringBuilder, you may write the following code using a collect method like this:

String[] names = {"John", "Peter", "Susan", "Kim", "Jen", "George", "Alan", "Stacy", "Michelle", "john"}; StringBuilder sb = Stream.of(names).collect(() −> new StringBuilder(), (c, e) −> c.append(e), (c1, c2) −> c1.append(c2));

Point Key

M30_LIAN1878_11_GE_C30.indd 1160 12/29/17 4:53 PM

30.6 Stream Reduction Using the collect Method 1161

The lambda expression () −> new StringBuilder() creates a StringBuilder object for storing the result, which can be simplified using the method reference StringBuilder::new. The lambda expression (c, e) −> c.append(e) adds a string e to a StringBuilder c, which can be simplified using a method reference StringBuilder::append. The lambda expression (c1, c2) −> c1.append(c2) merges the contents in c2 into c1, which also can be simplified using a method reference StringBuilder::append. So, you can simplify the preceding statement as follows:

StringBuilder sb = Stream.of(names).collect(StringBuilder::new, StringBuilder::append, StringBuilder::append);

The sequential foreach loop implementation for this collect method might be as follows:

StringBuilder sb = new StringBuilder(); for (String s: Stream.of(names)) { sb.append(s); }

Note that the combiner function (c1, c2) −> c1.append(c2) is not used in the sequential implementation. It is used when the stream pipeline is executed in parallel. When executing a collect method in parallel, multiple result StringBuilder are created and then merged using a combiner function. So, the purpose of the combiner function is for parallel processing.

Here is another example that creates an ArrayList from strings in the stream:

ArrayList<String> list = Stream.of(names).collect(ArrayList::new, ArrayList::add, ArrayList::addAll);

The supplier function is the ArrayList constructor. The accumulator is the add method that adds an element to the ArrayList. The combiner function merges an ArrayList into another ArrayList. The three arguments—supplier, accumulator, and combiner—are tightly cou- pled and are defined using standard methods. For simplicity, Java provides another collect method that takes an argument of the Collector type, called a collector. The Collector interface defines the methods for returning a supplier, an accumulator, and a combiner. You can use a static factory method toList() in the Collectors class to create an instance of the Collector interface. So, the preceding statement can be simplified using a standard col- lector as follows:

List<String> list = Stream.of(names).collect(Collectors.toList());

Listing 30.5 gives an example of using the collect methods and the Collectors’ factory methods.

Listing 30.5 CollectDemo.java 1 import java.util.ArrayList; 2 import java.util.List; 3 import java.util.Map; 4 import java.util.Set; 5 import java.util.stream.Collectors; 6 import java.util.stream.Stream; 7 8 public class CollectDemo { 9 public static void main(String[] args) { 10 String[] names = {"John", "Peter", "Susan", "Kim", "Jen", 11 "George", "Alan", "Stacy", "Michelle", "john"}; 12 System.out.println("The number of characters for all names: " + 13 Stream.of(names).collect(StringBuilder::new, 14 StringBuilder::append, StringBuilder::append).length());

create an array

collect into StringBuilder

M30_LIAN1878_11_GE_C30.indd 1161 12/29/17 4:53 PM

1162 Chapter 30 Aggregate Operations for Collection Streams

15 16 List<String> list = Stream.of(names).collect(ArrayList::new, 17 ArrayList::add, ArrayList::addAll); 18 System.out.println(list); 19 20 list = Stream.of(names).collect(Collectors.toList()); 21 System.out.println(list); 22 23 Set<String> set = Stream.of(names).map(e −> e.toUpperCase()). 24 collect(Collectors.toSet()); 25 System.out.println(set); 26 27 Map<String, Integer> map = Stream.of(names).collect( 28 Collectors.toMap(e −> e, e −> e.length())); 29 System.out.println(map); 30 31 System.out.println("The total number of characters is " + 32 Stream.of(names). 33 collect(Collectors.summingInt(e −> e.length()))); 34 35 java.util.IntSummaryStatistics stats = Stream.of(names). 36 collect(Collectors.summarizingInt(e −> e.length())); 37 System.out.println("Max is " + stats.getMax()); 38 System.out.println("Min is " + stats.getMin()); 39 System.out.println("Average is " + stats.getAverage()); 40 } 41 }

The number of characters for all names: 47 [John, Peter, Susan, Kim, Jen, George, Alan, Stacy, Michelle, john] [John, Peter, Susan, Kim, Jen, George, Alan, Stacy, Michelle, john] [JEN, GEORGE, ALAN, SUSAN, JOHN, PETER, MICHELLE, KIM, STACY] {Michelle=8, Stacy=5, Jen=3, George=6, Susan=5, Alan=4, John=4, john=4, Peter=5, Kim=3} The total number of characters is 47 Max is 8 Min is 3 Average is 4.7

The program creates an array of strings, names (lines 10–11). The collect method (line 13) specifies a supplier (StringBuilder::new) for creating a StringBuilder, an accumula- tor (StringBuilder::append) for adding a string to the StringBuilder, and a combiner (StringBuilder::append) for combining two StringBuilders (lines 13–14). The stream pipeline obtains a StringBuilder that contains all strings in the stream. The length() method returns the length of the characters in the StringBuilder.

The collect method (line 16) specifies a supplier (ArrayList::new) for creating an ArrayList, an accumulator (ArrayList::add) for adding a string to the ArrayList, and a combiner (ArrayList:addAll) for combining two ArrayLists (lines 16–17). The stream pipeline obtains an ArrayList that contains all strings in the stream. This statement is simpli- fied using a standard collector Collectors.toList() for a list (line 20).

The program creates a string stream, maps each string to uppercase (line 23), and creates a set using the collect method with a standard collector Collectors.toSet() for a set (lines 24). There are two uppercase strings JOHN in the stream. Since a set contains no duplicate elements, only one JOHN is stored in the set.

collect into ArrayList

Collectors.toList()

Collectors.toSet()

Collectors.toMap()

sum of integers summary information

M30_LIAN1878_11_GE_C30.indd 1162 12/29/17 4:53 PM

30.7 Grouping Elements Using the groupingby Collector 1163

The program creates a string stream and creates a map using the collect method with a standard collector for a map (lines 28). The key of the map is the string and the value is the length of the string. Note key must be unique. If two strings are identical in the stream, a runtime exception would occur.

The Collectors class also contains the method for returning collectors that produce summary information. For example, Collectors.summingInt produces the sum of integer values in the stream (line 33), and Collectors.summarizingInt produces an IntSummaryStatistics for the integer values in the stream (lines 35–36).

30.6.1 Show the output of the following code:

int[] values = {1, 2, 3, 4, 1}; List<Integer> list = IntStream.of(values).mapToObj(e −> e) .collect(Collectors.toList()); System.out.println(list); Set<Integer> set = IntStream.of(values).mapToObj(e −> e) .collect(Collectors.toSet()); System.out.println(set); Map<Integer, Integer> map = IntStream.of(values).distinct() .mapToObj(e −> e) .collect(Collectors.toMap(e −> e, e −> e.hashCode())); System.out.println(map); System.out.println( IntStream.of(values).mapToObj(e −> e) .collect(Collectors.summingInt(e −> e)));

System.out.println( IntStream.of(values).mapToObj(e −> e) .collect(Collectors.averagingDouble(e −> e)));

30.7 Grouping Elements Using the groupingby Collector

You can use the groupingBy collector along with the collect method to collect the elements by groups.

The elements in a stream can be divided into groups using the groupingBy collector and then applying aggregate collectors on each group. For example, you can group all the strings by their first letter and obtain the count of the elements in the group as follows:

String[] names = {"John", "Peter", "Susan", "Kim", "Jen", "George", "Alan", "Stacy", "Steve", "john"}; Map<Character, Long> map = Stream.of(names).collect( Collectors.groupingBy(e −> e.charAt(0), Collectors.counting()));

The first argument in the groupingBy method specifies the criteria for grouping, known as a classifier. The second argument specifies how the elements in a group are processed, known as a group processor. A processor is commonly a summary collector such as counting(). Using the groupingBy collector, the collect method returns a map with the classifier as the key. You may also specify a supplier in the groupingBy method such as the following:

Map<Character, Long> map = Stream.of(names).collect( Collectors.groupingBy(e −> e.charAt(0), TreeMap::new, Collectors.counting()));

Point Check

Point Key

group classifier group processor

M30_LIAN1878_11_GE_C30.indd 1163 12/29/17 4:53 PM

1164 Chapter 30 Aggregate Operations for Collection Streams

In this case, a tree map is used to store the map entries. Listing 30.6 gives an example of using the groupingBy method.

Listing 30.6 CollectGroupDemo.java 1 import java.util.Map; 2 import java.util.TreeMap; 3 import java.util.stream.Collectors; 4 import java.util.stream.IntStream; 5 import java.util.stream.Stream; 6 7 public class CollectGroupDemo { 8 public static void main(String[] args) { 9 String[] names = {"John", "Peter", "Susan", "Kim", "Jen", 10 "George", "Alan", "Stacy", "Steve", "john"}; 11 12 Map<String, Long> map1 = Stream.of(names). 13 map(e −> e.toUpperCase()).collect( 14 Collectors.groupingBy(e −> e, Collectors.counting())); 15 System.out.println(map1); 16 17 Map<Character, Long> map2 = Stream.of(names).collect( 18 Collectors.groupingBy(e −> e.charAt(0), TreeMap::new, 19 Collectors.counting())); 20 System.out.println(map2); 21 22 int[] values = {2, 3, 4, 1, 2, 3, 2, 3, 4, 5, 1, 421}; 23 IntStream.of(values).mapToObj(e −> e).collect( 24 Collectors.groupingBy(e −> e, TreeMap::new, 25 Collectors.counting())). 26 forEach((k, v) −> System.out.println(k + " occurs " + v + 27 (v > 1 ? " times " : " time "))); 28 29 MyStudent[] students = {new MyStudent("John", "Lu", "CS", 32, 78), 30 new MyStudent("Susan", "Yao", "Math", 31, 85.4), 31 new MyStudent("Kim", "Johnson", "CS", 30, 78.1)}; 32 33 System.out.printf("%10s%10s\n", "Department", "Average"); 34 Stream.of(students).collect(Collectors. 35 groupingBy(MyStudent::getMajor, TreeMap::new, 36 Collectors.averagingDouble(MyStudent::getScore))). 37 forEach((k, v) −> System.out.printf("%10s%10.2f\n", k, v)); 38 } 39 } 40 41 class MyStudent { 42 private String firstName; 43 private String lastName; 44 private String major; 45 private int age; 46 private double score; 47 48 public MyStudent(String firstName, String lastName, String major, 49 int age, double score) { 50 this.firstName = firstName; 51 this.lastName = lastName; 52 this.major = major; 53 this.age = age; 54 this.score = score; 55 }

create an array

groupingBy

groupingby

Student class

M30_LIAN1878_11_GE_C30.indd 1164 12/29/17 4:53 PM

56 57 public String getFirstName() { 58 return firstName; 59 } 60 61 public String getLastName() { 62 return lastName; 63 } 64 65 public String getMajor() { 66 return major; 67 } 68 69 public int getAge() { 70 return age; 71 } 72 73 public double getScore() { 74 return score; 75 } 76 }

{JEN=1, ALAN=1, GEORGE=1, SUSAN=1, JOHN=2, STEVE=1, PETER=1, STACY=1, KIM=1} {A=1, G=1, J=2, K=1, P=1, S=3, j=1} 1 occurs 2 times 2 occurs 3 times 3 occurs 3 times 4 occurs 2 times 5 occurs 1 time 421 occurs 1 time Department Average CS 78.05 Math 85.40

The program creates a string stream, maps its elements to uppercase (lines 12–13), and col- lects the elements into a map with the string as the key and the occurrence of the string as the value (line 14). Note the counting() collector uses a value of the Long type. So, the Map is declared Map<String, Long>.

The program creates a string stream and collects the elements into a map with the first character in the string as the key and the occurrence of the first character as the value (lines 17–18). The key and value entries are stored in a TreeMap.

The program creates an int array (line 22). A stream is created from this array and the ele- ments are mapped to Integer objects using the mapToObj method (line 23). The collect method returns a map with the Integer value as the key and the occurrence of the integer as the value (lines 24–25). The forEach method displays the key and value entries. Note the collect method is a terminal method. It returns an instance of TreeMap in this case. TreeMap has the forEach method for performing an action on each element in the collection.

The program creates an array of Student objects (lines 29–31). The Student class is defined with properties firstName, lastName, major, age, and score in lines 41–76. The program creates a stream for the array and the collect method groups the students by their major and returns a map with the major as the key and the average scores for the group as the value (lines 34-36). The method reference MyStudent::getMajor is used to specify the group by classifier. The method reference TreeMap::new specifies a supplier for the result map. The method reference MyStudent:getScore specifies the value for averaging.

30.7 Grouping Elements Using the groupingby Collector 1165

M30_LIAN1878_11_GE_C30.indd 1165 12/29/17 4:53 PM

1166 Chapter 30 Aggregate Operations for Collection Streams

30.7.1 Show the output of the following code:

int[] values = {1, 2, 2, 3, 4, 2, 1}; IntStream.of(values).mapToObj(e −> e).collect( Collectors.groupingBy(e −> e, TreeMap::new, Collectors.counting())). forEach((k, v) −> System.out.println(k + " occurs " + v + (v > 1 ? " times " : " time ")));

IntStream.of(values).mapToObj(e −> e).collect( Collectors.groupingBy(e −> e, TreeMap::new, Collectors.summingInt(e −> e))). forEach((k, v) −> System.out.println(k + ": " + v));

MyStudent[] students = { new MyStudent("John", "Johnson", "CS", 23, 89.2), new MyStudent("Susan", "Johnson", "Math", 21, 89.1), new MyStudent("John", "Peterson", "CS", 21, 92.3), new MyStudent("Kim", "Yao", "Math", 22, 87.3), new MyStudent("Jeff", "Johnson", "CS", 23, 78.5)};

Stream.of(students) .sorted(Comparator.comparing(MyStudent::getLastName) .thenComparing(MyStudent::getFirstName)) .forEach(e −> System.out.println(e.getLastName() + ", " + e.getFirstName()));

Stream.of(students).collect(Collectors. groupingBy(MyStudent::getAge, TreeMap::new, Collectors.averagingDouble(MyStudent::getScore))). forEach((k, v) −> System.out.printf("%10s%10.2f\n", k, v));

30.8 Case Studies Many programs for processing arrays and collections can now be simplified and run faster using aggregate methods on streams.

You can write programs without using streams. However, using streams enables you to write shorter and simpler programs that can be executed faster in parallel by utilizing multiple pro- cessors. Many of the programs that involve arrays and collections in the early chapters can be simplified using streams. This section presents several case studies.

30.8.1 Case Study: Analyzing Numbers Section 7.3 gives a program that prompts the user to enter values, obtains their average, and displays the number of values greater than the average. The program can be simplified using a DoubleStream as shown in Listing 30.7.

Listing 30.7 AnalyzeNumbersUsingStream.java 1 import java.util.stream.*; 2 3 public class AnalyzeNumbersUsingStream { 4 public static void main(String[] args) { 5 java.util.Scanner input = new java.util.Scanner(System.in); 6 System.out.print("Enter the number of items: "); 7 int n = input.nextInt(); 8 double[] numbers = new double[n]; 9 double sum = 0; 10

Point Check

Point Key

create array

M30_LIAN1878_11_GE_C30.indd 1166 12/29/17 4:53 PM

30.8 Case Studies 1167

11 System.out.print("Enter the numbers: "); 12 for (int i = 0; i < n; i++) { 13 numbers[i] = input.nextDouble(); 14 } 15 16 double average = DoubleStream.of(numbers).average().getAsDouble(); 17 System.out.println("Average is " + average); 18 System.out.println("Number of elements above the average is " 19 + DoubleStream.of(numbers).filter(e −> e > average).count()); 20 } 21 }

Enter the number of items: 10

Enter the numbers: 3.4 5 6 1 6.5 7.8 3.5 8.5 6.3 9.5 Average is 5.75 Number of elements above the average is 6

The program obtains the input from the user and stores the values in an array (lines 8–14), and obtains the average of the values using a stream (line 16), and finds the number of values greater than the average using a filtered stream (line 19).

30.8.2 Case Study: Counting the Occurrences of Each Letter Listing 7.4, CountLettersInArrays.java gives a program that randomly generates 100 lowercase letters and counts the occurrences of each letter.

The program can be simplified using a Stream as shown in Listing 30.8.

Listing 30.8 CountLettersUsingStream.java 1 import java.util.Random; 2 import java.util.TreeMap; 3 import java.util.stream.Collectors; 4 import java.util.stream.Stream; 5 6 public class CountLettersUsingStream { 7 private static int count = 0; 8 9 public static void main(String[] args) { 10 Random random = new Random(); 11 Object[] chars = random.ints(100, (int)'a', (int)'z' + 1). 12 mapToObj(e −> (char)e).toArray(); 13 14 System.out.println("The lowercase letters are:"); 15 Stream.of(chars).forEach(e −> { 16 System.out.print(e + (++count % 20 == 0 ? "\n" : " ")); 17 }); 18 19 count = 0; // Reset the count for columns 20 System.out.println("\nThe occurrences of each letter are:"); 21 Stream.of(chars).collect(Collectors.groupingBy(e −> e, 22 TreeMap::new, Collectors.counting())).forEach((k, v) −> { 23 System.out.print(v + " " + k 24 + (++count % 10 == 0 ? "\n" : " ")); 25 }); 26 } 27 }

store number in array

get average

above average?

create array

display array

count occurrence

M30_LIAN1878_11_GE_C30.indd 1167 12/29/17 4:53 PM

1168 Chapter 30 Aggregate Operations for Collection Streams

The program generates a stream of 100 random integers. These integers are in the range between (char)'a' and (char)'z' (line 11). They are ASCII code for the lowercase letters. The mapToObj method maps the integers to their corresponding lowercase letters (line 12). The toArray() method returns an array consisting of these lowercase letters.

The program creates a stream of lowercase letters (line 15) and uses the forEach method to display each letter (line 16). The letters are displayed 20 per line. The static variable count is used to count the letters printed.

The program resets the count to 0 (line 19) , creates a stream of lowercase letters (line 21), returns a map with lowercase letters as the key and the occurrences of each letter as the value (lines 21–22), and invokes the forEach method to display each key and value 10 per line (lines 23–24).

The code has 66 lines in Listing 7.4. The new code has only 27 lines, which greatly simplified coding. Furthermore, using streams is more efficient.

30.8.3 Case Study: Counting the Occurrences of Each Letter in a String

The preceding example randomly generates lowercase letters and counts the occurrence of each letter. This example counts the occurrence of each letter in a string. The program given in Listing 30.9 prompts the user to enter the string, converts all letters to uppercase, and displays the count of each letter in the string.

Listing 30.9 CountOccurrenceOfLettersInAString.java 1 import java.util.*; 2 import java.util.stream.Stream; 3 import java.util.stream.Collectors; 4 5 public class CountOccurrenceOfLettersInAString { 6 private static int count = 0; 7 8 public static void main(String[] args) { 9 Scanner input = new Scanner(System.in); 10 System.out.print("Enter a string: "); 11 String s = input.nextLine(); 12 13 count = 0; // Reset the count for columns 14 System.out.println("The occurrences of each letter are:"); 15 Stream.of(toCharacterArray(s.toCharArray())) 16 .filter(ch −> Character.isLetter(ch)) 17 .map(ch −> Character.toUpperCase(ch)) 18 .collect(Collectors.groupingBy(e −> e,

The lowercase letters are:

e y l s r i b k j v j h a b z n w b t v s c c k r d w a m p w v u n q a m p l o a z g d e g f i n d x m z o u l o z j v h w i w n t g x w c d o t x h y v z y z q e a m f w p g u q t r e n n w f c r f

The occurrences of each letter are:

5 a 3 b 4 c 4 d 4 e 4 f 4 g 3 h 3 i 3 j 2 k 3 l 4 m 6 n 4 o 3 p 3 q 4 r 2 s 4 t 3 u 5 v 8 w 3 x 3 y 6 z

read a string

convert char[] to Character[]

select letters map to uppercase collect and count

M30_LIAN1878_11_GE_C30.indd 1168 12/29/17 4:53 PM

30.8 Case Studies 1169

19 TreeMap::new, Collectors.counting())) 20 .forEach((k, v) −> { System.out.print(v + " " + k 21 + (++count % 10 == 0 ? "\n" : " ")); 22 }); 23 } 24 25 public static Character[] toCharacterArray(char[] list) { 26 Character[] result = new Character[list.length]; 27 for (int i = 0; i < result.length; i++) { 28 result[i] = list[i]; 29 } 30 return result; 31 } 32 }

Enter a string: Welcome to JavaAA

The occurrences of each letter are: 4 A 1 C 2 E 1 J 1 L 1 M 2 O 1 T 1 V 1 W

The program reads a string s and obtains an array of char from the string by invoking s.toCharArray(). To create a stream of characters, you need to convert char[] into Character[]. So, the program defines the toCharacterArray method (lines 25–31) for obtaining a Character[] from char[].

The program creates a stream of Character object (line 15), eliminates nonletters from the stream using the filter method (line 16), coverts all letters to uppercase using the map method (line 17), and obtains a TreeMap using the collect method (lines 18–19). In the TreeMap, the key is the letter and the value is the count for the letter. The forEach method (lines 20–21) in TreeMap is used to display the value and key.

30.8.3 Case Study: Processing All Elements in a Two-Dimensional Array

You can create a stream from a one-dimensional array. Can you create a stream for process- ing two-dimensional arrays? Listing 30.10 gives an example of processing two-dimensional arrays using streams.

Listing 30.10 TwoDimensionalArrayStream.java

1 import java.util.IntSummaryStatistics; 2 import java.util.stream.IntStream; 3 import java.util.stream.Stream; 4 5 public class TwoDimensionalArrayStream { 6 private static int i = 0; 7 public static void main(String[] args) { 8 int[][] m = {{1, 2}, {3, 4}, {4, 5}, {1, 3}}; 9 10 int[] list = Stream.of(m).map(e −> IntStream.of(e)). 11 reduce((e1, e2) −> IntStream.concat(e1, e2)).get().toArray(); 12 13 IntSummaryStatistics stats = 14 IntStream.of(list).summaryStatistics(); 15 System.out.println("Max: " + stats.getMax()); 16 System.out.println("Min: " + stats.getMin());

char[] to Character[]

display value and key

create an array

reduce to onedimensional

obtain statistical information

M30_LIAN1878_11_GE_C30.indd 1169 12/29/17 4:53 PM

1170 Chapter 30 Aggregate Operations for Collection Streams

17 System.out.println("Sum: " + stats.getSum()); 18 System.out.println("Average: " + stats.getAverage()); 19 20 System.out.println("Sum of row "); 21 Stream.of(m).mapToInt(e −> IntStream.of(e).sum()) 22 .forEach(e −> 23 System.out.println("Sum of row " + i++ + ": " + e)); 24 } 25 }

Max: 5 Min: 1 Sum: 23 Average: 2.875 Sum of row Sum of row 0: 3 Sum of row 1: 7 Sum of row 2: 9 Sum of row 3: 4

The program creates a two-dimensional array m in line 8. Invoking Stream.of(m) creates a stream consisting of rows as elements (line 10). The map method maps each row to an IntStream. The reduce method concatenates these streams into one large stream (line 11). This large stream now contains all the elements in m. Invoking toArray() from the stream returns an array consisting of all the integers in the stream (line 11).

The program obtains a statistical summary for an IntStream created from the array (lines 13–14) and displays the maximum, minimum, sum, and average of the integers in the stream (lines 15–18).

Finally, the program creates a stream from m that consists of rows in m (line 21). Each row is mapped to an int value using the mapToInt method (line 21). The int value is the sum of the elements in the row. The forEach method displays the sum for each row (lines 22–23).

30.8.4 Case Study: Finding the Directory Size Listing 18.7 gives a recursive program that finds the size of a directory. The size of a directory is the sum of the sizes of all the files in the directory. The program can be implemented using streams as shown in Listing 30.11.

Listing 30.11 DirectorySizeStream.java 1 import java.io.File; 2 import java.nio.file.Files; 3 import java.util.Scanner; 4 5 public class DirectorySizeStream { 6 public static void main(String[] args) throws Exception { 7 // Prompt the user to enter a directory or a file 8 System.out.print("Enter a directory or a file: "); 9 Scanner input = new Scanner(System.in); 10 String directory = input.nextLine(); 11 12 // Display the size 13 System.out.println(getSize(new File(directory)) + " bytes"); 14 } 15

sum of each row

invoke method

M30_LIAN1878_11_GE_C30.indd 1170 12/29/17 4:53 PM

30.8 Case Studies 1171

16 public static long getSize(File file) { 17 if (file.isFile()) { 18 return file.length(); 19 } 20 else { 21 try { 22 return Files.list(file.toPath()).parallel(). 23 mapToLong(e −> getSize(e.toFile())).sum(); 24 } catch (Exception ex) { 25 return 0; 26 } 27 } 28 } 29 }

Enter a directory or a file: c:\book

48619631 bytes

Enter a directory or a file: c:\book\Welcome.java

172 bytes

Enter a directory or a file: c:\book\NonExistentFile

0 bytes

The program prompts the user to enter a file or a directory name (lines 8–10) and invokes the getSize(File file) method to return the size of the file or a directory (line 13). The File object can be a directory or a file. If it is a file, the getSize method returns the size of the file (lines 17–19). If it is a directory, invoking Files.list(file.toPath()) returns a stream consisting of the Path objects. Each Path object represents a subpath in the directory (line 22). Since JDK 1.8, new methods are added in some existing classes to return streams. The Files class now has the static list(Path) method that returns a stream of subpaths in the path. For each subpath e, the mapToLong method maps e into the size of e by invoking getSize(e). Finally, the terminal sum() method returns the size of the whole directory.

Both getSize methods in Listing 18.7 and Listing 30.9 are recursive. The getSize method in Listing 30.9 can be executed in a parallel stream. So, Listing 30.9 has better performance.

30.8.5 Case Study: Counting Keywords Listing 21.7 gives a program to count the keywords in a Java source file. The program reads the words from a text file and tests if the word is a keyword. You can rewrite the code using streams as shown in Listing 30.12.

Listing 30.12 CountKeywordStream.java 1 import java.util.*; 2 import java.io.*; 3 import java.nio.file.Files; 4 import java.util.stream.Stream; 5 6 public class CountKeywordStream { 7 public static void main(String[] args) throws Exception { 8 Scanner input = new Scanner(System.in);

getSize method is file? return file length

get subpaths recursive call

M30_LIAN1878_11_GE_C30.indd 1171 12/29/17 4:53 PM

1172 Chapter 30 Aggregate Operations for Collection Streams

9 System.out.print("Enter a Java source file: "); 10 String filename = input.nextLine(); 11 12 File file = new File(filename); 13 if (file.exists()) { 14 System.out.println("The number of keywords in " + filename 15 + " is " + countKeywords(file)); 16 } 17 else { 18 System.out.println("File " + filename + " does not exist"); 19 } 20 } 21 22 public static long countKeywords(File file) throws Exception { 23 // Array of all Java keywords + true, false and null 24 String[] keywordString = {"abstract", "assert", "boolean", 25 "break", "byte", "case", "catch", "char", "class", "const", 26 "continue", "default", "do", "double", "else", "enum", 27 "extends", "for", "final", "finally", "float", "goto", 28 "if", "implements", "import", "instanceof", "int", 29 "interface", "long", "native", "new", "package", "private", 30 "protected", "public", "return", "short", "static", 31 "strictfp", "super", "switch", "synchronized", "this", 32 "throw", "throws", "transient", "try", "void", "volatile", 33 "while", "true", "false", "null"}; 34 35 Set<String> keywordSet = 36 new HashSet<>(Arrays.asList(keywordString)); 37 38 return Files.lines(file.toPath()).parallel().mapToLong(line −> 39 Stream.of(line.split("[\\s++]")). 40 filter(word −> keywordSet.contains(word)).count()).sum(); 41 } 42 }

Enter a Java source file: c:\Welcome.java

The number of keywords in c:\Welcome.java is 5

Enter a Java source file: c:\TTT.java

File c:\TTT.java does not exist

The program prompts the user to enter a filename (lines 9–10). If the file exists, it invokes conutKeywords(file) to return the number of keywords in the file (line 15). The keywords are stored in a set keywordSet (lines 35–36). Invoking Files.lines(file.toPath()) returns a streams of lines form the file (line 38). Each line in the stream is mapped to a long value that counts the number of the keywords in the line. The line is split into an array of words using line.split("[\\s++]") and this array is used to create a stream (line 39). The filter method is applied to select the keyword from the stream. Invoking count() returns the number of the keywords in a line. The sum() method returns the total number of keywords in all lines (line 40).

The real benefit of using parallel streams in this example is for improving performance (line 38).

enter a filename

file exists?

count keywords

keywords

keyword set

count keyword

M30_LIAN1878_11_GE_C30.indd 1172 12/29/17 4:53 PM

30.8 Case Studies 1173

30.8.6 Case Study: Occurrences of Words Listing 21.9 gives a program that counts the occurrences of words in a text. You can rewrite the code using streams as shown in Listing 30.13.

Listing 30.13 CountOccurrenceOfWordsStream.java 1 import java.util.*; 2 import java.util.stream.Collectors; 3 import java.util.stream.Stream; 4 5 public class CountOccurrenceOfWordsStream { 6 public static void main(String[] args) { 7 // Set text in a string 8 String text = "Good morning. Have a good class. " 9 + "Have a good visit. Have fun!"; 10 11 Stream.of(text.split("[\\s+\\p{P}]")).parallel() 12 .filter(e −> e.length() > 0).collect( 13 Collectors.groupingBy(String::toLowerCase, TreeMap::new, 14 Collectors.counting())) 15 .forEach((k, v) −> System.out.println(k + " " + v)); 16 } 17 }

a 2 class 1 fun 1 good 3 have 3 morning 1 visit 1

The text is split into words using a whitespace \s or punctuation \p{P} as a delimiter (line 11) and a stream is created for the words. The filter method is used to select nonempty words (line 12). The collect method groups the words by converting each into lowercase and returns a TreeMap with the lowercase words as the key and their count as the value (lines 13–14). The forEach method displays the key and its value (line 15).

The code in Listing 30.11 is about half in size as the code in Listing 21.9. The new program greatly simplifies coding and improves performance.

30.8.1 Can the following code be used to replace line 19 in Listing 30.7?

DoubleStream.of(numbers).filter(e −> e > DoubleStream.of(numbers).average()).count());

30.8.2 Can the following code be used to replace lines 15–16 in Listing 30.8?

Stream.of(chars).forEach(e −> { int count = 0; System.out.print(e + (++count % 20 == 0 ? "\n" : " ")); });

30.8.3 Show the output of the following code?

String s = "ABC"; Stream.of(s.toCharArray()).forEach(ch −> System.out.println(ch));

Point Check

text

split into words filter empty words group by words

display counts

M30_LIAN1878_11_GE_C30.indd 1173 12/29/17 4:53 PM

1174 Chapter 30 Aggregate Operations for Collection Streams

30.8.4 Show the output of the following code (The toCharacterArray method is presented in Listing 30.9.)

String s = "ABC"; Stream.of(toCharacterArray(s.toCharArray())).forEach(ch −> System.out.println(ch));

30.8.5 Write the code to obtain a one-dimensional array list of strings from a two- dimensional array matrix of strings.

ChapTer summary 1. Java 8 introduces aggregate methods on collection streams to simplify coding and

improve performance.

2. A stream pipeline creates a stream from a data source, consists of zero or more intermediate methods (skip, limit, filter, distinct, sorted, map, and mapToInt), and a final terminal method (count, sum, average, max,min, forEach, findFirst, firstAny, anyMatch, allMatch, noneMatch, and toArray).

3. The execution of a stream is lazy, which means that the methods in the stream are not executed until the final terminal method is initiated.

4. The streams are transient objects. They are destroyed once the terminal method is executed.

5. The Stream<T> class defines the streams for a sequence of objects of the T type. IntStream, LongStream, and DoubleStream are the streams for a sequence of primitive int, long, and double values.

6. An important benefit of using streams is for performance. Streams can be executed in parallel mode to take advantages of multi-core architecture. You can switch a stream into a parallel or sequential mode by invoking the parallel() or sequential() method.

7. You can use the reduce method to reduce a stream into a single value and use the collect method to place the elements in the stream into a collection.

8. You can use the groupingBy collector to group the elements in the stream and apply aggregate methods for the elements in the group.

Quiz Answer the quiz for this chapter online at the book Companion Website.

programming exerCises 30.1 (Assign grades) Rewrite Programming Exercise 7.1 using streams. 30.2 (Count occurrence of numbers) Rewrite Programming Exercise 7.3 using streams. 30.3 (Analyze scores) Rewrite Programming Exercise 7.4 using streams.

M30_LIAN1878_11_GE_C30.indd 1174 12/29/17 4:53 PM

30.4 (Print distinct numbers) Rewrite Programming Exercise 7.5 using streams. Dis- play the numbers in increasing order.

30.5 (Count single digits) Rewrite Programming Exercise 7.7 using streams. 30.6 (Average an array) Rewrite Programming Exercise 7.8 using streams. 30.7 (Find the smallest element) Rewrite Programming Exercise 7.9 using streams. 30.8 (Eliminate duplicates) Rewrite Programming Exercise 7.15 using streams and

sort the elements in the new array in increasing order.

30.9 (Sort students) Rewrite Programming Exercise 7.17 using streams. Define a class named Student with data fields name and score and their getter meth- ods. Store each student in a Student object.

30.10 (Convert binary to decimal) Write a program that prompts the user to enter a binary number in string and displays its decimal value. Use Stream’s reduce method to convert a binary number to decimal.

30.11 (Convert hex to decimal) Write a program that prompts the user to enter a hex number in string and displays its decimal value. Use Stream’s reduce method to convert a hex number to decimal.

30.12 (Sum the digits in an integer) Rewrite Programming Exercise 6.2 using streams. 30.13 (Count the letters in a string) Rewrite Programming Exercise 6.20 using streams. 30.14 (Occurrences of a specified character) Rewrite Programming Exercise 6.23

using streams.

30.15 (Display words in ascending alphabetical order) Rewrite Programming Exer- cise 20.1 using streams.

30.16 (Distinct scores) Use streams to write a program that displays the distinct scores in the scores array in Section 8.8. Display the scores in increasing order, sepa- rated by exactly one space, five numbers per line.

30.17 (Count consonants and vowels) Rewrite Programming Exercise 21.4 using streams.

30.18 (Count the occurrences of words in a text file) Rewrite Programming Exercise 21.8 using streams.

30.19 (Summary information) Suppose the file test.txt contains floating-point numbers separated by spaces. Write a program to obtain the sum, average, maximum, and minimum of the numbers.

Programming Exercises 1175

M30_LIAN1878_11_GE_C30.indd 1175 12/29/17 4:53 PM

A01_GIPE3898_09_SE_FM.indd 3 12/20/17 5:56 PM

This page intentionally left blank

Objectives ■■ To specify styles for UI nodes using JavaFX CSS (§31.2).

■■ To create quadratic curve, cubic curve, and path using the QuadCurve, CubicCurve, and Path classes (§31.3).

■■ To translate, rotate, and scale to perform coordinate transformations for nodes (§31.4).

■■ To define a shape’s border using various types of strokes (§31.5).

■■ To create menus using the Menu, MenuItem, CheckMenuItem, and RadioMemuItem classes (§31.6).

■■ To create context menus using the ContextMenu class (§31.7).

■■ To use SplitPane to create adjustable horizontal and vertical panes (§31.8).

■■ To create tab panes using the TabPane control (§31.9).

■■ To create and display tables using the TableView and TableColumn classes (§31.10).

■■ To create JavaFX user interfaces using FMXL and the visual Scene Builder (§31.11).

Advanced JavaFX and FXML

Chapter

M31_LIAN1878_11_GE_C31.indd 1 12/29/17 5:43 PM

31-2 Chapter 31 Advanced JavaFX and FXML

31.1 Introduction JavaFX can be used to develop comprehensive rich Internet applications.

Chapters 14–16 introduced basics of JavaFX, event-driven programming, animations, and simple UI controls. This chapter introduces some advanced features for developing compre- hensive GUI applications.

31.2 JavaFX CSS JavaFX cascading style sheets can be used to specify styles for UI nodes.

JavaFX cascading style sheets are based on CSS with some extensions. CSS defines the style for webpages. It separates the contents of webpages from its style. JavaFX CSS can be used to define the style for the UI and separates the contents of the UI from the style. You can define the look and feel of the UI in a JavaFX CSS file and use the style sheet to set the color, font, margin, and border of the UI components. A JavaFX CSS file makes it easy to modify the style without modifying the Java source code.

A JavaFX style property is defined with a prefix –fx- to distinquish it from a property in CSS. All the available JavaFX properties are defined in http://docs.oracle.com/javafx/2/api/ javafx/scene/doc-files/cssref.html. Listing 31.1 gives an example of a style sheet.

lisTing 31.1 mystyle.css .plaincircle { -fx-fill: white; -fx-stroke: black; } .circleborder { -fx-stroke-width: 5; -fx-stroke-dash-array: 12 2 4 2; } .border { -fx-border-color: black; -fx-border-width: 5; } #redcircle { -fx-fill: red; -fx-stroke: red; } #greencircle { -fx-fill: green; -fx-stroke: green; }

A style sheet uses the style class or style id to define styles. Multiple style classes can be applied to a single node, and a style id to a unique node. The syntax .styleclass defines a style class. Here, the style classes are named plaincircle, circleborder, and circleborder. The syntax #styleid defines a style id. Here, the style ids are named red- circle and greencircle.

Each node in JavaFX has a styleClass variable of the List<String> type, which can be obtained from invoking getStyleClass(). You can add multiple style classes to a node and only one id to a node. Each node in JavaFX has an id variable of the String type, which can be set using the setID(String id) method. You can set only one id to a node.

The Scene and Parent classes have the stylesheets property, which can be obtained from invoking the getStylesheets() method. This property is of the

Point Key

M31_LIAN1878_11_GE_C31.indd 2 12/29/17 5:43 PM

31.2 JavaFX CSS 31-3

ObservableList<String> type. You can add multiple style sheets into this property. You can load a style sheet into a Scene or a Parent. Note that Parent is the superclass for containers and UI control.

Listing 31.2 gives an example that uses the style sheet defined in Listing 31.1.

Listing 31.2 StyleSheetDemo.java 1 import javafx.application.Application; 2 import javafx.scene.Scene; 3 import javafx.scene.layout.HBox; 4 import javafx.scene.layout.Pane; 5 import javafx.scene.shape.Circle; 6 import javafx.stage.Stage; 7 8 public class StyleSheetDemo extends Application { 9 @Override // Override the start method in the Application class 10 public void start(Stage primaryStage) { 11 HBox hBox = new HBox(5); 12 Scene scene = new Scene(hBox, 300, 250); 13 scene.getStylesheets().add("mystyle.css"); // Load the stylesheet 14 15 Pane pane1 = new Pane(); 16 Circle circle1 = new Circle(50, 50, 30); 17 Circle circle2 = new Circle(150, 50, 30); 18 Circle circle3 = new Circle(100, 100, 30); 19 pane1.getChildren().addAll(circle1, circle2, circle3); 20 pane1.getStyleClass().add("border"); 21 22 circle1.getStyleClass().add("plaincircle"); // Add a style class 23 circle2.getStyleClass().add("plaincircle"); // Add a style class 24 circle3.setId("redcircle"); // Add a style id 25 26 Pane pane2 = new Pane(); 27 Circle circle4 = new Circle(100, 100, 30); 28 circle4.getStyleClass().addAll("circleborder", "plainCircle"); 29 circle4.setId("greencircle"); // Add a style class 30 pane2.getChildren().add(circle4); 31 pane2.getStyleClass().add("border"); 32 33 hBox.getChildren().addAll(pane1, pane2); 34 35 primaryStage.setTitle("StyleSheetDemo"); // Set the window title 36 primaryStage.setScene(scene); // Place the scene in the window 37 primaryStage.show(); // Display the window 38 } 39 }

Figure 31.1 The style sheet is used to style the nodes in the scene.

(a) (b)

M31_LIAN1878_11_GE_C31.indd 3 12/29/17 5:43 PM

31-4 Chapter 31 Advanced JavaFX and FXML

The program loads the style sheet from the file mystyle.css by adding it to the stylesheets property (line 13). The file should be placed in the same directory with the source code for it to run correctly. After the style sheet is loaded, the program sets the style class plaincircle for circle1 and circle2 (lines 22 and 23) and sets the style id redcircle for circle3 (line 24). The program sets style classes circleborder and plaincircle and an id greencircle for circle4 (lines 28 and 29). The style class border is set for both pane1 and pane2 (lines 20 and 31).

The style sheet is set in the scene (line 13). All the nodes inside the scene can use this style sheet. What would happen if line 13 is deleted and the following line is inserted after line 15?

pane1.getStylesheets().add("mystyle.css");

In this case, only pane1 and the nodes inside pane1 can access the style sheet, but pane2 and circle4 cannot use this style sheet. Therefore, everything in pane1 is displayed the same as before the change, and pane2 and circle4 are displayed without applying the style class and id, as shown in Figure 31.1b.

Note the style class plaincircle and id greencircle both are applied to circle4 (lines 28 and 29). plaincircle sets fill to white and greencircle sets fill to green. The property settings in id take precedence over the ones in classes. Thus, circle4 is displayed in green in this program.

31.2.1 How do you load a style sheet to a Scene or a Parent? Can you load multiple style sheets?

31.2.2 If a style sheet is loaded from a node, can the pane and all its containing nodes access the style sheet?

31.2.3 Can a node add multiple style classes? Can a node set multiple style ids? 31.2.4 If the same property is defined in both a style class and a style id and applied to a

node, which one has the precedence?

31.3 QuadCurve, CubicCurve, and Path JavaFX provides the QuadCurve, CubicCurve, and Path classes for creating advanced shapes.

Section 14.11 introduced drawing simple shapes using the Line, Rectangle, Circle, Ellipse, Arc, Polygon, and Polyline classes. This section introduces drawing advanced shapes using the CubicCurve, QuadCurve, and Path classes.

31.3.1 QuadCurve and CubicCurve JavaFX provides the QuadCurve and CubicCurve classes for modeling quadratic curves and cubic curves. A quadratic curve is mathematically defined as a quadratic polynomial. To create a QuadCurve, use its no-arg constructor or the following constructor:

QuadCurve(double startX, double startY, double controlX, double controlY, double endX, double endY)

where (startX, startY) and (endX, endY) specify two endpoints and (controlX, controlY) is a control point. The control point is usually not on the curve instead of defining the trend of the curve, as shown in Figure 31.2a. Figure 31.3 shows the UML diagram for the QuadCurve class.

Point Check

Point Key

M31_LIAN1878_11_GE_C31.indd 4 12/29/17 5:43 PM

31.3 QuadCurve, CubicCurve, and Path 31-5

A cubic curve is mathematically defined as a cubic polynomial. To create a CubicCurve, use its no-arg constructor or the following constructor:

CubicCurve(double startX, double startY, double controlX1, double controlY1, double controlX2, double controlY2, double endX, double endY)

where (startX, startY) and (endX, endY) specify two endpoints and (controlX1, controlY1) and (controlX2, controlY2) are two control points. The control points are usually not on the curve, instead define the trend of the curve, as shown in Figure 31.2b. Figure 31.4 shows the UML diagram for the CubicCurve class.

Figure 31.2 (a) A quadratic curve is specified using three points. (b) A cubic curve is specified using four points.

(controlX, controlY)

(startX, startY) (endX, endY)

(controlX1, controlY1)

(startX, startY) (endX, endY)

(controlX2, controlY2)

(a) (b)

Figure 31.3 QuadCurve defines a quadratic curve.

The getter and setter methods for property values and a getter for prop- erty itself are provided in the class, but omitted in the UML diagram for brevity.javafx.scene.shape.QuadCurve

-startX: DoubleProperty -startY: DoubleProperty -endX: DoubleProperty -endY: DoubleProperty -controlX: DoubleProperty -controlY: DoubleProperty

The x-coordinate of the start point (default 0). The y-coordinate of the start point (default 0). The x-coordinate of the end point (default 0). The y-coordinate of the end point (default 0). The x-coordinate of the control point (default 0). The y-coordinate of the control point (default 0). Creates an empty quad curve. Creates a quad curve with the specified arguments.

+QuadCurve() +QuadCurve(startX: double, startY: double, controlX: double, controlY: double, endX: double, endY: double)

The getter and setter methods for property values and a getter for property itself are provided in the class, but omitted in the UML diagram for brevity.javafx.scene.shape.CubicCurve

-startX: DoubleProperty -startY: DoubleProperty -endX: DoubleProperty -endY: DoubleProperty -controlX1: DoubleProperty -controlY1: DoubleProperty -controlX2: DoubleProperty -controlY2: DoubleProperty

The x-coordinate of the start point (default 0). The y-coordinate of the start point (default 0). The x-coordinate of the end point (default 0). The y-coordinate of the end point (default 0). The x-coordinate of the first control point (default 0). The y-coordinate of the first control point (default 0). The x-coordinate of the second control point (default 0). The y-coordinate of the second control point (default 0). Creates an empty cubic curve. Creates a cubic curve with the specified arguments.

+CubicCurve() +CubicCurve(startX: double, startY: double, controlX1: double, controlY1: double, controlX2: double, controlY2: double, endX: double, endY: double)

Figure 31.4 CubicCurve defines a quadratic curve.

M31_LIAN1878_11_GE_C31.indd 5 12/29/17 5:43 PM

31-6 Chapter 31 Advanced JavaFX and FXML

Listing 31.3 gives a program that demonstrates how to draw quadratic and cubic curves. Figure 31.5a shows a sample run of the program.

Listing 31.3 CurveDemo.java 1 import javafx.application.Application; 2 import javafx.scene.Scene; 3 import javafx.scene.layout.Pane; 4 import javafx.scene.text.Text; 5 import javafx.scene.shape.Circle; 6 import javafx.scene.paint.Color; 7 import javafx.scene.shape.*; 8 import javafx.stage.Stage; 9 10 public class CurveDemo extends Application { 11 @Override // Override the start method in the Application class 12 public void start(Stage primaryStage) { 13 Pane pane = new Pane(); 14 15 // Create a QuadCurve 16 QuadCurve quadCurve = new QuadCurve(10, 80, 40, 20, 150, 56); 17 quadCurve.setFill(Color.WHITE); 18 quadCurve.setStroke(Color.BLACK); 19 20 pane.getChildren().addAll(quadCurve, new Circle(40, 20, 6), 21 new Text(40 + 5, 20 - 5, "Control point")); 22 23 // Create a CubicCurve 24 CubicCurve cubicCurve = new CubicCurve 25 (200, 80, 240, 20, 350, 156, 450, 80); 26 cubicCurve.setFill(Color.WHITE); 27 cubicCurve.setStroke(Color.BLACK); 28 29 pane.getChildren().addAll(cubicCurve, new Circle(240, 20, 6), 30 new Text(240 + 5, 20 − 5, "Control point 1"), 31 new Circle(350, 156, 6), 32 new Text(350 + 5, 156 − 5, "Control point 2")); 33 34 Scene scene = new Scene(pane, 300, 250); 35 primaryStage.setTitle("CurveDemo"); // Set the window title 36 primaryStage.setScene(scene); // Place the scene in the window 37 primaryStage.show(); // Display the window 38 } 39 }

Figure 31.5 You can draw quadratic and cubic curves using QuadCurve and CubicCurve.

(a) (b)

The program creates a QuadCurve with the specified start, control, and end points (line 16) and places the QuadCurve to the pane (line 20). To illustrate the control point, the program also displays the control point as a solid circle (line 21).

M31_LIAN1878_11_GE_C31.indd 6 12/29/17 5:43 PM

31.3 QuadCurve, CubicCurve, and Path 31-7

The program creates a CubicCurve with the specified start, first control, second control, and end points (lines 24 and 25) and places the CubicCurve to the pane (line 29). To illustrate the control points, the program also displays the control points in the pane (lines 29–32).

Note the curves are filled with color. The program sets the color to white and stroke to black in order to display the curves (lines 17 and 18, 26 and 27). If these code lines are removed from the program, the sample run would look like the one in Figure 31.5b.

31.3.2 Path The Path class models an arbitrary geometric path. A path is constructed by adding path elements into the path. The PathElement is the root class for the path elements MoveTo, HLineTo, VLineTo, LineTo, ArcTo, QuadCurveTo, CubicCurveTo, and ClosePath.

You can create a Path using its no-arg constructor. The process of the path construction can be viewed as drawing with a pen. The path does not have a default initial position. You need to set an initial position by adding a MoveTo(startX, startY) path element to the path. Adding a HLineTo(newX) element draws a horizontal line from the current position to the new x-coordinate. Adding a VLineTo(newY) element draws a vertical line from the current position to the new y-coordinate. Adding a LineTo(newX, newY) element draws a line from the current position to the new position. Adding an ArcTo(radiusX, radiusY, xAxisRotation, newX, newY, largeArcFlag, sweepArcFlag) element draws an arc from the previous position to the new position with the specified radius. Adding a QuadCurveTo(controlX, controlY, newX, newY) element draws a quadratic curve from the previous position to the new position with the specified control point. Adding a CubicCurveTo(controlX1, controlY1, controlX2, controlY2, newX, newY) element draws a cubic curve from the previous position to the new position with the specified control points. Adding a ClosePath() element closes the path by drawing a line that connects the starting point with the end point of the path.

Listing 31.4 gives an example that creates a path. A sample run of the program is shown in Figure 31.6.

Listing 31.4 PathDemo.java 1 import javafx.application.Application; 2 import javafx.scene.Scene; 3 import javafx.scene.layout.Pane; 4 import javafx.scene.paint.Color; 5 import javafx.scene.shape.*; 6 import javafx.stage.Stage; 7 8 public class PathDemo extends Application { 9 @Override // Override the start method in the Application class 10 public void start(Stage primaryStage) { 11 Pane pane = new Pane(); 12 13 // Create a Path 14 Path path = new Path(); 15 path.getElements().add(new MoveTo(50.0, 50.0)); 16 path.getElements().add(new HLineTo(150.5)); 17 path.getElements().add(new VLineTo(100.5)); 18 path.getElements().add(new LineTo(200.5, 150.5)); 19 20 ArcTo arcTo = new ArcTo(45, 45, 250, 100.5, 21 false, true); 22 path.getElements().add(arcTo); 23 24 path.getElements().add(new QuadCurveTo(50, 50, 350, 100)); 25 path.getElements().add( 26 new CubicCurveTo(250, 100, 350, 250, 450, 10));

M31_LIAN1878_11_GE_C31.indd 7 12/29/17 5:43 PM

31-8 Chapter 31 Advanced JavaFX and FXML

27 28 path.getElements().add(new ClosePath()); 29 30 pane.getChildren().add(path); 31 path.setFill(null); 32 Scene scene = new Scene(pane, 300, 250); 33 primaryStage.setTitle("PathDemo"); // Set the window title 34 primaryStage.setScene(scene); // Place the scene in the window 35 primaryStage.show(); // Display the window 36 } 37 }

Figure 31.6 You can draw a path by adding path elements.

HLineTo

VLineTo

LineTo ArcTo QuadCurveTo CubicCurveTo

ClosePath

The program creates a Path (line 14), moves its position (line 15), and adds a horizontal line (line 16), a vertical line (line 17), and a line (line 18). The getElements() method returns an ObservableList<PathElement>.

The program creates an ArcTo object (lines 20 and 21). The ArcTo class contains the largeArcFlag and sweepFlag properties. By default, these property values are false. You may set these properties to ture to display a large arc in the opposite direction.

The program adds a quadratic curve (line 24) and a cubic curve (lines 25 and 26) and closes the path (line 28).

By default, the path is not filled. You may change the fill property in the path to specify a color to fill the path.

31.3.1 Create a QuadCurve with starting point (100, 75.5), control point (40, 55.5), and end point (56, 80). Set its fill property to white and stroke to green.

31.3.2 Create CubicCurve object with starting point (100, 75.5), control point 1 (40, 55.5), control point 2 (78.5, 25.5), and end point (56, 80). Set its fill property to white and stroke to green.

31.3.3 Does a path have a default initial position? How do you set a position for a path? 31.3.4 How do you close a path? 31.3.5 How do you display a filled path?

31.4 Coordinate Transformations JavaFX supports coordinate transformations using translation, rotation, and scaling.

You have used the rotate method to rotate a node. You can also perform translations and scaling.

31.4.1 Translations You can use the setTranslateX(double x), setTranslateY(double y), and setTranslateZ(double z) methods in the Node class to translate the coordinates for a

Point Check

Point Key

M31_LIAN1878_11_GE_C31.indd 8 12/29/17 5:43 PM

31.4 Coordinate Transformations 31-9

node. For example, setTranslateX(5) moves the node 5 pixels to the right and setTrans- lateY(−10) 10 pixels up from the previous position. Figure 31.7 shows a rectangle displayed before and after applying translation. After invoking rectangle.setTranslateX(−6) and rectangle.setTranslateY(4), the rectangle is moved 6 pixels to the left and 4 pixels down from the previous position. Note the coordinate transformation using translation, rota- tion, and scaling does not change the contents of the shape being transferred. For example, if a rectangle’s x is 30 and width is 100, after applying transformations to the rectangle, its x is still 30 and width is still 100.

Figure 31.7 After applying translation of (-6, 4), the rectangle is moved by the specified distance relative to the previous position.

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14

3 4 5 6

previous position translation of (– 6, 4) current position

Listing 31.5 TranslationDemo.java 1 import javafx.application.Application; 2 import javafx.scene.Scene; 3 import javafx.scene.layout.Pane; 4 import javafx.scene.paint.Color; 5 import javafx.scene.shape.Rectangle; 6 import javafx.stage.Stage; 7 8 public class TranslationDemo extends Application { 9 @Override // Override the start method in the Application class 10 public void start(Stage primaryStage) { 11 Pane pane = new Pane(); 12 13 double x = 10; 14 double y = 10; 15 java.util.Random random = new java.util.Random(); 16 for (int i = 0; i < 10; i++) { 17 Rectangle rectangle = new Rectangle(10, 10, 50, 60); 18 rectangle.setFill(Color.WHITE); 19 rectangle.setStroke(Color.color(random.nextDouble(), 20 random.nextDouble(), random.nextDouble())); 21 rectangle.setTranslateX(x += 20); 22 rectangle.setTranslateY(y += 5); 23 pane.getChildren().add(rectangle); 24 } 25 26 Scene scene = new Scene(pane, 300, 250); 27 primaryStage.setTitle("TranslationDemo"); // Set the window title 28 primaryStage.setScene(scene); // Place the scene in the window 29 primaryStage.show(); // Display the window 30 } 31 }

M31_LIAN1878_11_GE_C31.indd 9 12/29/17 5:43 PM

31-10 Chapter 31 Advanced JavaFX and FXML

The program repeatedly creates 10 rectangles (line 17). For each rectangle, it sets its fill property to white (line 18) and its stroke property to a random color (lines 19 and 20), and translates it to a new location (lines 21 and 22). The variables x and y are used to set the translateX and translateY properties. These two variable values are changed every time it is applied to a rectangle (see Figure 31.8).

31.4.2 Rotations Rotation was introduced in Chapter 14. This section discusses it in more depth. You can use the rotate(double theta) method in the Node class to rotate a node by theta degrees from its pivot point clockwise, where theta is a double value in degrees. The pivot point is automatically computed based on the bounds of the node. For a circle, ellipse, and a rectangle, the pivot point is the center point of these nodes. For example, rectangle.rotate(45) rotates the rectangle 45 degrees clockwise along the eastern direction from the center, as shown in Figure 31.9.

Figure 31.9 After performing rectangle.rotate(45), the rectangle is rotated in 45 degrees from the center.

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 0 1

3 4

5 6

After rotate

Before rotate

Rectangle.rotate(45)

Figure 31.8 The rectangles are displayed successively in new locations.

Listing 31.6 gives a program that demonstrates the effect of rotation of coordinates. Figure 31.10 shows a sample run of the program.

Listing 31.6 RotateDemo.java 1 import javafx.application.Application; 2 import javafx.scene.Scene; 3 import javafx.scene.layout.Pane; 4 import javafx.scene.paint.Color; 5 import javafx.scene.shape.Rectangle; 6 import javafx.stage.Stage; 7 8 public class RotateDemo extends Application { 9 @Override // Override the start method in the Application class 10 public void start(Stage primaryStage) { 11 Pane pane = new Pane(); 12 java.util.Random random = new java.util.Random();

M31_LIAN1878_11_GE_C31.indd 10 12/29/17 5:43 PM

31.4 Coordinate Transformations 31-11

13 // The radius of the circle for anchoring rectangles 14 double radius = 90; 15 double width = 20; // Width of the rectangle 16 double height = 40; // Height of the rectangle 17 for (int i = 0; i < 8; i++) { 18 // Center of a rectangle 19 double x = 150 + radius * Math.cos(i * 2 * Math.PI / 8); 20 double y = 150 + radius * Math.sin(i * 2 * Math.PI / 8); 21 Rectangle rectangle = new Rectangle( 22 x − width / 2, y − height / 2, width, height); 23 rectangle.setFill(Color.WHITE); 24 rectangle.setStroke(Color.color(random.nextDouble(), 25 random.nextDouble(), random.nextDouble())); 26 rectangle.setRotate(i * 360 / 8); // Rotate the rectangle 27 pane.getChildren().add(rectangle); 28 } 29 30 Scene scene = new Scene(pane, 300, 300); 31 primaryStage.setTitle("RotateDemo"); // Set the window title 32 primaryStage.setScene(scene); // Place the scene in the window 33 primaryStage.show(); // Display the window 34 } 35 }

Figure 31.10 The rotate method rotates a node.

The program creates eight rectangles in a loop (lines 17–28). The center of each rectangle is located on the circle centered as (150, 150) (lines 19 and 20). A rectangle is created by speci- fying its upper left corner position with width and height (lines 21 and 22). The rectangle is rotated in line 26 and added to the pane in line 27.

31.4.3 Scaling You can use the setScaleX(double sx), setScaleY(double sy), and setScaleY(double sy) methods in the Node class to specify a scaling factor. The node will appear larger or smaller depending on the scaling factor. Scaling alters the coordinate space of the node such that each unit of distance along the axis is multiplied by the scale factor. As with rotation transformations, scaling transformations are applied to enlarge or shrink the node around the pivot point. For a node of the rectangle shape, the pivot point is the center of the rectangle. For example, if you apply a scaling factor (x = 2, y = 2), the entire rectangle including the stroke will double in size, growing to the left, right, up, and down from the center, as shown in Figure 31.11.

M31_LIAN1878_11_GE_C31.indd 11 12/29/17 5:43 PM

31-12 Chapter 31 Advanced JavaFX and FXML

Listing 31.7 gives a program that demonstrates the effect of using scaling. Figure 31.12 shows a sample run of the program.

Listing 31.7 ScaleDemo.java 1 import javafx.application.Application; 2 import javafx.scene.Scene; 3 import javafx.scene.layout.Pane; 4 import javafx.scene.shape.Line; 5 import javafx.scene.text.Text; 6 import javafx.scene.shape.Polyline; 7 import javafx.stage.Stage; 8 9 public class ScaleDemo extends Application { 10 @Override // Override the start method in the Application class 11 public void start(Stage primaryStage) { 12 // Create a polyline to draw a sine curve 13 Polyline polyline = new Polyline(); 14 for (double angle = −360; angle <= 360; angle++) { 15 polyline.getPoints().addAll( 16 angle, Math.sin(Math.toRadians(angle))); 17 } 18 polyline.setTranslateY(100); 19 polyline.setTranslateX(200); 20 polyline.setScaleX(0.5); 21 polyline.setScaleY(50); 22 polyline.setStrokeWidth(1.0 / 25); 23 24 // Draw x-axis 25 Line line1 = new Line(10, 100, 420, 100); 26 Line line2 = new Line(420, 100, 400, 90); 27 Line line3 = new Line(420, 100, 400, 110); 28 29 // Draw y-axis 30 Line line4 = new Line(200, 10, 200, 200); 31 Line line5 = new Line(200, 10, 190, 30); 32 Line line6 = new Line(200, 10, 210, 30); 33 34 // Draw x, y axis labels 35 Text text1 = new Text(380, 70, "X"); 36 Text text2 = new Text(220, 20, "Y"); 37 38 // Add nodes to a pane 39 Pane pane = new Pane(); 40 pane.getChildren().addAll(polyline, line1, line2, line3, line4, 41 line5, line6, text1, text2); 42

Figure 31.11 After applying scaling (x = 2, y = 2), the node is doubled in size.

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 0 1

3 4

5 6

new size after applying scaling factor (x = 2, y = 2)

original size

M31_LIAN1878_11_GE_C31.indd 12 12/29/17 5:43 PM

31.5 Strokes 31-13

43 Scene scene = new Scene(pane, 450, 200); 44 primaryStage.setTitle("ScaleDemo"); // Set the window title 45 primaryStage.setScene(scene); // Place the scene in the window 46 primaryStage.show(); // Display the window 47 } 48 }

Figure 31.12 The scale method scales the coordinates in the node.

The program creates a polyline (line 13) and adds the points for a sine curve into the polyline (lines 14–17). Since � sin(x) � 6 = 1, the y-coordinates are too small. To see the sine curve, the program scales the y-coordinates up by 50 times (line 21) and shrinks the x-coordinates by half (line 20).

Note scaling also causes the stroke width to change. To compensate it, the stroke width is purposely set to 1.0 / 25 (line 22).

31.4.1 Can you perform a coordinate transformation on any node? Does a coordinate trans- formation change the contents of a Shape object?

31.4.2 Does the method setTranslateX(6) move the node’s x-coordinate to 6? Does the method setTranslateX(6) move the node’s x-coordinate 6 pixel right from its current location?

31.4.3 Does the method rotate(Math.PI / 2) rotate a node 90 degrees? Does the method rotate(90) rotate a node 90 degrees?

31.4.4 How is the pivot point determined for performing a rotation? 31.4.5 What method do you use to scale a node two times on its x-axis?

31.5 Strokes Stroke defines a shape’s border line style.

JavaFX allows you to specify the attributes of a shape’s boundary using the methods in Figure 31.13.

Point Check

Point Key

The setStroke(paint) method sets a paint for the stroke. The width of the stroke can be specified using the setStrokeWidth(width) method.

Figure 31.13 The Shape class contains the methods for setting stroke properties.

javafx.scene.shape.Shape

+setStroke(paint: Paint): void +setStrokeWidth(width: double): void +setStrokeType(type: StrokeType): void

+setStrokeLineCap(type: StrokeLineCap): void +setStrokeLineJoin(type: StrokeLineJoin): void +getStrokeDashArray(): ObservableList<Double> +setStrokeDashOffset(distance: double): void

Sets a paint for the stroke. Sets a width for the stroke (default 1). Sets a type for the stroke to indicate whether the stroke is placed

inside, centered, or outside of the border (default: CENTERED). Specifies the end cap style for the stroke (default: BUTT). Specifies how two line segments are joined (default: MITER). Returns a list that specifies a dashed pattern for line segments.

Specifies the offset to the first segment in the dashed pattern.

M31_LIAN1878_11_GE_C31.indd 13 12/29/17 5:43 PM

31-14 Chapter 31 Advanced JavaFX and FXML

The setStrokeType(type) method sets a type for the stroke. The type defines whether the stroke is inside, outside, or in the center of the border using the constants StrokeType.INSIDE, StrokeType.OUTSIDE, or StrokeType.CENTERED (default), as shown in Figure 31.14.

Figure 31.15 (a) No decoration for a BUTT line cap. (b) A half circle is added to an unclosed path. (c) A square with half of the stroke width is extended to an unclosed path.

(a) BUTT (b) ROUND (c) SQUARE

Figure 31.16 Path segments can be joined in three ways: (a) MITER, (b) BEVEL, and (c) ROUND.

(a) MITER (b) BEVEL (c) ROUND

Figure 31.17 The numbers in the list specify the opaque and transparent segments of the stroke alternately.

10 20 30 40

… [10.0, 20.0, 30.0, 40.0]

10 20 30 40

… [10.0, 20.0, 30.0, 40.0, 50]

50 10 20 30 40 50

Figure 31.14 (a) No stroke is used. (b) A stroke is placed inside the border. (c) A stroke is placed in the center of the border. (d) A stroke is placed outside of the border.

(a) (b) (c) (d)

Note for the centered style, the stroke is applied by extending the boundary of the node by a distance of half of the strokeWidth on either side (inside and outside) of the boundary.

The setStrokeLineCap(capType) method sets an end cap style for the stroke. The styles are defined as StrokeLineCap.BUTT (default), StrokeLineCap.ROUND, and StrokeLine- Cap.SQUARE, as illustrated in Figure 31.15. The BUTT stroke ends an unclosed path with no added decoration. The ROUND stroke ends an unclosed side of a path with an added half circle whose radius is half of the stroke width. The SQUARE stroke ends an unclosed side of a path with an added square that extends half of the stroke width.

The setStrokeLineJoin method defines the decoration applied where path segments meet. You can specify three types of line join using the constants StrokeLineJoin.MITER (default), StrokeLineJoin.BEVEL, and StrokeLineJoin.ROUND, as shown in Figure 31.16.

The Shape class has a property named strokeDashArray of the ObservableList<Double> type. This property is used to define a dashed pattern for the stroke. Alternate numbers in the list specify the lengths of the opaque and transparent segments of the dashes. For example, the list [10.0, 20.0, 30.0, 40.0] specifies a pattern as shown in Figure 31.17.

M31_LIAN1878_11_GE_C31.indd 14 12/29/17 5:43 PM

31.5 Strokes 31-15

Listing 31.8 gives a program that demonstrates the methods to set attributes for a stroke. Figure 31.19 shows a sample run of the program.

Listing 31.8 StrokeDemo.java 1 import javafx.application.Application; 2 import javafx.scene.Scene; 3 import javafx.scene.layout.Pane; 4 import javafx.scene.paint.Color; 5 import javafx.stage.Stage; 6 import javafx.scene.shape.Rectangle; 7 import javafx.scene.shape.*; 8 9 public class StrokeDemo extends Application { 10 @Override // Override the start method in the Application class 11 public void start(Stage primaryStage) { 12 Rectangle rectangle1 = new Rectangle(20, 20, 70, 120); 13 rectangle1.setFill(Color.WHITE); 14 rectangle1.setStrokeWidth(15); 15 rectangle1.setStroke(Color.ORANGE); 16 17 Rectangle rectangle2 = new Rectangle(20, 20, 70, 120); 18 rectangle2.setFill(Color.WHITE); 19 rectangle2.setStrokeWidth(15); 20 rectangle2.setStroke(Color.ORANGE); 21 rectangle2.setTranslateX(100); 22 rectangle2.setStrokeLineJoin(StrokeLineJoin.BEVEL); 23 24 Rectangle rectangle3 = new Rectangle(20, 20, 70, 120); 25 rectangle3.setFill(Color.WHITE); 26 rectangle3.setStrokeWidth(15); 27 rectangle3.setStroke(Color.ORANGE); 28 rectangle3.setTranslateX(200); 29 rectangle3.setStrokeLineJoin(StrokeLineJoin.ROUND); 30 31 Line line1 = new Line(320, 20, 420, 20); 32 line1.setStrokeLineCap(StrokeLineCap.BUTT); 33 line1.setStrokeWidth(20); 34 35 Line line2 = new Line(320, 70, 420, 70); 36 line2.setStrokeLineCap(StrokeLineCap.ROUND); 37 line2.setStrokeWidth(20); 38 39 Line line3 = new Line(320, 120, 420, 120); 40 line3.setStrokeLineCap(StrokeLineCap.SQUARE); 41 line3.setStrokeWidth(20); 42 43 Line line4 = new Line(460, 20, 560, 120); 44 line4.getStrokeDashArray().addAll(10.0, 20.0, 30.0, 40.0); 45

Figure 31.18 The dash offset specifies on offset for the first segment.

5 20 30 40

… [10.0, 20.0, 30.0, 40.0]

10 20 30 405 Dash offset is 5

The setStrokeDashOffset(distance) method defines the offset to the first segment in the dash pattern. Figure 31.18 illustrates the offset 5 for the dash list [10.0, 20.0, 30.0, 40.0].

M31_LIAN1878_11_GE_C31.indd 15 12/29/17 5:43 PM

31-16 Chapter 31 Advanced JavaFX and FXML

46 Pane pane = new Pane(); 47 pane.getChildren().addAll(rectangle1, rectangle2, rectangle3, 48 line1, line2, line3, line4); 49 50 Scene scene = new Scene(pane, 610, 180); 51 primaryStage.setTitle("StrokeDemo"); // Set the window title 52 primaryStage.setScene(scene); // Place the scene in the window 53 primaryStage.show(); // Display the window 54 } 55 56 // Launch the program from command-line 57 public static void main(String[] args) { 58 launch(args); 59 } 60 }

Figure 31.19 You can specify the attributes for strokes.

miter join bevel join round join square cap

butt cap round cap

dash line

The program creates three rectangles (lines 12–29). Rectangle 1 uses default miter join, rec- tangle 2 uses bevel join (line 22), and rectangle 3 uses round join (line 29).

The program creates three lines with butt, round, and square end cap (lines 31–41). The program creates a line and sets dash pattern for this line (line 44). Note the

strokeDashArray property is of the ObservableList<Double> type. You have to add Double values to the list. Adding a number such as 10 would cause an error.

31.5.1 Are the methods for setting a stroke and its attributes defined in the Node or Shape class?

31.5.2 How do you set a stroke width to 3 pixels? 31.5.3 What are the stroke types? What is the default stroke type? How do you set a

stroke type?

31.5.4 What are the stroke line join types? What is the default stroke line join type? How do you set a stroke line join type?

31.5.5 What are the stroke cap types? What is the default stroke cap type? How do you set a stroke cap type?

31.5.6 How do you specify a dashed pattern for strokes?

31.6 Menus You can create menus in JavaFX.

Menus make selection easier and are widely used in window applications. JavaFX provides five classes that implement menus: MenuBar, Menu, MenuItem, CheckMenuItem, and RadioButtonMenuItem.

Point Check

Point Key

M31_LIAN1878_11_GE_C31.indd 16 12/29/17 5:43 PM

31.6 Menus 31-17

MenuBar is a top-level menu component used to hold the menus. A menu consists of menu items that the user can select (or toggle on or off). A menu item can be an instance of MenuItem, CheckMenuItem, or RadioButtonMenuItem. Menu items can be associated with nodes and keyboard accelerators.

31.6.1 Creating Menus The sequence of implementing menus in JavaFX is as follows:

1. Create a menu bar and add it to a pane. For example, the following code creates a pane and a menu bar, and adds the menu bar to the pane:

MenuBar menuBar = new MenuBar(); Pane pane = new Pane(); pane.getChildren().add(menuBar);

2. Create menus and add them under the menu bar. For example, the following creates two menus and adds them to a menu bar, as shown in Figure 31.20a:

Menu menuFile = new Menu("File"); Menu menuHelp = new Menu("Help"); menuBar.getMenus().addAll(menuFile, menuHelp);

Figure 31.20 (a) The menus are placed under a menu bar. (b) Clicking a menu on the menu bar reveals the items under the menu. (c) Clicking a menu item reveals the submenu items under the menu item.

(a) (b) (c)

3. Create menu items and add them to the menus.

menuFile.getItems().addAll(new MenuItem("New"), new MenuItem("Open"), new MenuItem("Print"), new MenuItem("Exit"));

This code adds the menu items New, Open, Print, and Exit, in this order, to the File menu, as shown in Figure 31.20b.

3.1. Creating submenu items. You can also embed menus inside menus so the embedded menus become submenus. Here is an example:

Menu softwareHelpSubMenu = new Menu("Software"); Menu hardwareHelpSubMenu = new Menu("Hardware"); menuHelp.getItems().add(softwareHelpSubMenu); menuHelp.getItems().add(hardwareHelpSubMenu); softwareHelpSubMenu.getItems().add(new MenuItem("Unix")); softwareHelpSubMenu.getItems().add(new MenuItem("Windows")); softwareHelpSubMenu.getItems().add(new MenuItem("Mac OS"));

M31_LIAN1878_11_GE_C31.indd 17 12/29/17 5:43 PM

31-18 Chapter 31 Advanced JavaFX and FXML

This code adds two submenus, softwareHelpSubMenu and hardwareHelpSubMenu, in MenuHelp. The menu items Unix, NT, and Win95 are added to softwareHelpSub- Menu (see Figure 31.20c).

3.2. Creating check-box menu items. You can also add a CheckMenuItem to a Menu. CheckMenuItem is a subclass of Menu- Item that adds a Boolean state to the MenuItem and displays a check when its state is true. You can click a menu item to turn it on or off. For example, the following statement adds the check-box menu item Check it (see Figure 31.21a).

menuHelp.getItems().add(new CheckMenuItem("Check it"));

Figure 31.21 (a) A check box menu item lets you check or uncheck a menu item just like a check box. (b) You can use RadioMenuItem to choose among mutually exclusive menu choices. (c) You can set image icons and keyboard accelerators in menus.

(c)(a) (b)

3.3. Creating radio menu items. You can also add radio menu items to a menu, using the RadioMenuItem class. This is often useful when you have a group of mutually exclusive choices in the menu. For example, the following statements add a submenu named Color and a set of radio buttons for choosing a color (see Figure 31.21b):

RadioMenuItem rmiBlue, rmiYellow, rmiRed; colorHelpSubMenu.getItems().add(rmiBlue = new RadioMenuItem("Blue")); colorHelpSubMenu.getItems().add(rmiYellow = new RadioMenuItem("Yellow")); colorHelpSubMenu.getItems().add(rmiRed = new RadioMenuItem("Red"));

ToggleGroup group = new ToggleGroup(); rmiBlue.setToggleGroup(group); rmiYellow.setToggleGroup(group); rmiRed.setToggleGroup(group);

4. The menu items generate ActionEvent. To handle ActionEvent, implement the setOn Action method.

5. Image Icons and Keyboard Accelerators

The Menu, CheckMenuItem, and RadioMenuItem are the subclasses of MenuItem. The MenuItem has a graphic property for specifying a node to be displayed in the menu item. Usually, the graphic is an image view. The classes Menu, MenuItem, Check- MenuItem, and RadioMenuItem have another constructor that you can use to specify a graphic. For example, the following code adds an image to the menu, menu item, check menu item, and radio menu item (see Figure 31.21c).

M31_LIAN1878_11_GE_C31.indd 18 12/29/17 5:43 PM

31.6 Menus 31-19

Menu menuFile = new Menu("File", new ImageView("image/usIcon.gif")); MenuItem menuItemOpen = new MenuItem("New", new ImageView("image/new.gif")); CheckMenuItem checkMenuItem = new CheckMenuItem("Check it", new ImageView("image/us.gif")); RadioMenuItem rmiBlue = new RadioMenuItem("Blue", new ImageView("image/us.gif"));

6. A key accelerator lets you select a menu item directly by pressing the CTRL and the accelerator key. For example, by using the following code, you can attach the accelerator key CTRL+N to the Open menu item:

menuItemOpen.setAccelerator( KeyCombination.keyCombination("Ctrl+O"));

31.6.2 Example: Using Menus This section gives an example that creates a user interface to perform arithmetic. The inter- face contains labels and text fields for Number 1, Number 2, and Result. The Result text field displays the result of the arithmetic operation between Number 1 and Number 2. Figure 31.22 contains a sample run of the program.

Figure 31.22 Arithmetic operations can be performed by clicking buttons or by choosing menu items from the Operation menu.

Here are the major steps in the program (Listing 31.9):

1. Create a menu bar and add it into a VBox. Create the menus Operation and Exit, and add them to the menu bar. Add the menu items Add, Subtract, Multiply, and Divide under the Operation menu and add the menu item Close under the Exit menu.

2. Create an HBox to hold labels and text fields and place it into the VBox.

3. Create an HBox to hold the four buttons labeled Add, Subtract, Multiply, and Divide and place it into the VBox.

4. Implement the handlers to process the events from the menu items and the buttons.

Listing 31.9 MenuDemo.java 1 import javafx.application.Application; 2 import javafx.geometry.Pos; 3 import javafx.scene.Scene; 4 import javafx.scene.control.Button; 5 import javafx.scene.control.Label; 6 import javafx.scene.control.Menu; 7 import javafx.scene.control.MenuBar; 8 import javafx.scene.control.MenuItem; 9 import javafx.scene.control.TextField; 10 import javafx.scene.input.KeyCombination;

M31_LIAN1878_11_GE_C31.indd 19 12/29/17 5:44 PM

31-20 Chapter 31 Advanced JavaFX and FXML

11 import javafx.scene.layout.HBox; 12 import javafx.scene.layout.VBox; 13 import javafx.stage.Stage; 14 15 public class MenuDemo extends Application { 16 private TextField tfNumber1 = new TextField(); 17 private TextField tfNumber2 = new TextField(); 18 private TextField tfResult = new TextField(); 19 20 @Override // Override the start method in the Application class 21 public void start(Stage primaryStage) { 22 MenuBar menuBar = new MenuBar(); 23 24 Menu menuOperation = new Menu("Operation"); 25 Menu menuExit = new Menu("Exit"); 26 menuBar.getMenus().addAll(menuOperation, menuExit); 27 28 MenuItem menuItemAdd = new MenuItem("Add"); 29 MenuItem menuItemSubtract = new MenuItem("Subtract"); 30 MenuItem menuItemMultiply = new MenuItem("Multiply"); 31 MenuItem menuItemDivide = new MenuItem("Divide"); 32 menuOperation.getItems().addAll(menuItemAdd, menuItemSubtract, 33 menuItemMultiply, menuItemDivide); 34 35 MenuItem menuItemClose = new MenuItem("Close"); 36 menuExit.getItems().add(menuItemClose); 37 38 menuItemAdd.setAccelerator( 39 KeyCombination.keyCombination("Ctrl+A")); 40 menuItemSubtract.setAccelerator( 41 KeyCombination.keyCombination("Ctrl+S")); 42 menuItemMultiply.setAccelerator( 43 KeyCombination.keyCombination("Ctrl+M")); 44 menuItemDivide.setAccelerator( 45 KeyCombination.keyCombination("Ctrl+D")); 46 47 HBox hBox1 = new HBox(5); 48 tfNumber1.setPrefColumnCount(2); 49 tfNumber2.setPrefColumnCount(2); 50 tfResult.setPrefColumnCount(2); 51 hBox1.getChildren().addAll(new Label("Number 1:"), tfNumber1, 52 new Label("Number 2:"), tfNumber2, new Label("Result:"), 53 tfResult); 54 hBox1.setAlignment(Pos.CENTER); 55 56 HBox hBox2 = new HBox(5); 57 Button btAdd = new Button("Add"); 58 Button btSubtract = new Button("Subtract"); 59 Button btMultiply = new Button("Multiply"); 60 Button btDivide = new Button("Divide"); 61 hBox2.getChildren().addAll(btAdd, btSubtract, btMultiply, btDivide); 62 hBox2.setAlignment(Pos.CENTER); 63 64 VBox vBox = new VBox(10); 65 vBox.getChildren().addAll(menuBar, hBox1, hBox2); 66 Scene scene = new Scene(vBox, 300, 250); 67 primaryStage.setTitle("MenuDemo"); // Set the window title 68 primaryStage.setScene(scene); // Place the scene in the window 69 primaryStage.show(); // Display the window 70

M31_LIAN1878_11_GE_C31.indd 20 12/29/17 5:44 PM

31.6 Menus 31-21

71 // Handle menu actions 72 menuItemAdd.setOnAction(e -> perform('+')); 73 menuItemSubtract.setOnAction(e -> perform('−')); 74 menuItemMultiply.setOnAction(e -> perform('*')); 75 menuItemDivide.setOnAction(e -> perform('/')); 76 menuItemClose.setOnAction(e -> System.exit(0)); 77 78 // Handle button actions 79 btAdd.setOnAction(e -> perform('+')); 80 btSubtract.setOnAction(e -> perform('−')); 81 btMultiply.setOnAction(e -> perform('*')); 82 btDivide.setOnAction(e -> perform('/')); 83 } 84 85 private void perform(char operator) { 86 double number1 = Double.parseDouble(tfNumber1.getText()); 87 double number2 = Double.parseDouble(tfNumber2.getText()); 88 89 double result = 0; 90 switch (operator) { 91 case '+': result = number1 + number2; break; 92 case '−': result = number1 − number2; break; 93 case '*': result = number1 * number2; break; 94 case '/': result = number1 / number2; break; 95 } 96 97 tfResult.setText(result + ""); 98 } 100 }

The program creates a menu bar (line 22), which holds two menus: menuOperation and menuExit (lines 24–36). The menuOperation contains four menu items for doing arithmetic: Add, Subtract, Multiply, and Divide. The menuExit contains the menu item Close for exiting the program. The menu items in the Operation menu are created with keyboard accelerators (lines 38–45).

The labels and text fields are placed in an HBox (lines 47–54) and four buttons are placed in another HBox (lines 56–62). The menu bar and these two HBoxes are added to a VBox (line 65), which is placed in the scene (line 66).

The user enters two numbers in the number fields. When an operation is chosen from the menu, its result, involving two numbers, is displayed in the Result field. The user can also click the buttons to perform the same operation.

The program sets actions for the menu items and buttons in lines 72–82. The private method perform(char operator) (lines 85–98) retrieves operands from the text fields in Number 1 and Number 2, applies the binary operator on the operands, and sets the result in the Result text field.

31.6.1 How do you create a menu bar, menu, menu item, check menu item, and radio menu item?

31.6.2 How do you place a menu into a menu bar? How do you place a menu item, check menu item, and radio menu item into a menu?

31.6.3 Can you place a menu item into another menu item or a check menu or a radio menu item into a menu item?

31.6.4 How do you associate an image with a menu, menu item, check menu item, and radio menu item?

31.6.5 How do you associate an accelerator CTRL+O with a menu item, check menu item, and radio menu item?

Point Check

M31_LIAN1878_11_GE_C31.indd 21 12/29/17 5:44 PM

31-22 Chapter 31 Advanced JavaFX and FXML

31.7 Context Menus You can create context menus in JavaFX.

A context menu, also known as a popup menu, is like a regular menu, but does not have a menu bar and can float anywhere on the screen. Creating a context menu is similar to creating a regular menu. First, you create an instance of ContextMenu, and then you can add MenuItem, CheckMenuItem, and RadioMenuItem to the context menu. For example, the following code creates a ContextMenu, then adds MenuItems into it:

ContextMenu contextMenu = new ContextMenu(); ContextMenu.getItems().add(new MenuItem("New")); ContextMenu.getItems().add(new MenuItem("Open"));

A regular menu is always added to a menu bar, but a context menu is associated with a parent node and is displayed using the show method in the ContextMenu class. You specify the parent node and the location of the context menu, using the coordinate system of the parent like this:

contextMenu.show(node, x, y);

Customarily, you display a context menu by pointing to a GUI component and clicking a certain mouse button, the so-called popup trigger. Popup triggers are system dependent. In Windows, the context menu is displayed when the right mouse button is released. In Motif, the context menu is displayed when the third mouse button is pressed and held down.

Listing 31.10 gives an example that creates a pane. When the mouse points to the pane, clicking a mouse button displays a context menu, as shown in Figure 31.23.

Point Key

Figure 31.23 A context menu is displayed when the popup trigger is issued on the pane.

Here are the major steps in the program (Listing 31.10):

1. Create a context menu using ContextMenu. Create menu items for New, Open, Print, and Exit using MenuItem.

2. Add the menu items into the context menu.

3. Create a pane and place it in the scene.

4. Implement the handler to process the events from the menu items.

5. Implement the mousePressed handler to display the context menu.

Listing 31.10 ContextMenuDemo.java 1 import javafx.application.Application; 2 import javafx.scene.Scene; 3 import javafx.scene.control.ContextMenu; 4 import javafx.scene.control.MenuItem; 5 import javafx.scene.image.ImageView; 6 import javafx.scene.layout.Pane; 7 import javafx.stage.Stage; 8

M31_LIAN1878_11_GE_C31.indd 22 12/29/17 5:44 PM

31.8 SplitPane 31-23

9 public class ContextMenuDemo extends Application { 10 @Override // Override the start method in the Application class 11 public void start(Stage primaryStage) { 12 ContextMenu contextMenu = new ContextMenu(); 13 MenuItem menuItemNew = new MenuItem("New", 14 new ImageView("image/new.gif")); 15 MenuItem menuItemOpen = new MenuItem("Open", 16 new ImageView("image/open.gif")); 17 MenuItem menuItemPrint = new MenuItem("Print", 18 new ImageView("image/print.gif")); 19 MenuItem menuItemExit = new MenuItem("Exit"); 20 contextMenu.getItems().addAll(menuItemNew, menuItemOpen, 21 menuItemPrint, menuItemExit); 22 23 Pane pane = new Pane(); 24 Scene scene = new Scene(pane, 300, 250); 25 primaryStage.setTitle("ContextMenuDemo"); // Set the window title 26 primaryStage.setScene(scene); // Place the scene in the window 27 primaryStage.show(); // Display the window 28 29 pane.setOnMousePressed( 30 e -> contextMenu.show(pane, e.getScreenX(), e.getScreenY())); 31 32 menuItemNew.setOnAction(e -> System.out.println("New")); 33 menuItemOpen.setOnAction(e -> System.out.println("Open")); 34 menuItemPrint.setOnAction(e -> System.out.println("Print")); 35 menuItemExit.setOnAction(e -> System.exit(0)); 36 } 37 }

The process of creating context menus is similar to the process for creating regular menus. To create a context menu, create a ContextMenu as the basis (line 12) and add MenuItems to it (lines 13–21).

To show a context menu, use the show method by specifying the parent node and the loca- tion for the context menu (lines 29 and 30). The show method is invoked when the context menu is triggered by a mouse click on the pane (line 30).

31.7.1 How do you create a context menu? How do you add menu items, check menu items, and radio menu items into a context menu?

31.7.2 How do you show a context menu?

31.8 SplitPane The SplitPane class can be used to display multiple panes and allow the user to adjust the size of the panes.

The SplitPane is a control that contains two components with a separate bar known as a divider, as shown in Figure 31.24.

Point Check

Point Key

Figure 31.24 SplitPane divides a container into two parts.

(a) Horizontal orientation (b) Vertical orientation

M31_LIAN1878_11_GE_C31.indd 23 12/29/17 5:44 PM

31-24 Chapter 31 Advanced JavaFX and FXML

Listing 31.11 gives an example that uses radio buttons to let the user select a country and displays the country’s flag and description in separate sides, as shown in Figure 31.26. The description of the currently selected layout manager is displayed in a text area. The radio but- tons, buttons, and text area are placed in two split panes.

Listing 31.11 SplitPaneDemo.java 1 import javafx.application.Application; 2 import javafx.geometry.Orientation; 3 import javafx.scene.Scene; 4 import javafx.scene.control.RadioButton; 5 import javafx.scene.control.ScrollPane; 6 import javafx.scene.control.SplitPane; 7 import javafx.scene.control.TextArea; 8 import javafx.scene.control.ToggleGroup; 9 import javafx.scene.image.Image; 10 import javafx.scene.image.ImageView; 11 import javafx.scene.layout.StackPane; 12 import javafx.scene.layout.VBox; 13 import javafx.stage.Stage; 14 15 public class SplitPaneDemo extends Application { 16 private Image usImage = new Image( 17 "http://www.cs.armstrong.edu/liang/common/image/us.gif"); 18 private Image ukImage = new Image( 19 "http://www.cs.armstrong.edu/liang/common/image/uk.gif"); 20 private Image caImage = new Image( 21 "http://www.cs.armstrong.edu/liang/common/image/ca.gif"); 22 private String usDescription = "Description for US ..."; 23 private String ukDescription = "Description for UK ..."; 24 private String caDescription = "Description for CA ..."; 25 26 @Override // Override the start method in the Application class 27 public void start(Stage primaryStage) { 28 VBox vBox = new VBox(10); 29 RadioButton rbUS = new RadioButton("US");

Figure 31.25 SplitPane provides methods to specify the properties of a split pane and for manipulating the components in a split pane.

The getter and setter methods for property values and a getter for property itself are provided in the class, but omitted in the UML diagram for brevity.

javafx.scene.control.Control

javax.scene.control.SplitPane

−orientation: ObjectProperty<Orientation> Specifies the orientation of the pane.

Constructs a default split pane with horizontal orientation. Returns a list of items in the pane.

+SplitPane() +getItems(): ObservableList<Node>

The two sides separated by the divider can appear in horizontal or vertical orientation. The divider separating two sides can be dragged to change the amount of space occupied by each side. Figure 31.25 shows the frequently used properties, constructors, and methods in SplitPane.

M31_LIAN1878_11_GE_C31.indd 24 12/29/17 5:44 PM

31.8 SplitPane 31-25

30 RadioButton rbUK = new RadioButton("UK"); 31 RadioButton rbCA = new RadioButton("CA"); 32 vBox.getChildren().addAll(rbUS, rbUK, rbCA); 33 34 SplitPane content = new SplitPane(); 35 content.setOrientation(Orientation.VERTICAL); 36 ImageView imageView = new ImageView(usImage); 37 StackPane imagePane = new StackPane(); 38 imagePane.getChildren().add(imageView); 39 TextArea taDescription = new TextArea(); 40 taDescription.setText(usDescription); 41 content.getItems().addAll( 42 imagePane, new ScrollPane(taDescription)); 43 44 SplitPane sp = new SplitPane(); 45 sp.getItems().addAll(vBox, content); 46 47 Scene scene = new Scene(sp, 300, 250); 48 primaryStage.setTitle("SplitPaneDemo"); // Set the window title 49 primaryStage.setScene(scene); // Place the scene in the window 50 primaryStage.show(); // Display the window 51 52 // Group radio buttons 53 ToggleGroup group = new ToggleGroup(); 54 rbUS.setToggleGroup(group); 55 rbUK.setToggleGroup(group); 56 rbCA.setToggleGroup(group); 57 58 rbUS.setSelected(true); 59 rbUS.setOnAction(e -> { 60 imageView.setImage(usImage); 61 taDescription.setText(usDescription); 62 }); 63 64 rbUK.setOnAction(e -> { 65 imageView.setImage(ukImage); 66 taDescription.setText(ukDescription); 67 }); 68 69 rbCA.setOnAction(e -> { 70 imageView.setImage(caImage); 71 taDescription.setText(caDescription); 72 }); 73 } 74 }

Figure 31.26 You can adjust the component size in the split panes.

M31_LIAN1878_11_GE_C31.indd 25 12/29/17 5:44 PM

31-26 Chapter 31 Advanced JavaFX and FXML

The program places three radio buttons in a VBox (lines 28–32) and creates a vertical split pane for holding an image view and a text area (lines 34–42). Split panes can be embedded. The program creates a horizontal split pane and places the VBox and the vertical split pane into it (lines 44 and 45).

Adding a split pane to an existing split pane results in three split panes. The program creates two split panes (lines 34, 42) to hold a panel for radio buttons, a panel for buttons, and a scroll pane.

The program groups radio buttons (lines 53–56) and processes the action for radio buttons (lines 59–72).

31.8.1 How do you create a horizontal SplitPane? How do you create a vertical SplitPane?

31.8.2 How do you add items into a SplitPane? Can you add a SplitPane to another SplitPane?

31.9 TabPane The TabPane class can be used to display multiple panes with tabs.

TabPane is a useful control that provides a set of mutually exclusive tabs, as shown in Figure 31.27. You can switch between a group of tabs. Only one tab is visible at a time. A Tab can be added to a TabPane. Tabs in a TabPane can be placed in the position top, left, bottom, or right.

Point Check

Point Key

Figure 31.27 TabPane holds a group of tabs.

Each tab represents a single page. Tabs are defined in the Tab class. Tabs can contain any Node such as a pane, a shape, or a control. A tab can contain another pane. Therefore, you can create a multilayered tab pane. Figures 31.28 and 31.29 show the frequently used properties, constructors, and methods in TabPane and Tab.

Figure 31.28 TabPane displays and manages the tabs.

The getter and setter methods for property values and a getter for property itself are provided in the class, but omitted in the UML diagram for brevity.

javafx.scene.control.Control

javafx.scene.control.TabPane

-side: ObjectProperty<Side> The position of the tab in the tab pane. Possible values are: Side.TOP, Side.BOTTOM, Side.LEFT, and Side.RIGHT (default: Side.TOP).

Creates a default Tab Pane. Returns a list of tabs in this Tab Pane.

+TabPane() +getTabs(): ObservableList<Tab>

M31_LIAN1878_11_GE_C31.indd 26 12/29/17 5:44 PM

31.9 TabPane 31-27

Listing 31.12 gives an example that uses a tab pane with four tabs to display four types of fig- ures: line, rectangle, rounded rectangle, and oval. You can select a figure to display by clicking the corresponding tab, as shown in Figure 31.27.

Listing 31.12 TabPaneDemo.java 1 import javafx.application.Application; 2 import javafx.scene.Scene; 3 import javafx.scene.control.Tab; 4 import javafx.scene.control.TabPane; 5 import javafx.scene.layout.StackPane; 6 import javafx.scene.shape.Circle; 7 import javafx.scene.shape.Ellipse; 8 import javafx.scene.shape.Line; 9 import javafx.scene.shape.Rectangle; 10 import javafx.stage.Stage; 11 12 public class TabPaneDemo extends Application { 13 @Override // Override the start method in the Application class 14 public void start(Stage primaryStage) { 15 TabPane tabPane = new TabPane(); 16 Tab tab1 = new Tab("Line"); 17 StackPane pane1 = new StackPane(); 18 pane1.getChildren().add(new Line(10, 10, 80, 80)); 19 tab1.setContent(pane1); 20 Tab tab2 = new Tab("Rectangle"); 21 tab2.setContent(new Rectangle(10, 10, 200, 200)); 22 Tab tab3 = new Tab("Circle"); 23 tab3.setContent(new Circle(50, 50, 20)); 24 Tab tab4 = new Tab("Ellipse"); 25 tab4.setContent(new Ellipse(10, 10, 100, 80)); 26 tabPane.getTabs().addAll(tab1, tab2, tab3, tab4); 27 28 Scene scene = new Scene(tabPane, 300, 250); 29 primaryStage.setTitle("DisplayFigure"); // Set the window title 30 primaryStage.setScene(scene); // Place the scene in the window 31 primaryStage.show(); // Display the window 32 } 33 }

Figure 31.29 Tab contains a node.

The getter and setter methods for property values and a getter for property itself are provided in the class, but omitted in the UML diagram for brevity.

java.lang.Object

javafx.scene.control.Tab

-content: ObjectProperty<Node>

-contextMenu:

ObjectProperty<ContextMenu>

-graphics: ObjectProperty<Node>

-id: StringProperty

-text: StringProperty

-tooltip: StringProperty

The content associated with the tab.

The context menu associated with the tab.

The graphics in the tab.

The id for the tab.

The text shown in the tab.

The tooltip associated with the tab.

Constructs a default tab.

Constructs a tab with the specified string.

+Tab()

+Tab(text: String)

M31_LIAN1878_11_GE_C31.indd 27 12/29/17 5:44 PM

31-28 Chapter 31 Advanced JavaFX and FXML

The program creates a tab pane (line 15) and four tabs (lines 16, 20, 22, and 24). A stack pane is created to hold a line (line 18) and placed into tab1 (line 19). A rectangle, circle, and oval are created and placed into tab2, tab3, and tab4. Note the line is centered in tab1 because it is placed in a stack pane. The other shapes are directly placed into the tab. They are displayed at the upper left corner of the tab.

By default, the tabs are placed at the top of the tab pane. You can use the setSide method to change its location.

31.9.1 How do you create a tab pane? How do you create a tab? How do you add a tab to a tab pane?

31.9.2 How do you place the tabs on the left of the tab pane? 31.9.3 Can a tab have a text as well as an image? Write the code to set an image for tab1

in Listing 31.12.

31.10 TableView You can display tables using the TableView class.

TableView is a control that displays data in rows and columns in a two-dimensional grid, as shown in Figure 31.30.

Point Check

Point Key

Figure 31.30 TableView displays data in a table.

TableView, TableColumn, and TableCell are used to display and manipulate a table. TableView displays a table. TableColumn defines the columns in a table. TableCell rep- resents a cell in the table. Creating a TableView is a multistep process. First, you need to create an instance of TableView and associate data with the TableView. Second, you need to create columns using the TableColumn class and set a column cell value factory to specify how to populate all cells within a single TableColumn.

Listing 31.13 gives a simple example to demonstrate using TableView and TableColumn. A sample run of the program is shown in Figure 31.31.

Listing 31.13 TableViewDemo.java 1 import javafx.application.Application; 2 import javafx.beans.property.SimpleBooleanProperty; 3 import javafx.beans.property.SimpleDoubleProperty; 4 import javafx.beans.property.SimpleStringProperty; 5 import javafx.collections.FXCollections; 6 import javafx.collections.ObservableList; 7 import javafx.scene.Scene; 8 import javafx.scene.control.TableColumn; 9 import javafx.scene.control.TableView; 10 import javafx.scene.control.cell.PropertyValueFactory; 11 import javafx.scene.layout.Pane; 12 import javafx.stage.Stage;

M31_LIAN1878_11_GE_C31.indd 28 12/29/17 5:44 PM

31.10 TableView 31-29

13 14 public class TableViewDemo extends Application { 15 @Override // Override the start method in the Application class 16 public void start(Stage primaryStage) { 17 TableView<Country> tableView = new TableView<>(); 18 ObservableList<Country> data = 19 FXCollections.observableArrayList( 20 new Country("USA", "Washington DC", 280, true), 21 new Country("Canada", "Ottawa", 32, true), 22 new Country("United Kingdom", "London", 60, true), 23 new Country("Germany", "Berlin", 83, true), 24 new Country("France", "Paris", 60, true)); 25 tableView.setItems(data); 26 27 TableColumn countryColumn = new TableColumn("Country"); 28 countryColumn.setMinWidth(100); 29 countryColumn.setCellValueFactory( 30 new PropertyValueFactory<Country, String>("country")); 31 32 TableColumn capitalColumn = new TableColumn("Capital"); 33 capitalColumn.setMinWidth(100); 34 capitalColumn.setCellValueFactory( 35 new PropertyValueFactory<Country, String>("capital")); 36 37 TableColumn populationColumn = 38 new TableColumn("Population (million)"); 39 populationColumn.setMinWidth(200); 40 populationColumn.setCellValueFactory( 41 new PropertyValueFactory<Country, Double>("population")); 42 43 TableColumn democraticColumn = 44 new TableColumn("Is Democratic?"); 45 democraticColumn.setMinWidth(200); 46 democraticColumn.setCellValueFactory( 47 new PropertyValueFactory<Country, Boolean>("democratic")); 48 49 tableView.getColumns().addAll(countryColumn, capitalColumn, 50 populationColumn, democraticColumn); 51 52 Pane pane = new Pane(); 53 pane.getChildren().add(tableView); 54 Scene scene = new Scene(pane, 300, 250); 55 primaryStage.setTitle("TableViewDemo"); // Set the window title 56 primaryStage.setScene(scene); // Place the scene in t he window 57 primaryStage.show(); // Display the window 58 } 59 60 public static class Country { 61 private final SimpleStringProperty country; 62 private final SimpleStringProperty capital; 63 private final SimpleDoubleProperty population; 64 private final SimpleBooleanProperty democratic; 65 66 private Country(String country, String capital, 67 double population, boolean democratic) { 68 this.country = new SimpleStringProperty(country); 69 this.capital = new SimpleStringProperty(capital); 70 this.population = new SimpleDoubleProperty(population); 71 this.democratic = new SimpleBooleanProperty(democratic); 72 }

M31_LIAN1878_11_GE_C31.indd 29 12/29/17 5:44 PM

31-30 Chapter 31 Advanced JavaFX and FXML

73 74 public String getCountry() { 75 return country.get(); 76 } 77 78 public void setCountry(String country) { 79 this.country.set(country); 80 } 81 82 public String getCapital() { 83 return capital.get(); 84 } 85 86 public void setCapital(String capital) { 87 this.capital.set(capital); 88 } 89 90 public double getPopulation() { 91 return population.get(); 92 } 93 94 public void setPopulation(double population) { 95 this.population.set(population); 96 } 97 98 public boolean isDemocratic() { 99 return democratic.get(); 100 } 101 102 public void setDemocratic(boolean democratic) { 103 this.democratic.set(democratic); 104 } 105 } 106 }

The program creates a TableView (line 17). The TableView class is a generic class whose concrete type is Country. Therefore, this TableView is for displaying Country. The table data is an ObservableList<Country>. The program creates the list (lines 18–24) and associates the list with the TableView (line 25).

The program creates a TableColumn for each column in the table (lines 27–47). A Prop- ertyValueFactory object is created and set for each column (line 30). This object is used to populate the data in the column. The PropertyValueFactory<S, T> class is a generic class. S is for the class displayed in the TableView and T is the class for the values in the column. The PropertyValueFactory object associates a property in class S with a column.

When you create a table in a JavaFX application, it is a best practice to define the data model in a class. The Country class defines the data for TableView. Each property in the class defines a column in the table. This property should be defined as binding property with the getter and setter methods for the value.

The program adds the columns into the TableView (lines 49 and 50), adds the TableView in a pane (line 53), and places the pane in the scene (line 54). Note line 31 can be simplified using the following code:

new PropertyValueFactory<>("country");

From this example, you see how to display data in a table using the TableView and TableCol- umn classes. The frequently used properties and methods for the TableView and TableColumn classes are given in Figures 31.32 and 31.33.

M31_LIAN1878_11_GE_C31.indd 30 12/29/17 5:44 PM

31.10 TableView 31-31

Figure 31.33 You can add subcolumns in a column.

You can create nested columns. For example, the following code creates two subcolumns under Location, as shown in Figures 31.33.

TableColumn locationColumn = new TableColumn("Location"); locationColumn.getColumns().addAll(new TableColumn("latitude"), new TableColumn("longitude"));

Figure 31.31 TableView displays a table.

The getter and setter methods for property values and a getter for property itself are provided in the class, but omitted in the UML diagram for brevity.

javafx.scene.control.Control

javafx.scene.control.TableView<S>

-editable: BooleanProperty

-items: ObjectProperty<ObservableList<S>> -placeholder: ObjectProperty<Node> -selectionModel: ObjectProperty< TableViewSelectionModel<S>>

Specifies whether this TableView is editable. For a cell to be editable, TableView, TableColumn, and TableCell for the cell should all be true.

The data model for the TableView.

This Node is shown when table has no contents. Specifies single or multiple selections.

Creates a default TableView with no content.

Creates a default TableView with the specified content.

+TableView()

+TableView(items: ObservableList<S>)

Figure 31.32 TableColumn defines a column in the TableView.

The getter and setter methods for property values and a getter for property itself are provided in the class, but omitted in the UML diagram for brevity.

java.lang.Object

javafx.scene.control.TableColumn<S,T>

-editable: BooleanProperty

-cellValueFactory:

ObjectProperty<Callback<TableColumn.

CellDataFeatures<S,T>,ObservableValue

<T>>>

-graphic: ObjectProperty<Node>

-id: StringProperty

-resizable: BooleanProperty

-sortable: BooleanProperty

-text: StringProperty

-style: StringProperty

-visible: BooleanProperty

Specifies whether this TableColumn allows editing.

The cell value factory to specify how to populate all cells within a single column.

The graphic for this TableColumn.

The id for this TableColumn.

Indicates whether the column is resizable.

Indicates whether the column is sortable.

The text in the table column header.

Specifies the CSS style for the column.

Specifies whether the column is visible (default: true).

Creates a default TableColumn.

Creates a TableView with the specified header text.

+TableColumn()

+TableColumn(text: String)

M31_LIAN1878_11_GE_C31.indd 31 12/29/17 5:44 PM

31-32 Chapter 31 Advanced JavaFX and FXML

The TableView data model is an observable list. When data is changed, the change is auto- matically shown in the table. Listing 31.14 gives an example that lets the user add new rows to the table.

Listing 31.14 AddNewRowDemo.java 1 import javafx.application.Application; 2 import javafx.beans.property.SimpleBooleanProperty; 3 import javafx.beans.property.SimpleDoubleProperty; 4 import javafx.beans.property.SimpleStringProperty; 5 import javafx.collections.FXCollections; 6 import javafx.collections.ObservableList; 7 import javafx.scene.Scene; 8 import javafx.scene.control.Button; 9 import javafx.scene.control.CheckBox; 10 import javafx.scene.control.Label; 11 import javafx.scene.control.TableColumn; 12 import javafx.scene.control.TableView; 13 import javafx.scene.control.TextField; 14 import javafx.scene.control.cell.PropertyValueFactory; 15 import javafx.scene.layout.BorderPane; 16 import javafx.scene.layout.FlowPane; 17 import javafx.stage.Stage; 18 19 public class AddNewRowDemo extends Application { 20 @Override // Override the start method in the Application class 21 public void start(Stage primaryStage) { 22 TableView<Country> tableView = new TableView<>(); 23 ObservableList<Country> data = 24 FXCollections.observableArrayList( 25 new Country("USA", "Washington DC", 280, true), 26 new Country("Canada", "Ottawa", 32, true), 27 new Country("United Kingdom", "London", 60, true), 28 new Country("Germany", "Berlin", 83, true), 29 new Country("France", "Paris", 60, true)); 30 tableView.setItems(data); 31 32 TableColumn countryColumn = new TableColumn("Country"); 33 countryColumn.setMinWidth(100); 34 countryColumn.setCellValueFactory( 35 new PropertyValueFactory<Country, String>("country")); 36 37 TableColumn capitalColumn = new TableColumn("Capital"); 38 capitalColumn.setMinWidth(100); 39 capitalColumn.setCellValueFactory( 40 new PropertyValueFactory<Country, String>("capital")); 41 42 TableColumn populationColumn = 43 new TableColumn("Population (million)"); 44 populationColumn.setMinWidth(100); 45 populationColumn.setCellValueFactory( 46 new PropertyValueFactory<Country, Double>("population")); 47 48 TableColumn democraticColumn = 49 new TableColumn("Is Democratic?"); 50 democraticColumn.setMinWidth(100); 51 democraticColumn.setCellValueFactory( 52 new PropertyValueFactory<Country, Boolean>("democratic")); 53 54 tableView.getColumns().addAll(countryColumn, capitalColumn,

M31_LIAN1878_11_GE_C31.indd 32 12/29/17 5:44 PM

31.10 TableView 31-33

55 populationColumn, democraticColumn); 56 57 FlowPane flowPane = new FlowPane(3, 3); 58 TextField tfCountry = new TextField(); 59 TextField tfCapital = new TextField(); 60 TextField tfPopulation = new TextField(); 61 CheckBox chkDemocratic = new CheckBox("Is democratic?"); 62 Button btAddRow = new Button("Add new row"); 63 tfCountry.setPrefColumnCount(5); 64 tfCapital.setPrefColumnCount(5); 65 tfPopulation.setPrefColumnCount(5); 66 flowPane.getChildren().addAll(new Label("Country: "), 67 tfCountry, new Label("Capital"), tfCapital, 68 new Label("Population"), tfPopulation, chkDemocratic, 69 btAddRow); 70 71 btAddRow.setOnAction(e -> { 72 data.add(new Country(tfCountry.getText(), tfCapital.getText(), 73 Double.parseDouble(tfPopulation.getText()), 74 chkDemocratic.isSelected())); 75 tfCountry.clear(); 76 tfCapital.clear(); 77 tfPopulation.clear(); 78 }); 79 80 BorderPane pane = new BorderPane(); 81 pane.setCenter(tableView); 82 pane.setBottom(flowPane); 83 84 Scene scene = new Scene(pane, 500, 250); 85 primaryStage.setTitle("AddNewRowDemo"); // Set the window title 86 primaryStage.setScene(scene); // Place the scene in the window 87 primaryStage.show(); // Display the window 88 } 89 90 public static class Country { 91 private final SimpleStringProperty country; 92 private final SimpleStringProperty capital; 93 private final SimpleDoubleProperty population; 94 private final SimpleBooleanProperty democratic; 95 96 private Country(String country, String capital, 97 double population, boolean democratic) { 98 this.country = new SimpleStringProperty(country); 99 this.capital = new SimpleStringProperty(capital); 100 this.population = new SimpleDoubleProperty(population); 101 this.democratic = new SimpleBooleanProperty(democratic); 102 } 103 104 public String getCountry() { 105 return country.get(); 106 } 107 108 public void setCountry(String country) { 109 this.country.set(country); 110 } 111 112 public String getCapital() { 113 return capital.get(); 114 }

M31_LIAN1878_11_GE_C31.indd 33 12/29/17 5:44 PM

31-34 Chapter 31 Advanced JavaFX and FXML

115 116 public void setCapital(String capital) { 117 this.capital.set(capital); 118 } 119 120 public double getPopulation() { 121 return population.get(); 122 } 123 124 public void setPopulation(double population) { 125 this.population.set(population); 126 } 127 128 public boolean isDemocratic() { 129 return democratic.get(); 130 } 131 132 public void setDemocratic(boolean democratic) { 133 this.democratic.set(democratic); 134 } 135 } 136 }

The program is the same in Listing 31.13 except that a new code is added to let the user enter a new row (lines 57–82). The user enters the new row from the text fields and a check box and presses the Add New Row button to add a new row to the data. Since data is an observable list, the change in data is automatically updated in the table.

As shown in Figure 31.34a, a new country information is entered in the text fields. After clicking the Add New Row button, the new country is displayed in the table view.

Figure 31.34 Change in the table data model is automatically displayed in the table view.

(a)

(b)

M31_LIAN1878_11_GE_C31.indd 34 12/29/17 5:44 PM

31.11 Developing JavaFX Programs Using FXML 31-35

TableView not only displays data, but also allows data to be edited. To enable data editing in the table, write the code as follows:

1. Set the TableView’s editable to true.

2. Set the column’s cell factory to a text field table cell.

3. Implement the column’s setOnEditCommit method to assign the edited value to the data model.

Here is the example of enabling editing for the countryColumn.

tableView.setEditable(true); countryColumn.setCellFactory(TextFieldTableCell.forTableColumn()); countryColumn.setOnEditCommit( new EventHandler<CellEditEvent<Country, String>>() { @Override public void handle(CellEditEvent<Country, String> t) { t.getTableView().getItems().get( t.getTablePosition().getRow()) .setCountry(t.getNewValue()); } } );

31.10.1 How do you create a table view? How do you create a table column? How do you add a table column to a table view?

31.10.2 What is the data type for a TableView’s data model? How do you associate a data model with a TableView?

31.10.3 How do you set a cell value factory for a TableColumn? 31.10.4 How do you set an image in a table column header?

31.11 Developing JavaFX Programs Using FXML You can create JavaFX user interfaces using FXML scripts.

There are two ways to develop JavaFX applications. One way is to write everything in Java source code as you have done so far. The other way is to use FXML. FXML is an XML-based script language for describing the user interface. Using FXML enables you to separate user interface from the logic of the Java code. JavaFX Scene Builder is a visual design tool for creating the user interface without manually writing the FXML script. You drag and drop the UI components to the content pane and set properties for the components in the Inspector. The Scene Builder automatically generates the FXML scripts for the user interface. This section demonstrates how to use the Scene Builder to create JavaFX applications.

NOTE It is important that you first learn how to write the JavaFX code without using FXML to grasp the fundamentals of JavaFX programming before learning FXML. Once you understand the basics of JavaFX, it is easy to create JavaFX programs using FXML. For this reason, FXML is introduced after you have learned the basics of JavaFX programming.

31.11.1 Installing JavaFX Scene Builder You can use the JavaFX Scene Builder standalone or with an IDE such as NetBeans or Eclipse. This section demonstrates using the JavaFX Scene Builder with NetBeans. You can download the latest version of Scene Builder from http://gluonhq.com/open-source/scene-builder/.

Point Check

Point Key

M31_LIAN1878_11_GE_C31.indd 35 12/29/17 5:44 PM

31-36 Chapter 31 Advanced JavaFX and FXML

31.11.2 Creating a JavaFX FXML Project To use JavaFX FXML, you need to create a JavaFX FXML in NetBeans. Here are the steps of creating a JavaFX FXML project:

1. Choose File, New Project to display the New Project dialog box, as shown in Figure 31.35.

2. Choose JavaFX in the Categories and JavaFX FXML Application in the Projects. Click Next to display the New JavaFX Application dialog box, as shown in Figure 31.36.

3. Enter Calculator as the project name and click Finish to create the project. You will see the project created as shown in Figure 31.37.

Three files, Calculator.java, FXMLDocument.fxml, and FXMLDocumentController. java, are created in the project. Their source codes are shown in Listings 31.15, 31.16, and 31.17. From the perspective of the MVC architecture, these three files correspond to model, view, and controller. You can define data model in the Calculator.java class. The .fxml file describes the user interface. The controller file defines the actions for processing the events for the user interface.

Listing 31.15 Calculator.java 1 package calculator; 2 3 import javafx.application.Application; 4 import javafx.fxml.FXMLLoader; 5 import javafx.scene.Parent; 6 import javafx.scene.Scene; 7 import javafx.stage.Stage; 8 9 public class Calculator extends Application { 10 @Override 11 public void start(Stage stage) throws Exception { 12 Parent root = 13 FXMLLoader.load(getClass().getResource("FXMLDocument.fxml")); 14 Scene scene = new Scene(root); 15 stage.setScene(scene); 16 stage.show(); 17 } 18 19 /** 20 * @param args the command line arguments 21 */ 22 public static void main(String[] args) { 23 launch(args); 24 } 25 }

Listing 31.16 FXMLDocument.fxml <?xml version="1.0" encoding="UTF-8"?>

<?import java.lang.*?> <?import java.util.*?> <?import javafx.scene.*?> <?import javafx.scene.control.*?> <?import javafx.scene.layout.*?>

<AnchorPane id="AnchorPane" prefHeight="200" prefWidth="320" xmlns:fx="http://javafx.com/fxml/1"

M31_LIAN1878_11_GE_C31.indd 36 12/29/17 5:44 PM

31.11 Developing JavaFX Programs Using FXML 31-37

fx:controller="calculator.FXMLDocumentController"> <children> <Button layoutX="126" layoutY="90" text="Click Me!" onAction="#handleButtonAction" fx:id="button" /> <Label layoutX="126" layoutY="120" minHeight="16" minWidth="69" fx:id="label" /> </children> </AnchorPane>

Listing 31.17 FXMLDocumentController.java package calculator;

import java.net.URL; import java.util.ResourceBundle; import javafx.event.ActionEvent; import javafx.fxml.FXML; import javafx.fxml.Initializable; import javafx.scene.control.Label;

public class FXMLDocumentController implements Initializable { @FXML private Label label;

@FXML private void handleButtonAction(ActionEvent event) { System.out.println("You clicked me!"); label.setText("Hello World!"); }

@Override public void initialize(URL url, ResourceBundle rb) { // TODO } }

Figure 31.35 You can choose JavaFX in the Categories and JavaFX FXML Application in the Project to create a FXML project.

M31_LIAN1878_11_GE_C31.indd 37 12/29/17 5:44 PM

31-38 Chapter 31 Advanced JavaFX and FXML

When you create a JavaFX FXML project, NetBeans creates a default .fxml file that contains the contents for a simple sample user interface. To view the user interface, double-click the .fxml file to open the Scene Builder, as shown in Figure 31.39. Note NetBeans can automati- cally detect the Scene Builder after it is installed on your machine.

31.11.3 Creating User Interfaces We now turn our attention to developing a simple calculator as shown in Figure 31.38. The Calculator program enables the user to enter numbers and perform addition, subtraction, multiplication, and division.

Figure 31.38 The application performs arithmetic operations.

Figure 31.36 You can enter project information in the New JavaFX Application dialog.

Figure 31.37 A FXML project is created.

M31_LIAN1878_11_GE_C31.indd 38 12/29/17 5:44 PM

31.11 Developing JavaFX Programs Using FXML 31-39

Figure 31.40 You can open the Library pane by clicking the Library icon and choose View as List.

Figure 31.39 Double-click the .fxml file to open the Scene Builder.

To start a new user interface, delete the default user interface in the .fxml file from the content pane, as shown in Figure 31.41. Here are the steps to create a new user interface:

1. (Optional) On some systems, the components in the Library pane are not visible by sections. Click the Library icon to open the context menu as shown in Figure 31.40 and choose View as List.

2. Drag a BorderPane into the user interface and drag an HBox to the center of the BorderPane and another HBox to the bottom of the BorderPane. Set the alignment of both HBox to CENTER as shown in Figure 31.42. Set the Spacing property in the Layout section of the Inspector to 5. When you select a component in the visual layout, the properties of the component are displayed in the Inspector pane, where you can set the properties.

M31_LIAN1878_11_GE_C31.indd 39 12/29/17 5:44 PM

31-40 Chapter 31 Advanced JavaFX and FXML

3. Drag and drop a Label, a TextField, a Label, a TextField, a Label, and a Tex- tField and change the label’s text to Number 1, Number 2, and Result, as shown in Figure 31.43. Set the Pref Column Count property for each text field to 2 in the Layout section of the Inspector. In the Code section of the Inspector, set the id for the text fields to tfNumber1, tfNumber2, and tfResult, as shown in Figure 31.44. These ids are useful to reference the text fields and obtain their values in the controller.

4. Drag and drop four Buttons to the second HBox and set their text property to Add, Subtract, Multiply, and Divide, as shown in Figure 31.45.

Figure 31.41 The UI is empty after deleting the default button in the pane.

Figure 31.42 A BorderPane is dropped to the UI and an HBox is placed at the bottom of the BorderPane.

Figure 31.43 The labels and text fields are dropped to the UI.

M31_LIAN1878_11_GE_C31.indd 40 12/29/17 5:44 PM

31-41

After you create and make changes to the user interface in the content pane, you need to save the changes by choosing File, Save from the menu bar in the Scene Builder. The .fxml file is updated and synchronized with the changes in the content pane. You can view the contents in the .fxml file from NetBeans, as shown in Figure 31.46.

Figure 31.44 Set the appropriate id for the text fields.

Figure 31.45 The buttons are dropped to the HBox.

Figure 31.46 You can view the contents of the FXML file.

M31_LIAN1878_11_GE_C31.indd 41 12/29/17 5:44 PM

31-42 Chapter 31 Advanced JavaFX and FXML

31.11.4 Handling Events in the Controller The .fxml file describes the user interface. You write the code to implement the logic in the controller file, as shown in Listing 31.18.

Listing 31.18 FXMLDocumentController.java 1 package calculator; 2 3 import javafx.event.ActionEvent; 4 import javafx.fxml.FXML; 5 import javafx.scene.control.TextField; 6 7 public class FXMLDocumentController { 8 @FXML 9 private TextField tfNumber1, tfNumber2, tfResult; 10 11 @FXML 12 private void addButtonAction(ActionEvent event) { 13 tfResult.setText(getResult('+') + ""); 14 } 15 16 @FXML 17 private void subtractButtonAction(ActionEvent event) { 18 tfResult.setText(getResult('−') + ""); 19 } 20 21 @FXML 22 private void multiplyButtonAction(ActionEvent event) { 23 tfResult.setText(getResult('*') + ""); 24 } 25 26 @FXML 27 private void divideButtonAction(ActionEvent event) { 28 tfResult.setText(getResult('/') + ""); 29 } 30 31 private double getResult(char op) { 32 double number1 = Double.parseDouble(tfNumber1.getText()); 33 double number2 = Double.parseDouble(tfNumber2.getText()); 34 switch (op) { 35 case '+': return number1 + number2; 36 case '−': return number1 − number2; 37 case '*': return number1 * number2; 38 case '/': return number1 / number2; 39 } 40 return Double.NaN; 41 } 42 }

The controller class declares three TextFields, tfNumber1, tfNumber2, and tfResult (line 9). The @FXML annotation denotes that these data fields are linked to the text fields in the user interface. Recall in the user interface, we set the id for the three text fields as tfNumber1, tfNumber2, and tfResult.

The codes for handling the events from the buttons are defined in the methods addButton- Action, subtractButtonAction, multiplyButtonAction, and divideButtonAction (lines 11–29). The @FXML annotation is used to denote that these methods will be tied to the button actions in the view.

Through the @FXML annotation, the data fields and methods in the controller are linked to the components and actions defined in the .fxml file.

M31_LIAN1878_11_GE_C31.indd 42 12/29/17 5:44 PM

31.11 Developing JavaFX Programs Using FXML 31-43

31.11.5 Linking View with Controller You can now link the actions from the components in the view with the processing methods in the controller. Here are the steps to accomplish it:

1. Add the following attribute in the <BorderPane> tag for using a controller with the view.

fx:controller="calculator.FXMLDocumentController"

2. Double-click the .fxml file in the project to display the visual layout window. In the Inspector for the Add button, choose addButtonAction from a list of action process- ing methods, as shown in Figure 31.47. The complete code for the .fxml file is shown in Listing 31.19.

Listing 31.19 FXMLDocument.fxml <?xml version="1.0" encoding="UTF-8"?>

<?import java.lang.*?> <?import java.util.*?> <?import javafx.scene.*?> <?import javafx.scene.control.*?> <?import javafx.scene.layout.*?>

M31_LIAN1878_11_GE_C31.indd 43 12/29/17 5:44 PM

31-44 Chapter 31 Advanced JavaFX and FXML

31.11.6 Running the Project The code in the model is automatically generated as shown in Listing 31.15. This is the main program that loads the FXML to create the user interface in a Parent object (lines 12 and 13). The parent object is then added to the scene (line 14). The scene is set to the stage (line 15). The stage is displayed in line 16.

Figure 31.47 Choosing addButtonAction to generate the code for handling action for the Add button.

ChapTer summary 1. JavaFX provides the cascading style sheets based on CSS. You can use the

getStylesheets method to load a style sheet and use the setStyle, setStyleClass, and setId methods to set JavaFX CSS for nodes.

2. JavaFX provides the QuadCurve, CubicCurve, and Path classes for creating advanced shapes.

3. JavaFX supports coordinate transformations using translation, rotation, and scaling.

4. You can specify the pattern for a stroke, how the lines are joined in a stroke, the width of a stroke, and the type of a stroke.

5. You can create menus using the Menu, MenuItem, CheckMenuItem, and RadioMenuItem classes.

6. You can create context menus using the ContextMenu class.

7. The SplitPane can be used to display multiple panes horizontally or vertically and allows the user to adjust the sizes of the panes.

8. The TabPane can be used to display multiple panes with tabs for selecting panes.

9. You can create and display tables using the TableView and TableColumn classes.

10. You can create JavaFX user interfaces using FXML. FXML is XML-based script lan- guage for describing the user interface. Using FXML enables you to separate user inter- face from the logic of Java code.

11. JavaFX Scene Builder is a visual tool for creating the user interface without manually writing the FXML scripts.

M31_LIAN1878_11_GE_C31.indd 44 12/29/17 5:44 PM

Programming Exercises 31-45

Quiz Answer the quiz for this chapter online at the book Companion Website.

programming exerCiSeS

Sections 31.2 31.1 (Use JavaFX CSS) Create a CSS style sheet that defines a class for white fill and

black stroke color and an id for red stroke and green color. Write a program that displays four circles and uses the style class and id. The sample run of the program is shown in Figure 31.48a.

Figure 31.49 (a) Five texts are displayed with a random color and a specified font. (b) A path is displayed inside the circle. (c) Two circles are displayed in an oval.

(a) (b) (c)

Figure 31.48 (a) The border and the color style for the shapes are defined in a style class. (b) Exercise 31.2 displays a tic-tac-toe board with images using style sheet for border. (c) Three cards are randomly selected.

(a) (b) (c)

*31.2 (Tic-tac-toe board) Write a program that displays a tic-tac-toe board, as shown in Figure 31.48b. A cell may be X, O, or empty. What to display at each cell is randomly decided. The X and O are images in the files x.gif and o.gif. Use the style sheet for border.

*31.3 (Display three cards) Write a program that displays three cards randomly selected from a deck of 52, as shown in Figure 31.48c. The card image files are named 1.png, 2.png, . . ., 52.png and stored in the image/card directory. All the three cards are distinct and selected randomly. Hint: You can select random cards by stor- ing the numbers 1–52 to an array, perform a random shuffle using Section 7.2.6, and use the first three numbers in the array as the file names for the image. Use the style sheet for border.

Sections 31.3 31.4 (Color and font) Write a program that displays five texts vertically, as shown in

Figure 31.49a. Set a random color and opacity for each text and set the font of each text to Times Roman, bold, italic, and 22 pixels.

M31_LIAN1878_11_GE_C31.indd 45 12/29/17 5:44 PM

31-46 Chapter 31 Advanced JavaFX and FXML

*31.5 (Cubic curve) Write a program that creates two shapes: a circle and a path consist- ing of two cubic curves, as shown in Figure 31.49b.

*31.6 (Eyes) Write a program that displays two eyes in an oval, as shown in Figure 31.49c.

Sections 31.4 *31.7 (Translation) Write a program that displays a rectangle with upper-left corner point

at (40, 40), width 50, and height 40. Enter the values in the text fields x and y and press the Translate button to translate the rectangle to a new location, as shown in Figure 31.50a.

Figure 31.50 (a) Exercise 31.7 translates coordinates. (b) Exercise 31.8 rotates coordinates. (c) Exercise 31.9 scales coordinates.

(a) (b) (c)

*31.8 (Rotation) Write a program that displays an ellipse. The ellipse is centered in the pane with width 60 and height 40. Enter the value in the text field Angle and press the Rotate button to rotate the ellipse, as shown in Figure 31.50b.

*31.9 (Scale graphics) Write a program that displays an ellipse. The ellipse is centered in the pane with width 60 and height 40. Enter the scaling factors in the text fields and press the Scale button to scale the ellipse, as shown in Figure 31.50c.

*31.10 (Plot the sine function) Write a program that plots the sine function, as shown in Figure 31.51a.

Figure 31.51 (a) Exercise 31.10 displays a sine function. (b) Exercise 31.11 displays the log function.

(a) (b)

*31.11 (Plot the log function) Write a program that plots the log function, as shown in Figure 31.51a.

*31.12 (Plot the n2 function) Write a program that plots the n2 function, as shown in Figure 31.51b 2a.

M31_LIAN1878_11_GE_C31.indd 46 12/29/17 5:44 PM

Programming Exercises 31-47

*31.13 (Plot the log, n, nlogn, and n2 functions) Write a program that plots the log, n, nlogn, and n2 functions, as shown in Figure 31.52b.

*31.14 (Scale and rotate graphics) Write a program that enables the user to scale and rotate the STOP sign, as shown in Figure 31.53. The user can press the UP/DOWN arrow key to increase/decrease the size and press the RIGHT/LEFT arrow key to rotate left or right.

Figure 31.52 (a) Exercise 31.13 displays the n2 function. (b) Exercise 31.13 displays several functions.

(a) (b)

Figure 31.53 The program can rotate and scale the painting.

Sections 31.5 *31.15 (Sunshine) Write a program that displays a circle filled with a gradient color to

animate a sun and display light rays coming out from the sun using dashed lines, as shown in Figure 31.54a.

Figure 31.54 (a) Exercise 31.15 displays the sunshine. (b) Exercise 31.16 displays a cylinder.

(a) (b)

*31.16 (Display a cylinder) Write a program that displays a cylinder, as shown in Figure 31.54b. Use dashed strokes to draw the dashed arc.

M31_LIAN1878_11_GE_C31.indd 47 12/29/17 5:44 PM

31-48 Chapter 31 Advanced JavaFX and FXML

Sections 31.6 * 31.17 (Create an investment value calculator) Write a program that calculates the future

value of an investment at a given interest rate for a specified number of years. The formula for the calculation is as follows:

futureValue = investmentAmount × (1 + monthlyInterestRate)years×12

Use text fields for interest rate, investment amount, and years. Display the future amount in a text field when the user clicks the Calculate button or chooses Calcu- late from the Operation menu (see Figure 31.55). Click the Exit menu to exit the program.

Figure 31.56 The popup menu contains the commands to perform arithmetic operations.

Figure 31.55 The user enters the investment amount, years, and interest rate to compute future value.

Sections 31.8 *31.18 (Use popup menus) Modify Listing 31.9, MenuDemo.java, to create a popup

menu that contains the menus Operations and Exit, as shown in Figure 31.56. The popup is displayed when you click the right mouse button on the panel that contains the labels and the text fields.

*31.19 (Use SplitPane) Create a program that displays four shapes in split panes, as shown in Figure 31.57a.

Sections 31.9 *31.20 (Use tab panes) Modify Listing 31.12, TabPaneDemo.java, to add a pane

of radio buttons for specifying the tab placement of the tab pane, as shown in Figure 31.57b and c.

M31_LIAN1878_11_GE_C31.indd 48 12/29/17 5:44 PM

Programming Exercises 31-49

*31.21 (Use tab panes) Write a program using tab panes for performing integer and rational number arithmetic as shown in Figure 31.58.

Figure 31.57 (a) Four shapes are displayed in split panes. (b and c) The radio buttons let you choose the tab placement of the tabbed pane.

(a) (b) (c)

Figure 31.58 A tab pane is used to select panes that perform integer operations and rational number operations.

Figure 31.59 Clicking the Delete Selected Row button removes the selected row from the table.

Sections 31.10 *31.22 (Use table view) Revise Listing 31.14 to add a button to delete the selected row

from the table, as shown in Figure 31.59.

M31_LIAN1878_11_GE_C31.indd 49 12/29/17 5:44 PM

Objectives ■■ To get an overview of multithreading (§32.2).

■■ To develop task classes by implementing the Runnable interface (§32.3).

■■ To create threads to run tasks using the Thread class (§32.3).

■■ To control threads using the methods in the Thread class (§32.4).

■■ To control animations using threads and use Platform.runLater to run the code in the application thread (§32.5).

■■ To execute tasks in a thread pool (§32.6).

■■ To use synchronized methods or blocks to synchronize threads to avoid race conditions (§32.7).

■■ To synchronize threads using locks (§32.8).

■■ To facilitate thread communications using conditions on locks (§§32.9 and 32.10).

■■ To use blocking queues (ArrayBlockingQueue, LinkedBlocking- Queue, and PriorityBlockingQueue) to synchronize access to a queue (§32.11).

■■ To restrict the number of concurrent accesses to a shared resource using semaphores (§32.12).

■■ To use the resource-ordering technique to avoid deadlocks (§32.13).

■■ To describe the life cycle of a thread (§32.14).

■■ To create synchronized collections using the static methods in the Collections class (§32.15).

■■ To develop parallel programs using the Fork/Join Framework (§32.16).

Multithreading and Parallel Programming

Chapter

M32_LIAN1878_11_GE_C32.indd 1 12/29/17 5:46 PM

32-2 Chapter 32 Multithreading and Parallel Programming

32.1 Introduction Multithreading enables multiple tasks in a program to be executed concurrently.

One of the powerful features of Java is its built-in support for multithreading—the concurrent running of multiple tasks within a program. In many programming languages, you have to invoke system-dependent procedures and functions to implement multithreading. This chapter introduces the concepts of threads and how multithreading programs can be developed in Java.

32.2 Thread Concepts A program may consist of many tasks that can run concurrently. A thread is the flow of execution, from beginning to end, of a task.

A thread provides the mechanism for running a task. With Java, you can launch multiple threads from a program concurrently. These threads can be executed simultaneously in multi- processor systems, as shown in Figure 32.1a.

Point Key

multithreading

Point Key

thread

task

Figure 32.1 (a) Multiple threads running on multiple CPUs. (b) Multiple threads share a single CPU.

Thread 1

Thread 3

Thread 2

(a)

Thread 1

Thread 3

Thread 2

(b)

In single-processor systems, as shown in Figure 32.1b, the multiple threads share CPU time, known as time sharing, and the operating system is responsible for scheduling and allocating resources to them. This arrangement is practical because most of the time the CPU is idle. It does nothing, for example, while waiting for the user to enter data.

Multithreading can make your program more responsive and interactive as well as enhance performance. For example, a good word processor lets you print or save a file while you are typing. In some cases, multithreaded programs run faster than single-threaded programs even on single-processor systems. Java provides exceptionally good support for creating and running threads, and for locking resources to prevent conflicts.

You can create additional threads to run concurrent tasks in the program. In Java, each task is an instance of the Runnable interface, also called a runnable object. A thread is essentially an object that facilitates the execution of a task.

32.2.1 Why is multithreading needed? How can multiple threads run simultaneously in a single-processor system?

32.2.2 What is a runnable object? What is a thread?

32.3 Creating Tasks and Threads A task class must implement the Runnable interface. A task must be run from a thread.

Tasks are objects. To create tasks, you have to first define a class for tasks, which implements the Runnable interface. The Runnable interface is rather simple. All it contains is the run() method. You need to implement this method to tell the system how your thread is going to run. A template for developing a task class is shown in Figure 32.2a.

time sharing

task

runnable object

thread

Point Check

Point Key

Runnable interface

run() method

M32_LIAN1878_11_GE_C32.indd 2 12/29/17 5:46 PM

32.3 Creating Tasks and Threads 32-3

Once you have defined a TaskClass, you can create a task using its constructor. For example,

TaskClass task = new TaskClass(...);

A task must be executed in a thread. The Thread class contains the constructors for creating threads and many useful methods for controlling threads. To create a thread for a task, use

Thread thread = new Thread(task);

You can then invoke the start() method to tell the JVM that the thread is ready to run, as follows:

thread.start();

The JVM will execute the task by invoking the task’s run() method. Figure 32.2b outlines the major steps for creating a task, a thread, and starting the thread.

Listing 32.1 gives a program that creates three tasks and three threads to run them.

■■ The first task prints the letter a 100 times.

■■ The second task prints the letter b 100 times.

■■ The third task prints the integers 1 through 100.

When you run this program, the three threads will share the CPU and take turns printing letters and numbers on the console. Figure 32.3 shows a sample run of the program.

Thread class create a task

create a thread

start a thread

Figure 32.2 Define a task class by implementing the Runnable interface.

// Client class public class Client { ... public void someMethod() { ... // Create an instance of TaskClass TaskClass task = new TaskClass(...);

// Create a thread Thread thread = new Thread(task);

// Start a thread thread.start(); ... } ... }

// Custom task class public class TaskClass implements Runnable { ... public TaskClass(...) { ... } // Implement the run method in Runnable public void run() { // Tell system how to run custom thread ... } ... }

TaskClassjava.lang.Runnable

(a) (b)

Figure 32.3 Tasks printA, printB, and print100 are executed simultaneously to display the letter a 100 times, the letter b 100 times, and the numbers from 1 to 100.

M32_LIAN1878_11_GE_C32.indd 3 12/29/17 5:46 PM

32-4 Chapter 32 Multithreading and Parallel Programming

Listing 32.1 TaskThreadDemo.java 1 public class TaskThreadDemo { 2 public static void main(String[] args) { 3 // Create tasks 4 Runnable printA = new PrintChar('a', 100); 5 Runnable printB = new PrintChar('b', 100); 6 Runnable print100 = new PrintNum(100); 7 8 // Create threads 9 Thread thread1 = new Thread(printA); 10 Thread thread2 = new Thread(printB); 11 Thread thread3 = new Thread(print100); 12 13 // Start threads 14 thread1.start(); 15 thread2.start(); 16 thread3.start(); 17 } 18 } 19 20 // The task for printing a character a specified number of times 21 class PrintChar implements Runnable { 22 private char charToPrint; // The character to print 23 private int times; // The number of times to repeat 24 25 /** Construct a task with a specified character and number of 26 * times to print the character 27 */ 28 public PrintChar(char c, int t) { 29 charToPrint = c; 30 times = t; 31 } 32 33 @Override /** Override the run() method to tell the system 34 * what task to perform 35 */ 36 public void run() { 37 for (int i = 0; i < times; i++) { 38 System.out.print(charToPrint); 39 } 40 } 41 } 42 43 // The task class for printing numbers from 1 to n for a given n 44 class PrintNum implements Runnable { 45 private int lastNum; 46 47 /** Construct a task for printing 1, 2, ..., n */ 48 public PrintNum(int n) { 49 lastNum = n; 50 } 51 52 @Override /** Tell the thread how to run */ 53 public void run() { 54 for (int i = 1; i <= lastNum; i++) { 55 System.out.print(" " + i); 56 } 57 } 58 }

create tasks

create threads

start threads

task class

run

task class

run

M32_LIAN1878_11_GE_C32.indd 4 12/29/17 5:46 PM

32.3 Creating Tasks and Threads 32-5

The program creates three tasks (lines 4–6). To run them concurrently, three threads are created (lines 9–11). The start() method (lines 14–16) is invoked to start a thread that causes the run() method in the task to be executed. When the run() method completes, the thread terminates.

Because the first two tasks, printA and printB, have similar functionality, they can be defined in one task class PrintChar (lines 21–41). The PrintChar class implements Runnable and overrides the run() method (lines 36–40) with the print-character action. This class provides a framework for printing any single character a given number of times. The runnable objects, printA and printB, are instances of the PrintChar class.

The PrintNum class (lines 44–58) implements Runnable and overrides the run() method (lines 53–57) with the print-number action. This class provides a framework for printing num- bers from 1 to n, for any integer n. The runnable object print100 is an instance of the class printNum class.

Note If you don’t see the effect of these three threads running concurrently, increase the number of characters to be printed. For example, change line 4 to

Runnable printA = new PrintChar('a', 10000);

Important Note The run()method in a task specifies how to perform the task. This method is automati- cally invoked by the JVM. You should not invoke it. Invoking run()directly merely executes this method in the same thread; no new thread is started.

32.3.1 How do you define a task class? How do you create a thread for a task? 32.3.2 What would happen if you replace the start() method with the run() method in

lines 14–16 in Listing 32.1?

effect of concurrency

run() method

Point Check

print100.start();

printA.start();

printB.start();

Replaced by print100.run(); printA.run(); printB.run();

32.3.3 What is wrong in the following two programs? Correct the errors.

public class Test implements Runnable { public static void main(String[] args) { new Test(); } public Test() {

Test task = new Test(); new Thread(task).start(); }

public void run() { System.out.println("test"); } }

public class Test implements Runnable { public static void main(String[] args) { new Test(); }

public Test() { Thread t = new Thread(this); t.start(); t.start(); }

public void run() { System.out.println("test"); } }

(a) (b)

M32_LIAN1878_11_GE_C32.indd 5 12/29/17 5:46 PM

32-6 Chapter 32 Multithreading and Parallel Programming

32.4 The Thread Class The Thread class contains the constructors for creating threads for tasks and the methods for controlling threads.

Figure 32.4 shows the class diagram for the Thread class.

Point Key

Figure 32.4 The Thread class contains the methods for controlling threads.

java.lang.Thread

«interface» java.lang.Runnable

+Thread(task: Runnable)

+Thread()

+start(): void

+interrupt(): void

+isAlive(): boolean

+setPriority(p: int): void

+join(): void

Creates a Thread for a speci�ed task.

Creates an empty Thread.

Starts the thread that causes the run() method to be invoked by the JVM.

Interrupts this thread.

Sets priority p (ranging from 1 to 10) for this thread.

Waits for this thread to �nish.

Puts a thread to sleep for a speci�ed time in milliseconds.

Causes a thread to pause temporarily and allow other threads to execute.

Tests whether the thread is currently running.

+sleep(millis: long): void

+yield(): void

Figure 32.5 Define a thread class by extending the Thread class.

// Client class public class Client { ... public void someMethod() { ... // Create a thread CustomThread thread1 = new CustomThread(...);

// Start a thread thread1.start();

// Create another thread CustomThread thread2 = new CustomThread(...);

// Start a thread thread2.start(); } ... }

// Custom thread class public class CustomThread extends Thread { ... public CustomThread(...) { ... }

// Override the run method in Runnable public void run() { // Tell system how to perform this task ... } ... }

CustomThreadjava.lang.Thread

(a) (b)

...

Note Since the Thread class implements Runnable, you could define a class that extends Thread and implements the run method, as shown in Figure 32.5a, then create an object from the class and invoke its start method in a client program to start the thread, as shown in Figure 32.5b.separating task from thread

M32_LIAN1878_11_GE_C32.indd 6 12/29/17 5:46 PM

32.4 The Thread Class 32-7

This approach is, however, not recommended because it mixes the task and the mechanism of running the task. Separating the task from the thread is a preferred design.

Note The Thread class also contains the stop(), suspend(), and resume() methods. As of Java 2, these methods were deprecated (or outdated) because they are known to be inherently unsafe. Instead of using the stop() method, you should assign null to a Thread variable to indicate that it has stopped.

You can use the yield() method to temporarily release time for other threads. For example, suppose that you modify the code in the run() method in lines 53–57 for PrintNum in Listing 32.1 as follows:

public void run() { for (int i = 1; i <= lastNum; i++) { System.out.print(" " + i); Thread.yield(); } }

Every time a number is printed, the thread of the print100 task is yielded to other threads. The sleep(long millis) method puts the thread to sleep for a specified time in milli-

seconds to allow other threads to execute. For example, suppose that you modify the code in lines 53–57 in Listing 32.1 as follows:

public void run() { try { for (int i = 1; i <= lastNum; i++) { System.out.print(" " + i); if (i >= 50) Thread.sleep(1); } } catch (InterruptedException ex) { } }

Every time a number (>= 50) is printed, the thread of the print100 task is put to sleep for 1 millisecond.

The sleep method may throw an InterruptedException, which is a checked exception. Such an exception may occur when a sleeping thread’s interrupt() method is called. The interrupt() method is very rarely invoked on a thread, so an Interrupted Exception is unlikely to occur. But since Java forces you to catch checked exceptions, you have to put it in a try-catch block. If a sleep method is invoked in a loop, you should wrap the loop in a try-catch block, as shown in (a) below. If the loop is outside the try-catch block, as shown in (b), the thread may continue to execute even though it is being interrupted.

deprecated method

yield()

sleep(long)

InterruptedException

public void run() { try { while (...) { ... Thread.sleep(1000); } } catch (InterruptedException ex) { ex.printStackTrace(); } }

public void run() { while (...) { try { ... Thread.sleep(sleepTime); } catch (InterruptedException ex) { ex.printStackTrace(); } } }

(a) Correct (b) Incorrect

M32_LIAN1878_11_GE_C32.indd 7 12/29/17 5:46 PM

32-8 Chapter 32 Multithreading and Parallel Programming

You can use the join() method to force one thread to wait for another thread to finish. For example, suppose that you modify the code in lines 53–57 in Listing 32.1 as follows:

join()

Thread print100

Wait for thread4 to �nish

Thread thread4

thread4 �nished

public void run() { Thread thread4 = new Thread( new PrintChar('c', 40)); thread4.start(); try { for (int i = 1; i <= lastNum; i++) { System.out.print (" " + i); if (i == 50) thread4.join(); } } catch (InterruptedException ex) { } }

thread4.join()

A new thread4 is created and it prints character c 40 times. The numbers from 50 to 100 are printed after thread thread4 is finished.

Java assigns every thread a priority. By default, a thread inherits the priority of the thread that spawned it. You can increase or decrease the priority of any thread by using the setPriority method and you can get the thread’s priority by using the getPriority method. Priorities are numbers ranging from 1 to 10. The Thread class has the int constants MIN_PRIORITY, NORM_PRIORITY, and MAX_PRIORITY, representing 1, 5, and 10, respec- tively. The priority of the main thread is Thread.NORM_PRIORITY.

The JVM always picks the currently runnable thread with the highest priority. A lower priority thread can run only when no higher priority threads are running. If all runnable threads have equal priorities, each is assigned an equal portion of the CPU time in a circular queue. This is called round-robin scheduling. For example, suppose that you insert the following code in line 16 in Listing 32.1:

thread3.setPriority(Thread.MAX_PRIORITY);

The thread for the print100 task will be finished first.

Tip The priority numbers may be changed in a future version of Java. To minimize the impact of any changes, use the constants in the Thread class to specify thread priorities.

Tip A thread may never get a chance to run if there is always a higher priority thread running or a same-priority thread that never yields. This situation is known as contention or starvation. To avoid contention, the thread with higher priority must periodically invoke the sleep or yield method to give a thread with a lower or the same priority a chance to run.

32.4.1 Which of the following methods are instance methods in java.lang.Thread? Which method may throw an InterruptedException? Which of them are deprecated in Java?

run, start, stop, suspend, resume, sleep, interrupt, yield, join

32.4.2 If a loop contains a method that throws an InterruptedException, why should the loop be placed inside a try-catch block?

32.4.3 How do you set a priority for a thread? What is the default priority?

setPriority(int)

round-robin scheduling

contention or starvation

Point Check

M32_LIAN1878_11_GE_C32.indd 8 12/29/17 5:46 PM

32.5 Animation Using Threads and the Platform.runLater Method 32-9

32.5 Animation Using Threads and the Platform. runLater Method You can use a thread to control an animation and run the code in JavaFX GUI thread using the Platform.runLater method.

The use of a Timeline object to control animations was introduced in Section 15.11, Ani- mation. Alternatively, you can also use a thread to control animation. Listing 32.2 gives an example that displays flashing text on a label, as shown in Figure 32.6.

Point Key

Figure 32.6 The text “Welcome” blinks.

Listing 32.2 FlashText.java 1 import javafx.application.Application; 2 import javafx.application.Platform; 3 import javafx.scene.Scene; 4 import javafx.scene.control.Label; 5 import javafx.scene.layout.StackPane; 6 import javafx.stage.Stage; 7 8 public class FlashText extends Application { 9 private String text = ""; 10 11 @Override // Override the start method in the Application class 12 public void start(Stage primaryStage) { 13 StackPane pane = new StackPane(); 14 Label lblText = new Label("Programming is fun"); 15 pane.getChildren().add(lblText); 16 17 new Thread(new Runnable() { 18 @Override 19 public void run() { 20 try { 21 while (true) { 22 if (lblText.getText().trim().length() == 0) 23 text = "Welcome"; 24 else 25 text = ""; 26 27 Platform.runLater(new Runnable() { // Run from JavaFX GUI 28 @Override 29 public void run() { 30 lblText.setText(text); 31 } 32 }); 33 34 Thread.sleep(200); 35 } 36 } 37 catch (InterruptedException ex) { 38 } 39 } 40 }).start(); 41

create a label label in a pane

create a thread

run thread

change text

Platform.runLater

update GUI

sleep

M32_LIAN1878_11_GE_C32.indd 9 12/29/17 5:46 PM

32-10 Chapter 32 Multithreading and Parallel Programming

42 // Create a scene and place it in the stage 43 Scene scene = new Scene(pane, 200, 50); 44 primaryStage.setTitle("FlashText"); // Set the stage title 45 primaryStage.setScene(scene); // Place the scene in the stage 46 primaryStage.show(); // Display the stage 47 } 48 }

The program creates a Runnable object in an anonymous inner class (lines 17–40). This object is started in line 40 and runs continuously to change the text in the label. It sets a text in the label if the label is blank (line 23) and sets its text blank (line 25) if the label has a text. The text is set and unset to simulate a flashing effect.

JavaFX GUI is run from the JavaFX application thread. The flashing control is run from a separate thread. The code in a nonapplication thread cannot update GUI in the application thread. To update the text in the label, a new Runnable object is created in lines 27–32. Invok- ing Platform.runLater(Runnable r) tells the system to run this Runnable object in the application thread.

The anonymous inner classes in this program can be simplifed using lambda expressions as follows:

new Thread(() -> { // lambda expression try { while (true) { if (lblText.getText().trim().length() == 0) text = "Welcome"; else text = "";

Platform.runLater(() -> lblText.setText(text)); // lambda exp

Thread.sleep(200); } } catch (InterruptedException ex) { } }).start();

32.5.1 What causes the text to flash? 32.5.2 Is an instance of FlashText a runnable object? 32.5.3 What is the purpose of using Platform.runLater? 32.5.4 Can you replace the code in lines 27–32 using the following code?

Platform.runLater(e -> lblText.setText(text));

32.5.5 What happens if line 34 (Thread.sleep(200)) is not used? 32.5.6 There is an issue in Listing 16.9, ListViewDemo. If you press the CTRL key and

select Canada, Demark, and China in this order, you will get an ArrayIndex- OutBoundsException. What is the reason for this error and how do you fix it? (Thanks to Henri Heimonen of Finland for contributing this question).

32.6 Thread Pools A thread pool can be used to execute tasks efficiently.

In Section 32.3, Creating Tasks and Threads, you learned how to define a task class by imple- menting java.lang.Runnable, and how to create a thread to run a task like this:

Runnable task = new TaskClass(...); new Thread(task).start();

JavaFX application thread

Platform.runLater

Point Check

Point Key

M32_LIAN1878_11_GE_C32.indd 10 12/29/17 5:46 PM

32.6 Thread Pools 32-11

To create an Executor object, use the static methods in the Executors class, as shown in Figure 32.8. The newFixedThreadPool(int) method creates a fixed number of threads in a pool. If a thread completes executing a task, it can be reused to execute another task. If a thread terminates due to a failure prior to shutdown, a new thread will be created to replace it if all the threads in the pool are not idle and there are tasks waiting for execution. The newCachedThreadPool() method creates a new thread if all the threads in the pool are not idle and there are tasks waiting for execution. A thread in a cached pool will be terminated if it has not been used for 60 seconds. A cached pool is efficient for many short tasks.

Figure 32.7 The Executor interface executes threads and the ExecutorService subinterface manages threads.

+shutdown(): void

+shutdownNow(): List<Runnable>

+isShutdown(): boolean +isTerminated(): boolean

«interface» java.util.concurrent.ExecutorService

+execute(Runnable object): void

«interface» java.util.concurrent.Executor

Executes the runnable task.

Shuts down the executor, but allows the tasks in the executor to complete. Once shut down, it cannot accept new tasks. Shuts down the executor immediately even though there are un�nished threads in the pool. Returns a list of un�nished tasks. Returns true if the executor has been shut down. Returns true if all tasks in the pool are terminated.

This approach is convenient for a single task execution, but it is not efficient for a large number of tasks because you have to create a thread for each task. Starting a new thread for each task could limit throughput and cause poor performance. Using a thread pool is an ideal way to manage the number of tasks executing concurrently. Java provides the Executor interface for executing tasks in a thread pool and the ExecutorService interface for man- aging and controlling tasks. ExecutorService is a subinterface of Executor, as shown in Figure 32.7.

Figure 32.8 The Executors class provides static methods for creating Executor objects.

Creates a thread pool with a �xed number of threads executing concurrently. A thread may be reused to execute another task after its current task is �nished.

Creates a thread pool that creates new threads as needed, but will reuse previously constructed threads when they are available.

java.util.concurrent.Executors

+newFixedThreadPool(numberOfThreads: int): ExecutorService

+newCachedThreadPool(): ExecutorService

Listing 32.3 shows how to rewrite Listing 32.1 using a thread pool.

Listing 32.3 ExecutorDemo.java 1 import java.util.concurrent.*; 2 3 public class ExecutorDemo {

M32_LIAN1878_11_GE_C32.indd 11 12/29/17 5:46 PM

32-12 Chapter 32 Multithreading and Parallel Programming

4 public static void main(String[] args) { 5 // Create a fixed thread pool with maximum three threads 6 ExecutorService executor = Executors.newFixedThreadPool(3); 7 8 // Submit runnable tasks to the executor 9 executor.execute(new PrintChar('a', 100)); 10 executor.execute(new PrintChar('b', 100)); 11 executor.execute(new PrintNum(100)); 12 13 // Shut down the executor 14 executor.shutdown(); 15 } 16 }

Line 6 creates a thread pool executor with a total of three threads maximum. Classes PrintChar and PrintNum are defined in Listing 32.1. Line 9 creates a task, new PrintChar('a', 100), and adds it to the pool. Similarly, another two runnable tasks are created and added to the same pool in lines 10 and 11. The executor creates three threads to execute three tasks concurrently.

Suppose you replace line 6 with

ExecutorService executor = Executors.newFixedThreadPool(1);

What will happen? The three runnable tasks will be executed sequentially because there is only one thread in the pool.

Suppose you replace line 6 with

ExecutorService executor = Executors.newCachedThreadPool();

What will happen? New threads will be created for each waiting task, so all the tasks will be executed concurrently.

The shutdown() method in line 14 tells the executor to shut down. No new tasks can be accepted, but any existing tasks will continue to finish.

Tip If you need to create a thread for just one task, use the Thread class. If you need to create threads for multiple tasks, it is better to use a thread pool.

32.6.1 What are the benefits of using a thread pool? 32.6.2 How do you create a thread pool with three fixed threads? How do you submit a

task to a thread pool? How do you know that all the tasks are finished?

32.7 Thread Synchronization Thread synchronization is to coordinate the execution of the dependent threads.

A shared resource may become corrupted if it is accessed simultaneously by multiple threads. The following example demonstrates the problem.

Suppose that you create and launch 100 threads, each of which adds a penny to an account. Define a class named Account to model the account, a class named AddAPennyTask to add a penny to the account, and a main class that creates and launches threads. The relationships of these classes are shown in Figure 32.9. The program is given in Listing 32.4.

create executor

submit task

shut down executor

Point Check

Point Key

M32_LIAN1878_11_GE_C32.indd 12 12/29/17 5:46 PM

32.7 Thread Synchronization 32-13

Figure 32.9 AccountWithoutSync contains an instance of Account and 100 threads of AddAPennyTask.

100 1 AddAPennyTask

+run(): void

AccountWithoutSync

-account: Account

+main(args: String[]): void

1 1 Account

+getBalance(): int +deposit(amount: int): void

-balance: int

«interface» java.lang.Runnable

Listing 32.4 AccountWithoutSync.java 1 import java.util.concurrent.*; 2 3 public class AccountWithoutSync { 4 private static Account account = new Account(); 5 6 public static void main(String[] args) { 7 ExecutorService executor = Executors.newCachedThreadPool(); 8 9 // Create and launch 100 threads 10 for (int i = 0; i < 100; i++) { 11 executor.execute(new AddAPennyTask()); 12 } 13 14 executor.shutdown(); 15 16 // Wait until all tasks are finished 17 while (!executor.isTerminated()) { 18 } 19 20 System.out.println("What is balance? " + account.getBalance()); 21 } 22 23 // A thread for adding a penny to the account 24 private static class AddAPennyTask implements Runnable { 25 public void run() { 26 account.deposit(1); 27 } 28 } 29 30 // An inner class for account 31 private static class Account { 32 private int balance = 0; 33 34 public int getBalance() { 35 return balance; 36 } 37 38 public void deposit(int amount) { 39 int newBalance = balance + amount; 40 41 // This delay is deliberately added to magnify the

create executor

submit task

shut down executor

wait for all tasks to terminate

M32_LIAN1878_11_GE_C32.indd 13 12/29/17 5:46 PM

32-14 Chapter 32 Multithreading and Parallel Programming

42 // data-corruption problem and make it easy to see. 43 try { 44 Thread.sleep(5); 45 } 46 catch (InterruptedException ex) { 47 } 48 49 balance = newBalance; 50 } 51 } 52 }

The classes AddAPennyTask and Account in lines 24–51 are inner classes. Line 4 creates an Account with initial balance 0. Line 11 creates a task to add a penny to the account and submits the task to the executor. Line 11 is repeated 100 times in lines 10–12. The program repeatedly checks whether all tasks are completed in lines 17 and 18. The account balance is displayed in line 20 after all tasks are completed.

The program creates 100 threads executed in a thread pool executor (lines 10–12). The isTerminated() method (line 17) is used to test whether all the threads in the pool are terminated.

The balance of the account is initially 0 (line 32). When all the threads are finished, the balance should be 100 but the output is unpredictable. As can be seen in Figure 32.10, the answers are wrong in the sample run. This demonstrates the data-corruption problem that occurs when all the threads have access to the same data source simultaneously.

Figure 32.10 The AccountWithoutSync program causes data inconsistency.

Lines 39–49 could be replaced by one statement:

balance = balance + amount;

It is highly unlikely, although plausible, that the problem can be replicated using this single statement. The statements in lines 39–49 are deliberately designed to magnify the data- corruption problem and make it easy to see. If you run the program several times but still do not see the problem, increase the sleep time in line 44. This will increase the chances for showing the problem of data inconsistency.

What, then, caused the error in this program? A possible scenario is shown in Figure 32.11.

Figure 32.11 Task 1 and Task 2 both add 1 to the same balance.

Step Balance Task 1 Task 2

1 0 newBalance = balance + 1;

newBalance = balance + 1;2 0 3 1 balance = newBalance; 4 1 balance = newBalance;

M32_LIAN1878_11_GE_C32.indd 14 12/29/17 5:46 PM

32.7 Thread Synchronization 32-15

In Step 1, Task 1 gets the balance from the account. In Step 2, Task 2 gets the same balance from the account. In Step 3, Task 1 writes a new balance to the account. In Step 4, Task 2 writes a new balance to the account.

The effect of this scenario is that Task 1 does nothing because in Step 4, Task 2 overrides Task 1’s result. Obviously, the problem is that Task 1 and Task 2 are accessing a common resource in a way that causes a conflict. This is a common problem, known as a race condi- tion, in multithreaded programs. A class is said to be thread-safe if an object of the class does not cause a race condition in the presence of multiple threads. As demonstrated in the preced- ing example, the Account class is not thread-safe.

32.7.1 The synchronized Keyword To avoid race conditions, it is necessary to prevent more than one thread from simultaneously entering a certain part of the program, known as the critical region. The critical region in Listing 32.4 is the entire deposit method. You can use the keyword synchronized to synchronize the method so that only one thread can access the method at a time. There are several ways to correct the problem in Listing 32.4. One approach is to make Account thread-safe by adding the keyword synchronized in the deposit method in line 38, as follows:

public synchronized void deposit(double amount)

A synchronized method acquires a lock before it executes. A lock is a mechanism for exclu- sive use of a resource. In the case of an instance method, the lock is on the object for which the method was invoked. In the case of a static method, the lock is on the class. If one thread invokes a synchronized instance method (respectively, static method) on an object, the lock of that object (respectively, class) is acquired first, then the method is executed, and finally the lock is released. Another thread invoking the same method of that object (respectively, class) is blocked until the lock is released.

With the deposit method synchronized, the preceding scenario cannot happen. If Task 1 enters the method, Task 2 is blocked until Task 1 finishes the method, as shown in Figure 32.12.

race condition thread-safe

critical region

Figure 32.12 Task 1 and Task 2 are synchronized.

Acquire a lock on the object account

Execute the deposit method

Release the lock

Task 1

Execute the deposit method

Task 2

Wait to acquire the lock

Acquire a lock on the object account

32.7.2 Synchronizing Statements Invoking a synchronized instance method of an object acquires a lock on the object, and invok- ing a synchronized static method of a class acquires a lock on the class. A synchronized state- ment can be used to acquire a lock on any object, not just this object, when executing a block

M32_LIAN1878_11_GE_C32.indd 15 12/29/17 5:46 PM

32-16 Chapter 32 Multithreading and Parallel Programming

of the code in a method. This block is referred to as a synchronized block. The general form of a synchronized statement is as follows:

synchronized (expr) { statements; }

The expression expr must evaluate to an object reference. If the object is already locked by another thread, the thread is blocked until the lock is released. When a lock is obtained on the object, the statements in the synchronized block are executed and then the lock is released.

Synchronized statements enable you to synchronize part of the code in a method instead of the entire method. This increases concurrency. You can make Listing 32.4 thread-safe by placing the statement in line 26 inside a synchronized block:

synchronized (account) { account.deposit(1); }

Note Any synchronized instance method can be converted into a synchronized statement. For example, the following synchronized instance method in (a) is equivalent to (b):

synchronized block

public synchronized void xMethod() { // method body }

public void xMethod() { synchronized (this) { // method body } }

(a) (b)

32.7.1 Give some examples of possible resource corruption when running multiple threads. How do you synchronize conflicting threads?

32.7.2 Suppose you place the statement in line 26 of Listing 32.4 inside a synchronized block to avoid race conditions, as follows:

synchronized (this) { account.deposit(1); }

Will it work?

32.8 Synchronization Using Locks Locks and conditions can be explicitly used to synchronize threads.

Recall that in Listing 32.4, 100 tasks deposit a penny to the same account concurrently, which causes conflicts. To avoid it, you use the synchronized keyword in the deposit method, as follows:

public synchronized void deposit(double amount)

A synchronized instance method implicitly acquires a lock on the instance before it executes the method.

Java enables you to acquire locks explicitly, which give you more control for coordinating threads. A lock is an instance of the Lock interface, which defines the methods for acquiring and releasing locks, as shown in Figure 32.13. A lock may also use the newCondition() method to create any number of Condition objects, which can be used for thread communications.

Point Check

Point Key

lock

M32_LIAN1878_11_GE_C32.indd 16 12/29/17 5:46 PM

32.8 Synchronization Using Locks 32-17

Figure 32.13 The ReentrantLock class implements the Lock interface to represent a lock.

«interface»

java.util.concurrent.locks.Lock

Acquires the lock. Releases the lock. Returns a new Condition instance that is bound to this Lock instance.

java.util.concurrent.locks.ReentrantLock

+lock(): void +unlock(): void +newCondition(): Condition

+ReentrantLock() +ReentrantLock(fair: boolean)

Same as ReentrantLock(false). Creates a lock with the given fairness policy. When the

fairness is true, the longest-waiting thread will get the lock. Otherwise, there is no particular access order.

ReentrantLock is a concrete implementation of Lock for creating mutually exclusive locks. You can create a lock with the specified fairness policy. True fairness policies guarantee that the longest waiting thread will obtain the lock first. False fairness policies grant a lock to a waiting thread arbitrarily. Programs using fair locks accessed by many threads may have poorer overall performance than those using the default setting, but they have smaller variances in times to obtain locks and prevent starvation.

Listing 32.5 revises the program in Listing 32.7 to synchronize the account modification using explicit locks.

Listing 32.5 AccountWithSyncUsingLock.java 1 import java.util.concurrent.*; 2 import java.util.concurrent.locks.*; 3 4 public class AccountWithSyncUsingLock { 5 private static Account account = new Account(); 6 7 public static void main(String[] args) { 8 ExecutorService executor = Executors.newCachedThreadPool(); 9 10 // Create and launch 100 threads 11 for (int i = 0; i < 100; i++) { 12 executor.execute(new AddAPennyTask()); 13 } 14 15 executor.shutdown(); 16 17 // Wait until all tasks are finished 18 while (!executor.isTerminated()) { 19 } 20 21 System.out.println("What is balance? " + account.getBalance()); 22 } 23 24 // A thread for adding a penny to the account 25 public static class AddAPennyTask implements Runnable { 26 public void run() { 27 account.deposit(1); 28 } 29 } 30

fairness policy

package for locks

M32_LIAN1878_11_GE_C32.indd 17 12/29/17 5:46 PM

32-18 Chapter 32 Multithreading and Parallel Programming

31 // An inner class for Account 32 public static class Account { 33 private static Lock lock = new ReentrantLock(); // Create a lock 34 private int balance = 0; 35 36 public int getBalance() { 37 return balance; 38 } 39 40 public void deposit(int amount) { 41 lock.lock(); // Acquire the lock 42 43 try { 44 int newBalance = balance + amount; 45 46 // This delay is deliberately added to magnify the 47 // data-corruption problem and make it easy to see. 48 Thread.sleep(5); 49 50 balance = newBalance; 51 } 52 catch (InterruptedException ex) { 53 } 54 finally { 55 lock.unlock(); // Release the lock 56 } 57 } 58 } 59 }

Line 33 creates a lock, line 41 acquires the lock, and line 55 releases the lock.

Tip It is a good practice to always immediately follow a call to lock() with a try-catch block and release the lock in the finally clause, as shown in lines 41–56, to ensure that the lock is always released.

Listing 32.5 can be implemented using a synchronize method for deposit rather than using a lock. In general, using synchronized methods or statements is simpler than using explicit locks for mutual exclusion. However, using explicit locks is more intuitive and flexible to synchronize threads with conditions, as you will see in the next section.

32.8.1 How do you create a lock object? How do you acquire a lock and release a lock?

32.9 Cooperation among Threads Conditions on locks can be used to coordinate thread interactions.

Thread synchronization suffices to avoid race conditions by ensuring the mutual exclusion of multiple threads in the critical region, but sometimes you also need a way for threads to cooper- ate. Conditions can be used to facilitate communications among threads. A thread can specify what to do under a certain condition. Conditions are objects created by invoking the newCon- dition() method on a Lock object. Once a condition is created, you can use its await(), signal(), and signalAll() methods for thread communications, as shown in Figure 32.14. The await() method causes the current thread to wait until the condition is signaled. The signal() method wakes up one waiting thread, and the signalAll() method wakes all waiting threads.

release the lock

Point Check

Point Key

condition

create a lock

acquire the lock

M32_LIAN1878_11_GE_C32.indd 18 12/29/17 5:46 PM

32.9 Cooperation among Threads 32-19

Let us use an example to demonstrate thread communications. Suppose you create and launch two tasks: one that deposits into an account, and one that withdraws from the same account. The withdraw task has to wait if the amount to be withdrawn is more than the current balance. Whenever new funds are deposited into the account, the deposit task notifies the withdraw thread to resume. If the amount is still not enough for a withdrawal, the withdraw thread has to continue to wait for a new deposit.

To synchronize the operations, use a lock with a condition: newDeposit (i.e., new deposit added to the account). If the balance is less than the amount to be withdrawn, the withdraw task will wait for the newDeposit condition. When the deposit task adds money to the account, the task signals the waiting withdraw task to try again. The interaction between the two tasks is shown in Figure 32.15.

thread cooperation example

Figure 32.14 The Condition interface defines the methods for performing synchronization.

«interface» java.util.concurrent.Condition

+await(): void +signal(): void +signalAll(): Condition

Causes the current thread to wait until the condition is signaled. Wakes up one waiting thread. Wakes up all waiting threads.

Figure 32.15 The condition newDeposit is used for communications between the two threads.

while (balance < withdrawAmount) newDeposit.await();

Withdraw Task

balance – = withdrawAmount

lock.unlock();

Deposit Task

lock.lock();

newDeposit.signalAll();

balance += depositAmount

lock.unlock();

lock.lock();

You create a condition from a Lock object. To use a condition, you have to first obtain a lock. The await() method causes the thread to wait and automatically releases the lock on the condition. Once the condition is right, the thread reacquires the lock and continues executing.

Assume the initial balance is 0 and the amount to deposit and withdraw are randomly gen- erated. Listing 32.6 gives the program. A sample run of the program is shown in Figure 32.16.

Figure 32.16 The withdraw task waits if there are not sufficient funds to withdraw.

M32_LIAN1878_11_GE_C32.indd 19 12/29/17 5:46 PM

32-20 Chapter 32 Multithreading and Parallel Programming

Listing 32.6 ThreadCooperation.java 1 import java.util.concurrent.*; 2 import java.util.concurrent.locks.*; 3 4 public class ThreadCooperation { 5 private static Account account = new Account(); 6 7 public static void main(String[] args) { 8 // Create a thread pool with two threads 9 ExecutorService executor = Executors.newFixedThreadPool(2); 10 executor.execute(new DepositTask()); 11 executor.execute(new WithdrawTask()); 12 executor.shutdown(); 13 14 System.out.println("Thread 1\t\tThread 2\t\tBalance"); 15 } 16 17 public static class DepositTask implements Runnable { 18 @Override // Keep adding an amount to the account 19 public void run() { 20 try { // Purposely delay it to let the withdraw method proceed 21 while (true) { 22 account.deposit((int)(Math.random() * 10) + 1); 23 Thread.sleep(1000); 24 } 25 } 26 catch (InterruptedException ex) { 27 ex.printStackTrace(); 28 } 29 } 30 } 31 32 public static class WithdrawTask implements Runnable { 33 @Override // Keep subtracting an amount from the account 34 public void run() { 35 while (true) { 36 account.withdraw((int)(Math.random() * 10) + 1); 37 } 38 } 39 } 40 41 // An inner class for account 42 private static class Account { 43 // Create a new lock 44 private static Lock lock = new ReentrantLock(); 45 46 // Create a condition 47 private static Condition newDeposit = lock.newCondition(); 48 49 private int balance = 0; 50 51 public int getBalance() { 52 return balance; 53 } 54 55 public void withdraw(int amount) { 56 lock.lock(); // Acquire the lock 57 try { 58 while (balance < amount) {

create a lock

create a condition

acquire the lock

create two threads

M32_LIAN1878_11_GE_C32.indd 20 12/29/17 5:46 PM

32.9 Cooperation among Threads 32-21

59 System.out.println("\t\t\tWait for a deposit"); 60 newDeposit.await(); 61 } 62 63 balance -= amount; 64 System.out.println("\t\t\tWithdraw " + amount + 65 "\t\t" + getBalance()); 66 } 67 catch (InterruptedException ex) { 68 ex.printStackTrace(); 69 } 70 finally { 71 lock.unlock(); // Release the lock 72 } 73 } 74 75 public void deposit(int amount) { 76 lock.lock(); // Acquire the lock 77 try { 78 balance += amount; 79 System.out.println("Deposit " + amount + 80 "\t\t\t\t\t" + getBalance()); 81 82 // Signal thread waiting on the condition 83 newDeposit.signalAll(); 84 } 85 finally { 86 lock.unlock(); // Release the lock 87 } 88 } 89 } 90 }

The example creates a new inner class named Account to model the account with two meth- ods, deposit(int) and withdraw(int), a class named DepositTask to add an amount to the balance, a class named WithdrawTask to withdraw an amount from the balance, and a main class that creates and launches two threads.

The program creates and submits the deposit task (line 10) and the withdraw task (line 11). The deposit task is purposely put to sleep (line 23) to let the withdraw task run. When there are not enough funds to withdraw, the withdraw task waits (line 59) for notification of the balance change from the deposit task (line 83).

A lock is created in line 44. A condition named newDeposit on the lock is created in line 47. A condition is bound to a lock. Before waiting or signaling the condition, a thread must first acquire the lock for the condition. The withdraw task acquires the lock in line 56, waits for the newDeposit condition (line 60) when there is not a sufficient amount to withdraw, and releases the lock in line 71. The deposit task acquires the lock in line 76 and signals all waiting threads (line 83) for the newDeposit condition after a new deposit is made.

What will happen if you replace the while loop in lines 58–61 with the following if statement?

if (balance < amount) { System.out.println("\t\t\tWait for a deposit"); newDeposit.await(); }

The deposit task will notify the withdraw task whenever the balance changes. (balance < amount) may still be true when the withdraw task is awakened. Using the if statement would

acquire the lock

signal threads

release the lock

wait on the condition

release the lock

M32_LIAN1878_11_GE_C32.indd 21 12/29/17 5:46 PM

32-22 Chapter 32 Multithreading and Parallel Programming

lead to incorrect withdraw. Using the loop statement, the withdraw task will have a chance to recheck the condition.

Caution Once a thread invokes await() on a condition, the thread waits for a signal to resume. If you forget to call signal() or signalAll() on the condition, the thread will wait forever.

Caution A condition is created from a Lock object. To invoke its method (e.g., await(), signal(), and signalAll()), you must first own the lock. If you invoke these methods without acquiring the lock, an IllegalMonitorStateException will be thrown.

Locks and conditions were introduced in Java 5. Prior to Java 5, thread communications were programmed using the object’s built-in monitors. Locks and conditions are more powerful and flexible than the built-in monitor, so will not need to use monitors. However, if you are work- ing with legacy Java code, you may encounter Java’s built-in monitor.

A monitor is an object with mutual exclusion and synchronization capabilities. Only one thread can execute a method at a time in the monitor. A thread enters the monitor by acquiring a lock on it and exits by releasing the lock. Any object can be a monitor. An object becomes a monitor once a thread locks it. Locking is implemented using the synchronized keyword on a method or a block. A thread must acquire a lock before executing a synchronized method or block. A thread can wait in a monitor if the condition is not right for it to continue executing in the monitor. You can invoke the wait() method on the monitor object to release the lock so some other thread can get in the monitor and perhaps change the monitor’s state. When the condition is right, the other thread can invoke the notify() or notifyAll() method to signal one or all waiting threads to regain the lock and resume execution. The template for invoking these methods is shown in Figure 32.17.

ever-waiting threads

IllegalMonitorState Exception

Java’s built-in monitor monitor

Figure 32.17 The wait(), notify(), and notifyAll() methods coordinate thread communication.

synchronized (anObject) { try {

// Wait for the condition to become true while (!condition)

anObject.wait();

// Do something when condition is true } catch (InterruptedException ex) {

ex.printStackTrace(); }

}

Task 1

synchronized (anObject) { // When condition becomes true anObject.notify(); or anObject.notifyAll(); ...

}

Task 2

resume

The wait(), notify(), and notifyAll() methods must be called in a synchronized method or a synchronized block on the receiving object of these methods. Otherwise, an IllegalMonitorStateException will occur.

When wait() is invoked, it pauses the thread and simultaneously releases the lock on the object. When the thread is restarted after being notified, the lock is automatically reacquired.

The wait(), notify(), and notifyAll() methods on an object are analogous to the await(), signal(), and signalAll() methods on a condition.

M32_LIAN1878_11_GE_C32.indd 22 12/29/17 5:46 PM

32.10 Case Study: Producer/Consumer 32-23

32.9.1 How do you create a condition on a lock? What are the await(), signal(), and signalAll() methods for?

32.9.2 What would happen if the while loop in line 58 of Listing 32.6 was changed to an if statement?

Point Check

while (balance < amount) Replaced by

if (balance < amount)

32.9.3 Why does the following class have a syntax error? public class Test implements Runnable { public static void main(String[] args) { new Test(); } public Test() throws InterruptedException { Thread thread = new Thread(this); thread.sleep(1000); }

public synchronized void run() { } }

32.9.4 What is a possible cause for IllegalMonitorStateException? 32.9.5 Can wait(), notify(), and notifyAll() be invoked from any object? What is

the purpose of these methods?

32.9.6 What is wrong in the following code? synchronized (object1) { try { while (!condition) object2.wait(); } catch (InterruptedException ex) { } }

32.10 Case Study: Producer/Consumer This section gives the classic Consumer/Producer example for demonstrating thread coordination.

Suppose that you use a buffer to store integers and that the buffer size is limited. The buffer provides the method write(int) to add an int value to the buffer and the method read() to read and delete an int value from the buffer. To synchronize the operations, use a lock with two conditions: notEmpty (i.e., the buffer is not empty) and notFull (i.e., the buffer is not full). When a task adds an int to the buffer, if the buffer is full, the task will wait for the notFull condition. When a task reads an int from the buffer, if the buffer is empty, the task will wait for the notEmpty condition. The interaction between the two tasks is shown in Figure 32.18.

Listing 32.7 presents the complete program. The program contains the Buffer class (lines 50–101) and two tasks for repeatedly adding and consuming numbers to and from the buffer (lines 16–47). The write(int) method (lines 62–79) adds an integer to the buffer. The read() method (lines 81–100) deletes and returns an integer from the buffer.

Point Key

M32_LIAN1878_11_GE_C32.indd 23 12/29/17 5:46 PM

32-24 Chapter 32 Multithreading and Parallel Programming

The buffer is actually a first-in, first-out queue (lines 52 and 53). The conditions notEmpty and notFull on the lock are created in lines 59 and 60. The conditions are bound to a lock. A lock must be acquired before a condition can be applied. If you use the wait() and notify() methods to rewrite this example, you have to designate two objects as monitors.

Listing 32.7 ConsumerProducer.java 1 import java.util.concurrent.*; 2 import java.util.concurrent.locks.*; 3 4 public class ConsumerProducer { 5 private static Buffer buffer = new Buffer(); 6 7 public static void main(String[] args) { 8 // Create a thread pool with two threads 9 ExecutorService executor = Executors.newFixedThreadPool(2); 10 executor.execute(new ProducerTask()); 11 executor.execute(new ConsumerTask()); 12 executor.shutdown(); 13 } 14 15 // A task for adding an int to the buffer 16 private static class ProducerTask implements Runnable { 17 public void run() { 18 try { 19 int i = 1; 20 while (true) { 21 System.out.println("Producer writes " + i); 22 buffer.write(i++); // Add a value to the buffer 23 // Put the thread into sleep 24 Thread.sleep((int)(Math.random() * 10000)); 25 } 26 } 27 catch (InterruptedException ex) { 28 ex.printStackTrace(); 29 } 30 } 31 } 32 33 // A task for reading and deleting an int from the buffer 34 private static class ConsumerTask implements Runnable { 35 public void run() {

create a buffer

create two threads

producer task

consumer task

Figure 32.18 The conditions notFull and notEmpty are used to coordinate task interactions.

while (count == CAPACITY) notFull.await();

Task for adding an int

Add an int to the buffer

notEmpty.signal();

while (count == 0) notEmpty.await();

Task for deleting an int

Delete an int from the buffer

notFull.signal();

M32_LIAN1878_11_GE_C32.indd 24 12/29/17 5:46 PM

32.10 Case Study: Producer/Consumer 32-25

36 try { 37 while (true) { 38 System.out.println("\t\t\tConsumer reads " + buffer.read()); 39 // Put the thread into sleep 40 Thread.sleep((int)(Math.random() * 10000)); 41 } 42 } 43 catch (InterruptedException ex) { 44 ex.printStackTrace(); 45 } 46 } 47 } 48 49 // An inner class for buffer 50 private static class Buffer { 51 private static final int CAPACITY = 1; // buffer size 52 private java.util.LinkedList<Integer> queue = 53 new java.util.LinkedList<>(); 54 55 // Create a new lock 56 private static Lock lock = new ReentrantLock(); 57 58 // Create two conditions 59 private static Condition notEmpty = lock.newCondition(); 60 private static Condition notFull = lock.newCondition(); 61 62 public void write(int value) { 63 lock.lock(); // Acquire the lock 64 try { 65 while (queue.size() == CAPACITY) { 66 System.out.println("Wait for notFull condition"); 67 notFull.await(); 68 } 69 70 queue.offer(value); 71 notEmpty.signal(); // Signal notEmpty condition 72 } 73 catch (InterruptedException ex) { 74 ex.printStackTrace(); 75 } 76 finally { 77 lock.unlock(); // Release the lock 78 } 79 } 80 81 public int read() { 82 int value = 0; 83 lock.lock(); // Acquire the lock 84 try { 85 while (queue.isEmpty()) { 86 System.out.println("\t\t\tWait for notEmpty condition"); 87 notEmpty.await(); 88 } 89 90 value = queue.remove(); 91 notFull.signal(); // Signal notFull condition 92 } 93 catch (InterruptedException ex) { 94 ex.printStackTrace(); 95 }

signal notFull

wait for notEmpty

acquire the lock

release the lock

signal notEmpty

wait for notFull

acquire the lock

create a condition create a condition

create a lock

M32_LIAN1878_11_GE_C32.indd 25 12/29/17 5:46 PM

32-26 Chapter 32 Multithreading and Parallel Programming

96 finally { 97 lock.unlock(); // Release the lock 98 return value; 99 } 100 } 101 } 102 }

A sample run of the program is shown in Figure 32.19.

release the lock

Figure 32.19 Locks and conditions are used for communications between the Producer and Consumer threads.

Figure 32.20 BlockingQueue is a subinterface of Queue.

Inserts an element to the tail of the queue. Waits if the queue is full.

Retrieves and removes the head of this queue. Waits if the queue is empty.

+put(element: E): void

+take(): E

«interface» java.util.Collection<E>

«interface» java.util.Queue<E>

«interface» java.util.concurrent.BlockingQueue<E>

32.10.1 Can the read and write methods in the Buffer class be executed concurrently? 32.10.2 When invoking the read method, what happens if the queue is empty? 32.10.3 When invoking the write method, what happens if the queue is full?

32.11 Blocking Queues Java Collections Framework provides ArrayBlockingQueue, LinkedBlocking- Queue, and PriorityBlockingQueue for supporting blocking queues.

Queues and priority queues were introduced in Section 20.9. A blocking queue causes a thread to block when you try to add an element to a full queue or to remove an element from an empty queue. The BlockingQueue interface extends java.util.Queue and provides the synchro- nized put and take methods for adding an element to the tail of the queue and for removing an element from the head of the queue, as shown in Figure 32.20.

Point Check

Point Key

blocking queue

M32_LIAN1878_11_GE_C32.indd 26 12/29/17 5:46 PM

32.11 Blocking Queues 32-27

Note The put method will never block an unbounded LinkedBlockingQueue or PriorityBlockingQueue.

Listing 32.8 gives an example of using an ArrayBlockingQueue to simplify the Con- sumer/Producer example in Listing 32.10. Line 5 creates an ArrayBlockingQueue to store integers. The Producer thread puts an integer into the queue (line 22) and the Consumer thread takes an integer from the queue (line 38).

Listing 32.8 ConsumerProducerUsingBlockingQueue.java 1 import java.util.concurrent.*; 2 3 public class ConsumerProducerUsingBlockingQueue { 4 private static ArrayBlockingQueue<Integer> buffer = 5 new ArrayBlockingQueue<>(2); 6 7 public static void main(String[] args) { 8 // Create a thread pool with two threads 9 ExecutorService executor = Executors.newFixedThreadPool(2); 10 executor.execute(new ProducerTask()); 11 executor.execute(new ConsumerTask()); 12 executor.shutdown(); 13 } 14 15 // A task for adding an int to the buffer 16 private static class ProducerTask implements Runnable { 17 public void run() { 18 try { 19 int i = 1; 20 while (true) { 21 System.out.println("Producer writes " + i); 22 buffer.put(i++); // Add any value to the buffer, say, 1 23 // Put the thread into sleep

unbounded queue

create a buffer

create two threads

producer task

put

consumer task

Figure 32.21 ArrayBlockingQueue, LinkedBlockingQueue, and PriorityBlockingQueue are concrete blocking queues.

«interface» java.util.concurrent.BlockingQueue<E>

+ArrayBlockingQueue(capacity: int)

+ArrayBlockingQueue(capacity: int, fair: boolean)

ArrayBlockingQueue<E>

+LinkedBlockingQueue()

+LinkedBlockingQueue(capacity: int)

LinkedBlockingQueue<E>

+PriorityBlockingQueue()

+PriorityBlockingQueue(capacity: int)

PriorityBlockingQueue<E>

Three concrete blocking queues—ArrayBlockingQueue, LinkedBlockingQueue, and PriorityBlockingQueue—are provided in Java, as shown in Figure 32.21. All are in the java.util.concurrent package. ArrayBlockingQueue implements a blocking queue using an array. You have to specify a capacity or an optional fairness to construct an Array- BlockingQueue. LinkedBlockingQueue implements a blocking queue using a linked list. You can create an unbounded or bounded LinkedBlockingQueue. PriorityBlocking- Queue is a priority queue. You can create an unbounded or bounded priority queue.

M32_LIAN1878_11_GE_C32.indd 27 12/29/17 5:46 PM

32-28 Chapter 32 Multithreading and Parallel Programming

38 System.out.println("\t\t\tConsumer reads " + buffer.take());

Figure 32.22 A limited number of threads can access a shared resource controlled by a semaphore.

Acquire a permit from a semaphore. Wait if the permit is not available.

Release the permit to the semaphore.

A thread accessing a shared resource.

Access the resource

semaphore.acquire();

semaphore.release();

39 // Put the thread into sleep 40 Thread.sleep((int)(Math.random() * 10000)); 41 } 42 } 43 catch (InterruptedException ex) { 44 ex.printStackTrace(); 45 } 46 } 47 } 48 }

In Listing 32.7, you used locks and conditions to synchronize the Producer and Consumer threads. This program does not use locks and conditions because synchronization is already implemented in ArrayBlockingQueue.

32.11.1 What is a blocking queue? What blocking queues are supported in Java? 32.11.2 What method do you use to add an element to an ArrayBlockingQueue? What

happens if the queue is full?

32.11.3 What method do you use to retrieve an element from an ArrayBlockingQueue? What happens if the queue is empty?

32.12 Semaphores Semaphores can be used to restrict the number of threads that access a shared resource.

In computer science, a semaphore is an object that controls the access to a common resource. Before accessing the resource, a thread must acquire a permit from the semaphore. After finishing with the resource, the thread must return the permit back to the semaphore, as shown in Figure 32.22.

take

Consumer task

Point Check

Point Key

semaphore

24 Thread.sleep((int)(Math.random() * 10000)); 25 } 26 } 27 catch (InterruptedException ex) { 28 ex.printStackTrace(); 29 } 30 } 31 } 32 33 // A task for reading and deleting an int from the buffer 34 private static class ConsumerTask implements Runnable { 35 public void run() { 36 try { 37 while (true) {

M32_LIAN1878_11_GE_C32.indd 28 12/29/17 5:46 PM

32.12 Semaphores 32-29

A semaphore with just one permit can be used to simulate a mutually exclusive lock. Listing 32.9 revises the Account inner class in Listing 32.9 using a semaphore to ensure that only one thread at a time can access the deposit method.

Listing 32.9 New Account Inner Class 1 // An inner class for Account 2 private static class Account { 3 // Create a semaphore 4 private static Semaphore semaphore = new Semaphore(1); 5 private int balance = 0; 6 7 public int getBalance() { 8 return balance; 9 } 10 11 public void deposit(int amount) { 12 try { 13 semaphore.acquire(); // Acquire a permit 14 int newBalance = balance + amount; 15 16 // This delay is deliberately added to magnify the 17 // data-corruption problem and make it easy to see 18 Thread.sleep(5); 19 20 balance = newBalance; 21 } 22 catch (InterruptedException ex) { 23 } 24 finally { 25 semaphore.release(); // Release a permit 26 } 27 } 28 }

A semaphore with one permit is created in line 4. A thread first acquires a permit when execut- ing the deposit method in line 13. After the balance is updated, the thread releases the permit in line 25. It is a good practice to always place the release() method in the finally clause to ensure that the permit is finally released even in the case of exceptions.

create a semaphore

acquire a permit

release a permit

Figure 32.23 The Semaphore class contains the methods for accessing a semaphore.

java.util.concurrent.Semaphore

+Semaphore(numberOfPermits: int)

+Semaphore(numberOfPermits: int, fair: boolean)

+acquire(): void

+release(): void

Creates a semaphore with the speci�ed number of permits. The fairness policy is false.

Creates a semaphore with the speci�ed number of permits and the fairness policy.

Acquires a permit from this semaphore. If no permit is available, the thread is blocked until one is available.

Releases a permit back to the semaphore.

To create a semaphore, you have to specify the number of permits with an optional fair- ness policy, as shown in Figure 32.23. A task acquires a permit by invoking the semaphore’s acquire() method and releases the permit by invoking the semaphore’s release() method. Once a permit is acquired, the total number of available permits in a semaphore is reduced by 1. Once a permit is released, the total number of available permits in a semaphore is increased by 1.

M32_LIAN1878_11_GE_C32.indd 29 12/29/17 5:46 PM

32-30 Chapter 32 Multithreading and Parallel Programming

32.12.1 What are the similarities and differences between a lock and a semaphore? 32.12.2 How do you create a semaphore that allows three concurrent threads? How do

you acquire a semaphore? How do you release a semaphore?

32.13 Avoiding Deadlocks Deadlocks can be avoided by using a proper resource ordering.

Sometimes two or more threads need to acquire the locks on several shared objects. This could cause a deadlock, in which each thread has the lock on one of the objects and is waiting for the lock on the other object. Consider the scenario with two threads and two objects, as shown in Figure 32.24. Thread 1 has acquired a lock on object1, and Thread 2 has acquired a lock on object2. Now Thread 1 is waiting for the lock on object2, and Thread 2 for the lock on object1. Each thread waits for the other to release the lock it needs and until that happens, neither can continue to run.

Point Check

Point Key

deadlock

Figure 32.24 Thread 1 and Thread 2 are deadlocked.

synchronized (object1) {

// do something here

synchronized (object2) {

// do something here }

}

Thread 1

synchronized (object2) {

// do something here

synchronized (object1) { // do something here

} }

Thread 2 Step

1 2 3 4 5 6

Wait for Thread 2 to release the lock on object2

Wait for Thread 1 to release the lock on object1

Deadlock is easily avoided by using a simple technique known as resource ordering. With this technique, you assign an order to all the objects whose locks must be acquired and ensure that each thread acquires the locks in that order. For example in Figure 32.24, suppose the objects are ordered as object1 and object2. Using the resource-ordering technique, Thread 2 must acquire a lock on object1 first, then on object2. Once Thread 1 acquires a lock on object1, Thread 2 has to wait for a lock on object1. Thus, Thread 1 will be able to acquire a lock on object2 and no deadlock will occur.

32.13.1 What is a deadlock? How can you avoid deadlock?

32.14 Thread States A thread state indicates the status of thread.

Tasks are executed in threads. Threads can be in one of the five states: New, Ready, Running, Blocked, or Finished (see Figure 32.25).

When a thread is newly created, it enters the New state. After a thread is started by calling its start() method, it enters the Ready state. A ready thread is runnable but may not be run- ning yet. The operating system has to allocate CPU time to it.

When a ready thread begins executing, it enters the Running state. A running thread can enter the Ready state if its given CPU time expires or its yield() method is called.

resource ordering

Point Check

Point Key

M32_LIAN1878_11_GE_C32.indd 30 12/29/17 5:46 PM

32.15 Synchronized Collections 32-31

A thread can enter the Blocked state (i.e., become inactive) for several reasons. It may have invoked the join(), sleep(), or wait() method. It may be waiting for an I/O operation to finish. A blocked thread may be reactivated when the action inactivating it is reversed. For example, if a thread has been put to sleep and the sleep time has expired, the thread is reactivated and enters the Ready state.

Finally, a thread is Finished if it completes the execution of its run() method. The isAlive() method is used to find out the state of a thread. It returns true if a thread

is in the Ready, Blocked, or Running state; it returns false if a thread is new and has not started or if it is finished.

The interrupt() method interrupts a thread in the following way: If a thread is currently in the Ready or Running state, its interrupted flag is set; if a thread is currently blocked, it is awakened and enters the Ready state, and a java.lang.InterruptedException is thrown.

32.14.1 What is a thread state? Describe the states for a thread.

32.15 Synchronized Collections Java Collections Framework provides synchronized collections for lists, sets, and maps.

The classes in the Java Collections Framework are not thread-safe; that is, their contents may become corrupted if they are accessed and updated concurrently by multiple threads. You can protect the data in a collection by locking the collection or by using synchronized collections.

The Collections class provides six static methods for wrapping a collection into a syn- chronized version, as shown in Figure 32.26. The collections created using these methods are called synchronization wrappers.

Point Check

Point Key

synchronized collection

synchronization wrapper

Figure 32.25 A thread can be in one of the five states: New, Ready, Running, Blocked, or Finished.

Thread created start() run()

Wait for target to �nish

Ready

Running

FinishedNew

Wait for time out

Wait to be noti�ed

run() completed yield(), or time out

sleep()join() wait() Target �nished

SignaledTime out

Blocked

Figure 32.26 You can obtain synchronized collections using the methods in the Collections class.

java.util.Collections

+synchronizedCollection(c: Collection): Collection

+synchronizedList(list: List): List

+synchronizedMap(m: Map): Map

+synchronizedSet(s: Set): Set

+synchronizedSortedMap(s: SortedMap): SortedMap

+synchronizedSortedSet(s: SortedSet): SortedSet

Returns a synchronized collection.

Returns a synchronized list from the speci�ed list.

Returns a synchronized map from the speci�ed map.

Returns a synchronized set from the speci�ed set.

Returns a synchronized sorted map from the speci�ed sorted map. Returns a synchronized sorted set.

M32_LIAN1878_11_GE_C32.indd 31 12/29/17 5:46 PM

32-32 Chapter 32 Multithreading and Parallel Programming

Invoking synchronizedCollection(Collection c) returns a new Collection object, in which all the methods that access and update the original collection c are synchro- nized. These methods are implemented using the synchronized keyword. For example, the add method is implemented like this:

public boolean add(E o) { synchronized (this) { return c.add(o); } }

Synchronized collections can be safely accessed and modified by multiple threads concurrently.

Note The methods in java.util.Vector, java.util.Stack, and java.util. Hashtable are already synchronized. These are old classes introduced in JDK 1.0. Starting with JDK 1.5, you should use java.util.ArrayList to replace Vector, java.util.LinkedList to replace Stack, and java.util.Map to replace Hashtable. If synchronization is needed, use a synchronization wrapper.

The synchronization wrapper classes are thread-safe, but the iterator is fail-fast. This means that if you are using an iterator to traverse a collection while the underlying collection is being modified by another thread, then the iterator will immediately fail by throwing java.util. ConcurrentModificationException, which is a subclass of RuntimeException. To avoid this error, you need to create a synchronized collection object and acquire a lock on the object when traversing it. For example, to traverse a set, you have to write the code like this:

Set hashSet = Collections.synchronizedSet(new HashSet()); syn chronized (hashSet) { // Must synchronize it Iterator iterator = hashSet.iterator(); while (iterator.hasNext()) { System.out.println(iterator.next()); } }

Failure to do so may result in nondeterministic behavior, such as a Con current- Modi ficationException.

32.15.1 What is a synchronized collection? Is ArrayList synchronized? How do you make it synchronized?

32.15.2 Explain why an iterator is fail-fast.

32.16 Parallel Programming The Fork/Join Framework is used for parallel programming in Java.

Section 7.12 introduced the Arrays.sort and Arrays.parallelSort method for sorting an array. The parallelSort method utilizes multiple processors to reduce sort time. Chapter 22 introduced parallel streams for executing stream operations in parallel to speed up processing using multiple processors. The parallel processing are implemented using the Fork/ Join Framework. This section, introduces the new Fork/Join Framework so you can write own code for parallel programming.

The Fork/Join Framework is illustrated in Figure 32.27 (the diagram resembles a fork, hence its name). A problem is divided into nonoverlapping subproblems, which can be solved indepen- dently in parallel. The solutions to all subproblems are then joined to obtain an overall solution for the problem. This is the parallel implementation of the divide-and-conquer approach. In JDK 7’s Fork/Join Framework, a fork can be viewed as an independent task that runs on a thread.

fail-fast

Point Check

Point Key

JDK 7 feature

Fork/Join Framework

M32_LIAN1878_11_GE_C32.indd 32 12/29/17 5:46 PM

32.16 Parallel Programming 32-33

The framework defines a task using the ForkJoinTask class, as shown in Figure 32.28 and executes a task in an instance of ForkJoinPool, as shown in Figure 32.29.

ForkJoinTask

ForkJoinPool

Figure 32.27 The nonoverlapping subproblems are solved in parallel.

Subproblem

Problem

Fork

Solution

Join

Figure 32.28 The ForkJoinTask class defines a task for asynchronous execution.

«interface» java.util.concurrent.Future<V>

+cancel(interrupt: boolean): boolean +get(): V

+isDone(): boolean

Attempts to cancel this task. Waits if needed for the computation to complete and returns the result. Returns true if this task is completed.

java.util.concurrent.ForkJoinTask<V>

+adapt(Runnable task): ForkJoinTask<V> +fork(): ForkJoinTask<V> +join(): V +invoke(): V

+invokeAll(tasks ForkJoinTask<?>…): void

Returns a ForkJoinTask from a runnable task. Arranges asynchronous execution of the task. Returns the result of computations when it is done. Performs the task and awaits for its completion, and returns its result. Forks the given tasks and returns when all tasks are completed.

java.util.concurrent.RecursiveAction<V>

#compute(): void De�nes how task is performed.

java.util.concurrent.RecursiveTask<V>

#compute(): V De�nes how task is performed. Returns the value after the task is completed.

Figure 32.29 The ForkJoinPool executes Fork/Join tasks.

«interface» java.util.concurrent.ExecutorService

java.util.concurrent.ForkJoinPool

+ForkJoinPool() +ForkJoinPool(parallelism: int) +invoke(ForkJoinTask<T>): T

Creates a ForkJoinPool with all available processors. Creates a ForkJoinPool with the speci�ed number of processors. Performs the task and returns its result upon completion.

See Figure 32.8

M32_LIAN1878_11_GE_C32.indd 33 12/29/17 5:46 PM

32-34 Chapter 32 Multithreading and Parallel Programming

ForkJoinTask is the abstract base class for tasks. A ForkJoinTask is a thread-like entity, but it is much lighter than a normal thread because huge numbers of tasks and subtasks can be executed by a small number of actual threads in a ForkJoinPool. The tasks are primar- ily coordinated using fork() and join(). Invoking fork() on a task arranges asynchro- nous execution and invoking join() waits until the task is completed. The invoke() and invokeAll(tasks) methods implicitly invoke fork() to execute the task and join() to wait for the tasks to complete and return the result, if any. Note the static method invokeAll takes a variable number of ForkJoinTask arguments using the ... syntax, which is intro- duced in Section 7.9.

The Fork/Join Framework is designed to parallelize divide-and-conquer solutions, which are naturally recursive. RecursiveAction and RecursiveTask are two subclasses of ForkJoinTask. To define a concrete task class, your class should extend RecursiveAction or RecursiveTask. RecursiveAction is for a task that doesn’t return a value and RecursiveTask is for a task that does return a value. Your task class should override the compute() method to specify how a task is performed.

We now use a merge sort to demonstrate how to develop parallel programs using the Fork/ Join Framework. The merge sort algorithm (introduced in Section 25.3) divides the array into two halves and applies a merge sort on each half recursively. After the two halves are sorted, the algorithm merges them. Listing 32.10 gives a parallel implementation of the merge sort algorithm and compares its execution time with a sequential sort.

Listing 32.10 ParallelMergeSort.java 1 import java.util.concurrent.RecursiveAction; 2 import java.util.concurrent.ForkJoinPool; 3 4 public class ParallelMergeSort { 5 public static void main(String[] args) { 6 final int SIZE = 7000000; 7 int[] list1 = new int[SIZE]; 8 int[] list2 = new int[SIZE]; 9 10 for (int i = 0; i < list1.length; i++) 11 list1[i] = list2[i] = (int)(Math.random() * 10000000); 12 13 long startTime = System.currentTimeMillis(); 14 parallelMergeSort(list1); // Invoke parallel merge sort 15 long endTime = System.currentTimeMillis(); 16 System.out.println("\nParallel time with " 17 + Runtime.getRuntime().availableProcessors() + 18 " processors is " + (endTime - startTime) + " milliseconds"); 19 20 startTime = System.currentTimeMillis(); 21 MergeSort.mergeSort(list2); // MergeSort is in Listing 23.5 22 endTime = System.currentTimeMillis(); 23 System.out.println("\nSequential time is " + 24 (endTime - startTime) + " milliseconds"); 25 } 26 27 public static void parallelMergeSort(int[] list) { 28 RecursiveAction mainTask = new SortTask(list); 29 ForkJoinPool pool = new ForkJoinPool(); 30 pool.invoke(mainTask); 31 } 32 33 private static class SortTask extends RecursiveAction { 34 private final int THRESHOLD = 500;

RecursiveAction

RecursiveTask

invoke parallel sort

invoke sequential sort

create a ForkJoinTask create a ForkJoinPool execute a task

define concrete ForkJoinTask

M32_LIAN1878_11_GE_C32.indd 34 12/29/17 5:46 PM

32.16 Parallel Programming 32-35

35 private int[] list; 36 37 SortTask(int[] list) { 38 this.list = list; 39 } 40 41 @Override 42 protected void compute() { 43 if (list.length < THRESHOLD) 44 java.util.Arrays.sort(list); 45 else { 46 // Obtain the first half 47 int[] firstHalf = new int[list.length / 2]; 48 System.arraycopy(list, 0, firstHalf, 0, list.length / 2); 49 50 // Obtain the second half 51 int secondHalfLength = list.length - list.length / 2; 52 int[] secondHalf = new int[secondHalfLength]; 53 System.arraycopy(list, list.length / 2, 54 secondHalf, 0, secondHalfLength); 55 56 // Recursively sort the two halves 57 invokeAll(new SortTask(firstHalf), 58 new SortTask(secondHalf)); 59 60 // Merge firstHalf with secondHalf into list 61 MergeSort.merge(firstHalf, secondHalf, list); 62 } 63 } 64 } 65 }

perform the task

sort a small list

split into two parts

solve each part

merge two parts

Parallel time with two processors is 2829 milliseconds

Sequential time is 4751 milliseconds

Since the sort algorithm does not return a value, we define a concrete ForkJoinTask class by extending RecursiveAction (lines 33–64). The compute method is overridden to imple- ment a recursive merge sort (lines 42–63). If the list is small, it is more efficient to be solved sequentially (line 44). For a large list, it is split into two halves (lines 47–54). The two halves are sorted concurrently (lines 57 and 58) and then merged (line 61).

The program creates a main ForkJoinTask (line 28), a ForkJoinPool (line 29), and places the main task for execution in a ForkJoinPool (line 30). The invoke method will return after the main task is completed.

When executing the main task, the task is split into subtasks, and the subtasks are invoked using the invokeAll method (lines 57 and 58). The invokeAll method will return after all the subtasks are completed. Note each subtask is further split into smaller tasks recursively. Huge numbers of subtasks may be created and executed in the pool. The Fork/Join Framework automatically executes and coordinates all the tasks efficiently.

The MergeSort class is defined in Listing 23.5. The program invokes MergeSort.merge to merge two sorted sublists (line 61). The program also invokes MergeSort.mergeSort (line 21) to sort a list using merge sort sequentially. You can see that the parallel sort is much faster than the sequential sort.

Note the loop for initializing the list can also be parallelized. However, you should avoid using Math.random() in the code because it is synchronized and cannot be executed in parallel (see Programming Exercise 32.12). The parallelMergeSort method only sorts an

M32_LIAN1878_11_GE_C32.indd 35 12/29/17 5:46 PM

32-36 Chapter 32 Multithreading and Parallel Programming

array of int values, but you can modify it to become a generic method (see Programming Exercise 32.13).

In general, a problem can be solved in parallel using the following pattern:

if (the program is small) solve it sequentially; else { divide the problem into nonoverlapping subproblems; solve the subproblems concurrently; combine the results from subproblems to solve the whole problem; }

Listing 32.11 develops a parallel method that finds the maximal number in a list.

Listing 32.11 ParallelMax.java 1 import java.util.concurrent.*; 2 3 public class ParallelMax { 4 public static void main(String[] args) { 5 // Create a list 6 final int N = 9000000; 7 int[] list = new int[N]; 8 for (int i = 0; i < list.length; i++) 9 list[i] = i; 10 11 long startTime = System.currentTimeMillis(); 12 System.out.println("\nThe maximal number is " + max(list)); 13 long endTime = System.currentTimeMillis(); 14 System.out.println("The number of processors is " + 15 Runtime.getRuntime().availableProcessors()); 16 System.out.println("Time is " + (endTime − startTime) 17 + " milliseconds"); 18 } 19 20 public static int max(int[] list) { 21 RecursiveTask<Integer> task = new MaxTask(list, 0, list.length); 22 ForkJoinPool pool = new ForkJoinPool(); 23 return pool.invoke(task); 24 } 25 26 private static class MaxTask extends RecursiveTask<Integer> { 27 private final static int THRESHOLD = 1000; 28 private int[] list; 29 private int low; 30 private int high; 31 32 public MaxTask(int[] list, int low, int high) { 33 this.list = list; 34 this.low = low; 35 this.high = high; 36 } 37 38 @Override 39 public Integer compute() { 40 if (high − low < THRESHOLD) { 41 int max = list[0]; 42 for (int i = low; i < high; i++) 43 if (list[i] > max) 44 max = list[i]; 45 return new Integer(max);

invoke max

create a ForkJoinTask create a ForkJoinPool execute a task

define concrete ForkJoinTask

perform the task

solve a small problem

M32_LIAN1878_11_GE_C32.indd 36 12/29/17 5:46 PM

Chapter Summary 32-37

46 } 47 else { 48 int mid = (low + high) / 2; 49 RecursiveTask<Integer> left = new MaxTask(list, low, mid); 50 RecursiveTask<Integer> right = new MaxTask(list, mid, high); 51 52 right.fork(); 53 left.fork(); 54 return new Integer(Math.max(left.join().intValue(), 55 right.join().intValue())); 56 } 57 } 58 } 59 }

split into two parts

fork right fork left join tasks

The maximal number is 8999999

The number of processors is 2

Time is 44 milliseconds

Since the algorithm returns an integer, we define a task class for fork join by extending RecursiveTask<Integer> (lines 26–58). The compute method is overridden to return the max element in a list[low..high] (lines 39–57). If the list is small, it is more efficient to be solved sequentially (lines 40–46). For a large list, it is split into two halves (lines 48–50). The tasks left and right find the maximal element in the left half and right half, respectively. Invoking fork() on the task causes the task to be executed (lines 52 and 53). The join() method awaits for the task to complete and then returns the result (lines 54 and 55).

32.16.1 How do you define a ForkJoinTask? What are the differences between RecursiveAction and RecursiveTask?

32.16.2 How do you tell the system to execute a task? 32.16.3 What method can you use to test if a task has been completed? 32.16.4 How do you create a ForkJoinPool? How do you place a task into a

ForkJoinPool?

Point Check

condition 32-18 deadlock 32-30 fail-fast 32-32 fairness policy 32-17 Fork/Join Framework 32-32 lock 32-16 monitor 32-22

multithreading 32-2 race condition 32-15 semaphore 32-28 synchronization wrapper 32-31 synchronized block 32-16 thread 32-2 thread-safe 32-15

Key Terms

ChapTer summary 1. Each task is an instance of the Runnable interface. A thread is an object that facilitates

the execution of a task. You can define a task class by implementing the Runnable interface and create a thread by wrapping a task using a Thread constructor.

2. After a thread object is created, use the start() method to start a thread, and the sleep(long) method to put a thread to sleep so other threads get a chance to run.

M32_LIAN1878_11_GE_C32.indd 37 12/29/17 5:46 PM

32-38 Chapter 32 Multithreading and Parallel Programming

3. A thread object never directly invokes the run method. The JVM invokes the run method when it is time to execute the thread. Your class must override the run method to tell the system what the thread will do when it runs.

4. To prevent threads from corrupting a shared resource, use synchronized methods or blocks. A synchronized method acquires a lock before it executes. In the case of an instance method, the lock is on the object for which the method was invoked. In the case of a static method, the lock is on the class.

5. A synchronized statement can be used to acquire a lock on any object, not just this object, when executing a block of the code in a method. This block is referred to as a synchronized block.

6. You can use explicit locks and conditions to facilitate communications among threads, as well as using the built-in monitor for objects.

7. The blocking queues (ArrayBlockingQueue, LinkedBlockingQueue, and Priority BlockingQueue) provided in the Java Collections Framework provide auto- matical synchronization for the access to a queue.

8. You can use semaphores to restrict the number of concurrent accesses to a shared resource.

9. Deadlock occurs when two or more threads acquire locks on multiple objects and each has a lock on one object and is waiting for the lock on the other object. The resource- ordering technique can be used to avoid deadlock.

10. The JDK 7’s Fork/Join Framework is designed for developing parallel programs. You can define a task class that extends RecursiveAction or RecursiveTask and execute the tasks concurrently in ForkJoinPool and obtain the overall solution after all tasks are completed.

Quiz Answer the quiz for this chapter online at the book Companion Website.

programming exerCises

Sections 32.1–32.5 *32.1 (Revise Listing 32.1) Rewrite Listing 32.1 to display the output in a text area, as

shown in Figure 32.30.

Figure 32.30 The output from three threads is displayed in a text area.

M32_LIAN1878_11_GE_C32.indd 38 12/29/17 5:46 PM

Programming Exercises 32-39

32.2 (Racing cars) Rewrite Programming Exercise 15.29 using a thread to control car racing. Compare the program with Programming Exercise 15.29 by setting the delay time to 10 in both the programs. Which one runs the animation faster?

32.3 (Raise flags) Rewrite Listing 15.13 using a thread to animate a flag being raised. Compare the program with Listing 15.13 by setting the delay time to 10 in both programs. Which one runs the animation faster?

Sections 32.8–32.12 32.4 (Synchronize threads) Write a program that launches 1,000 threads. Each thread

adds 1 to a variable sum that initially is 0. You need to pass sum by reference to each thread. In order to pass it by reference, define an Integer wrapper object to hold sum. Run the program with and without synchronization to see its effect.

32.5 (Display a running fan) Rewrite Programming Exercise 15.28 using a thread to control the fan animation.

32.6 (Bouncing balls) Rewrite Listing 15.17, BallPane.java using a thread to animate bouncing ball movements.

32.7 (Control a clock) Rewrite Programming Exercise 15.32 using a thread to control the clock animation.

32.8 (Account synchronization) Rewrite Listing 32.6, ThreadCooperation.java, using the object’s wait() and notifyAll() methods.

32.9 (Demonstrate ConcurrentModificationException) The iterator is fail-fast. Write a program to demonstrate it by creating two threads that concurrently access and modify a set. The first thread creates a hash set filled with numbers and adds a new number to the set every second. The second thread obtains an iterator for the set and traverses the set back and forth through the iterator every second. You will receive a ConcurrentModificationException because the underlying set is being modified in the first thread while the set in the second thread is being traversed.

*32.10 (Use synchronized sets) Using synchronization, correct the problem in the preceding exercise so that the second thread does not throw a Concurrent ModificationException.

Section 32.15 *32.11 (Demonstrate deadlock) Write a program that demonstrates deadlock.

Section 32.18 *32.12 (Parallel array initializer) Implement the following method using the Fork/Join

Framework to assign random values to the list.

public static void parallelAssignValues(double[] list)

Write a test program that creates a list with 9,000,000 elements and invokes parallelAssignValues to assign random values to the list. Also implement a sequential algorithm and compare the execution time of the two. Note if you use Math.random(), your parallel code execution time will be worse than the sequen- tial code execution time because Math.random() is synchronized and cannot be executed in parallel. To fix this problem, create a Random object for assigning random values to a small list.

M32_LIAN1878_11_GE_C32.indd 39 12/29/17 5:46 PM

32-40 Chapter 32 Multithreading and Parallel Programming

32.13 (Generic parallel merge sort) Revise Listing 32.10, ParallelMergeSort.java, to define a generic parallelMergeSort method as follows:

public static <E extends Comparable<E>> void parallelMergeSort(E[] list)

*32.14 (Parallel quick sort) Implement the following method in parallel to sort a list using quick sort (see Listing 23.7):

public static void parallelQuickSort(int[] list)

Write a test program that times the execution time for a list of size 9,000,000 using this parallel method and a sequential method.

*32.15 (Parallel sum) Implement the following method using Fork/Join to find the sum of a list.

public static double parallelSum(double[] list)

Write a test program that finds the sum in a list of 9,000,000 double values.

*32.16 (Parallel matrix addition) Programming Exercise 8.5 describes how to perform matrix addition. Suppose you have multiple processors, so you can speed up the matrix addition. Implement the following method in parallel:

public static double[][] parallelAddMatrix( double[][] a, double[][] b)

Write a test program that measures the execution time for adding two 2,000 * 2,000 matrices using the parallel method and sequential method, respectively.

*32.17 (Parallel matrix multiplication) Programming Exercise 8.6 describes how to perform matrix multiplication. Suppose that you have multiple processors, so you can speed up the matrix multiplication. Implement the following method in parallel:

public static double[][] parallelMultiplyMatrix( double[][] a, double[][] b)

Write a test program that measures the execution time for multiplying two 2,000 * 2,000 matrices using the parallel method and sequential method, respectively.

*32.18 (Parallel Eight Queens) Revise Listing 22.11, EightQueens.java, to develop a parallel algorithm that finds all solutions for the Eight Queens problem. (Hint: Launch eight subtasks, each of which places the queen in a different column in the first row.)

Comprehensive ***32.19 (Sorting animation) Write an animation for selection sort, insertion sort, and

bubble sort, as shown in Figure 32.31. Create an array of integers 1, 2, . . . , 50. Shuffle it randomly. Create a pane to display the array in a histogram. You should invoke each sort method in a separate thread. Each algorithm uses two nested loops. When the algorithm completes an iteration in the outer loop, put the thread to sleep for 0.5 seconds and redisplay the array in the histogram. Color the last bar in the sorted subarray.

M32_LIAN1878_11_GE_C32.indd 40 12/29/17 5:46 PM

Programming Exercises 32-41

Figure 32.31 Three sorting algorithms are illustrated in the animation.

***32.20 (Sudoku search animation) Modify Programming Exercise 22.21 to display the intermediate results of the search. Figure 32.32 gives a snapshot of an animation in progress with number 2 placed in the cell in Figure 32.32a, number 3 placed in the cell in Figure 32.32b, and number 3 placed in the cell in Figure 32.32c. The animation displays all the search steps.

Figure 32.32 The intermediate search steps are displayed in the animation for the Sudoku problem.

(a) (b) (c)

M32_LIAN1878_11_GE_C32.indd 41 12/29/17 5:46 PM

32-42 Chapter 32 Multithreading and Parallel Programming

Figure 32.33 The intermediate search steps are displayed in the animation for the Eight Queens problem.

32.21 (Combine colliding bouncing balls) Rewrite Programming Exercise 20.5 using a thread to animate bouncing ball movements.

***32.22 (Eight Queens animation) Modify Listing 22.11, EightQueens.java, to display the intermediate results of the search. As shown in Figure 32.33, the current row being searched is highlighted. Every one second, a new state of the chess board is displayed.

M32_LIAN1878_11_GE_C32.indd 42 12/29/17 5:46 PM

Objectives ■■ To explain the terms: TCP, IP, domain name, domain name server,

stream-based communications, and packet-based communications (§33.2).

■■ To create servers using server sockets (§33.2.1) and clients using client sockets (§33.2.2).

■■ To implement Java networking programs using stream sockets (§33.2.3).

■■ To develop an example of a client/server application (§33.2.4).

■■ To obtain Internet addresses using the InetAddress class (§33.3).

■■ To develop servers for multiple clients (§33.4).

■■ To send and receive objects on a network (§33.5).

■■ To develop an interactive tic-tac-toe game played on the Internet (§33.6).

Networking

Chapter

M33_LIAN1878_11_GE_C33.indd 1 12/29/17 5:48 PM

33-2 Chapter 33 Networking

33.1 Introduction Computer networking is used to send and receive messages among computers on the Internet.

To browse the Web or send an email, your computer must be connected to the Internet. The Internet is the global network of millions of computers. Your computer can connect to the Internet through an Internet Service Provider (ISP) using a dialup, DSL, or cable modem, or through a local area network (LAN).

When a computer needs to communicate with another computer, it needs to know the other computer’s address. An Internet Protocol (IP) address uniquely identifies the computer on the Internet. An IP address consists of four dotted decimal numbers between 0 and 255, such as 130.254.204.33. Since it is not easy to remember so many numbers, they are often mapped to meaningful names called domain names, such as liang.armstrong.edu. Special servers called Domain Name Servers (DNS) on the Internet translate host names into IP addresses. When a computer contacts liang.armstrong.edu, it first asks the DNS to translate this domain name into a numeric IP address then sends the request using the IP address.

The Internet Protocol is a low-level protocol for delivering data from one computer to another across the Internet in packets. Two higher-level protocols used in conjunction with the IP are the Transmission Control Protocol (TCP) and the User Datagram Protocol (UDP). TCP enables two hosts to establish a connection and exchange streams of data. TCP guarantees delivery of data and also guarantees that packets will be delivered in the same order in which they were sent. UDP is a standard, low-overhead, connectionless, host-to-host protocol that is used over the IP. UDP allows an application program on one computer to send a datagram to an application program on another computer.

Java supports both stream-based and packet-based communications. Stream-based communications use TCP for data transmission, whereas packet-based communications use UDP. Since TCP can detect lost transmissions and resubmit them, transmissions are lossless and reli- able. UDP, in contrast, cannot guarantee lossless transmission. Stream-based communications are used in most areas of Java programming and are the focus of this chapter. Packet-based com- munications are introduced in Supplement III.P, Networking Using Datagram Protocol.

33.2 Client/Server Computing Java provides the ServerSocket class for creating a server socket, and the Socket class for creating a client socket. Two programs on the Internet communicate through a server socket and a client socket using I/O streams.

Networking is tightly integrated in Java. The Java API provides the classes for creating sockets to facilitate program communications over the Internet. Sockets are the endpoints of logical connections between two hosts and can be used to send and receive data. Java treats socket communications much as it treats I/O operations; thus, programs can read from or write to sockets as easily as they can read from or write to files.

Network programming usually involves a server and one or more clients. The client sends requests to the server, and the server responds. The client begins by attempting to establish a connection to the server. The server can accept or deny the connection. Once a connection is established, the client and the server communicate through sockets.

The server must be running when a client attempts to connect to the server. The server waits for a connection request from the client. The statements needed to create sockets on a server and on a client are shown in Figure 33.1.

33.2.1 Server Sockets To establish a server, you need to create a server socket and attach it to a port, which is where the server listens for connections. The port identifies the TCP service on the socket. Port num- bers range from 0 to 65536, but port numbers 0 to 1024 are reserved for privileged services.

Point Key

IP address

domain name domain name server

TCP UDP

stream-based communication packet-based communication

Point Key

socket

server socket

port

M33_LIAN1878_11_GE_C33.indd 2 12/29/17 5:48 PM

33.2 Client/Server Computing 33-3

For instance, the email server runs on port 25, and the Web server usually runs on port 80. You can choose any port number that is not currently used by other programs. The following state- ment creates a server socket serverSocket:

ServerSocket serverSocket = new ServerSocket(port);

Note Attempting to create a server socket on a port already in use would cause a java.net. BindException.

33.2.2 Client Sockets After a server socket is created, the server can use the following statement to listen for connections:

Socket socket = serverSocket.accept();

This statement waits until a client connects to the server socket. The client issues the following statement to request a connection to a server:

Socket socket = new Socket(serverName, port);

This statement opens a socket so that the client program can communicate with the server. serverName is the server’s Internet host name or IP address. The following statement creates a socket on the client machine to connect to the host 130.254.204.33 at port 8000:

Socket socket = new Socket("130.254.204.33", 8000);

Alternatively, you can use the domain name to create a socket, as follows:

Socket socket = new Socket("liang.armstrong.edu", 8000);

When you create a socket with a host name, the JVM asks the DNS to translate the host name into the IP address.

Note A program can use the host name localhost or the IP address 127.0.0.1 to refer to the machine on which a client is running.

BindException

connect to client

client socket

use IP address

use domain name

localhost

Figure 33.1 The server creates a server socket and, once a connection to a client is established, connects to the client with a client socket.

Server Host

Step 1: Create a server socket on a port, e.g., 8000, using the following statement:

ServerSocket serverSocket = new ServerSocket(8000);

Step 2: Create a socket to connect to a client, using the following statement:

Socket socket = serverSocket.accept();

Client Host

Step 3: A client program uses the following statement to connect to the server:

Socket socket = new Socket(serverHost, 8000);

I/O Stream

Network

M33_LIAN1878_11_GE_C33.indd 3 12/29/17 5:48 PM

33-4 Chapter 33 Networking

Note The Socket constructor throws a java.net.UnknownHostException if the host cannot be found.

33.2.3 Data Transmission through Sockets After the server accepts the connection, the communication between the server and the client is conducted in the same way as for I/O streams. The statements needed to create the streams and to exchange data between them are shown in Figure 33.2.

UnknownHostException

Figure 33.2 The server and client exchange data through I/O streams on top of the socket.

int port = 8000; int port = 8000;

Connection Request

I/O Streams

Server Client

DataInputStream in; DataOutputStream out; ServerSocket server; Socket socket;

server = new ServerSocket(port); socket = server.accept(); in = new DataInputStream (socket.getInputStream()); out = new DataOutputStream (socket.getOutputStream()); System.out.println(in.readDouble()); out.writeDouble(aNumber);

String host = "localhost" DataInputStream in; DataOutputStream out; Socket socket;

socket = new Socket(host, port); in = new DataInputStream (socket.getInputStream()); out = new DataOutputStream (socket.getOutputStream());

System.out.println(in.readDouble()); out.writeDouble(aNumber);

To get an input stream and an output stream, use the getInputStream() and getOutputStream() methods on a socket object. For example, the following statements create an InputStream stream called input and an OutputStream stream called output from a socket:

InputStream input = socket.getInputStream(); OutputStream output = socket.getOutputStream();

The InputStream and OutputStream streams are used to read or write bytes. You can use DataInputStream, DataOutputStream, BufferedReader, and PrintWriter to wrap on the InputStream and OutputStream to read or write data, such as int, double, or String. The following statements, for instance, create the DataInputStream stream input and the DataOutputstream output to read and write primitive data values:

DataInputStream input = new DataInputStream (socket.getInputStream()); DataOutputStream output = new DataOutputStream (socket.getOutputStream());

The server can use input.readDouble() to receive a double value from the client, and output.writeDouble(d) to send the double value d to the client.

Tip Recall that binary I/O is more efficient than text I/O because text I/O requires encoding and decoding. Therefore, it is better to use binary I/O for transmitting data between a server and a client to improve performance.

M33_LIAN1878_11_GE_C33.indd 4 12/29/17 5:48 PM

33.2 Client/Server Computing 33-5

The client sends the radius through a DataOutputStream on the output stream socket, and the server receives the radius through the DataInputStream on the input stream socket, as shown in Figure 33.4a. The server computes the area and sends it to the client through a DataOutput- Stream on the output stream socket, and the client receives the area through a DataInputStream on the input stream socket, as shown in Figure 33.4b. The server and client programs are given in Listings 33.1 and 33.2. Figure 33.5 contains a sample run of the server and the client.

Figure 33.3 The client sends the radius to the server; the server computes the area and sends it to the client.

compute area

Server Client radius

area

33.2.4 A Client/Server Example This example presents a client program and a server program. The client sends data to a server. The server receives the data, uses it to produce a result, and then sends the result back to the client. The client displays the result on the console. In this example, the data sent from the client comprise the radius of a circle, and the result produced by the server is the area of the circle (see Figure 33.3).

Figure 33.4 (a) The client sends the radius to the server. (b) The server sends the area to the client.

Server radius

DataInputStream

socket.getInputStream

socket

Network

radius

DataOutputStream

socket.getOutputStream

socket

Client

(a)

Server area

DataOutputStream

socket.getOutputStream

socket

Network

area

DataInputStream

socket.getInputStream

socket

Client

(b)

Figure 33.5 The client sends the radius to the server. The server receives it, computes the area, and sends the area to the client.

Listing 33.1 Server.java 1 import java.io.*; 2 import java.net.*; 3 import java.util.Date;

M33_LIAN1878_11_GE_C33.indd 5 12/29/17 5:48 PM

33-6 Chapter 33 Networking

4 import javafx.application.Application; 5 import javafx.application.Platform; 6 import javafx.scene.Scene; 7 import javafx.scene.control.ScrollPane; 8 import javafx.scene.control.TextArea; 9 import javafx.stage.Stage; 10 11 public class Server extends Application { 12 @Override // Override the start method in the Application class 13 public void start(Stage primaryStage) { 14 // Text area for displaying contents 15 TextArea ta = new TextArea(); 16 17 // Create a scene and place it in the stage 18 Scene scene = new Scene(new ScrollPane(ta), 450, 200); 19 primaryStage.setTitle("Server"); // Set the stage title 20 primaryStage.setScene(scene); // Place the scene in the stage 21 primaryStage.show(); // Display the stage 22 23 new Thread(() -> { 24 try { 25 // Create a server socket 26 ServerSocket serverSocket = new ServerSocket(8000); 27 Platform.runLater(() -> 28 ta.appendText("Server started at " + new Date() + '\n')); 29 30 // Listen for a connection request 31 Socket socket = serverSocket.accept(); 32 33 // Create data input and output streams 34 DataInputStream inputFromClient = new DataInputStream( 35 socket.getInputStream()); 36 DataOutputStream outputToClient = new DataOutputStream( 37 socket.getOutputStream()); 38 39 while (true) { 40 // Receive radius from the client 41 double radius = inputFromClient.readDouble(); 42 43 // Compute area 44 double area = radius * radius * Math.PI; 45 46 // Send area back to the client 47 outputToClient.writeDouble(area); 48 49 Platform.runLater(() -> { 50 ta.appendText("Radius received from client: " 51 + radius + '\n'); 52 ta.appendText("Area is: " + area + '\n'); 53 }); 54 } 55 } 56 catch(IOException ex) { 57 ex.printStackTrace(); 58 } 59 }).start(); 60 } 61 }

create server UI

server socket

update UI

connect client

input from client

output to client

read radius

write area

update UI

M33_LIAN1878_11_GE_C33.indd 6 12/29/17 5:48 PM

33.2 Client/Server Computing 33-7

Listing 33.2 Client.java 1 import java.io.*; 2 import java.net.*; 3 import javafx.application.Application; 4 import javafx.geometry.Insets; 5 import javafx.geometry.Pos; 6 import javafx.scene.Scene; 7 import javafx.scene.control.Label; 8 import javafx.scene.control.ScrollPane; 9 import javafx.scene.control.TextArea; 10 import javafx.scene.control.TextField; 11 import javafx.scene.layout.BorderPane; 12 import javafx.stage.Stage; 13 14 public class Client extends Application { 15 // IO streams 16 DataOutputStream toServer = null; 17 DataInputStream fromServer = null; 18 19 @Override // Override the start method in the Application class 20 public void start(Stage primaryStage) { 21 // Panel p to hold the label and text field 22 BorderPane paneForTextField = new BorderPane(); 23 paneForTextField.setPadding(new Insets(5, 5, 5, 5)); 24 paneForTextField.setStyle("-fx-border-color: green"); 25 paneForTextField.setLeft(new Label("Enter a radius: ")); 26 27 TextField tf = new TextField(); 28 tf.setAlignment(Pos.BOTTOM_RIGHT); 29 paneForTextField.setCenter(tf); 30 31 BorderPane mainPane = new BorderPane(); 32 // Text area to display contents 33 TextArea ta = new TextArea(); 34 mainPane.setCenter(new ScrollPane(ta)); 35 mainPane.setTop(paneForTextField); 36 37 // Create a scene and place it in the stage 38 Scene scene = new Scene(mainPane, 450, 200); 39 primaryStage.setTitle("Client"); // Set the stage title 40 primaryStage.setScene(scene); // Place the scene in the stage 41 primaryStage.show(); // Display the stage 42 43 tf.setOnAction(e -> { 44 try { 45 // Get the radius from the text field 46 double radius = Double.parseDouble(tf.getText().trim()); 47 48 // Send the radius to the server 49 toServer.writeDouble(radius); 50 toServer.flush(); 51 52 // Get area from the server 53 double area = fromServer.readDouble(); 54 55 // Display to the text area 56 ta.appendText("Radius is " + radius + "\n"); 57 ta.appendText("Area received from the server is " 58 + area + '\n');

handle action event

read radius

write radius

read area

create UI

M33_LIAN1878_11_GE_C33.indd 7 12/29/17 5:48 PM

33-8 Chapter 33 Networking

59 } 60 catch (IOException ex) { 61 System.err.println(ex); 62 } 63 }); 64 65 try { 66 // Create a socket to connect to the server 67 Socket socket = new Socket("localhost", 8000); 68 // Socket socket = new Socket("130.254.204.36", 8000); 69 // Socket socket = new Socket("drake.Armstrong.edu", 8000); 70 71 // Create an input stream to receive data from the server 72 fromServer = new DataInputStream(socket.getInputStream()); 73 74 // Create an output stream to send data to the server 75 toServer = new DataOutputStream(socket.getOutputStream()); 76 } 77 catch (IOException ex) { 78 ta.appendText(ex.toString() + '\n'); 79 } 80 } 81 }

You start the server program first then start the client program. In the client program, enter a radius in the text field and press Enter to send the radius to the server. The server computes the area and sends it back to the client. This process is repeated until one of the two programs terminates.

The networking classes are in the package java.net. You should import this package when writing Java network programs.

The Server class creates a ServerSocket serverSocket and attaches it to port 8000 using this statement (line 26 in Server.java):

ServerSocket serverSocket = new ServerSocket(8000);

The server then starts to listen for connection requests, using the following statement (line 31 in Server.java):

Socket socket = serverSocket.accept();

The server waits until the client requests a connection. After it is connected, the server reads the radius from the client through an input stream, computes the area, and sends the result to the client through an output stream. The ServerSocket accept() method takes time to execute. It is not appropriate to run this method in the JavaFX application thread. So, we place it in a separate thread (lines 23–59). The statements for updating GUI needs to run from the JavaFX application thread using the Platform.runLater method (lines 27–28, 49–53).

The Client class uses the following statement to create a socket that will request a con- nection to the server on the same machine (localhost) at port 8000 (line 67 in Client.java).

Socket socket = new Socket("localhost", 8000);

If you run the server and the client on different machines, replace localhost with the server machine’s host name or IP address. In this example, the server and the client are running on the same machine.

If the server is not running, the client program terminates with a java.net. ConnectException. After it is connected, the client gets input and output streams—wrapped by data input and output streams—in order to receive and send data to the server.

request connection

input from server

output to server

M33_LIAN1878_11_GE_C33.indd 8 12/29/17 5:48 PM

33.3 The InetAddress Class 33-9

If you receive a java.net.BindException when you start the server, the server port is currently in use. You need to terminate the process that is using the server port then restart the server.

Note When you create a server socket, you have to specify a port (e.g., 8000) for the socket. When a client connects to the server (line 67 in Client.java), a socket is created on the client. This socket has its own local port. This port number (e.g., 2047) is automatically chosen by the JVM, as shown in Figure 33.6.

client socket port

Figure 33.6 The JVM automatically chooses an available port to create a socket for the client.

Server 0

socket

Client

port number

8000 . . .

0 1

. 2047

socket

To see the local port on the client, insert the following statement in line 70 in Client.java.

System.out.println("local port: " + socket.getLocalPort());

33.2.1 How do you create a server socket? What port numbers can be used? What happens if a requested port number is already in use? Can a port connect to multiple clients?

33.2.2 What are the differences between a server socket and a client socket? 33.2.3 How does a client program initiate a connection? 33.2.4 How does a server accept a connection? 33.2.5 How are data transferred between a client and a server?

33.3 The InetAddress Class The server program can use the InetAddress class to obtain the information about the IP address and host name for the client.

Occasionally, you would like to know who is connecting to the server. You can use the InetAddress class to find the client’s host name and IP address. The InetAddress class models an IP address. You can use the following statement in the server program to get an instance of InetAddress on a socket that connects to the client:

InetAddress inetAddress = socket.getInetAddress();

Next, you can display the client’s host name and IP address, as follows:

System.out.println("Client's host name is " + inetAddress.getHostName());

Point Check

Point Key

M33_LIAN1878_11_GE_C33.indd 9 12/29/17 5:48 PM

33-10 Chapter 33 Networking

System.out.println("Client's IP Address is " + inetAddress.getHostAddress());

You can also create an instance of InetAddress from a host name or IP address using the static getByName method. For example, the following statement creates an InetAddress for the host liang.armstrong.edu.

InetAddress address = InetAddress.getByName("liang.armstrong.edu");

Listing 33.3 gives a program that identifies the host name and IP address of the arguments you pass in from the command line. Line 7 creates an InetAddress using the getByName method. Lines 8 and 9 use the getHostName and getHostAddress methods to get the host’s name and IP address. Figure 33.7 shows a sample run of the program.

Figure 33.7 The program identifies host names and IP addresses.

Listing 33.3 IdentifyHostNameIP.java 1 import java.net.*; 2 3 public class IdentifyHostNameIP { 4 public static void main(String[] args) { 5 for (int i = 0; i < args.length; i++) { 6 try { 7 InetAddress address = InetAddress.getByName(args[i]); 8 System.out.print("Host name: " + address.getHostName() + " "); 9 System.out.println("IP address: " + address.getHostAddress()); 10 } 11 catch (UnknownHostException ex) { 12 System.err.println("Unknown host or IP address " + args[i]); 13 } 14 } 15 } 16 }

33.3.1 How do you obtain an instance of InetAddress? 33.3.2 What methods can you use to get the IP address and hostname from an InetAddress?

33.4 Serving Multiple Clients A server can serve multiple clients. The connection to each client is handled by one thread.

Multiple clients are quite often connected to a single server at the same time. Typically, a server runs continuously on a server computer, and clients from all over the Internet can con- nect to it. You can use threads to handle the server’s multiple clients simultaneously—simply

get an InetAddress get host name get host IP

Point Check

Point Key

M33_LIAN1878_11_GE_C33.indd 10 12/29/17 5:48 PM

33.4 Serving Multiple Clients 33-11

create a thread for each connection. Here is how the server handles the establishment of a connection:

while (true) { Socket socket = serverSocket.accept(); // Connect to a client Thread thread = new ThreadClass(socket); thread.start(); }

The server socket can have many connections. Each iteration of the while loop creates a new connection. Whenever a connection is established, a new thread is created to handle commu- nication between the server and the new client, and this allows multiple connections to run at the same time.

Listing 33.4 creates a server class that serves multiple clients simultaneously. For each con- nection, the server starts a new thread. This thread continuously receives input (the radius of a circle) from clients and sends the results (the area of the circle) back to them (see Figure 33.8). The client program is the same as in Listing 33.2. A sample run of the server with two clients is shown in Figure 33.9.

Figure 33.9 The server spawns a thread in order to serve a client.

Figure 33.8 Multithreading enables a server to handle multiple independent clients.

Server

Client n. . .Client 1

A server socket on a port

A socket for a client A socket for a

client

Listing 33.4 MultiThreadServer.java 1 import java.io.*; 2 import java.net.*; 3 import java.util.Date; 4 import javafx.application.Application; 5 import javafx.application.Platform; 6 import javafx.scene.Scene; 7 import javafx.scene.control.ScrollPane; 8 import javafx.scene.control.TextArea; 9 import javafx.stage.Stage; 10

M33_LIAN1878_11_GE_C33.indd 11 12/29/17 5:48 PM

33-12 Chapter 33 Networking

11 public class MultiThreadServer extends Application { 12 // Text area for displaying contents 13 private TextArea ta = new TextArea(); 14 15 // Number a client 16 private int clientNo = 0; 17 18 @Override // Override the start method in the Application class 19 public void start(Stage primaryStage) { 20 // Create a scene and place it in the stage 21 Scene scene = new Scene(new ScrollPane(ta), 450, 200); 22 primaryStage.setTitle("MultiThreadServer"); // Set the stage title 23 primaryStage.setScene(scene); // Place the scene in the stage 24 primaryStage.show(); // Display the stage 25 26 new Thread( () -> { 27 try { 28 // Create a server socket 29 ServerSocket serverSocket = new ServerSocket(8000); 30 ta.appendText("MultiThreadServer started at " 31 + new Date() + '\n'); 32 33 while (true) { 34 // Listen for a new connection request 35 Socket socket = serverSocket.accept(); 36 37 // Increment clientNo 38 clientNo++; 39 40 Platform.runLater( () -> { 41 // Display the client number 42 ta.appendText("Starting thread for client " + clientNo + 43 " at " + new Date() + '\n'); 44 45 // Find the client's host name, and IP address 46 InetAddress inetAddress = socket.getInetAddress(); 47 ta.appendText("Client " + clientNo + "'s host name is " 48 + inetAddress.getHostName() + "\n"); 49 ta.appendText("Client " + clientNo + "'s IP Address is" 50 + inetAddress.getHostAddress() + "\n"); 51 }); 52 53 // Create and start a new thread for the connection 54 new Thread(new HandleAClient(socket)).start(); 55 } 56 } 57 catch(IOException ex) { 58 System.err.println(ex); 59 } 60 }).start(); 61 } 62 63 // Define the thread class for handling new connection 64 class HandleAClient implements Runnable { 65 private Socket socket; // A connected socket 66 67 /** Construct a thread */ 68 public HandleAClient(Socket socket) { 69 this.socket = socket; 70 } 71

server socket

connect client

update GUI

network information

create task

start thread

task class

M33_LIAN1878_11_GE_C33.indd 12 12/29/17 5:48 PM

33.5 Sending and Receiving Objects 33-13

72 /** Run a thread */ 73 public void run() { 74 try { 75 // Create data input and output streams 76 DataInputStream inputFromClient = new DataInputStream( 77 socket.getInputStream()); 78 DataOutputStream outputToClient = new DataOutputStream( 79 socket.getOutputStream()); 80 81 // Continuously serve the client 82 while (true) { 83 // Receive radius from the client 84 double radius = inputFromClient.readDouble(); 85 86 // Compute area 87 double area = radius * radius * Math.PI; 88 89 // Send area back to the client 90 outputToClient.writeDouble(area); 91 92 Platform.runLater(() -> { 93 ta.appendText("radius received from client: " + 94 radius + '\n'); 95 ta.appendText("Area found: " + area + '\n'); 96 }); 97 } 98 } 99 catch(IOException ex) { 100 ex.printStackTrace(); 101 } 102 } 103 } 104 }

The server creates a server socket at port 8000 (line 29) and waits for a connection (line 35). When a connection with a client is established, the server creates a new thread to handle the communication (line 54). It then waits for another connection in an infinite while loop (lines 33–55).

The threads, which run independently of one another, communicate with designated clients. Each thread creates data input and output streams that receive and send data to a client.

33.4.1 How do you make a server serve multiple clients?

33.5 Sending and Receiving Objects A program can send and receive objects from another program.

In the preceding examples, you learned how to send and receive data of primitive types. You can also send and receive objects using ObjectOutputStream and ObjectInputStream on socket streams. To enable passing, the objects must be serializable. The following example demonstrates how to send and receive objects.

The example consists of three classes: StudentAddress.java (Listing 33.5), StudentClient.java (Listing 33.6), and StudentServer.java (Listing 33.7). The client program collects student information from the client and sends it to a server, as shown in Figure 33.10.

The StudentAddress class contains the student information: name, street, city, state, and zip. The StudentAddress class implements the Serializable interface. Therefore, a StudentAddress object can be sent and received using the object output and input streams.

Point Check

Point Key

I/O

update GUI

M33_LIAN1878_11_GE_C33.indd 13 12/29/17 5:48 PM

33-14 Chapter 33 Networking

Listing 33.5 StudentAddress.java 1 public class StudentAddress implements java.io.Serializable { 2 private String name; 3 private String street; 4 private String city; 5 private String state; 6 private String zip; 7 8 public StudentAddress(String name, String street, String city, 9 String state, String zip) { 10 this.name = name; 11 this.street = street; 12 this.city = city; 13 this.state = state; 14 this.zip = zip; 15 } 16 17 public String getName() { 18 return name; 19 } 20 21 public String getStreet() { 22 return street; 23 } 24 25 public String getCity() { 26 return city; 27 } 28 29 public String getState() { 30 return state; 31 } 32 33 public String getZip() { 34 return zip; 35 } 36 }

The client sends a StudentAddress object through an ObjectOutputStream on the output stream socket, and the server receives the Student object through the ObjectInputStream on the input stream socket, as shown in Figure 33.11. The client uses the writeObject method in the ObjectOutputStream class to send data about a student to the server, and the server receives the student’s information using the read Object method in the ObjectInputStream class. The server and client programs are given in Listings 33.6 and 33.7.

Figure 33.10 The client sends the student information in an object to the server.

serialized

M33_LIAN1878_11_GE_C33.indd 14 12/29/17 5:48 PM

33.5 Sending and Receiving Objects 33-15

Listing 33.6 StudentClient.java 1 import java.io.*; 2 import java.net.*; 3 import javafx.application.Application; 4 import javafx.event.ActionEvent; 5 import javafx.event.EventHandler; 6 import javafx.geometry.HPos; 7 import javafx.geometry.Pos; 8 import javafx.scene.Scene; 9 import javafx.scene.control.Button; 10 import javafx.scene.control.Label; 11 import javafx.scene.control.TextField; 12 import javafx.scene.layout.GridPane; 13 import javafx.scene.layout.HBox; 14 import javafx.stage.Stage; 15 16 public class StudentClient extends Application { 17 private TextField tfName = new TextField(); 18 private TextField tfStreet = new TextField(); 19 private TextField tfCity = new TextField(); 20 private TextField tfState = new TextField(); 21 private TextField tfZip = new TextField(); 22 23 // Button for sending a student to the server 24 private Button btRegister = new Button("Register to the Server"); 25 26 // Host name or ip 27 String host = "localhost"; 28 29 @Override // Override the start method in the Application class 30 public void start(Stage primaryStage) { 31 GridPane pane = new GridPane(); 32 pane.add(new Label("Name"), 0, 0); 33 pane.add(tfName, 1, 0); 34 pane.add(new Label("Street"), 0, 1); 35 pane.add(tfStreet, 1, 1); 36 pane.add(new Label("City"), 0, 2); 37

create UI

Figure 33.11 The client sends a StudentAddress object to the server.

Server

student object

in: ObjectInputStream

socket.getInputStream()

socket

Network

student object

out: ObjectOutputStream

in.readObject() out.writeObject(Object)

socket.getOutputStream()

socket

Client

M33_LIAN1878_11_GE_C33.indd 15 12/29/17 5:48 PM

33-16 Chapter 33 Networking

38 HBox hBox = new HBox(2); 39 pane.add(hBox, 1, 2); 40 hBox.getChildren().addAll(tfCity, new Label("State"), tfState, 41 new Label("Zip"), tfZip); 42 pane.add(btRegister, 1, 3); 43 GridPane.setHalignment(btRegister, HPos.RIGHT); 44 45 pane.setAlignment(Pos.CENTER); 46 tfName.setPrefColumnCount(15); 47 tfStreet.setPrefColumnCount(15); 48 tfCity.setPrefColumnCount(10); 49 tfState.setPrefColumnCount(2); 50 tfZip.setPrefColumnCount(3); 51 52 btRegister.setOnAction(new ButtonListener()); 53 54 // Create a scene and place it in the stage 55 Scene scene = new Scene(pane, 450, 200); 56 primaryStage.setTitle("StudentClient"); // Set the stage title 57 primaryStage.setScene(scene); // Place the scene in the stage 58 primaryStage.show(); // Display the stage 59 } 60 61 /** Handle button action */ 62 private class ButtonListener implements EventHandler<ActionEvent> { 63 @Override 64 public void handle(ActionEvent e) { 65 try { 66 // Establish connection with the server 67 Socket socket = new Socket(host, 8000); 68 69 // Create an output stream to the server 70 ObjectOutputStream toServer = 71 new ObjectOutputStream(socket.getOutputStream()); 72 73 // Get text field 74 String name = tfName.getText().trim(); 75 String street = tfStreet.getText().trim(); 76 String city = tfCity.getText().trim(); 77 String state = tfState.getText().trim(); 78 String zip = tfZip.getText().trim(); 79 80 // Create a Student object and send to the server 81 StudentAddress s = 82 new StudentAddress(name, street, city, state, zip); 83 toServer.writeObject(s); 84 } 85 catch (IOException ex) { 86 ex.printStackTrace(); 87 } 88 } 89 } 90 }

Listing 33.7 StudentServer.java 1 import java.io.*; 2 import java.net.*; 3 4 public class StudentServer {

server socket

output stream

send to server

M33_LIAN1878_11_GE_C33.indd 16 12/29/17 5:48 PM

33.5 Sending and Receiving Objects 33-17

5 private ObjectOutputStream outputToFile; 6 private ObjectInputStream inputFromClient; 7 8 public static void main(String[] args) { 9 new StudentServer(); 10 } 11 12 public StudentServer() { 13 try { 14 // Create a server socket 15 ServerSocket serverSocket = new ServerSocket(8000); 16 System.out.println("Server started "); 17 18 // Create an object output stream 19 outputToFile = new ObjectOutputStream( 20 new FileOutputStream("student.dat", true)); 21 22 while (true) { 23 // Listen for a new connection request 24 Socket socket = serverSocket.accept(); 25 26 // Create an input stream from the socket 27 inputFromClient = 28 new ObjectInputStream(socket.getInputStream()); 29 30 // Read from input 31 Object object = inputFromClient.readObject(); 32 33 // Write to the file 34 outputToFile.writeObject(object); 35 System.out.println("A new student object is stored"); 36 } 37 } 38 catch(ClassNotFoundException ex) { 39 ex.printStackTrace(); 40 } 41 catch(IOException ex) { 42 ex.printStackTrace(); 43 } 44 finally { 45 try { 46 inputFromClient.close(); 47 outputToFile.close(); 48 } 49 catch (Exception ex) { 50 ex.printStackTrace(); 51 } 52 } 53 } 54 }

On the client side, when the user clicks the Register to the Server button, the client creates a socket to connect to the host (line 67), creates an ObjectOutputStream on the output stream of the socket (lines 70 and 71), and invokes the writeObject method to send the StudentAddress object to the server through the object output stream (line 83).

On the server side, when a client connects to the server, the server creates an Object InputStream on the input stream of the socket (lines 27 and 28), invokes the readObject method to receive the StudentAddress object through the object input stream (line 31), and writes the object to a file (line 34).

server socket

output to file

connect to client

input stream

get from client

write to file

M33_LIAN1878_11_GE_C33.indd 17 12/29/17 5:48 PM

33-18 Chapter 33 Networking

Figure 33.12 The server can create many sessions, each of which facilitates a tic-tac-toe game for two players.

Server

Player 2

Session n...

Player 1 Player 2Player 1

Session 1

...

33.5.1 How does a server receive connection from a client? How does a client connect to a server?

33.5.2 How do you find the host name of a client program from the server? 33.5.3 How do you send and receive an object?

33.6 Case Study: Distributed Tic-Tac-Toe Games This section develops a program that enables two players to play the tic-tac-toe game on the Internet.

In Section 16.12, Case Study: Developing a Tic-Tac-Toe Game, you developed a program for a tic-tac-toe game that enables two players to play the game on the same machine. In this section, you will learn how to develop a distributed tic-tac-toe game using multithreads and networking with socket streams. A distributed tic-tac-toe game enables users to play on dif- ferent machines from anywhere on the Internet.

You need to develop a server for multiple clients. The server creates a server socket and accepts connections from every two players to form a session. Each session is a thread that communicates with the two players and determines the status of the game. The server can establish any number of sessions, as shown in Figure 33.12.

For each session, the first client connecting to the server is identified as player 1 with token X, and the second client connecting is identified as player 2 with token O. The server notifies the players of their respective tokens. Once two clients are connected to it, the server starts a thread to facilitate the game between the two players by performing the steps repeatedly, as shown in Figure 33.13.

Point Check

Point Key

The server does not have to be a graphical component, but creating it in a GUI in which game information can be viewed is user friendly. You can create a scroll pane to hold a text area in the GUI and display game information in the text area. The server creates a thread to handle a game session when two players are connected to the server.

The client is responsible for interacting with the players. It creates a user interface with nine cells and displays the game title and status to the players in the labels. The client class is very similar to the TicTacToe class presented in the case study in Listing 16.13. However, the client in this example does not determine the game status (win or draw); it simply passes the moves to the server and receives the game status from the server.

Based on the foregoing analysis, you can create the following classes:

■■ TicTacToeServer serves all the clients in Listing 33.9.

■■ HandleASession facilitates the game for two players. This class is defined in Listing 33.9, TicTacToeServer.java.

M33_LIAN1878_11_GE_C33.indd 18 12/29/17 5:48 PM

33.6 Case Study: Distributed Tic-Tac-Toe Games 33-19

■■ TicTacToeClient models a player in Listing 33.10.

■■ Cell models a cell in the game. It is an inner class in TicTacToeClient.

■■ TicTacToeConstants is an interface that defines the constants shared by all the classes in the example in Listing 33.8.

The relationships of these classes are shown in Figure 33.14.

Listing 33.8 TicTacToeConstants.java 1 public interface TicTacToeConstants { 2 public static int PLAYER1 = 1; // Indicate player 1 3 public static int PLAYER2 = 2; // Indicate player 2 4 public static int PLAYER1_WON = 1; // Indicate player 1 won 5 public static int PLAYER2_WON = 2; // Indicate player 2 won 6 public static int DRAW = 3; // Indicate a draw 7 public static int CONTINUE = 4; // Indicate to continue 8 }

Listing 33.9 TicTacToeServer.java 1 import java.io.*; 2 import java.net.*;

Figure 33.13 The server starts a thread to facilitate communications between the two players.

Player 1

1. Initialize user interface.

2. Request connection to the server

and learn which token to use from the

server.

3. Get the start signal from the server.

4. Wait for the player to mark a cell,

send the cell's row and column index to

the server.

5. Receive status from the server.

6. If WIN, display the winner; if Player 2

wins, receive the last move from

Player 2. Break the loop.

7. If DRAW, display game is over;

break the loop.

8. If CONTINUE, receive Player 2's

selected row and column index and

mark the cell for Player 2.

Server

Create a server socket.

Accept connection from the �rst player and notify

the player who is Player 1 with token X.

Accept connection from the second player and

notify the player who is Player 2 with token O.

Start a thread for the session.

Player 2

1. Initialize user interface.

2. Request connection to the server and

learn which token to use from the server.

3. Receive status from the server.

4. If WIN, display the winner. If Player 1

wins, receive Player 1's last move, and

break the loop.

5. If DRAW, display game is over, and

receive Player 1's last move, and break

the loop.

6. If CONTINUE, receive Player 1's

selected row and index and mark the cell

for Player 1.

7. Wait for the player to move, and send

the selected row and column to the

server.

Handle a session:

1. Tell Player 1 to start.

2. Receive row and column of the selected cell from

Player 1.

3. Determine the game status (WIN, DRAW,

CONTINUE). If Player 1 wins, or draws, send the status

(PLAYER1_WON, DRAW) to both players and send

Player 1's move to Player 2. Exit.

4. If CONTINUE, notify Player 2 to take the turn, and

send Player 1's newly selected row and column index to

Player 2.

5. Receive row and column of the selected cell from

Player 2.

6. If Player 2 wins, send the status (PLAYER2_WON) to

both players, and send Player 2's move to Player 1. Exit.

7. If CONTINUE, send the status, and send Player 2's

newly selected row and column index to Player 1.

M33_LIAN1878_11_GE_C33.indd 19 12/29/17 5:48 PM

33-20 Chapter 33 Networking

3 import java.util.Date; 4 import javafx.application.Application; 5 import javafx.application.Platform; 6 import javafx.scene.Scene; 7 import javafx.scene.control.ScrollPane; 8 import javafx.scene.control.TextArea; 9 import javafx.stage.Stage; 10 11 public class TicTacToeServer extends Application 12 implements TicTacToeConstants { 13 private int sessionNo = 1; // Number a session 14 15 @Override // Override the start method in the Application class 16 public void start(Stage primaryStage) { 17 TextArea taLog = new TextArea(); 18 19 // Create a scene and place it in the stage 20 Scene scene = new Scene(new ScrollPane(taLog), 450, 200); 21 primaryStage.setTitle("TicTacToeServer"); // Set the stage title 22 primaryStage.setScene(scene); // Place the scene in the stage 23 primaryStage.show(); // Display the stage 24 25 new Thread( () -> { 26 try { 27 // Create a server socket

Figure 33.14 TicTacToeServer creates an instance of HandleASession for each session of two players. TicTacToeClient creates nine cells in the UI.

TicTacToeConstants

Runnable

TicTacToeServer

TicTacToeClient CellJApplet

JFrame

Similar to Listing 18.10

TicTacToeServer

+PLAYER1 = 1: int +PLAYER2 = 2: int +PLAYER1_WON = 1: int +PLAYER2_WON = 2: int +DRAW = 3: int +CONTINUE = 4: int

HandleASession TicTacToeClient

-player1: Socket -player2: Socket -cell: char[][] -continueToPlay: boolean

+run(): void -isWon(): boolean -isFull(): boolean -sendMove(out: DataOutputStream, row: int, column: int): void

-myTurn: boolean -myToken: char -otherToken: char -cell: Cell[][] -continueToPlay: boolean -rowSelected: int -columnSelected: int -fromServer: DataInputStream -toServer: DataOutputStream -waiting: boolean

+run(): void -connectToServer(): void -receiveMove(): void -sendMove(): void -receiveInfoFromServer(): void -waitForPlayerAction(): void

HandleASession

«interface» TicTacToeConstants

+main(args: String[]):void

create UI

M33_LIAN1878_11_GE_C33.indd 20 12/29/17 5:48 PM

33.6 Case Study: Distributed Tic-Tac-Toe Games 33-21

28 ServerSocket serverSocket = new ServerSocket(8000); 29 Platform.runLater(() -> taLog.appendText(new Date() + 30 ": Server started at socket 8000\n")); 31 32 // Ready to create a session for every two players 33 while (true) { 34 Platform.runLater(() -> taLog.appendText(new Date() + 35 ": Wait for players to join session " + sessionNo + '\n')); 36 37 // Connect to player 1 38 Socket player1 = serverSocket.accept(); 39 40 Platform.runLater(() -> { 41 taLog.appendText(new Date() + ": Player 1 joined session " 42 + sessionNo + '\n'); 43 taLog.appendText("Player 1's IP address" + 44 player1.getInetAddress().getHostAddress() + '\n'); 45 }); 46 47 // Notify that the player is Player 1 48 new DataOutputStream( 49 player1.getOutputStream()).writeInt(PLAYER1); 50 51 // Connect to player 2 52 Socket player2 = serverSocket.accept(); 53 54 Platform.runLater(() -> { 55 taLog.appendText(new Date() + 56 ": Player 2 joined session " + sessionNo + '\n'); 57 taLog.appendText("Player 2's IP address" + 58 player2.getInetAddress().getHostAddress() + '\n'); 59 }); 60 61 // Notify that the player is Player 2 62 new DataOutputStream( 63 player2.getOutputStream()).writeInt(PLAYER2); 64 65 // Display this session and increment session number 66 Platform.runLater(() -> 67 taLog.appendText(new Date() + 68 ": Start a thread for session " + sessionNo++ + '\n')); 69 70 // Launch a new thread for this session of two players 71 new Thread(new HandleASession(player1, player2)).start(); 72 } 73 } 74 catch(IOException ex) { 75 ex.printStackTrace(); 76 } 77 }).start(); 78 } 79 80 // Define the thread class for handling a new session for two players 81 class HandleASession implements Runnable, TicTacToeConstants { 82 private Socket player1; 83 private Socket player2; 84 85 // Create and initialize cells 86 private char[][] cell = new char[3][3]; 87 88 private DataInputStream fromPlayer1;

server socket

connect to client

to player1

connect to client

to player2

a session for two players

M33_LIAN1878_11_GE_C33.indd 21 12/29/17 5:48 PM

33-22 Chapter 33 Networking

89 private DataOutputStream toPlayer1; 90 private DataInputStream fromPlayer2; 91 private DataOutputStream toPlayer2; 92 93 // Continue to play 94 private boolean continueToPlay = true; 95 96 /** Construct a thread */ 97 public HandleASession(Socket player1, Socket player2) { 98 this.player1 = player1; 99 this.player2 = player2; 100 101 // Initialize cells 102 for (int i = 0; i < 3; i++) 103 for (int j = 0; j < 3; j++) 104 cell[i][j] = ' '; 105 } 106 107 /** Implement the run() method for the thread */ 108 public void run() { 109 try { 110 // Create data input and output streams 111 DataInputStream fromPlayer1 = new DataInputStream( 112 player1.getInputStream()); 113 DataOutputStream toPlayer1 = new DataOutputStream( 114 player1.getOutputStream()); 115 DataInputStream fromPlayer2 = new DataInputStream( 116 player2.getInputStream()); 117 DataOutputStream toPlayer2 = new DataOutputStream( 118 player2.getOutputStream()); 119 120 // Write anything to notify player 1 to start 121 // This is just to let player 1 know to start 122 toPlayer1.writeInt(1); 123 124 // Continuously serve the players and determine and report 125 // the game status to the players 126 while (true) { 127 // Receive a move from player 1 128 int row = fromPlayer1.readInt(); 129 int column = fromPlayer1.readInt(); 130 cell[row][column] = 'X'; 131 132 // Check if Player 1 wins 133 if (isWon('X')) { 134 toPlayer1.writeInt(PLAYER1_WON); 135 toPlayer2.writeInt(PLAYER1_WON); 136 sendMove(toPlayer2, row, column); 137 break; // Break the loop 138 } 139 else if (isFull()) { // Check if all cells are filled 140 toPlayer1.writeInt(DRAW); 141 toPlayer2.writeInt(DRAW); 142 sendMove(toPlayer2, row, column); 143 break; 144 } 145 else { 146 // Notify player 2 to take the turn 147 toPlayer2.writeInt(CONTINUE); 148

IO streams

X won?

Is full?

M33_LIAN1878_11_GE_C33.indd 22 12/29/17 5:48 PM

33.6 Case Study: Distributed Tic-Tac-Toe Games 33-23

149 // Send player 1's selected row and column to player 2 150 sendMove(toPlayer2, row, column); 151 } 152 153 // Receive a move from Player 2 154 row = fromPlayer2.readInt(); 155 column = fromPlayer2.readInt(); 156 cell[row][column] = 'O'; 157 158 // Check if Player 2 wins 159 if (isWon('O')) { 160 toPlayer1.writeInt(PLAYER2_WON); 161 toPlayer2.writeInt(PLAYER2_WON); 162 sendMove(toPlayer1, row, column); 163 break; 164 } 165 else { 166 // Notify player 1 to take the turn 167 toPlayer1.writeInt(CONTINUE); 168 169 // Send player 2's selected row and column to player 1 170 sendMove(toPlayer1, row, column); 171 } 172 } 173 } 174 catch(IOException ex) { 175 ex.printStackTrace(); 176 } 177 } 178 179 /** Send the move to other player */ 180 private void sendMove(DataOutputStream out, int row, int column) 181 throws IOException { 182 out.writeInt(row); // Send row index 183 out.writeInt(column); // Send column index 184 } 185 186 /** Determine if the cells are all occupied */ 187 private boolean isFull() { 188 for (int i = 0; i < 3; i++) 189 for (int j = 0; j < 3; j++) 190 if (cell[i][j] == ' ') 191 return false; // At least one cell is not filled 192 193 // All cells are filled 194 return true; 195 } 196 197 /** Determine if the player with the specified token wins */ 198 private boolean isWon(char token) { 199 // Check all rows 200 for (int i = 0; i < 3; i++) 201 if ((cell[i][0] == token) 202 && (cell[i][1] == token) 203 && (cell[i][2] == token)) { 204 return true; 205 } 206 207 /** Check all columns */ 208 for (int j = 0; j < 3; j++)

O won?

send a move

M33_LIAN1878_11_GE_C33.indd 23 12/29/17 5:48 PM

33-24 Chapter 33 Networking

209 if ((cell[0][j] == token) 210 && (cell[1][j] == token) 211 && (cell[2][j] == token)) { 212 return true; 213 } 214 215 /** Check major diagonal */ 216 if ((cell[0][0] == token) 217 && (cell[1][1] == token) 218 && (cell[2][2] == token)) { 219 return true; 220 } 221 222 /** Check subdiagonal */ 223 if ((cell[0][2] == token) 224 && (cell[1][1] == token) 225 && (cell[2][0] == token)) { 226 return true; 227 } 228 229 /** All checked, but no winner */ 230 return false; 231 } 232 } 233 }

Listing 33.10 TicTacToeClient.java 1 import java.io.*; 2 import java.net.*; 3 import java.util.Date; 4 import javafx.application.Application; 5 import javafx.application.Platform; 6 import javafx.scene.Scene; 7 import javafx.scene.control.Label; 8 import javafx.scene.control.ScrollPane; 9 import javafx.scene.control.TextArea; 10 import javafx.scene.layout.BorderPane; 11 import javafx.scene.layout.GridPane; 12 import javafx.scene.layout.Pane; 13 import javafx.scene.paint.Color; 14 import javafx.scene.shape.Ellipse; 15 import javafx.scene.shape.Line; 16 import javafx.stage.Stage; 17 18 public class TicTacToeClient extends Application 19 implements TicTacToeConstants { 20 // Indicate whether the player has the turn 21 private boolean myTurn = false; 22 23 // Indicate the token for the player 24 private char myToken = ' '; 25 26 // Indicate the token for the other player 27 private char otherToken = ' '; 28 29 // Create and initialize cells 30 private Cell[][] cell = new Cell[3][3]; 31

M33_LIAN1878_11_GE_C33.indd 24 12/29/17 5:48 PM

33.6 Case Study: Distributed Tic-Tac-Toe Games 33-25

32 // Create and initialize a title label 33 private Label lblTitle = new Label(); 34 35 // Create and initialize a status label 36 private Label lblStatus = new Label(); 37 38 // Indicate selected row and column by the current move 39 private int rowSelected; 40 private int columnSelected; 41 42 // Input and output streams from/to server 43 private DataInputStream fromServer; 44 private DataOutputStream toServer; 45 46 // Continue to play? 47 private boolean continueToPlay = true; 48 49 // Wait for the player to mark a cell 50 private boolean waiting = true; 51 52 // Host name or ip 53 private String host = "localhost"; 54 55 @Override // Override the start method in the Application class 56 public void start(Stage primaryStage) { 57 // Pane to hold cell 58 GridPane pane = new GridPane(); 59 for (int i = 0; i < 3; i++) 60 for (int j = 0; j < 3; j++) 61 pane.add(cell[i][j] = new Cell(i, j), j, i); 62 63 BorderPane borderPane = new BorderPane(); 64 borderPane.setTop(lblTitle); 65 borderPane.setCenter(pane); 66 borderPane.setBottom(lblStatus); 67 68 // Create a scene and place it in the stage 69 Scene scene = new Scene(borderPane, 320, 350); 70 primaryStage.setTitle("TicTacToeClient"); // Set the stage title 71 primaryStage.setScene(scene); // Place the scene in the stage 72 primaryStage.show(); // Display the stage 73 74 // Connect to the server 75 connectToServer(); 76 } 77 78 private void connectToServer() { 79 try { 80 // Create a socket to connect to the server 81 Socket socket = new Socket(host, 8000); 82 83 // Create an input stream to receive data from the server 84 fromServer = new DataInputStream(socket.getInputStream()); 85 86 // Create an output stream to send data to the server 87 toServer = new DataOutputStream(socket.getOutputStream()); 88 } 89 catch (Exception ex) { 90 ex.printStackTrace(); 91 }

create UI

connect to server

input from server

output to server

M33_LIAN1878_11_GE_C33.indd 25 12/29/17 5:48 PM

33-26 Chapter 33 Networking

92 93 // Control the game on a separate thread 94 new Thread(() -> { 95 try { 96 // Get notification from the server 97 int player = fromServer.readInt(); 98 99 // Am I player 1 or 2? 100 if (player == PLAYER1) { 101 myToken = 'X'; 102 otherToken = 'O'; 103 Platform.runLater(() -> { 104 lblTitle.setText("Player 1 with token 'X'"); 105 lblStatus.setText("Waiting for player 2 to join"); 106 }); 107 108 // Receive startup notification from the server 109 fromServer.readInt(); // Whatever read is ignored 110 111 // The other player has joined 112 Platform.runLater(() -> 113 lblStatus.setText("Player 2 has joined. I start first")); 114 115 // It is my turn 116 myTurn = true; 117 } 118 else if (player == PLAYER2) { 119 myToken = 'O'; 120 otherToken = 'X'; 121 Platform.runLater(() -> { 122 lblTitle.setText("Player 2 with token 'O'"); 123 lblStatus.setText("Waiting for player 1 to move"); 124 }); 125 } 126 127 // Continue to play 128 while (continueToPlay) { 129 if (player == PLAYER1) { 130 waitForPlayerAction(); // Wait for player 1 to move 131 sendMove(); // Send the move to the server 132 receiveInfoFromServer(); // Receive info from the server 133 } 134 else if (player == PLAYER2) { 135 receiveInfoFromServer(); // Receive info from the server 136 waitForPlayerAction(); // Wait for player 2 to move 137 sendMove(); // Send player 2's move to the server 138 } 139 } 140 } 141 catch (Exception ex) { 142 ex.printStackTrace(); 143 } 144 }).start(); 145 } 146 147 /** Wait for the player to mark a cell */ 148 private void waitForPlayerAction() throws InterruptedException { 149 while (waiting) { 150 Thread.sleep(100); 151 }

M33_LIAN1878_11_GE_C33.indd 26 12/29/17 5:48 PM

33.6 Case Study: Distributed Tic-Tac-Toe Games 33-27

152 153 waiting = true; 154 } 155 156 /** Send this player's move to the server */ 157 private void sendMove() throws IOException { 158 toServer.writeInt(rowSelected); // Send the selected row 159 toServer.writeInt(columnSelected); // Send the selected column 160 } 161 162 /** Receive info from the server */ 163 private void receiveInfoFromServer() throws IOException { 164 // Receive game status 165 int status = fromServer.readInt(); 166 167 if (status == PLAYER1_WON) { 168 // Player 1 won, stop playing 169 continueToPlay = false; 170 if (myToken == 'X') { 171 Platform.runLater(() -> lblStatus.setText("I won! (X)")); 172 } 173 else if (myToken == 'O') { 174 Platform.runLater(() -> 175 lblStatus.setText("Player 1 (X) has won!")); 176 receiveMove(); 177 } 178 } 179 else if (status == PLAYER2_WON) { 180 // Player 2 won, stop playing 181 continueToPlay = false; 182 if (myToken == 'O') { 183 Platform.runLater(() -> lblStatus.setText("I won! (O)")); 184 } 185 else if (myToken == 'X') { 186 Platform.runLater(() -> 187 lblStatus.setText("Player 2 (O) has won!")); 188 receiveMove(); 189 } 190 } 191 else if (status == DRAW) { 192 // No winner, game is over 193 continueToPlay = false; 194 Platform.runLater(() -> 195 lblStatus.setText("Game is over, no winner!")); 196 197 if (myToken == 'O') { 198 receiveMove(); 199 } 200 } 201 else { 202 receiveMove(); 203 Platform.runLater(() -> lblStatus.setText("My turn")); 204 myTurn = true; // It is my turn 205 } 206 } 207 208 private void receiveMove() throws IOException { 209 // Get the other player's move 210 int row = fromServer.readInt(); 211 int column = fromServer.readInt();

M33_LIAN1878_11_GE_C33.indd 27 12/29/17 5:48 PM

33-28 Chapter 33 Networking

212 Platform.runLater(() -> cell[row][column].setToken(otherToken)); 213 } 214 215 // An inner class for a cell 216 public class Cell extends Pane { 217 // Indicate the row and column of this cell in the board 218 private int row; 219 private int column; 220 221 // Token used for this cell 222 private char token = ' '; 223 224 public Cell(int row, int column) { 225 this.row = row; 226 this.column = column; 227 this.setPrefSize(2000, 2000); // What happens without this? 228 setStyle("-fx-border-color: black"); // Set cell's border 229 this.setOnMouseClicked(e -> handleMouseClick()); 230 } 231 232 /** Return token */ 233 public char getToken() { 234 return token; 235 } 236 237 /** Set a new token */ 238 public void setToken(char c) { 239 token = c; 240 repaint(); 241 } 242 243 protected void repaint() { 244 if (token == 'X') { 245 Line line1 = new Line(10, 10, 246 this.getWidth() − 10, this.getHeight() − 10); 247 line1.endXProperty().bind(this.widthProperty().subtract(10)); 248 line1.endYProperty().bind(this.heightProperty().subtract(10)); 249 Line line2 = new Line(10, this.getHeight() − 10, 250 this.getWidth() − 10, 10); 251 line2.startYProperty().bind( 252 this.heightProperty().subtract(10)); 253 line2.endXProperty().bind(this.widthProperty().subtract(10)); 254 255 // Add the lines to the pane 256 this.getChildren().addAll(line1, line2); 257 } 258 else if (token == 'O') { 259 Ellipse ellipse = new Ellipse(this.getWidth() / 2, 260 this.getHeight() / 2, this.getWidth() / 2 − 10, 261 this.getHeight() / 2 − 10); 262 ellipse.centerXProperty().bind( 263 this.widthProperty().divide(2)); 264 ellipse.centerYProperty().bind( 265 this.heightProperty().divide(2)); 266 ellipse.radiusXProperty().bind( 267 this.widthProperty().divide(2).subtract(10)); 268 ellipse.radiusYProperty().bind( 269 this.heightProperty().divide(2).subtract(10)); 270 ellipse.setStroke(Color.BLACK); 271 ellipse.setFill(Color.WHITE);

model a cell

draw X

draw O

M33_LIAN1878_11_GE_C33.indd 28 12/29/17 5:48 PM

33.6 Case Study: Distributed Tic-Tac-Toe Games 33-29

272 273 getChildren().add(ellipse); // Add the ellipse to the pane 274 } 275 } 276 277 /* Handle a mouse click event */ 278 private void handleMouseClick() { 279 // If cell is not occupied and the player has the turn 280 if (token == ' ' && myTurn) { 281 setToken(myToken); // Set the player's token in the cell 282 myTurn = false; 283 rowSelected = row; 284 columnSelected = column; 285 lblStatus.setText("Waiting for the other player to move"); 286 waiting = false; // Just completed a successful move 287 } 288 } 289 } 290 }

The server can serve any number of sessions simultaneously. Each session takes care of two players. The client can be deployed to run as a Java applet. To run a client as a Java applet from a Web browser, the server must run from a Web server. Figures 33.15 and 33.16 show sample runs of the server and the clients.

mouse clicked handler

Figure 33.15 TicTacToeServer accepts connection requests and creates sessions to serve pairs of players.

The TicTacToeConstants interface defines the constants shared by all the classes in the project. Each class that uses the constants needs to implement the interface. Centrally defining constants in an interface is a common practice in Java.

Once a session is established, the server receives moves from the players in alternation. Upon receiving a move from a player, the server determines the status of the game. If the game is not finished, the server sends the status (CONTINUE) and the player’s move to the other

Figure 33.16 TicTacToeClient can run as an applet or standalone.

M33_LIAN1878_11_GE_C33.indd 29 12/29/17 5:48 PM

33-30 Chapter 33 Networking

player. If the game is won or a draw, the server sends the status (PLAYER1_WON, PLAYER2_ WON, or DRAW) to both players.

The implementation of Java network programs at the socket level is tightly synchronized. An operation to send data from one machine requires an operation to receive data from the other machine. As shown in this example, the server and the client are tightly synchronized to send or receive data.

33.6.1 What would happen if the preferred size for a cell is not set in line 227 in Listing 33.10?

33.6.2 If a player does not have the turn but clicks on an empty cell, what will the client program in Listing 33.10 do?

Point Check

ChapTer summary 1. Java supports stream sockets and datagram sockets. Stream sockets use TCP (Trans-

mission Control Protocol) for data transmission, whereas datagram sockets use UDP (User Datagram Protocol). Since TCP can detect lost transmissions and resubmit them, transmissions are lossless and reliable. UDP, in contrast, cannot guarantee lossless transmission.

2. To create a server, you must first obtain a server socket, using new ServerSocket(port). After a server socket is created, the server can start to listen for connections, using the accept() method on the server socket. The client requests a connection to a server by using new Socket(serverName, port) to create a client socket.

3. Stream socket communication is very much like input/output stream communication after the connection between a server and a client is established. You can obtain an input stream using the getInputStream() method and an output stream using the getOut- putStream() method on the socket.

4. A server must often work with multiple clients at the same time. You can use threads to handle the server’s multiple clients simultaneously by creating a thread for each connection.

Quiz Answer the quiz for this chapter online at book Companion Website.

client socket 33-3 domain name 33-2 domain name server 33-2 localhost 33-3 IP address 33-2 port 33-2

packet-based communication 33-2 server socket 33-2 socket 33-2 stream-based communication 33-2 TCP 33-2 UDP 33-2

Key Terms

M33_LIAN1878_11_GE_C33.indd 30 12/29/17 5:48 PM

Programming Exercises 33-31

programming exerCiSeS

Section 33.2 *33.1 (Loan server) Write a server for a client. The client sends loan information (annual

interest rate, number of years, and loan amount) to the server (see Figure 33.17a). The server computes monthly payment and total payment, and sends them back to the client (see Figure 33.17b). Name the client Exercise33_01Client and the server Exercise33_01Server.

Figure 33.17 The client in (a) sends the annual interest rate, number of years, and loan amount to the server and receives the monthly payment and total payment from the server in (b).

(a) (b)

Figure 33.18 The client in (a) sends the weight and height of a person to the server and receives the BMI from the server in (b).

(a) (b)

*33.2 (BMI server) Write a server for a client. The client sends the weight and height for a person to the server (see Figure 33.18a). The server computes BMI (Body Mass Index) and sends back to the client a string that reports the BMI (see Figure 33.18b). See Section 3.8 for computing BMI. Name the client Exercise33_02Client and the server Exercise33_02Server.

Sections 33.3 and 33.4 *33.3 (Loan server for multiple clients) Revise Programming Exercise 33.1 to write a

server for multiple clients.

Section 33.5 33.4 (Count clients) Write a server that tracks the number of the clients connected to

the server. When a new connection is established, the count is incremented by 1. The count is stored using a random-access file. Write a client program that receives

M33_LIAN1878_11_GE_C33.indd 31 12/29/17 5:48 PM

33-32 Chapter 33 Networking

the count from the server and displays a message, such as “You are visitor number 11”, as shown in Figure 33.19. Name the client Exercise33_04Client and the server Exercise33_04Server.

■■ Use the StudentAddress class defined in Listing 33.5 to hold the name, street, city, state, and zip in an object.

■■ The user can use the buttons First, Next, Previous, and Last to view an address, and the Add button to add a new address.

■■ Limit the concurrent connections to two clients.

Name the client Exercise33_06Client and the server Exercise33_6Server.

*33.7 (Transfer last 100 numbers in an array) Programming Exercise 22.12 retrieves the last 100 prime numbers from a file PrimeNumbers.dat. Write a client program that requests the server to send the last 100 prime numbers in an array. Name the server program Exercise33_07Server and the client program Exercise33_07Client. Assume the numbers of the long type are stored in PrimeNumbers.dat in binary format.

*33.8 (Transfer last 100 numbers in an ArrayList) Programming Exercise 22.12 retrieves the last 100 prime numbers from a file PrimeNumbers.dat. Write a cli- ent program that requests the server to send the last 100 prime numbers in an ArrayList. Name the server program Exercise33_08Server and the client pro- gram Exercise33_08Client. Assume the numbers of the long type are stored in PrimeNumbers.dat in binary format.

Figure 33.20 You can view and add an address.

Figure 33.19 The client displays how many times the server has been accessed. The server stores the count.

33.5 (Send loan information in an object) Revise Exercise 33.1 for the client to send a loan object that contains annual interest rate, number of years, and loan amount and for the server to send the monthly payment and total payment.

Section 33.6 33.6 (Display and add addresses) Develop a client/server application to view and add

addresses, as shown in Figure 33.20.

M33_LIAN1878_11_GE_C33.indd 32 12/29/17 5:48 PM

Programming Exercises 33-33

Section 33.7 **33.9 (Chat) Write a program that enables two users to chat. Implement one user as the

server (see Figure 33.21a) and the other as the client (see Figure 33.21b). The server has two text areas: one for entering text, and the other (noneditable) for displaying text received from the client. When the user presses the Enter key, the current line is sent to the client. The client has two text areas: one (noneditable) for displaying text from the server and the other for entering text. When the user presses the Enter key, the current line is sent to the server. Name the client Exercise33_09Client and the server Exercise33_09Server.

Figure 33.21 The server and client send text to and receive text from each other.

(a) (b)

Figure 33.22 The server starts in (a) with three clients in (b) and (c).

(a) (b) (c)

***33.10 (Multiple client chat) Write a program that enables any number of clients to chat. Implement one server that serves all the clients, as shown in Figure 33.22. Name the client Exercise33_10Client and the server Exercise33_10Server.

M33_LIAN1878_11_GE_C33.indd 33 12/29/17 5:48 PM

Objectives ■■ To understand the concepts of databases and database management

systems (§34.2).

■■ To understand the relational data model: relational data structures, con- straints, and languages (§34.2).

■■ To use SQL to create and drop tables and to retrieve and modify data (§34.3).

■■ To learn how to load a driver, connect to a database, execute state- ments, and process result sets using JDBC (§34.4).

■■ To use prepared statements to execute precompiled SQL statements (§34.5).

■■ To use callable statements to execute stored SQL procedures and func- tions (§34.6).

■■ To explore database metadata using the DatabaseMetaData and ResultSetMetaData interfaces (§34.7).

Java Database Programming

Chapter

M34_LIAN1878_11_GE_C34.indd 1 12/29/17 5:48 PM

34-2 Chapter 34 Java Database Programming

34.1 Introduction Java provides the API for developing database applications that works with any relational database systems.

You may have heard a lot about database systems. Database systems are everywhere. Your social security information is stored in a database by the government. If you shop online, your purchase information is stored in a database by the company. If you attend a university, your academic information is stored in a database by the university. Database systems not only store data, they also provide means of accessing, updating, manipulating, and ana- lyzing data. Your social security information is updated periodically, and you can register for courses online. Database systems play an important role in society and in commerce.

This chapter introduces database systems, the SQL language, and how database applica- tions can be developed using Java. If you already know SQL, you can skip Sections 34.2 and 34.3.

34.2 Relational Database Systems SQL is the standard database language for defining and accessing databases.

A database system consists of a database, the software that stores and manages data in the database, and the application programs that present data and enable the user to interact with the database system, as shown in Figure 34.1.

Point Key

database system

Figure 34.1 A database system consists of data, database management software, and appli- cation programs.

database

Application Users

Application Programs

Database Management System (DBMS)

System Users

A database is a repository of data that form information. When you purchase a database system—such as MySQL, Oracle, IBM’s DB2 and Informix, Microsoft SQL Server, or Sybase—from a software vendor, you actually purchase the software comprising a database management system (DBMS). Database management systems are designed for use by profes- sional programmers and are not suitable for ordinary customers. Application programs are built on top of the DBMS for customers to access and update the database. Thus, application pro- grams can be viewed as the interfaces between the database system and its users. Application programs may be stand-alone GUI applications or Web applications and may access several different database systems in the network, as shown in Figure 34.2.

Most of today’s database systems are relational database systems. They are based on the relational data model, which has three key components: structure, integrity, and language.

DBMS

M34_LIAN1878_11_GE_C34.indd 2 12/29/17 5:48 PM

34.2 Relational Database Systems 34-3

Structure defines the representation of the data. Integrity imposes constraints on the data. Language provides the means for accessing and manipulating data.

34.2.1 Relational Structures The relational model is built around a simple and natural structure. A relation is actually a table that consists of nonduplicate rows. Tables are easy to understand and use. The relational model provides a simple yet powerful way to represent data.

A row of a table represents a record, and a column of a table represents the value of a single attribute of the record. In relational database theory, a row is called a tuple, and a column is called an attribute. Figure 34.3 shows a sample table that stores information about the courses offered by a university. The table has eight tuples, and each tuple has five attributes.

relational model

tuple attribute

Figure 34.2 An application program can access multiple database systems.

Database Management System

database

Application Programs

Application Users

Database Management System

……

…

Figure 34.3 A table has a table name, column names, and rows.

Columns/Attributes

Tuples/ Rows 11111 CSCI 1301 Introduction to Java I 4

11112 CSCI 1302 Introduction to Java II 3 11113 CSCI 3720 Database Systems 3 11114 CSCI 4750 Rapid Java Application 3 11115 MATH 2750 Calculus I 5 11116 MATH 3750 Calculus II 5 11117 EDUC 1111 Reading 3 11118 ITEC 1344 Database Administration 3

courseId subjectId courseNumber title numOfCreditsCourse Table

Relation/Table Name

Tables describe the relationship among data. Each row in a table represents a record of related data. For example, “11111,” “CSCI,” “1301,” “Introduction to Java I,” and “4” are related to form a record (the first row in Figure 34.3) in the Course table. Just as the data in the same row are related, so too data in different tables may be related through common attributes. Suppose the database has two other tables, Student and Enrollment, as shown in

M34_LIAN1878_11_GE_C34.indd 3 12/29/17 5:48 PM

34-4 Chapter 34 Java Database Programming

34.2.2 Integrity Constraints An integrity constraint imposes a condition that all the legal values in a table must satisfy. Figure 34.6 shows an example of some integrity constraints in the Subject and Course tables.

In general, there are three types of constraints: domain constraints, primary key con- straints, and foreign key constraints. Domain constraints and primary key constraints are known as intrarelational constraints, meaning that a constraint involves only one relation. The foreign key constraint is interrelational, meaning that a constraint involves more than one relation.

integrity constraint

Figure 34.4 A Student table stores student information.

deptID

444111110 Jacob R Smith 9129219434 1985-04-09 99 Kingston Street 31435 BIOL 444111111 John K Stevenson 9129219434 null 100 Main Street 31411 BIOL 444111112 George K Smith 9129213454 1974-10-10 1200 Abercorn St. 31419 CS 444111113 Frank E Jones 9125919434 1970-09-09 100 Main Street 31411 BIOL 444111114 Jean K Smith 9129219434 1970-02-09 100 Main Street 31411 CHEM 444111115 Josh R Woo 7075989434 1970-02-09 555 Franklin St. 31411 CHEM 444111116 Josh R Smith 9129219434 1973-02-09 100 Main Street 31411 BIOL 444111117 Joy P Kennedy 9129229434 1974-03-19 103 Bay Street 31412 CS 444111118 Toni R Peterson 9129229434 1964-04-29 103 Bay Street 31412 MATH 444111119 Patrick R Stoneman 9129229434 1969-04-29 101 Washington St. 31435 MATH 444111120 Rick R Carter 9125919434 1986-04-09 19 West Ford St. 31411 BIOL

Student Table

ssn �rstName mi lastName phone birthDate street zipCode

Figure 34.5 An Enrollment table stores student enrollment information.

Enrollment Table

444111110 11111 2004-03-19 A 444111110 11112 2004-03-19 B 444111110 11113 2004-03-19 C 444111111 11111 2004-03-19 D 444111111 11112 2004-03-19 F 444111111 11113 2004-03-19 A 444111112 11114 2004-03-19 B 444111112 11115 2004-03-19 C 444111112 11116 2004-03-19 D 444111113 11111 2004-03-19 A 444111113 11113 2004-03-19 A 444111114 11115 2004-03-19 B 444111115 11115 2004-03-19 F 444111115 11116 2004-03-19 F 444111116 11111 2004-03-19 D 444111117 11111 2004-03-19 D 444111118 11111 2004-03-19 A 444111118 11112 2004-03-19 D 444111118 11113 2004-03-19 B

ssn courseId dateRegistered grade

Figures 34.4 and 34.5. The Course table and the Enrollment table are related through their common attribute courseId, and the Enrollment table and the Student table are related through ssn.

M34_LIAN1878_11_GE_C34.indd 4 12/29/17 5:48 PM

34.2 Relational Database Systems 34-5

Domain Constraints

Domain constraints specify the permissible values for an attribute. Domains can be specified using standard data types, such as integers, floating-point numbers, fixed-length strings, and variant-length strings. The standard data type specifies a broad range of values. Additional con- straints can be specified to narrow the ranges. For example, you can specify that the numOfCredits attribute (in the Course table) must be greater than 0 and less than 5. If an attribute has different values for each tuple in a relation, you can specify the attribute to be unique. You can also specify whether an attribute can be null, which is a special value in a database meaning unknown or not applicable. As shown in the Student table, birthDate may be null.

Primary Key Constraints

A primary key is a set of attributes that uniquely identifyies the tuples in a relations. Why is it called a primary key, rather than simply key? To understand this, it is helpful to know superkeys, keys, and candidate keys. A superkey is an attribute or a set of attributes that uniquely identifies the relation. That is, no two tuples have the same values on a superkey. By definition, a relation consists of a set of distinct tuples. The set of all attributes in the relation forms a superkey.

A key K is a minimal superkey, meaning that any proper subset of K is not a superkey. A relation can have several keys. In this case, each of the keys is called a candidate key. The primary key is one of the candidate keys designated by the database designer. The primary key is often used to identify tuples in a relation. As shown in Figure 34.6, courseId is the primary key in the Course table, and ssn and courseId form a primary key in the Enrollment table.

Foreign Key Constraints

In a relational database, data are related. Tuples in a relation are related, and tuples in different relations are related through their common attributes. Informally speaking, the common attrib- utes are foreign keys. The foreign key constraints define the relationships among relations.

Formally, a set of attributes FK is a foreign key in a relation R that references relation T if it satisfies the following two rules:

■■ The attributes in FK have the same domain as the primary key in T.

■■ A nonnull value on FK in R must match a primary key value in T.

domain constraint

superkey

primary key

candidate key

relational database

foreign key constraint foreign key

Figure 34.6 The Enrollment table and the Course table have integrity constraints.

11111 CSCI 1301 Introduction to Java I 4 11112 CSCI 1302 Introduction to Java II 3 11113 CSCI 3720 Database Systems 3 ...

444111110 11111 2004-03-19 A 444111110 11112 2004-03-19 B 444111110 11113 2004-03-19 C ...

Course Table

Each value in the numOfCredits column must be greater than 0 and less than 5

Each value in courseId in the Enrollment table must match a value in courseId in the Course table

Each row must have a value for courseId, and the value must be unique

Enrollment Table ssn courseId dateRegistered grade

courseId subjectId courseNumber title numOfCredits

M34_LIAN1878_11_GE_C34.indd 5 12/29/17 5:48 PM

34-6 Chapter 34 Java Database Programming

As shown in Figure 34.6, courseId is the foreign key in Enrollment that references the primary key courseId in Course. Every courseId value must match a courseId value in Course.

Enforcing Integrity Constraints

The database management system enforces integrity constraints and rejects operations that would violate them. For example, if you attempt to insert the new record (“11115,” “CSCI,” “2490,” “C+ + Programming,” “0”) into the Course table, it would fail because the credit hours must be greater than 0; if you attempted to insert a record with the same primary key as an existing record in the table, the DBMS would report an error and reject the operation; if you attempted to delete a record from the Course table whose primary key value is referenced by the records in the Enrollment table, the DBMS would reject this operation.

Note All relational database systems support primary key constraints and foreign key constraints, but not all database systems support domain constraints. In the Microsoft Access database, for example, you cannot specify the constraint that numOfCredits is greater than 0 and less than 5.

34.2.1 What are superkeys, candidate keys, and primary keys? 34.2.2 What is a foreign key? 34.2.3 Can a relation have more than one primary key or foreign key? 34.2.4 Does a foreign key need to be a primary key in the same relation? 34.2.5 Does a foreign key need to have the same name as its referenced primary key? 34.2.6 Can a foreign key value be null?

34.3 SQL Structured Query Language (SQL) is the language for defining tables and integrity constraints, and for accessing and manipulating data.

SQL (pronounced “S-Q-L” or “sequel”) is the universal language for accessing relational database systems. Application programs may allow users to access a database without directly using SQL, but these applications themselves must use SQL to access the database. This section introduces some basic SQL commands.

Note There are many relational database management systems. They share the common SQL language but do not all support every feature of SQL. Some systems have their own extensions to SQL. This section introduces standard SQL supported by all systems.

SQL can be used on MySQL, Oracle, Sybase, IBM DB2, IBM Informix, MS Access, Apache Derby, or any other relational database system. Apache Derby is an open source rela- tional database management system developed using Java. Oracle distributes Apache Derby as Java DB and bundled with Java so you can use it in any Java application without installing a database. Java DB is ideal for supporting a small database in a Java application. This chapter uses MySQL to demonstrate SQL and Java database programming.

The Companion Website contains the following supplements on how to install and use three popular databases: MySQL, Oracle, and Java DB:

■■ Supplement IV.B: Tutorial for MySQL

■■ Supplement IV.C: Tutorial for Oracle

■■ Supplement IV.D: Tutorial for Java DB

auto enforcement

Point Check

Point Key

SQL

database language

standard SQL

MySQL Tutorial

Oracle Tutorial

Java DB Tutorial

M34_LIAN1878_11_GE_C34.indd 6 12/29/17 5:48 PM

34.3 SQL 34-7

34.3.1 Creating a User Account on MySQL Assume you have installed MySQL 5 with the default configuration. To match all the examples in this book, you should create a user named scott with the password tiger. You can perform the administrative tasks using the MySQL Workbench or using the command line. MySQL Workbench is a GUI tool for managing MySQL databases. Here are the steps to create a user from the command line:

1. From the DOS command prompt, type

mysql –uroot -p

You will be prompted to enter the root password, as shown in Figure 34.7.

2. At the mysql prompt, enter

use mysql;

3. To create user scott with password tiger, enter

create user 'scott'@'localhost' identified by 'tiger';

4. To grant privileges to scott, enter

grant select, insert, update, delete, create, create view, drop, execute, references on *.* to 'scott'@'localhost';

■■ If you want to enable remote access of the account from any IP address, enter

grant all privileges on *.* to 'scott'@'%' identified by 'tiger';

■■ If you want to restrict the account’s remote access to just one particular IP address, enter

grant all privileges on *.* to 'scott'@'ipAddress' identified by 'tiger';

5. Enter

exit;

to exit the MySQL console.

Figure 34.7 You can access a MySQL database server from the command window.

M34_LIAN1878_11_GE_C34.indd 7 12/29/17 5:48 PM

34-8 Chapter 34 Java Database Programming

Note On Windows, your MySQL database server starts every time your computer starts. You can stop it by typing the command net stop mysql and restart it by typing the command net start mysql.

By default, the server contains two databases named mysql and test. The mysql database contains the tables that store information about the server and its users. This database is intended for the server administrator to use. For example, the administrator can use it to cre- ate users and grant or revoke user privileges. Since you are the owner of the server installed on your system, you have full access to the mysql database. However, you should not cre- ate user tables in the mysql database. You can use the test database to store data or create new databases. You can also create a new database using the command create data- base databasename or delete an existing database using the command drop database databasename.

34.3.2 Creating a Database To match the examples in this book, you should create a database named javabook. Here are the steps to create it:

1. From the DOS command prompt, type

mysql –uscott -ptiger

to login to mysql, as shown in Figure 34.8.

2. At the mysql prompt, enter

create database javabook;

stop mysql start mysql

Figure 34.8 You can create databases in MySQL.

For your convenience, the SQL statements for creating and initializing tables used in this book are provided in Supplement IV.A. You can download the script for MySQL and save it to script.sql. To execute the script, first switch to the javabook database using the following command:

use javabook;

then type

source script.sql;

as shown in Figure 34.9.

run script file

M34_LIAN1878_11_GE_C34.indd 8 12/29/17 5:48 PM

34.3 SQL 34-9

Note You can populate the javabook database using the script from Supplement IV.A.

34.3.3 Creating and Dropping Tables Tables are the essential objects in a database. To create a table, use the create table state- ment to specify a table name, attributes, and types, as in the following example:

create table Course ( courseId char(5), subjectId char(4) not null, courseNumber integer, title varchar(50) not null, numOfCredits integer, primary key (courseId) );

This statement creates the Course table with attributes courseId, subjectId, courseNumber, title, and numOfCredits. Each attribute has a data type that specifies the type of data stored in the attribute. char(5) specifies that courseId consists of five charac- ters. varchar(50) specifies that title is a variant-length string with a maximum of 50 char- acters. integer specifies that courseNumber is an integer. The primary key is courseId.

The tables Student and Enrollment can be created as follows:

populating database

create table

Figure 34.9 You can run SQL commands in a script file.

create table Student ( ssn char(9), firstName varchar(25), mi char(1), lastName varchar(25), birthDate date, street varchar(25), phone char(11), zipCode char(5), deptId char(4), primary key (ssn) );

create table Enrollment ( ssn char(9), courseId char(5), dateRegistered date, grade char(1), primary key (ssn, courseId), foreign key (ssn) references Student(ssn), foreign key (courseId) references Course(courseId) );

Note SQL keywords are not case sensitive. This book adopts the following naming conventions: tables are named in the same way as Java classes, and attributes are named in the same way as Java variables. SQL keywords are named in the same way as Java keywords.

naming convention

M34_LIAN1878_11_GE_C34.indd 9 12/29/17 5:48 PM

34-10 Chapter 34 Java Database Programming

If a table is no longer needed, it can be dropped permanently using the drop table command. For example, the following statement drops the Course table:

drop table Course;

If a table to be dropped is referenced by other tables, you have to drop the other tables first. For example, if you have created the tables Course, Student, and Enrollment and want to drop Course, you have to first drop Enrollment, because Course is referenced by Enrollment.

Figure 34.10 shows how to enter the create table statement from the MySQL console.

drop table

Figure 34.10 A table is created using the create table statement.

Figure 34.11 (a) You can use Notepad to create a text file for SQL commands. (b) You can run the SQL commands in a script file from MySQL.

(a) (b)

If you make typing errors, you have to retype the whole command. To avoid retyping, you can save the command in a file, then run the command from the file. To do so, create a text file to contain commands, named, for example, test.sql. You can create the text file using any text editor, such as Notepad, as shown in Figure 34.11a. To comment a line, precede it with two dashes. You can now run the script file by typing source test.sql from the SQL command prompt, as shown in Figure 34.11b.

34.3.4 Simple Insert, Update, and Delete Once a table is created, you can insert data into it. You can also update and delete records. This section introduces simple insert, update, and delete statements.

The syntax to insert a record into a table is:

insert into tableName [(column1, column2, ..., column)] values (value1, value2, ..., valuen);

M34_LIAN1878_11_GE_C34.indd 10 12/29/17 5:48 PM

34.3 SQL 34-11

For example, the following statement inserts a record into the Course table. The new record has the courseId ‘11113’, subjectId ‘CSCI’, courseNumber ‘3720’, title ‘Database Systems’, and creditHours 3.

insert into Course (courseId, subjectId, courseNumber, title, numOfCredits) values ('11113', 'CSCI', '3720', 'Database Systems', 3);

The column names are optional. If they are omitted, all the column values for the record must be entered, even though the columns have default values. String values are case sensitive and enclosed inside single quotation marks in SQL.

The syntax to update a table is:

update tableName set column1 = newValue1 [, column2 = newValue2, ...] [where condition];

For example, the following statement changes the numOfCredits for the course whose title is Database Systems to 4.

update Course set numOfCredits = 4 where title = 'Database Systems';

The syntax to delete records from a table is:

delete from tableName [where condition];

For example, the following statement deletes the Database Systems course from the Course table:

delete from Course where title = 'Database Systems';

The following statement deletes all the records from the Course table:

delete from Course;

34.3.5 Simple Queries To retrieve information from tables, use a select statement with the following syntax:

select column-list from table-list [where condition];

The select clause lists the columns to be selected. The from clause refers to the tables involved in the query. The optional where clause specifies the conditions for the selected rows.

Query 1: Select all the students in the CS department, as shown in Figure 34.12.

select firstName, mi, lastName from Student where deptId = 'CS';

M34_LIAN1878_11_GE_C34.indd 11 12/29/17 5:48 PM

34-12 Chapter 34 Java Database Programming

34.3.6 Comparison and Boolean Operators SQL has six comparison operators, as shown in Table 34.1, and three Boolean operators, as shown in Table 34.2.

Operator Description

= Equal to

<> or != Not equal to

< Less than

<= Less than or equal to

> Greater than

>= Greater than or equal to

Table 34.1 Comparison Operators

Operator Description

not Logical negation

and Logical conjunction

or Logical disjunction

Table 34.2 Boolean Operators

Note The comparison and Boolean operators in SQL have the same meanings as in Java. In SQL the equal to operator is =, but in Java it is ==. In SQL the not equal to operator is <> or !=, but in Java it is !=. The not, and, and or operators are !, && (&), and || (|) in Java.

Query 2: Get the names of the students who are in the CS dept and live in the ZIP code 31411.

select firstName, mi, lastName from Student where deptId = 'CS' and zipCode = '31411';

Note To select all the attributes from a table, you don’t have to list all the attribute names in the select clause. Instead, you can just use an asterisk (*), which stands for all the attributes. For example, the following query displays all the attributes of the students who are in the CS dept and live in ZIP code 31411.

select * from Student where deptId = 'CS' and zipCode = '31411';

Figure 34.12 The result of the select statement is displayed in the MySQL console.

M34_LIAN1878_11_GE_C34.indd 12 12/29/17 5:48 PM

34.3 SQL 34-13

34.3.7 The like, between-and, and is null Operators SQL has a like operator that can be used for pattern matching. The syntax to check whether a string s has a pattern p is

s like p or s not like p

You can use the wildcard characters % (percent symbol) and _ (underline symbol) in the pattern p. % matches zero or more characters, and _ matches any single character in s. For example, lastName like '_mi%' matches any string whose second and third letters are m and i. lastName not like '_mi%' excludes any string whose second and third letters are m and i.

Note In earlier versions of MS Access, the wildcard character is *, and the character ? matches any single character.

The between-and operator checks whether a value v is between two other values, v1 and v2, using the following syntax:

v between v1 and v2 or v not between v1 and v2

v between v1 and v2 is equivalent to v >= v1 and v <= v2, and v not between v1 and v2 is equivalent to v < v1 or v > v2.

The is null operator checks whether a value v is null using the following syntax:

v is null or v is not null

Query 3: Get the Social Security numbers of the students whose grades are between ‘C’ and ‘A’.

select ssn from Enrollment where grade between 'C' and 'A';

34.3.8 Column Alias When a query result is displayed, SQL uses the column names as column headings. Usually the user gives abbreviated names for the columns, and the columns cannot have spaces when the table is created. Sometimes it is desirable to give more descriptive names in the result heading. You can use the column aliases with the following syntax:

select columnName [as] alias

Query 4: Get the last name and ZIP code of the students in the CS department. Display the column headings as “Last Name” for lastName and “Zip Code” for zipCode. The query result is shown in Figure 34.13.

Figure 34.13 You can use a column alias in the display.

M34_LIAN1878_11_GE_C34.indd 13 12/29/17 5:48 PM

34-14 Chapter 34 Java Database Programming

select lastName as "Last Name", zipCode as "Zip Code" from Student where deptId = 'CS';

Note The as keyword is optional in MySQL and Oracle, but it is required in MS Access.

34.3.9 The Arithmetic Operators You can use the arithmetic operators * (multiplication), / (division), + (addition), and − (sub- traction) in SQL.

Query 5: Assume a credit hour is 50 minutes of lectures and get the total minutes for each course with the subject CSCI. The query result is shown in Figure 34.14.

select title, 50 * numOfCredits as "Lecture Minutes Per Week" from Course where subjectId = 'CSCI';

Figure 34.15 (a) The duplicate tuples are displayed. (b) The distinct tuples are displayed.

(a) (b)

Figure 34.14 You can use arithmetic operators in SQL.

34.3.10 Displaying Distinct Tuples SQL provides the distinct keyword, which can be used to eliminate duplicate tuples in the result. Figure 34.15a displays all the subject IDs used by the courses, and Figure 34.15b dis- plays all the distinct subject IDs used by the courses using the following statement:

select distinct subjectId as "Subject ID" from Course;

M34_LIAN1878_11_GE_C34.indd 14 12/29/17 5:48 PM

34.3 SQL 34-15

When there is more than one column in the select clause, the distinct keyword applies to the whole tuple in the result. For example, the following statement displays all tuples with distinct subjectId and title, as shown in Figure 34.16. Note some tuples may have the same subjectId but different title. These tuples are distinct.

select distinct subjectId, title from Course;

Figure 34.16 The keyword distinct applies to the entire tuple.

34.3.11 Displaying Sorted Tuples SQL provides the order by clause to sort the output using the following syntax:

select column-list from table-list [where condition] [order by columns-to-be-sorted];

In the syntax, columns-to-be-sorted specifies a column or a list of columns to be sorted. By default, the order is ascending. To sort in a descending order, append the desc keyword. You could also append the asc keyword after columns-to-be-sorted, but it is not neces- sary. When multiple columns are specified, the rows are sorted based on the first column, then the rows with the same values on the first column are sorted based on the second column, and so on.

Query 6: List the full names of the students in the CS department, ordered primarily on their last names in descending order and secondarily on their first names in ascending order. The query result is shown in Figure 34.17.

Figure 34.17 You can sort results using the order by clause.

M34_LIAN1878_11_GE_C34.indd 15 12/29/17 5:48 PM

34-16 Chapter 34 Java Database Programming

select lastName, firstName, deptId from Student where deptId = 'CS' order by lastName desc, firstName asc;

34.3.12 Joining Tables Often you need to get information from multiple tables, as demonstrated in the next query.

Query 7: List the courses taken by the student Jacob Smith. To solve this query, you need to join tables Student and Enrollment, as shown in Figure 34.18.

Figure 34.18 Student and Enrollment are joined on ssn.

A tuple

Student Table

ssn lastName mi �rstName …

Enrollment Table

ssn courseId …

Equal

Figure 34.19 Query 7 demonstrates queries involving multiple tables.

You can write the query in SQL as follows:

select distinct lastName, firstName, courseId from Student, Enrollment where Student.ssn = Enrollment.ssn and lastName = 'Smith' and firstName = 'Jacob';

The tables Student and Enrollment are listed in the from clause. The query examines every pair of rows, each made of one item from Student and another from Enrollment and selects the pairs that satisfy the condition in the where clause. The rows in Student have the last name, Smith, and the first name, Jacob, and both rows from Student and Enrollment have the same ssn values. For each pair selected, lastName and firstName from Student and courseId from Enrollment are used to produce the result, as shown in Figure 34.19. Student and Enrollment have the same attribute ssn. To distinguish them in a query, use Student.ssn and Enrollment.ssn.

M34_LIAN1878_11_GE_C34.indd 16 12/29/17 5:48 PM

34.4 JDBC 34-17

For more features of SQL, see Supplements IV.H and IV.I.

34.3.1 Create the tables Course, Student, and Enrollment using the create table statements in Section 34.3.3, Creating and Dropping Tables. Insert rows into the Course, Student, and Enrollment tables using the data in Figures 34.3–34.5.

34.3.2 List all CSCI courses with at least four credit hours. 34.3.3 List all students whose last names contain the letter e two times. 34.3.4 List all students whose birthdays are null. 34.3.5 List all students who take Math courses. 34.3.6 List the number of courses in each subject. 34.3.7 Assume each credit hour is 50 minutes of lectures. Get the total minutes for the

courses that each student takes.

34.4 JDBC JDBC is the Java API for accessing relational database.

The Java API for developing Java database applications is called JDBC. JDBC is the trade- marked name of a Java API that supports Java programs that access relational databases. JDBC is not an acronym, but it is often thought to stand for Java Database Connectivity.

JDBC provides Java programmers with a uniform interface for accessing and manipulat- ing relational databases. Using the JDBC API, applications written in the Java programming language can execute SQL statements, retrieve results, present data in a user-friendly interface, and propagate changes back to the database. The JDBC API can also be used to interact with multiple data sources in a distributed, heterogeneous environment.

The relationships among Java programs, JDBC API, JDBC drivers, and relational databases are shown in Figure 34.20. The JDBC API is a set of Java interfaces and classes used to write Java programs for accessing and manipulating relational databases. Since a JDBC driver serves as the interface to facilitate communications between JDBC and a proprietary database, JDBC drivers are database specific and are normally provided by the database vendors. You need

Point Check

Point Key

Figure 34.20 Java programs access and manipulate databases through JDBC drivers.

Java Programs

JDBC API

DB2 JDBC Driver

Local or remote ORACLE DB

Local or remote DB2 DB

MySQL JDBC Driver

Local or remote MySQL DB

Oracle JDBC Driver

M34_LIAN1878_11_GE_C34.indd 17 12/29/17 5:48 PM

34-18 Chapter 34 Java Database Programming

MySQL JDBC drivers to access the MySQL database, Oracle JDBC drivers to access the Oracle database, and DB2 JDBC driver to access the DB2 database.

34.4.1 Developing Database Applications Using JDBC The JDBC API is a Java application program interface to generic SQL databases that enables Java developers to develop DBMS-independent Java applications using a uniform interface.

The JDBC API consists of classes and interfaces for establishing connections with data- bases, sending SQL statements to databases, processing the results of SQL statements, and obtaining database metadata. Four key interfaces are needed to develop any database applica- tion using Java: Driver, Connection, Statement, and ResultSet. These interfaces define a framework for generic SQL database access. The JDBC API defines these interfaces, and the JDBC driver vendors provide the implementation for the interfaces. Programmers use these interfaces.

The relationship of these interfaces is shown in Figure 34.21. A JDBC application loads an appropriate driver using the Driver interface, connects to the database using the Connection interface, creates and executes SQL statements using the Statement interface, and processes the result using the ResultSet interface if the statements return results. Note some state- ments, such as SQL data definition statements and SQL data modification statements, do not return results.

Figure 34.21 JDBC classes enable Java programs to connect to the database, send SQL statements, and process results.

Driver

Connection Connection

Statement

ResultSet

Statement

ResultSet

Statement

ResultSet

Statement

ResultSet

The JDBC interfaces and classes are the building blocks in the development of Java data- base programs. A typical Java program takes the following steps to access a database.

1. Loading drivers.

An appropriate driver must be loaded using the statement shown below before connecting to a database.

Class.forName("JDBCDriverClass");

A driver is a concrete class that implements the java.sql.Driver interface. The drivers for MySQL, Oracle, and Java DB are listed in Table 34.3. If your program accesses several differ- ent databases, all their respective drivers must be loaded.

The most recent platform independent version of MySQL JDBC driver is mysql- connector- java-5.1.26.jar. This file is contained in a ZIP file downloadable from dev.mysql.com/down- loads/connector/j/. The most recent version of Oracle JDBC driver is ojdbc6.jar (downloadable from www.oracle.com/technetwork/database/enterprise-edition/jdbc-112010-090769.html).

mysql-connector-java-5.1.26.jar ojdbc6.jar

M34_LIAN1878_11_GE_C34.indd 18 12/29/17 5:48 PM

34.4 JDBC 34-19

Java DB has two versions: embedded and networked. Embedded version is used when you access Java DB locally, while the network version enables you to access Java DB on the net- work. To use these drivers, you have to add their jar files in the classpath using the following DOS command on Windows:

set classpath=%classpath%;c:\book\lib\mysql-connector-java-5.1.26. jar;c:\book\lib\ojdbc6.jar;c:\program files\jdk1.8.0\db\lib\derby. jar

If you use an IDE such as Eclipse or NetBeans, you need to add these jar files into the library in the IDE.

Note com.mysql.jdbc.Driver is a class in mysql-connector-java-5.1.26.jar, and oracle.jdbc.driver.OracleDriver is a class in ojdbc6.jar. mysql- connector-java-5.1.26.jar, ojdbc6.jar, and derby.jar contains many classes to support the driver. These classes are used by JDBC but not directly by JDBC programmers. When you use a class explicitly in the program, it is automatically loaded by the JVM. The driver classes, however, are not used explicitly in the program, so you have to write the code to tell the JVM to load them.

Note Java supports automatic driver discovery, so you don’t have to load the driver explicitly. At the time of this writing, however, this feature is not supported for all database drivers. To be safe, load the driver explicitly.

2. Establishing connections.

To connect to a database, use the static method getConnection(databaseURL) in the DriverManager class, as follows:

Connection connection = DriverManager.getConnection(databaseURL);

where databaseURL is the unique identifier of the database on the Internet. Table 34.4 lists the URL patterns for the MySQL, Oracle, and Java DB.

why load a driver?

automatic driver discovery

Database Driver Class Source

MySQL com.mysql.jdbc.Driver mysql-connector-java-5.1.26.jar

Oracle oracle.jdbc.driver.OracleDriver ojdbc6.jar

Java DB (embedded) org.apache.derby.jdbc.EmbeddedDriver derby.jar

Java DB (network) org.apache.derby.jdbc.ClientDriver derbynet.jar

Table 34.3 JDBC Drivers

Database URL Pattern

MySQL jdbc:mysql://hostname/dbname

Oracle jdbc:oracle:thin:@hostname:port#:oracleDBSID

Java DB (embedded) jdbc:derby:dbname

Java DB (network) jdbc:derby://hostname/dbname

Table 34.4 JDBC URLs

M34_LIAN1878_11_GE_C34.indd 19 12/29/17 5:48 PM

34-20 Chapter 34 Java Database Programming

The databaseURL for a MySQL database specifies the host name and database name to locate a database. For example, the following statement creates a Connection object for the local MySQL database javabook with username scott and password tiger:

Connection connection = DriverManager.getConnection ("jdbc:mysql://localhost/javabook", "scott", "tiger");

Recall that by default, MySQL contains two databases named mysql and test. Section 34.3.2, Creating a Database, created a custom database named javabook. We will use javabook in the examples.

The databaseURL for an Oracle database specifies the hostname, the port# where the database listens for incoming connection requests, and the oracleDBSID database name to locate a database. For example, the following statement creates a Connection object for the Oracle database on liang.armstrong.edu with the username scott and password tiger:

Connection connection = DriverManager.getConnection ("jdbc:oracle:thin:@liang.armstrong.edu:1521:orcl", "scott", "tiger");

3. Creating statements.

If a Connection object can be envisioned as a cable linking your program to a database, an object of Statement can be viewed as a cart that delivers SQL statements for execution by the database and brings the result back to the program. Once a Connection object is created, you can create statements for executing SQL statements as follows:

Statement statement = connection.createStatement();

4. Executing statements.

SQL data definition language (DDL) and update statements can be executed using executeUpdate(String sql), and an SQL query statement can be executed using executeQuery(String sql). The result of the query is returned in ResultSet. For example, the following code executes the SQL statement create table Temp (col1 char(5), col2 char(5)):

statement.executeUpdate ("create table Temp (col1 char(5), col2 char(5))");

This next code executes the SQL query select firstName, mi, lastName from Student where lastName = 'Smith':

// Select the columns from the Student table ResultSet resultSet = statement.executeQuery ("select firstName, mi, lastName from Student where lastName " + " = 'Smith'");

5. Processing ResultSet.

The ResultSet maintains a table whose current row can be retrieved. The initial row position is null. You can use the next method to move to the next row and the various getter methods to retrieve values from a current row. For example, the following code displays all the results from the preceding SQL query:

// Iterate through the result and print the student names while (resultSet.next()) System.out.println(resultSet.getString(1) + " " + resultSet.getString(2) + " " + resultSet.getString(3));

connect MySQL DB

connect Oracle DB

M34_LIAN1878_11_GE_C34.indd 20 12/29/17 5:48 PM

34.4 JDBC 34-21

The getString(1), getString(2), and getString(3) methods retrieve the column values for firstName, mi, and lastName, respectively. Alternatively, you can use getString("firstName"), getString("mi"), and getString("lastName") to retrieve the same three column values. The first execution of the next() method sets the current row to the first row in the result set, and subsequent invocations of the next() method set the current row to the second row, third row, and so on, to the last row.

Listing 34.1 is a complete example that demonstrates connecting to a database, executing a simple query, and processing the query result with JDBC. The program connects to a local MySQL database and displays the students whose last name is Smith.

lisTing 34.1 SimpleJdbc.java 1 import java.sql.*; 2 3 public class SimpleJdbc { 4 public static void main(String[] args) 5 throws SQLException, ClassNotFoundException { 6 // Load the JDBC driver 7 Class.forName("com.mysql.jdbc.Driver"); 8 System.out.println("Driver loaded"); 9 10 // Connect to a database 11 Connection connection = DriverManager.getConnection 12 ("jdbc:mysql://localhost/javabook", "scott", "tiger"); 13 System.out.println("Database connected"); 14 15 // Create a statement 16 Statement statement = connection.createStatement(); 17 18 // Execute a statement 19 ResultSet resultSet = statement.executeQuery 20 ("select firstName, mi, lastName from Student where lastName " 21 + " = 'Smith'"); 22 23 // Iterate through the result and print the student names 24 while (resultSet.next()) 25 System.out.println(resultSet.getString(1) + "\t" + 26 resultSet.getString(2) + "\t" + resultSet.getString(3)); 27 28 // Close the connection 29 connection.close(); 30 } 31 }

The statement in line 7 loads a JDBC driver for MySQL, and the statement in lines 11–13 connects to a local MySQL database. You can change them to connect to an Oracle or other databases. The program creates a Statement object (line 16), executes an SQL statement and returns a ResultSet object (lines 19–21), and retrieves the query result from the ResultSet object (lines 24–26). The last statement (line 29) closes the connection and releases resources related to the connection. You can rewrite this program using the try-with-resources syntax. See www.cs.armstrong.edu/liang/intro11e/html/SimpleJdbcWithAutoClose.html.

Note If you run this program from the DOS prompt, specify the appropriate driver in the classpath, as shown in Figure 34.22.

load driver

connect database

create statement

execute statement

get result

close connection

run from DOS prompt

M34_LIAN1878_11_GE_C34.indd 21 12/29/17 5:48 PM

34-22 Chapter 34 Java Database Programming

The classpath directory and jar files are separated by commas. The period (.) represents the current directory. For convenience, the driver files are placed under the lib directory.

Caution Do not use a semicolon (;) to end the Oracle SQL command in a Java program. The semicolon may not work with the Oracle JDBC drivers. It does work, however, with the other drivers used in this book.

Note The Connection interface handles transactions and specifies how they are processed. By default, a new connection is in autocommit mode, and all its SQL statements are executed and committed as individual transactions. The commit occurs when the state- ment completes or the next execute occurs, whichever comes first. In the case of state- ments returning a result set, the statement completes when the last row of the result set has been retrieved or the result set has been closed. If a single statement returns multiple results, the commit occurs when all the results have been retrieved. You can use the setAutoCommit(false) method to disable autocommit, so all SQL statements are grouped into one transaction that is terminated by a call to either the commit() or the rollback() method. The rollback() method undoes all the changes made by the transaction.

34.4.2 Accessing a Database from JavaFX This section gives an example that demonstrates connecting to a database from a JavaFX pro- gram. The program lets the user enter the SSN and the course ID to find a student’s grade, as shown in Figure 34.23. The code in Listing 34.2 uses the MySQL database on the localhost.

the semicolon issue

autocommit

Figure 34.23 A JavaFX client can access the database on the server.

Figure 34.22 You must include the driver file to run Java database programs.

lisTing 34.2 FindGrade.java 1 import javafx.application.Application; 2 import javafx.scene.Scene; 3 import javafx.scene.control.Button; 4 import javafx.scene.control.Label; 5 import javafx.scene.control.TextField; 6 import javafx.scene.layout.HBox; 7 import javafx.scene.layout.VBox; 8 import javafx.stage.Stage; 9 import java.sql.*;

M34_LIAN1878_11_GE_C34.indd 22 12/29/17 5:48 PM

34.4 JDBC 34-23

10 11 public class FindGrade extends Application { 12 // Statement for executing queries 13 private Statement stmt; 14 private TextField tfSSN = new TextField(); 15 private TextField tfCourseId = new TextField(); 16 private Label lblStatus = new Label(); 17 18 @Override // Override the start method in the Application class 19 public void start(Stage primaryStage) { 20 // Initialize database connection and create a Statement object 21 initializeDB(); 22 23 Button btShowGrade = new Button("Show Grade"); 24 HBox hBox = new HBox(5); 25 hBox.getChildren().addAll(new Label("SSN"), tfSSN, 26 new Label("Course ID"), tfCourseId, (btShowGrade)); 27 28 VBox vBox = new VBox(10); 29 vBox.getChildren().addAll(hBox, lblStatus); 30 31 tfSSN.setPrefColumnCount(6); 32 tfCourseId.setPrefColumnCount(6); 33 btShowGrade.setOnAction(e -> showGrade()); 34 35 // Create a scene and place it in the stage 36 Scene scene = new Scene(vBox, 420, 80); 37 primaryStage.setTitle("FindGrade"); // Set the stage title 38 primaryStage.setScene(scene); // Place the scene in the stage 39 primaryStage.show(); // Display the stage 40 } 41 42 private void initializeDB() { 43 try { 44 // Load the JDBC driver 45 Class.forName("com.mysql.jdbc.Driver"); 46 // Class.forName("oracle.jdbc.driver.OracleDriver"); 47 System.out.println("Driver loaded"); 48 49 // Establish a connection 50 Connection connection = DriverManager.getConnection 51 ("jdbc:mysql://localhost/javabook", "scott", "tiger"); 52 // ("jdbc:oracle:thin:@liang.armstrong.edu:1521:orcl", 53 // "scott", "tiger"); 54 System.out.println("Database connected"); 55 56 // Create a statement 57 stmt = connection.createStatement(); 58 } 59 catch (Exception ex) { 60 ex.printStackTrace(); 61 } 62 } 63 64 private void showGrade() { 65 String ssn = tfSSN.getText(); 66 String courseId = tfCourseId.getText(); 67 try { 68 String queryString = "select firstName, mi, " + 69 "lastName, title, grade from Student, Enrollment, Course " +

button listener

load driver Oracle driver commented

connect to MySQL database

connect to Oracle commented

execute statement

show result

create statement

M34_LIAN1878_11_GE_C34.indd 23 12/29/17 5:48 PM

34-24 Chapter 34 Java Database Programming

70 "where Student.ssn = '" + ssn + "' and Enrollment.courseId " 71 + "= '" + courseId + 72 "' and Enrollment.courseId = Course.courseId " + 73 " and Enrollment.ssn = Student.ssn"; 74 75 ResultSet rset = stmt.executeQuery(queryString); 76 77 if (rset.next()) { 78 String lastName = rset.getString(1); 79 String mi = rset.getString(2); 80 String firstName = rset.getString(3); 81 String title = rset.getString(4); 82 String grade = rset.getString(5); 83 84 // Display result in a label 85 lblStatus.setText(firstName + " " + mi + 86 " " + lastName + "'s grade on course " + title + " is " + 87 grade); 88 } else { 89 lblStatus.setText("Not found"); 90 } 91 } 92 catch (SQLException ex) { 93 ex.printStackTrace(); 94 } 95 } 96 }

The initializeDB() method (lines 42–62) loads the MySQL driver (line 45), connects to the MySQL database on host liang.armstrong.edu (lines 50–55), and creates a state- ment (line 57).

Note There is a security hole in this program. If you enter 1' or true or '1 in the SSN field, you will get the first student’s score, because the query string now becomes

select firstName, mi, lastName, title, grade from Student, Enrollment, Course where Student.ssn = '1' or true or '1' and

Enrollment.courseId = ' ' and Enrollment.courseId = Course.courseId and Enrollment.ssn = Student.ssn;

You can avoid this problem by using the PreparedStatement interface, which will be discussed in the next section.

34.4.1 What are the advantages of developing database applications using Java? 34.4.2 Describe the following JDBC interfaces: Driver, Connection, Statement, and

ResultSet.

34.4.3 How do you load a JDBC driver? What are the driver classes for MySQL, Oracle, and Java DB?

34.4.4 How do you create a database connection? What are the URLs for MySQL, Oracle, and Java DB?

34.4.5 How do you create a Statement and execute an SQL statement? 34.4.6 How do you retrieve values in a ResultSet? 34.4.7 Does JDBC automatically commit a transaction? How do you set autocommit to false?

security hole

Point Check

M34_LIAN1878_11_GE_C34.indd 24 12/29/17 5:48 PM

34.5 PreparedStatement 34-25

34.5 PreparedStatement PreparedStatement enables you to create parameterized SQL statements.

Once a connection to a particular database is established, it can be used to send SQL statements from your program to the database. The Statement interface is used to execute static SQL statements that don’t contain any parameters. The PreparedStatement interface, extend- ing Statement, is used to execute a precompiled SQL statement with or without parameters. Since the SQL statements are precompiled, they are efficient for repeated executions.

A PreparedStatement object is created using the prepareStatement method in the Connection interface. For example, the following code creates a PreparedStatement for an SQL insert statement:

PreparedStatement preparedStatement = connection.prepareStatement ("insert into Student (firstName, mi, lastName) " + "values (?, ?, ?)");

This insert statement has three question marks as placeholders for parameters representing values for firstName, mi, and lastName in a record of the Student table.

As a subinterface of Statement, the PreparedStatement interface inherits all the meth- ods defined in Statement. It also provides the methods for setting parameters in the object of PreparedStatement. These methods are used to set the values for the parameters before executing statements or procedures. In general, the setter methods have the following name and signature:

setX(int parameterIndex, X value);

where X is the type of the parameter, and parameterIndex is the index of the parameter in the statement. The index starts from 1. For example, the method setString(int parameterIndex, String value) sets a String value to the specified parameter.

The following statements pass the parameters "Jack", "A", and "Ryan" to the placeholders for firstName, mi, and lastName in preparedStatement:

preparedStatement.setString(1, "Jack"); preparedStatement.setString(2, "A"); preparedStatement.setString(3, "Ryan");

After setting the parameters, you can execute the prepared statement by invoking execute- Query() for a SELECT statement and executeUpdate() for a DDL or update statement.

The executeQuery() and executeUpdate() methods are similar to the ones defined in the Statement interface except that they don’t have any parameters, because the SQL statements are already specified in the prepareStatement method when the object of PreparedStatement is created.

Using a prepared SQL statement, Listing 34.2 can be improved as in Listing 34.3.

lisTing 34.3 FindGradeUsingPreparedStatement.java 1 import javafx.application.Application; 2 import javafx.scene.Scene; 3 import javafx.scene.control.Button; 4 import javafx.scene.control.Label; 5 import javafx.scene.control.TextField; 6 import javafx.scene.layout.HBox; 7 import javafx.scene.layout.VBox; 8 import javafx.stage.Stage; 9 import java.sql.*; 10

Point Key

M34_LIAN1878_11_GE_C34.indd 25 12/29/17 5:48 PM

34-26 Chapter 34 Java Database Programming

11 public class FindGradeUsingPreparedStatement extends Application { 12 // PreparedStatement for executing queries 13 private PreparedStatement preparedStatement; 14 private TextField tfSSN = new TextField(); 15 private TextField tfCourseId = new TextField(); 16 private Label lblStatus = new Label(); 17 18 @Override // Override the start method in the Application class 19 public void start(Stage primaryStage) { 20 // Initialize database connection and create a Statement object 21 initializeDB(); 22 23 Button btShowGrade = new Button("Show Grade"); 24 HBox hBox = new HBox(5); 25 hBox.getChildren().addAll(new Label("SSN"), tfSSN, 26 new Label("Course ID"), tfCourseId, (btShowGrade)); 27 28 VBox vBox = new VBox(10); 29 vBox.getChildren().addAll(hBox, lblStatus); 30 31 tfSSN.setPrefColumnCount(6); 32 tfCourseId.setPrefColumnCount(6); 33 btShowGrade.setOnAction(e -> showGrade()); 34 35 // Create a scene and place it in the stage 36 Scene scene = new Scene(vBox, 420, 80); 37 primaryStage.setTitle("FindGrade"); // Set the stage title 38 primaryStage.setScene(scene); // Place the scene in the stage 39 primaryStage.show(); // Display the stage 40 } 41 42 private void initializeDB() { 43 try { 44 // Load the JDBC driver 45 Class.forName("com.mysql.jdbc.Driver"); 46 // Class.forName("oracle.jdbc.driver.OracleDriver"); 47 System.out.println("Driver loaded"); 48 49 // Establish a connection 50 Connection connection = DriverManager.getConnection 51 ("jdbc:mysql://localhost/javabook", "scott", "tiger"); 52 // ("jdbc:oracle:thin:@liang.armstrong.edu:1521:orcl", 53 // "scott", "tiger"); 54 System.out.println("Database connected"); 55 56 String queryString = "select firstName, mi, " + 57 "lastName, title, grade from Student, Enrollment, Course " + 58 "where Student.ssn = ? and Enrollment.courseId = ? " + 59 "and Enrollment.courseId = Course.courseId"; 60 61 // Create a statement 62 preparedStatement = connection.prepareStatement(queryString); 63 } 64 catch (Exception ex) { 65 ex.printStackTrace(); 66 } 67 } 68 69 private void showGrade() { 70 String ssn = tfSSN.getText();

prepare statement

placeholder

connect database

load driver

M34_LIAN1878_11_GE_C34.indd 26 12/29/17 5:48 PM

34.6 CallableStatement 34-27

71 String courseId = tfCourseId.getText(); 72 try { 73 preparedStatement.setString(1, ssn); 74 preparedStatement.setString(2, courseId); 75 ResultSet rset = preparedStatement.executeQuery(); 76 77 if (rset.next()) { 78 String lastName = rset.getString(1); 79 String mi = rset.getString(2); 80 String firstName = rset.getString(3); 81 String title = rset.getString(4); 82 String grade = rset.getString(5); 83 84 // Display result in a label 85 lblStatus.setText(firstName + " " + mi + 86 " " + lastName + "'s grade on course " + title + " is " + 87 grade); 88 } else { 89 lblStatus.setText("Not found"); 90 } 91 } 92 catch (SQLException ex) { 93 ex.printStackTrace(); 94 } 95 } 96 }

This example does exactly the same thing as Listing 34.2 except that it uses the prepared statement to dynamically set the parameters. The code in this example is almost the same as in the preceding example. The new code is highlighted.

A prepared query string is defined in lines 56–59 with ssn and courseId as parameters. An SQL prepared statement is obtained in line 62. Before executing the query, the actual values of ssn and courseId are set to the parameters in lines 73–74. Line 75 executes the prepared statement.

34.5.1 Describe prepared statements. How do you create instances of Prepared- Statement? How do you execute a PreparedStatement? How do you set parameter values in a PreparedStatement?

34.5.2 What are the benefits of using prepared statements?

34.6 CallableStatement CallableStatement enables you to execute SQL stored procedures.

The CallableStatement interface is designed to execute SQL-stored procedures. The pro- cedures may have IN, OUT, or IN OUT parameters. An IN parameter receives a value passed to the procedure when it is called. An OUT parameter returns a value after the procedure is completed, but it doesn’t contain any value when the procedure is called. An IN OUT parameter contains a value passed to the procedure when it is called and returns a value after it is com- pleted. For example, the following procedure in Oracle PL/SQL has IN parameter p1, OUT parameter p2, and IN OUT parameter p3:

create or replace procedure sampleProcedure (p1 in varchar, p2 out number, p3 in out integer) is begin /* do something */ end sampleProcedure; /

Point Check

Point Key

IN parameter

OUT parameter

IN OUT parameter

show result

execute statement

M34_LIAN1878_11_GE_C34.indd 27 12/29/17 5:48 PM

34-28 Chapter 34 Java Database Programming

Note The syntax of stored procedures is vendor specific. We use both Oracle and MySQL for demonstrations of stored procedures in this book.

A CallableStatement object can be created using the prepareCall(String call) method in the Connection interface. For example, the following code creates a Callable Statement cstmt on Connection connection for the procedure sampleProcedure:

CallableStatement callableStatement = connection.prepareCall( "{call sampleProcedure(?, ?, ?)}");

{call sampleProcedure(?, ?, ...)} is referred to as the SQL escape syntax, which signals the driver that the code within it should be handled differently. The driver parses the escape syntax and translates it into code that the database understands. In this exam- ple, sampleProcedure is an Oracle procedure. The call is translated to the string begin sampleProcedure(?, ?, ?); end and passed to an Oracle database for execution.

You can call procedures as well as functions. The syntax to create an SQL callable state- ment for a function is:

{? = call functionName(?, ?, ...)}

CallableStatement inherits PreparedStatement. Additionally, the Callable Statement interface provides methods for registering the OUT parameters and for getting values from the OUT parameters.

Before calling an SQL procedure, you need to use appropriate setter methods to pass values to IN and IN OUT parameters, and use registerOutParameter to register OUT and IN OUT parameters. For example, before calling procedure sampleProcedure, the following state- ments pass values to parameters p1 (IN) and p3 (IN OUT) and register parameters p2 (OUT) and p3 (IN OUT):

callableStatement.setString(1, "Dallas"); // Set Dallas to p1 callableStatement.setLong(3, 1); // Set 1 to p3 // Register OUT parameters callableStatement.registerOutParameter(2, java.sql.Types.DOUBLE); callableStatement.registerOutParameter(3, java.sql.Types.INTEGER);

You can use execute() or executeUpdate() to execute the procedure depending on the type of SQL statement, then use getter methods to retrieve values from the OUT parameters. For example, the next statements retrieve the values from parameters p2 and p3:

double d = callableStatement.getDouble(2); int i = callableStatement.getInt(3);

Let us define a MySQL function that returns the number of the records in the table that match the specified firstName and lastName in the Student table.

/* For the callable statement example. Use MySQL version 5 */ drop function if exists studentFound;

delimiter //

create function studentFound(first varchar(20), last varchar(20)) returns int begin declare result int;

select count(*) into result

M34_LIAN1878_11_GE_C34.indd 28 12/29/17 5:48 PM

34.6 CallableStatement 34-29

from Student where Student.firstName = first and Student.lastName = last;

return result; end; //

delimiter ; /* Please note that there is a space between delimiter and ; */

If you use an Oracle database, the function can be defined as follows:

create or replace function studentFound (first varchar2, last varchar2) /* Do not name firstName and lastName. */ return number is numberOfSelectedRows number := 0; begin select count(*) into numberOfSelectedRows from Student where Student.firstName = first and Student.lastName = last;

return numberOfSelectedRows; end studentFound; /

Suppose the function studentFound is already created in the database. Listing 34.4 gives an example that tests this function using callable statements.

lisTing 34.4 TestCallableStatement.java 1 import java.sql.*; 2 3 public class TestCallableStatement { 4 /** Creates new form TestTableEditor */ 5 public static void main(String[] args) throws Exception { 6 Class.forName("com.mysql.jdbc.Driver"); 7 Connection connection = DriverManager.getConnection( 8 "jdbc:mysql://localhost/javabook", 9 "scott", "tiger"); 10 // Connection connection = DriverManager.getConnection( 11 // ("jdbc:oracle:thin:@liang.armstrong.edu:1521:orcl", 12 // "scott", "tiger"); 13 14 // Create a callable statement 15 CallableStatement callableStatement = connection.prepareCall( 16 "{? = call studentFound(?, ?)}"); 17 18 java.util.Scanner input = new java.util.Scanner(System.in); 19 System.out.print("Enter student's first name: "); 20 String firstName = input.nextLine(); 21 System.out.print("Enter student's last name: "); 22 String lastName = input.nextLine(); 23 24 callableStatement.setString(2, firstName); 25 callableStatement.setString(3, lastName); 26 callableStatement.registerOutParameter(1, Types.INTEGER); 27 callableStatement.execute(); execute statement

enter lastName

enter firstName

create callable statement

connect database load driver

M34_LIAN1878_11_GE_C34.indd 29 12/29/17 5:48 PM

34-30 Chapter 34 Java Database Programming

28 29 if (callableStatement.getInt(1) >= 1) 30 System.out.println(firstName + " " + lastName + 31 " is in the database"); 32 else 33 System.out.println(firstName + " " + lastName + 34 " is not in the database"); 35 } 36 }

Enter student's first name: Jacob

Enter student's last name: Smith

Jacob Smith is in the database

Enter student's first name: John

Enter student's last name: Smith

John Smith is not in the database

get OUT parameter

The program loads a MySQL driver (line 6), connects to a MySQL database (lines 7–9), and creates a callable statement for executing the function studentFound (lines 15–16).

The function’s first parameter is the return value; its second and third parameters correspond to the first and last names. Before executing the callable statement, the program sets the first name and last name (lines 24–25) and registers the OUT parameter (line 26). The statement is executed in line 27.

The function’s return value is obtained in line 29. If the value is greater than or equal to 1, the student with the specified first and last name is found in the table.

34.6.1 Describe callable statements. How do you create instances of CallableStatement? How do you execute a CallableStatement? How do you register OUT param- eters in a CallableStatement?

34.7 Retrieving Metadata The database metadata such as database URL, username, and JDBC driver name can be obtained using the DatabaseMetaData interface and result set metadata such as table column count and column names can be obtained using the ResultSetMetaData interface.

JDBC provides the DatabaseMetaData interface for obtaining database-wide information, and the ResultSetMetaData interface for obtaining information on a specific ResultSet.

34.7.1 Database Metadata The Connection interface establishes a connection to a database. It is within the context of a connection that SQL statements are executed and results are returned. A connection also pro- vides access to database metadata information that describes the capabilities of the database, supported SQL grammar, stored procedures, and so on. To obtain an instance of Database- MetaData for a database, use the getMetaData method on a Connection object like this:

DatabaseMetaData dbMetaData = connection.getMetaData();

If your program connects to a local MySQL database, the program in Listing 34.5 displays the database information statements shown in Figure 34.24.

Point Check

Point Key

database metadata

M34_LIAN1878_11_GE_C34.indd 30 12/29/17 5:48 PM

34.7 Retrieving Metadata 34-31

lisTing 34.5 TestDatabaseMetaData.java 1 import java.sql.*; 2 3 public class TestDatabaseMetaData { 4 public static void main(String[] args) 5 throws SQLException, ClassNotFoundException { 6 // Load the JDBC driver 7 Class.forName("com.mysql.jdbc.Driver"); 8 System.out.println("Driver loaded"); 9 10 // Connect to a database 11 Connection connection = DriverManager.getConnection 12 ("jdbc:mysql://localhost/javabook", "scott", "tiger"); 13 System.out.println("Database connected"); 14 15 DatabaseMetaData dbMetaData = connection.getMetaData(); 16 System.out.println("database URL: " + dbMetaData.getURL()); 17 System.out.println("database username: " + 18 dbMetaData.getUserName()); 19 System.out.println("database product name: " + 20 dbMetaData.getDatabaseProductName()); 21 System.out.println("database product version: " + 22 dbMetaData.getDatabaseProductVersion()); 23 System.out.println("JDBC driver name: " + 24 dbMetaData.getDriverName()); 25 System.out.println("JDBC driver version: " + 26 dbMetaData.getDriverVersion()); 27 System.out.println("JDBC driver major version: " + 28 dbMetaData.getDriverMajorVersion()); 29 System.out.println("JDBC driver minor version: " + 30 dbMetaData.getDriverMinorVersion()); 31 System.out.println("Max number of connections: " + 32 dbMetaData.getMaxConnections()); 33 System.out.println("MaxTableNameLength: " + 34 dbMetaData.getMaxTableNameLength()); 35 System.out.println("MaxColumnsInTable: " + 36 dbMetaData.getMaxColumnsInTable()); 37 38 // Close the connection 39 connection.close(); 40 } 41 }

load driver

connect database

database metadata get metadata

Figure 34.24 The DatabaseMetaData interface enables you to obtain database information.

M34_LIAN1878_11_GE_C34.indd 31 12/29/17 5:48 PM

34-32 Chapter 34 Java Database Programming

34.7.2 Obtaining Database Tables You can identify the tables in the database through database metadata using the getTables method. Listing 34.6 displays all the user tables in the javabook database on a local MySQL database. Figure 34.25 shows a sample output of the program.

lisTing 34.6 FindUserTables.java 1 import java.sql.*; 2 3 public class FindUserTables { 4 public static void main(String[] args) 5 throws SQLException, ClassNotFoundException { 6 // Load the JDBC driver 7 Class.forName("com.mysql.jdbc.Driver"); 8 System.out.println("Driver loaded"); 9 10 // Connect to a database 11 Connection connection = DriverManager.getConnection 12 ("jdbc:mysql://localhost/javabook", "scott", "tiger"); 13 System.out.println("Database connected"); 14 15 DatabaseMetaData dbMetaData = connection.getMetaData(); 16 17 ResultSet rsTables = dbMetaData.getTables(null, null, null, 18 new String[] {"TABLE"}); 19 System.out.print("User tables: "); 20 while (rsTables.next()) 21 System.out.print(rsTables.getString("TABLE_NAME") + " "); 22 23 // Close the connection 24 connection.close(); 25 } 26 }

load driver

connect database

database metadata

obtain tables

get table names

Figure 34.25 You can find all the tables in the database.

Line 17 obtains table information in a result set using the getTables method. One of the columns in the result set is TABLE_NAME. Line 21 retrieves the table name from this result set column.

34.7.3 Result Set Metadata The ResultSetMetaData interface describes information pertaining to the result set. A ResultSetMetaData object can be used to find the types and properties of the columns in a ResultSet. To obtain an instance of ResultSetMetaData, use the getMetaData method on a result set like this:

ResultSetMetaData rsMetaData = resultSet.getMetaData();

M34_LIAN1878_11_GE_C34.indd 32 12/29/17 5:48 PM

34.7 Retrieving Metadata 34-33

You can use the getColumnCount() method to find the number of columns in the result and the getColumnName(int) method to get the column names. For example, Listing 34.7 dis- plays all the column names and contents resulting from the SQL SELECT statement select * from Enrollment. The output is shown in Figure 34.26.

lisTing 34.7 TestResultSetMetaData.java 1 import java.sql.*; 2 3 public class TestResultSetMetaData { 4 public static void main(String[] args) 5 throws SQLException, ClassNotFoundException { 6 // Load the JDBC driver 7 Class.forName("com.mysql.jdbc.Driver"); 8 System.out.println("Driver loaded"); 9 10 // Connect to a database 11 Connection connection = DriverManager.getConnection 12 ("jdbc:mysql://localhost/javabook", "scott", "tiger"); 13 System.out.println("Database connected"); 14 15 // Create a statement 16 Statement statement = connection.createStatement(); 17 18 // Execute a statement 19 ResultSet resultSet = statement.executeQuery 20 ("select * from Enrollment"); 21 22 ResultSetMetaData rsMetaData = resultSet.getMetaData(); 23 for (int i = 1; i <= rsMetaData.getColumnCount(); i++) 24 System.out.printf("%-12s\t", rsMetaData.getColumnName(i)); 25 System.out.println(); 26 27 // Iterate through the result and print the students' names 28 while (resultSet.next()) { 29 for (int i = 1; i <= rsMetaData.getColumnCount(); i++) 30 System.out.printf("%-12s\t", resultSet.getObject(i)); 31 System.out.println(); 32 } 33 34 // Close the connection 35 connection.close(); 36 } 37 }

load driver

connect database

create statement

create result set

result set metadata column count column name

Figure 34.26 The ResultSetMetaData interface enables you to obtain result set information.

M34_LIAN1878_11_GE_C34.indd 33 12/29/17 5:48 PM

34-34 Chapter 34 Java Database Programming

34.7.1 What is DatabaseMetaData for? Describe the methods in DatabaseMetaData. How do you get an instance of DatabaseMetaData?

34.7.2 What is ResultSetMetaData for? Describe the methods in ResultSet- MetaData. How do you get an instance of ResultSetMetaData?

34.7.3 How do you find the number of columns in a result set? How do you find the column names in a result set?

Point Check

candidate key 34-5 database system 34-2 domain constraint 34-5 foreign key 34-5 foreign key constraint 34-5

integrity constraint 34-4 primary key 34-5 relational database 34-5 Structured Query Language (SQL) 34-6 superkey 34-5

Key Terms

ChapTer summary 1. This chapter introduced the concepts of database systems, relational databases, rela-

tional data models, data integrity, and SQL. You learned how to develop database appli- cations using Java.

2. The Java API for developing Java database applications is called JDBC. JDBC provides Java programmers with a uniform interface for accessing and manipulating relational databases.

3. The JDBC API consists of classes and interfaces for establishing connections with data- bases, sending SQL statements to databases, processing the results of SQL statements, and obtaining database metadata.

4. Since a JDBC driver serves as the interface to facilitate communications between JDBC and a proprietary database, JDBC drivers are database specific. If you use a driver, make sure it is in the classpath before running the program.

5. Four key interfaces are needed to develop any database application using Java: Driver, Connection, Statement, and ResultSet. These interfaces define a framework for generic SQL database access. The JDBC driver vendors provide implementation for them.

6. A JDBC application loads an appropriate driver using the Driver interface, connects to the database using the Connection interface, creates and executes SQL statements using the Statement interface, and processes the result using the ResultSet interface if the statements return results.

7. The PreparedStatement interface is designed to execute dynamic SQL statements with parameters. These SQL statements are precompiled for efficient use when repeat- edly executed.

8. Database metadata is information that describes the database itself. JDBC provides the DatabaseMetaData interface for obtaining database-wide information and the ResultSetMetaData interface for obtaining information on the specific ResultSet.

M34_LIAN1878_11_GE_C34.indd 34 12/29/17 5:48 PM

Programming Exercises 34-35

Quiz Answer the quiz for this chapter online at the book Companion Website.

programming exerCises *34.1 (Access and update a Staff table) Write a program that views, inserts, and updates

staff information stored in a database, as shown in Figure 34.27a. The View button displays a record with a specified ID. The Insert button inserts a new record. The Update button updates the record for the specified ID. The Staff table is created as follows:

create table Staff ( id char(9) not null, lastName varchar(15), firstName varchar(15), mi char(1), address varchar(20), city varchar(20), state char(2), telephone char(10), email varchar(40), primary key (id) );

Figure 34.27 (a) The program lets you view, insert, and update staff information. (b) The PieChart and BarChart components display the query data obtained from the data module.

(a) (b)

**34.2 (Visualize data) Write a program that displays the number of students in each department in a pie chart and a bar chart, as shown in Figure 34.27b. The PieChart and BarChart classes are created in Programming Exercises 14.12 and 14.13. The number of students for each department can be obtained from the Student table (see Figure 34.4) using the following SQL statement:

select deptId, count(*) from Student where deptId is not null group by deptId;

*34.3 (Connection dialog) Develop a subclass of BorderPane named DBConnection- Pane that enables the user to select or enter a JDBC driver and a URL and to enter a username and password, as shown in Figure 34.28. When the Connect to DB but- ton is clicked, a Connection object for the database is stored in the connection property. You can then use the getConnection() method to return the connection.

M34_LIAN1878_11_GE_C34.indd 35 12/29/17 5:48 PM

34-36 Chapter 34 Java Database Programming

*34.4 (Find grades) Listing 34.2, FindGrade.java, presented a program that finds a stu- dent’s grade for a specified course. Rewrite the program to find all the grades for a specified student, as shown in Figure 34.29.

Figure 34.29 The program displays the grades for the courses for a specified student.

Figure 34.30 (a) Enter a table name to display the table contents. (b) Select a table name from the combo box to display its contents.

(a) (b)

Figure 34.28 The DBConnectionPane component enables the user to enter database information.

*34.5 (Display table contents) Write a program that displays the content for a given table. As shown in Figure 34.30a, you enter a table and click the Show Contents button to display the table contents in the text area.

*34.6 (Find tables and showing their contents) Write a program that fills in table names in a combo box, as shown in Figure 34.30b. You can select a table from the combo box to display its contents in the text area.

**34.7 (Populate Quiz table) Create a table named Quiz as follows:

create table Quiz( questionId int, question varchar(4000), choicea varchar(1000),

M34_LIAN1878_11_GE_C34.indd 36 12/29/17 5:48 PM

Programming Exercises 34-37

choiceb varchar(1000), choicec varchar(1000), choiced varchar(1000), answer varchar(5));

The Quiz table stores multiple-choice questions. Suppose the multiple-choice questions are stored in a text file accessible from http://www.cs.armstrong.edu/ liang/data/Quiz.txt in the following format:

1. question1 a. choice a b. choice b c. choice c d. choice d Answer:cd

2. question2 a. choice a b. choice b c. choice c d. choice d Answer:a

...

Write a program that reads the data from the file and populate it into the Quiz table.

*34.8 (Populate Salary table) Create a table named Salary as follows:

create table Salary( firstName varchar(100), lastName varchar(100), rank varchar(15), salary float);

Obtain the data for salary from http://cs.armstrong.edu/liang/data/Salary.txt and populate it into the Salary table in the database.

*34.9 (Copy table) Suppose the database contains a student table defined as follows:

create table Student1 ( username varchar(50) not null, password varchar(50) not null, fullname varchar(200) not null, constraint pkStudent primary key (username) );

Create a new table named Student2 as follows:

create table Student2 ( username varchar(50) not null, password varchar(50) not null, firstname varchar(100), lastname varchar(100), constraint pkStudent primary key (username) );

A full name is in the form of firstname mi lastname or firstname last- name. For example, John K Smith is a full name. Write a program that copies

M34_LIAN1878_11_GE_C34.indd 37 12/29/17 5:48 PM

34-38 Chapter 34 Java Database Programming

table Student1 into Student2. Your task is to split a full name into first- name, mi, and lastname for each record in Student1 and store a new record into Student2.

*34.10 (Record unsubmitted exercises) The following three tables store information on students, assigned exercises, and exercise submission in LiveLab. LiveLab is an automatic grading system for grading programming exercises.

create table AGSStudent ( username varchar(50) not null, password varchar(50) not null, fullname varchar(200) not null, instructorEmail varchar(100) not null, constraint pkAGSStudent primary key (username) );

create table ExerciseAssigned ( instructorEmail varchar(100), exerciseName varchar(100), maxscore double default 10, constraint pkCustomExercise primary key (instructorEmail, exerciseName) );

create table AGSLog ( username varchar(50), /* This is the student's user name */ exerciseName varchar(100), /* This is the exercise */ score double default null, submitted bit default 0, constraint pkLog primary key (username, exerciseName) );

The AGSStudent table stores the student information. The ExerciseAssigned table assigns the exercises by an instructor. The AGSLog table stores the grading results. When a student submits an exercise, a record is stored in the AGSLog table. However, there is no record in AGSLog if a student did not submit the exercise.

Write a program that adds a new record for each student and an assigned exercise to the student in the AGSLog table if a student has not submitted the exercise. The record should have 0 on score and submitted. For example, if the tables contain the following data in AGSLog before you run this program, the AGSLog table now contains the new records after the program runs.

AGSStudent

username password fullname instructorEmail

abc p1 John Roo [email protected] cde p2 Yao Mi [email protected] wbc p3 F3 [email protected]

ExerciseAssigned

instructorEmail exerciseName maxScore

[email protected] e1 10 [email protected] e2 10 [email protected] e1 4 [email protected] e4 20

M34_LIAN1878_11_GE_C34.indd 38 12/29/17 5:48 PM

Programming Exercises 34-39

AGSLog

username exerciseName score submitted

abc e1 9 1 wbc e2 7 1

AGSLog after the program runs

username exerciseName score submitted

abc e1 9 1 wbc e2 7 1 abc e2 0 wbc e1 0 cde e1 0 cde e4 0

*34.11 (Baby names) Create the following table:

create table Babyname ( year integer, name varchar(50), gender char(1), count integer, constraint pkBabyname primary key (year, name, gender) );

The baby name ranking data was described in Programming Exercise 12.31. Write a program to read data from the following URL and store into the Babyname table. https://liveexample.pearsoncmg.com/data/babynamesranking2001.txt,

. . .

https://liveexample.pearsoncmg.com/data/babynamesranking2010.txt.

M34_LIAN1878_11_GE_C34.indd 39 12/29/17 5:48 PM

Objectives ■■ To create a universal SQL client for accessing local or remote database

(§35.2).

■■ To execute SQL statements in a batch mode (§35.3).

■■ To process updatable and scrollable result sets (§35.4).

■■ To simplify Java database programming using RowSet (§35.5).

■■ To store and retrieve images in JDBC (§35.6).

Advanced Java Database Programming

Chapter

M35_LIAN1878_11_GE_C35.indd 1 12/29/17 5:50 PM

35-2 Chapter 35 Advanced Java Database Programming

35.1 Introduction This chapter introduces advanced features for Java database programming.

Chapter 34 introduced JDBC’s basic features. This chapter covers its advanced features. You will learn how to develop a universal SQL client for accessing any local or remote relational database, learn how to execute statements in a batch mode to improve performance, learn scrol- lable result sets and how to update a database through result sets, learn how to use RowSet to simplify database access, and learn how to store and retrieve images.

35.2 A Universal SQL Client This section develops a universal SQL client for connecting and accessing any SQL database.

In Chapter 34, you used various drivers to connect to the database, created statements for executing SQL statements, and processed the results from SQL queries. This section presents a universal SQL client that enables you to connect to any relational database and execute SQL commands interactively, as shown in Figure 35.1. The client can connect to any JDBC data source and can submit SQL SELECT commands and non-SELECT commands for execution. The execution result is displayed for the SELECT queries, and the execution status is displayed for the non-SELECT commands. Listing 35.1 gives the program.

Point Key

Figure 35.1 You can connect to any JDBC data source and execute SQL commands interactively.

lisTing 35.1 SQLClient.java 1 import java.sql.*; 2 import javafx.application.Application; 3 import javafx.collections.FXCollections; 4 import javafx.geometry.Pos; 5 import javafx.scene.Scene; 6 import javafx.scene.control.Button; 7 import javafx.scene.control.ComboBox; 8 import javafx.scene.control.Label; 9 import javafx.scene.control.PasswordField; 10 import javafx.scene.control.ScrollPane; 11 import javafx.scene.control.TextArea; 12 import javafx.scene.control.TextField; 13 import javafx.scene.layout.BorderPane; 14 import javafx.scene.layout.GridPane; 15 import javafx.scene.layout.HBox; 16 import javafx.scene.layout.VBox;

M35_LIAN1878_11_GE_C35.indd 2 12/29/17 5:50 PM

35.2 A Universal SQL Client 35-3

17 import javafx.stage.Stage; 18 19 public class SQLClient extends Application { 20 // Connection to the database 21 private Connection connection; 22 23 // Statement to execute SQL commands 24 private Statement statement; 25 26 // Text area to enter SQL commands 27 private TextArea tasqlCommand = new TextArea(); 28 29 // Text area to display results from SQL commands 30 private TextArea taSQLResult = new TextArea(); 31 32 // DBC info for a database connection 33 private TextField tfUsername = new TextField(); 34 private PasswordField pfPassword = new PasswordField(); 35 private ComboBox<String> cboURL = new ComboBox<>(); 36 private ComboBox<String> cboDriver = new ComboBox<>(); 37 38 private Button btExecuteSQL = new Button("Execute SQL Command"); 39 private Button btClearSQLCommand = new Button("Clear"); 40 private Button btConnectDB = new Button("Connect to Database"); 41 private Button btClearSQLResult = new Button("Clear Result"); 42 private Label lblConnectionStatus 43 = new Label("No connection now"); 44 45 @Override // Override the start method in the Application class 46 public void start(Stage primaryStage) { 47 cboURL.getItems().addAll(FXCollections.observableArrayList( 48 "jdbc:mysql://localhost/javabook", 49 "jdbc:mysql://liang.armstrong.edu/javabook", 50 "jdbc:odbc:exampleMDBDataSource", 51 "jdbc:oracle:thin:@liang.armstrong.edu:1521:orcl")); 52 cboURL.getSelectionModel().selectFirst(); 53 54 cboDriver.getItems().addAll(FXCollections.observableArrayList( 55 "com.mysql.jdbc.Driver", "sun.jdbc.odbc.dbcOdbcDriver", 56 "oracle.jdbc.driver.OracleDriver")); 57 cboDriver.getSelectionModel().selectFirst(); 58 59 // Create UI for connecting to the database 60 GridPane gridPane = new GridPane(); 61 gridPane.add(cboURL, 1, 0); 62 gridPane.add(cboDriver, 1, 1); 63 gridPane.add(tfUsername, 1, 2); 64 gridPane.add(pfPassword, 1, 3); 65 gridPane.add(new Label("JDBC Driver"), 0, 0); 66 gridPane.add(new Label("Database URL"), 0, 1); 67 gridPane.add(new Label("Username"), 0, 2); 68 gridPane.add(new Label("Password"), 0, 3); 69 70 HBox hBoxConnection = new HBox(); 71 hBoxConnection.getChildren().addAll( 72 lblConnectionStatus, btConnectDB); 73 hBoxConnection.setAlignment(Pos.CENTER_RIGHT); 74 75 VBox vBoxConnection = new VBox(5); 76 vBoxConnection.getChildren().addAll( 77 new Label("Enter Database Information"),

M35_LIAN1878_11_GE_C35.indd 3 12/29/17 5:50 PM

35-4 Chapter 35 Advanced Java Database Programming

78 gridPane, hBoxConnection); 79 80 gridPane.setStyle("-fx-border-color: black;"); 81 82 HBox hBoxSQLCommand = new HBox(5); 83 hBoxSQLCommand.getChildren().addAll( 84 btClearSQLCommand, btExecuteSQL); 85 hBoxSQLCommand.setAlignment(Pos.CENTER_RIGHT); 86 87 BorderPane borderPaneSqlCommand = new BorderPane(); 88 borderPaneSqlCommand.setTop( 89 new Label("Enter an SQL Command")); 90 borderPaneSqlCommand.setCenter( 91 new ScrollPane(tasqlCommand)); 92 borderPaneSqlCommand.setBottom( 93 hBoxSQLCommand); 94 95 HBox hBoxConnectionCommand = new HBox(10); 96 hBoxConnectionCommand.getChildren().addAll( 97 vBoxConnection, borderPaneSqlCommand); 98 99 BorderPane borderPaneExecutionResult = new BorderPane(); 100 borderPaneExecutionResult.setTop( 101 new Label("SQL Execution Result")); 102 borderPaneExecutionResult.setCenter(taSQLResult); 103 borderPaneExecutionResult.setBottom(btClearSQLResult); 104 105 BorderPane borderPane = new BorderPane(); 106 borderPane.setTop(hBoxConnectionCommand); 107 borderPane.setCenter(borderPaneExecutionResult); 108 109 // Create a scene and place it in the stage 110 Scene scene = new Scene(borderPane, 670, 400); 111 primaryStage.setTitle("SQLClient"); // Set the stage title 112 primaryStage.setScene(scene); // Place the scene in the stage 113 primaryStage.show(); // Display the stage 114 115 btConnectDB.setOnAction(e -> connectToDB()); 116 btExecuteSQL.setOnAction(e -> executeSQL()); 117 btClearSQLCommand.setOnAction(e -> tasqlCommand.setText(null)); 118 btClearSQLResult.setOnAction(e -> taSQLResult.setText(null)); 119 } 120 121 /** Connect to DB */ 122 private void connectToDB() { 123 // Get database information from the user input 124 String driver = cboDriver 125 .getSelectionModel().getSelectedItem(); 126 String url = cboURL.getSelectionModel().getSelectedItem(); 127 String username = tfUsername.getText().trim(); 128 String password = pfPassword.getText().trim(); 129 130 // Connection to the database 131 try { 132 Class.forName(driver); 133 connection = DriverManager.getConnection( 134 url, username, password); 135 lblConnectionStatus.setText("Connected to " + url); 136 } 137 catch (java.lang.Exception ex) { 138 ex.printStackTrace();

M35_LIAN1878_11_GE_C35.indd 4 12/29/17 5:50 PM

35.2 A Universal SQL Client 35-5

139 } 140 } 141 142 /** Execute SQL commands */ 143 private void executeSQL() { 144 if (connection == null) { 145 taSQLResult.setText("Please connect to a database first"); 146 return; 147 } 148 else { 149 String sqlCommands = tasqlCommand.getText().trim(); 150 String[] commands = sqlCommands.replace('\n', ' ').split(";"); 151 152 for (String aCommand: commands) { 153 if (aCommand.trim().toUpperCase().startsWith("SELECT")) { 154 processSQLSelect(aCommand); 155 } 156 else { 157 processSQLNonSelect(aCommand); 158 } 159 } 160 } 161 } 162 163 /** Execute SQL SELECT commands */ 164 private void processSQLSelect(String sqlCommand) { 165 try { 166 // Get a new statement for the current connection 167 statement = connection.createStatement(); 168 169 // Execute a SELECT SQL command 170 ResultSet resultSet = statement.executeQuery(sqlCommand); 171 172 // Find the number of columns in the result set 173 int columnCount = resultSet.getMetaData().getColumnCount(); 174 String row = ""; 175 176 // Display column names 177 for (int i = 1; i <= columnCount; i++) { 178 row += resultSet.getMetaData().getColumnName(i) + "\t"; 179 } 180 181 taSQLResult.appendText(row + '\n'); 182 183 while (resultSet.next()) { 184 // Reset row to empty 185 row = ""; 186 187 for (int i = 1; i <= columnCount; i++) { 188 // A non-String column is converted to a string 189 row += resultSet.getString(i) + "\t"; 190 } 191 192 taSQLResult.appendText(row + '\n'); 193 } 194 } 195 catch (SQLException ex) { 196 taSQLResult.setText(ex.toString()); 197 } 198 } 199

M35_LIAN1878_11_GE_C35.indd 5 12/29/17 5:50 PM

35-6 Chapter 35 Advanced Java Database Programming

200 /** Execute SQL DDL, and modification commands */ 201 private void processSQLNonSelect(String sqlCommand) { 202 try { 203 // Get a new statement for the current connection 204 statement = connection.createStatement(); 205 206 // Execute a non-SELECT SQL command 207 statement.executeUpdate(sqlCommand); 208 209 taSQLResult.setText("SQL command executed"); 210 } 211 catch (SQLException ex) { 212 taSQLResult.setText(ex.toString()); 213 } 214 } 215 }

The user selects or enters the JDBC driver, database URL, username, and password, and clicks the Connect to Database button to connect to the specified database using the connectToDB() method (lines 122–140).

When the user clicks the Execute SQL Command button, the executeSQL() method is invoked (lines 143–161) to get the SQL commands from the text area (jtaSQLCommand) and extract each command separated by a semicolon (;). It then determines whether the com- mand is a SELECT query or a DDL or data modification statement (lines 153–158). If the command is a SELECT query, the processSQLSelect method is invoked (lines 164–198). This method uses the executeQuery method (line 170) to obtain the query result. The result is displayed in the text area jtaSQLResult (line 181). If the command is a non-SELECT query, the processSQLNonSelect() method is invoked (lines 201–214). This method uses the executeUpdate method (line 207) to execute the SQL command.

The getMetaData method (lines 173, 178) in the ResultSet interface is used to obtain an instance of ResultSetMetaData. The getColumnCount method (line 173) returns the number of columns in the result set, and the getColumnName(i) method (line 178) returns the column name for the ith column.

35.3 Batch Processing You can send a batch of SQL statements to the database for execution at once to improve efficiency.

In all the preceding examples, SQL commands are submitted to the database for execution one at a time. This is inefficient for processing a large number of updates. For example, suppose you wanted to insert a thousand rows into a table. Submitting one INSERT com- mand at a time would take nearly a thousand times longer than submitting all the INSERT commands in a batch at once. To improve performance, JDBC introduced the batch update for processing nonselect SQL commands. A batch update consists of a sequence of nonselect SQL commands. These commands are collected in a batch and submitted to the database all together.

To use the batch update, you add nonselect commands to a batch using the addBatch method in the Statement interface. After all the SQL commands are added to the batch, use the executeBatch method to submit the batch to the database for execution.

For example, the following code adds a create table command, adds two insert statements in a batch, and executes the batch:

Statement statement = connection.createStatement();

// Add SQL commands to the batch statement.addBatch("create table T (C1 integer, C2 varchar(15))");

Point Key

M35_LIAN1878_11_GE_C35.indd 6 12/29/17 5:50 PM

35.3 Batch Processing 35-7

statement.addBatch("insert into T values (100, 'Smith')"); statement.addBatch("insert into T values (200, 'Jones')");

// Execute the batch int count[] = statement.executeBatch();

The executeBatch() method returns an array of counts, each of which counts the number of rows affected by the SQL command. The first count returns 0 because it is a DDL command. The other counts return 1 because only one row is affected.

Note To find out whether a driver supports batch updates, invoke supportsBatchUpdates() on a DatabaseMetaData instance. If the driver supports batch updates, it will return true. The JDBC drivers for MySQL, Access, and Oracle all support batch updates.

To demonstrate batch processing, consider writing a program that gets data from a text file and copies the data from the text file to a table, as shown in Figure 35.2. The text file consists of lines that each corresponds to a row in the table. The fields in a row are separated by commas. The string values in a row are enclosed in single quotes. You can view the text file by clicking the View File button and copy the text to the table by clicking the Copy button. The table must already be defined in the database. Figure 35.2 shows the text file table.txt copied to table Person. Person is created using the following statement:

create table Person ( firstName varchar(20), mi char(1), lastName varchar(20) )

Figure 35.2 The CopyFileToTable utility copies text files to database tables.

Listing 35.2 gives the solution to the problem.

Listing 35.2 CopyFileToTable.java 1 import java.io.File; 2 import java.io.FileNotFoundException; 3 import java.io.IOException; 4 import java.sql.*; 5 import java.util.Scanner; 6 import javafx.application.Application; 7 import javafx.collections.FXCollections; 8 import javafx.geometry.Pos; 9 import javafx.scene.Scene; 10 import javafx.scene.control.Button; 11 import javafx.scene.control.ComboBox; 12 import javafx.scene.control.Label; 13 import javafx.scene.control.PasswordField; 14 import javafx.scene.control.SplitPane;

M35_LIAN1878_11_GE_C35.indd 7 12/29/17 5:50 PM

35-8 Chapter 35 Advanced Java Database Programming

15 import javafx.scene.control.TextArea; 16 import javafx.scene.control.TextField; 17 import javafx.scene.layout.BorderPane; 18 import javafx.scene.layout.GridPane; 19 import javafx.scene.layout.HBox; 20 import javafx.scene.layout.VBox; 21 import javafx.stage.Stage; 22 23 public class CopyFileToTable extends Application { 24 // Text file info 25 private TextField tfFilename = new TextField(); 26 private TextArea taFile = new TextArea(); 27 28 // JDBC and table info 29 private ComboBox<String> cboURL = new ComboBox<>(); 30 private ComboBox<String> cboDriver = new ComboBox<>(); 31 private TextField tfUsername = new TextField(); 32 private PasswordField pfPassword = new PasswordField(); 33 private TextField tfTableName = new TextField(); 34 35 private Button btViewFile = new Button("View File"); 36 private Button btCopy = new Button("Copy"); 37 private Label lblStatus = new Label(); 38 39 @Override // Override the start method in the Application class 40 public void start(Stage primaryStage) { 41 cboURL.getItems().addAll(FXCollections.observableArrayList( 42 "jdbc:mysql://localhost/javabook", 43 "jdbc:mysql://liang.armstrong.edu/javabook", 44 "jdbc:odbc:exampleMDBDataSource", 45 "jdbc:oracle:thin:@liang.armstrong.edu:1521:orcl")); 46 cboURL.getSelectionModel().selectFirst(); 47 48 cboDriver.getItems().addAll(FXCollections.observableArrayList( 49 "com.mysql.jdbc.Driver", "sun.jdbc.odbc.dbcOdbcDriver", 50 "oracle.jdbc.driver.OracleDriver")); 51 cboDriver.getSelectionModel().selectFirst(); 52 53 // Create UI for connecting to the database 54 GridPane gridPane = new GridPane(); 55 gridPane.add(new Label("JDBC Driver"), 0, 0); 56 gridPane.add(new Label("Database URL"), 0, 1); 57 gridPane.add(new Label("Username"), 0, 2); 58 gridPane.add(new Label("Password"), 0, 3); 59 gridPane.add(new Label("Table Name"), 0, 4); 60 gridPane.add(cboURL, 1, 0); 61 gridPane.add(cboDriver, 1, 1); 62 gridPane.add(tfUsername, 1, 2); 63 gridPane.add(pfPassword, 1, 3); 64 gridPane.add(tfTableName, 1, 4); 65 66 HBox hBoxConnection = new HBox(10); 67 hBoxConnection.getChildren().addAll(lblStatus, btCopy); 68 hBoxConnection.setAlignment(Pos.CENTER_RIGHT); 69 70 VBox vBoxConnection = new VBox(5); 71 vBoxConnection.getChildren().addAll( 72 new Label("Target Database Table"), 73 gridPane, hBoxConnection); 74 75 gridPane.setStyle("-fx-border-color: black;");

M35_LIAN1878_11_GE_C35.indd 8 12/29/17 5:50 PM

35.3 Batch Processing 35-9

76 77 BorderPane borderPaneFileName = new BorderPane(); 78 borderPaneFileName.setLeft(new Label("Filename")); 79 borderPaneFileName.setCenter(tfFilename); 80 borderPaneFileName.setRight(btViewFile); 81 82 BorderPane borderPaneFileContent = new BorderPane(); 83 borderPaneFileContent.setTop(borderPaneFileName); 84 borderPaneFileContent.setCenter(taFile); 85 86 BorderPane borderPaneFileSource = new BorderPane(); 87 borderPaneFileSource.setTop(new Label("Source Text File")); 88 borderPaneFileSource.setCenter(borderPaneFileContent); 89 90 SplitPane sp = new SplitPane(); 91 sp.getItems().addAll(borderPaneFileSource, vBoxConnection); 92 93 // Create a scene and place it in the stage 94 Scene scene = new Scene(sp, 680, 230); 95 primaryStage.setTitle("CopyFileToTable"); // Set the stage title 96 primaryStage.setScene(scene); // Place the scene in the stage 97 primaryStage.show(); // Display the stage 98 99 btViewFile.setOnAction(e -> showFile()); 100 btCopy.setOnAction(e -> { 101 try { 102 copyFile(); 103 } 104 catch (Exception ex) { 105 lblStatus.setText(ex.toString()); 106 } 107 }); 108 } 109 110 /** Display the file in the text area */ 111 private void showFile() { 112 Scanner input = null; 113 try { 114 // Use a Scanner to read text from the file 115 input = new Scanner(new File(tfFilename.getText().trim())); 116 117 // Read a line and append the line to the text area 118 while (input.hasNext()) 119 taFile.appendText(input.nextLine() + '\n'); 120 } 121 catch (FileNotFoundException ex) { 122 System.out.println("File not found: " + tfFilename.getText()); 123 } 124 catch (IOException ex) { 125 ex.printStackTrace(); 126 } 127 finally { 128 if (input != null) input.close(); 129 } 130 } 131 132 private void copyFile() throws Exception { 133 // Load the JDBC driver 134 Class.forName(cboDriver.getSelectionModel() 135 .getSelectedItem().trim()); 136 System.out.println("Driver loaded");

M35_LIAN1878_11_GE_C35.indd 9 12/29/17 5:50 PM

35-10 Chapter 35 Advanced Java Database Programming

137 138 // Establish a connection 139 Connection conn = DriverManager.getConnection( 140 cboURL.getSelectionModel().getSelectedItem().trim(), 141 tfUsername.getText().trim(), 142 String.valueOf(pfPassword.getText()).trim()); 143 System.out.println("Database connected"); 144 145 // Read each line from the text file and insert it to the table 146 insertRows(conn); 147 } 148 149 private void insertRows(Connection connection) { 150 // Build the SQL INSERT statement 151 String sqlInsert = "insert into " + tfTableName.getText() 152 + " values ("; 153 154 // Use a Scanner to read text from the file 155 Scanner input = null; 156 157 // Get file name from the text field 158 String filename = tfFilename.getText().trim(); 159 160 try { 161 // Create a scanner 162 input = new Scanner(new File(filename)); 163 164 // Create a statement 165 Statement statement = connection.createStatement(); 166 167 System.out.println("Driver major version? " + 168 connection.getMetaData().getDriverMajorVersion()); 169 170 // Determine if batchUpdatesSupported is supported 171 boolean batchUpdatesSupported = false; 172 173 try { 174 if (connection.getMetaData().supportsBatchUpdates()) { 175 batchUpdatesSupported = true; 176 System.out.println("batch updates supported"); 177 } 178 else { 179 System.out.println("The driver " + 180 "does not support batch updates"); 181 } 182 } 183 catch (UnsupportedOperationException ex) { 184 System.out.println("The operation is not supported"); 185 } 186 187 // Determine if the driver is capable of batch updates 188 if (batchUpdatesSupported) { 189 // Read a line and add the insert table command to the batch 190 while (input.hasNext()) { 191 statement.addBatch(sqlInsert + input.nextLine() + ")"); 192 } 193 194 statement.executeBatch(); 195 196 lblStatus.setText("Batch updates completed"); 197 }

M35_LIAN1878_11_GE_C35.indd 10 12/29/17 5:50 PM

35.4 Scrollable and Updatable Result Set 35-11

198 else { 199 // Read a line and execute insert table command 200 while (input.hasNext()) { 201 statement.executeUpdate(sqlInsert + input.nextLine() + ")"); 202 } 203 204 lblStatus.setText("Single row update completed"); 205 } 206 } 207 catch (SQLException ex) { 208 System.out.println(ex); 209 } 210 catch (FileNotFoundException ex) { 211 System.out.println("File not found: " + filename); 212 } 213 finally { 214 if (input != null) input.close(); 215 } 216 } 217 }

The insertRows method (lines 149–216) uses the batch updates to submit SQL INSERT commands to the database for execution, if the driver supports batch updates. Lines 174–181 check whether the driver supports batch updates. If the driver does not support the opera- tion, an UnsupportedOperationException exception will be thrown (line 183) when the supportsBatchUpdates() method is invoked.

The tables must already be created in the database. The file format and contents must match the database table specification. Otherwise, the SQL INSERT command will fail.

In Exercise 35.1, you will write a program to insert a thousand records to a database and compare the performance with and without batch updates.

35.3.1 What is batch processing in JDBC? What are the benefits of using batch processing? 35.3.2 How do you add an SQL statement to a batch? How do you execute a batch? 35.3.3 Can you execute a SELECT statement in a batch? 35.3.4 How do you know whether a JDBC driver supports batch updates?

35.4 Scrollable and Updatable Result Set You can use scrollable and updatable result set to move the cursor anywhere in the result set to perform insertion, deletion, and update.

The result sets used in the preceding examples are read sequentially. A result set maintains a cursor pointing to its current row of data. Initially the cursor is positioned before the first row. The next() method moves the cursor forward to the next row. This is known as sequential forward reading.

A more powerful way of accessing database is to use a scrollable and updatable result, which enables you to scroll the rows both forward and backward and move the cursor to a desired location using the first, last, next, previous, absolute, or relative method. Additionally, you can insert, delete, or update a row in the result set and have the changes automatically reflected in the database.

To obtain a scrollable or updatable result set, you must first create a statement with an appropriate type and concurrency mode. For a static statement, use

Statement statement = connection.createStatement (int resultSetType, int resultSetConcurrency);

Point Check

Point Key

M35_LIAN1878_11_GE_C35.indd 11 12/29/17 5:50 PM

35-12 Chapter 35 Advanced Java Database Programming

For a prepared statement, use

PreparedStatement statement = connection.prepareStatement (String sql, int resultSetType, int resultSetConcurrency);

The possible values of resultSetType are the constants defined in the ResultSet:

■■ TYPE_FORWARD_ONLY: The result set is accessed forward sequentially.

■■ TYPE_SCROLL_INSENSITIVE: The result set is scrollable, but not sensitive to changes in the database.

■■ TYPE_SCROLL_SENSITIVE: The result set is scrollable and sensitive to changes made by others. Use this type if you want the result set to be scrollable and updatable.

The possible values of resultSetConcurrency are the constants defined in the ResultSet:

■■ CONCUR_READ_ONLY: The result set cannot be used to update the database.

■■ CONCUR_UPDATABLE: The result set can be used to update the database.

For example, if you want the result set to be scrollable and updatable, you can create a statement, as follows:

Statement statement = connection.createStatement (ResultSet.TYPE_SCROLL_SENSITIVE, ResultSet.CONCUR_UPDATABLE)

You use the executeQuery method in a Statement object to execute an SQL query that returns a result set as follows:

ResultSet resultSet = statement.executeQuery(query);

You can now use the methods first(), next(), previous(), and last() to move the cursor to the first row, next row, previous row, and last row. The absolute(int row) method moves the cursor to the specified row; and the getXxx(int columnIndex) or getXxx(String columnName) method is used to retrieve the value of a specified field at the current row. The methods insertRow(), deleteRow(), and updateRow() can also be used to insert, delete, and update the current row. Before applying insertRow or updateRow, you need to use the method updateXxx(int columnIndex, Xxx value) or update(String columnName, Xxx value) to write a new value to the field at the current row. The cancel- RowUpdates() method cancels the updates made to a row. The close() method closes the result set and releases its resource. The wasNull() method returns true if the last column read had a value of SQL NULL.

Listing 35.3 gives an example that demonstrates how to create a scrollable and updatable result set. The program creates a result set for the StateCapital table. The StateCapital table is defined as follows:

create table StateCapital ( state varchar(40), capital varchar(40) );

Listing 35.3 ScrollUpdateResultSet.java 1 import java.sql.*; 2 3 public class ScrollUpdateResultSet { 4 public static void main(String[] args) 5 throws SQLException, ClassNotFoundException {

M35_LIAN1878_11_GE_C35.indd 12 12/29/17 5:50 PM

35.4 Scrollable and Updatable Result Set 35-13

6 // Load the JDBC driver 7 Class.forName("oracle.jdbc.driver.OracleDriver"); 8 System.out.println("Driver loaded"); 9 10 // Connect to a database 11 Connection connection = DriverManager.getConnection 12 ("jdbc:oracle:thin:@liang.armstrong.edu:1521:orcl", 13 "scott", "tiger"); 14 connection.setAutoCommit(true); 15 System.out.println("Database connected"); 16 17 // Get a new statement for the current connection 18 Statement statement = connection.createStatement( 19 ResultSet.TYPE_SCROLL_SENSITIVE, ResultSet.CONCUR_UPDATABLE); 20 21 // Get ResultSet 22 ResultSet resultSet = statement.executeQuery 23 ("select state, capital from StateCapital"); 24 25 System.out.println("Before update "); 26 displayResultSet(resultSet); 27 28 // Update the second row 29 resultSet.absolute(2); // Move cursor to the second row 30 resultSet.updateString("state", "New S"); // Update the column 31 resultSet.updateString("capital", "New C"); // Update the column 32 resultSet.updateRow(); // Update the row in the data source 33 34 // Insert after the last row 35 resultSet.last(); 36 resultSet.moveToInsertRow(); // Move cursor to the insert row 37 resultSet.updateString("state", "Florida"); 38 resultSet.updateString("capital", "Tallahassee"); 39 resultSet.insertRow(); // Insert the row 40 resultSet.moveToCurrentRow(); // Move the cursor to the current row 41 42 // Delete fourth row 43 resultSet.absolute(4); // Move cursor to the 5th row 44 resultSet.deleteRow(); // Delete the second row 45 46 System.out.println("After update "); 47 resultSet = statement.executeQuery 48 ("select state, capital from StateCapital"); 49 displayResultSet(resultSet); 50 51 // Close the connection 52 resultSet.close(); 53 } 54 55 private static void displayResultSet(ResultSet resultSet) 56 throws SQLException { 57 ResultSetMetaData rsMetaData = resultSet.getMetaData(); 58 resultSet.beforeFirst(); 59 while (resultSet.next()) { 60 for (int i = 1; i <= rsMetaData.getColumnCount(); i++) 61 System.out.printf("%-12s\t", resultSet.getObject(i)); 62 System.out.println(); 63 } 64 } 65 }

M35_LIAN1878_11_GE_C35.indd 13 12/29/17 5:50 PM

35-14 Chapter 35 Advanced Java Database Programming

The code in lines 18–19 creates a Statement for producing scrollable and updatable result sets. The program moves the cursor to the second row in the result set (line 29), updates two

columns in this row (lines 30–31), and invokes the updateRow() method to update the row in the underlying database (line 32).

An updatable ResultSet object has a special row associated with it that serves as a staging area for building a row to be inserted. This special row is called the insert row. To insert a row, first invoke the moveToInsertRow() method to move the cursor to the insert row (line 36), then update the columns using the updateXxx method (lines 37–38), and finally insert the row using the insertRow() method (line 39). Invoking moveToCurrentRow()moves the cursor to the current inserted row (lines 40).

The program moves to the fourth row and invokes the deleteRow()method to delete the row from the database (lines 43–44).

Note Not all current drivers support scrollable and updatable result sets. The example is tested using Oracle ojdbc6 driver. You can use supportsResultSetType(int type) and supportsResultSetConcurrency(int type, int concurrency) in the DatabaseMetaData interface to find out which result type and currency modes are supported by the JDBC driver. But even if a driver supports the scrollable and updatable result set, a result set for a complex query might not be able to perform an update. For example, the result set for a query that involves several tables is likely not to support update operations.

Note The program may not work due to an issue in the Oracle JDBC driver if lines 22–23 are replaced by

ResultSet resultSet = statement.executeQuery ("select * from StateCapital");

35.4.1 What is a scrollable result set? What is an updatable result set? 35.4.2 How do you create a scrollable and updatable ResultSet? 35.4.3 How do you know whether a JDBC driver supports a scrollable and updatable

ResultSet?

Point Check

Driver loaded Database connected

Before update Indiana Indianapolis Illinois Springfield California Sacramento Georgia Atlanta Texas Austin

After update Indiana Indianapolis New S New C California Sacramento Texas Austin Florida Tallahassee

M35_LIAN1878_11_GE_C35.indd 14 12/29/17 5:50 PM

35.5 RowSet, JdbcRowSet, and CachedRowSet 35-15

35.5 RowSet, JdbcRowSet, and CachedRowSet The RowSet interface can be used to simplify database programming.

The RowSet interface extends java.sql.ResultSet with additional capabilities that allow a RowSet instance to be configured to connect to a JDBC url, username, and password, set an SQL command, execute the command, and retrieve the execution result. In essence, it com- bines Connection, Statement, and ResultSet into one interface.

Note Not all JDBC drivers support RowSet. Currently, the JDBC-ODBC driver does not sup- port all features of RowSet.

35.5.1 RowSet Basics There are two types of RowSet objects: connected and disconnected. A connected RowSet object makes a connection with a data source and maintains that connection throughout its life cycle. A disconnected RowSet object makes a connection with a data source, executes a query to get data from the data source, and then closes the connection. A disconnected rowset may make changes to its data while it is disconnected and then send the changes back to the original source of the data, but it must reestablish a connection to do so.

There are several versions of RowSet. Two frequently used are JdbcRowSet and Cached- RowSet. Both are subinterfaces of RowSet. JdbcRowSet is connected, while CachedRowSet is disconnected. Also, JdbcRowSet is neither serializable nor cloneable, while CachedRowSet is both. The database vendors are free to provide concrete implementations for these inter- faces. Oracle has provided the reference implementation JdbcRowSetImpl for JdbcRowSet and CachedRowSetImpl for CachedRowSet. Figure 35.3 shows the relationship of these components.

Point Key

Figure 35.3 The JdbcRowSetImpl and CachedRowSetImpl are concrete implementa- tions of RowSet.

«interface» java.sql.ResultSet

«interface» javax.sql.RowSet

«interface» javax.sql.rowset.JdbcRowSet

«interface» javax.sql.rowset.CachedRowSet

com.sun.rowset.JdbcRowSetImpl com.sun.rowset.CachedRowSetImpl

The RowSet interface contains the JavaBeans properties with getter and setter methods. You can use the setter methods to set a new url, username, password, and command for an SQL statement. Using a RowSet, Listing 34.1 can be simplified, as shown in Listing 35.4.

Listing 35.4 SimpleRowSet.java 1 import java.sql.SQLException; 2 import javax.sql.RowSet; 3 import com.sun.rowset.*; 4

M35_LIAN1878_11_GE_C35.indd 15 12/29/17 5:50 PM

35-16 Chapter 35 Advanced Java Database Programming

5 public class SimpleRowSet { 6 public static void main(String[] args) 7 throws SQLException, ClassNotFoundException { 8 // Load the JDBC driver 9 Class.forName("com.mysql.jdbc.Driver"); 10 System.out.println("Driver loaded"); 11 12 // Create a row set 13 RowSet rowSet = new JdbcRowSetImpl(); 14 15 // Set RowSet properties 16 rowSet.setUrl("jdbc:mysql://localhost/javabook"); 17 rowSet.setUsername("scott"); 18 rowSet.setPassword("tiger"); 19 rowSet.setCommand("select firstName, mi, lastName " + 20 "from Student where lastName = 'Smith'"); 21 rowSet.execute(); 22 23 // Iterate through the result and print the student names 24 while (rowSet.next()) 25 System.out.println(rowSet.getString(1) + "\t" + 26 rowSet.getString(2) + "\t" + rowSet.getString(3)); 27 28 // Close the connection 29 rowSet.close(); 30 } 31 }

Line 13 creates a RowSet object using JdbcRowSetImpl. The program uses the RowSet’s set method to set a URL, username, and password (lines 16–18) and a command for a query statement (line 19). Line 24 executes the command in the RowSet. The methods next() and getString(int) for processing the query result (lines 25–26) are inherited from ResultSet.

If you replace JdbcRowSet with CachedRowSet in line 13, the program will work just fine. Note, the JDBC-ODBC driver supports JdbcRowSetImpl, but not CachedRowSetImpl.

Tip Since RowSet is a subinterface of ResultSet, all the methods in ResultSet can be used in RowSet. For example, you can obtain ResultSetMetaData from a RowSet using the getMetaData() method.

35.5.2 RowSet for PreparedStatement The discussion in §34.5, “PreparedStatement,” introduced processing parameterized SQL statements using the PreparedStatement interface. RowSet has the capability to support parameterized SQL statements. The set methods for setting parameter values in Prepared- Statement are implemented in RowSet. You can use these methods to set parameter values for a parameterized SQL command. Listing 35.5 demonstrates how to use a parameterized statement in RowSet. Line 19 sets an SQL query statement with two parameters for lastName and mi in a RowSet. Since these two parameters are strings, the setString method is used to set actual values in lines 21–22.

Listing 35.5 RowSetPreparedStatement.java 1 import java.sql.*; 2 import javax.sql.RowSet; 3 import com.sun.rowset.*; 4

M35_LIAN1878_11_GE_C35.indd 16 12/29/17 5:50 PM

35.5 RowSet, JdbcRowSet, and CachedRowSet 35-17

5 public class RowSetPreparedStatement { 6 public static void main(String[] args) 7 throws SQLException, ClassNotFoundException { 8 // Load the JDBC driver 9 Class.forName("com.mysql.jdbc.Driver"); 10 System.out.println("Driver loaded"); 11 12 // Create a row set 13 RowSet rowSet = new JdbcRowSetImpl(); 14 15 // Set RowSet properties 16 rowSet.setUrl("jdbc:mysql://localhost/javabook"); 17 rowSet.setUsername("scott"); 18 rowSet.setPassword("tiger"); 19 rowSet.setCommand("select * from Student where lastName = ? " + 20 "and mi = ?"); 21 rowSet.setString(1, "Smith"); 22 rowSet.setString(2, "R"); 23 rowSet.execute(); 24 25 ResultSetMetaData rsMetaData = rowSet.getMetaData(); 26 for (int i = 1; i <= rsMetaData.getColumnCount(); i++) 27 System.out.printf("%-12s\t", rsMetaData.getColumnName(i)); 28 System.out.println(); 29 30 // Iterate through the result and print the student names 31 while (rowSet.next()) { 32 for (int i = 1; i <= rsMetaData.getColumnCount(); i++) 33 System.out.printf("%-12s\t", rowSet.getObject(i)); 34 System.out.println(); 35 } 36 37 // Close the connection 38 rowSet.close(); 39 } 40 }

35.5.3 Scrolling and Updating RowSet By default, a ResultSet object is neither scrollable nor updatable. However, a RowSet object is both. It is easier to scroll and update a database through a RowSet than through a ResultSet. Listing 35.6 rewrites Listing 35.3 using a RowSet. You can use methods such as absolute(int) to move the cursor and methods such as delete(), updateRow(), and insertRow() to update the database.

Listing 35.6 ScrollUpdateRowSet.java 1 import java.sql.*; 2 import javax.sql.RowSet; 3 import com.sun.rowset.JdbcRowSetImpl; 4 5 public class ScrollUpdateRowSet { 6 public static void main(String[] args) 7 throws SQLException, ClassNotFoundException { 8 // Load the JDBC driver 9 Class.forName("com.mysql.jdbc.Driver"); 10 System.out.println("Driver loaded"); 11

M35_LIAN1878_11_GE_C35.indd 17 12/29/17 5:50 PM

35-18 Chapter 35 Advanced Java Database Programming

12 // Create a row set 13 RowSet rowSet = new JdbcRowSetImpl(); 14 15 // Set RowSet properties 16 rowSet.setUrl("jdbc:mysql://localhost/javabook"); 17 rowSet.setUsername("scott"); 18 rowSet.setPassword("tiger"); 19 rowSet.setCommand("select state, capital from StateCapital"); 20 rowSet.execute(); 21 22 System.out.println("Before update "); 23 displayRowSet(rowSet); 24 25 // Update the second row 26 rowSet.absolute(2); // Move cursor to the 2nd row 27 rowSet.updateString("state", "New S"); // Update the column 28 rowSet.updateString("capital", "New C"); // Update the column 29 rowSet.updateRow(); // Update the row in the data source 30 31 // Insert after the second row 32 rowSet.last(); 33 rowSet.moveToInsertRow(); // Move cursor to the insert row 34 rowSet.updateString("state", "Florida"); 35 rowSet.updateString("capital", "Tallahassee"); 36 rowSet.insertRow(); // Insert the row 37 rowSet.moveToCurrentRow(); // Move the cursor to the current row 38 39 // Delete fourth row 40 rowSet.absolute(4); // Move cursor to the fifth row 41 rowSet.deleteRow(); // Delete the second row 42 43 System.out.println("After update "); 44 displayRowSet(rowSet); 45 46 // Close the connection 47 rowSet.close(); 48 } 49 50 private static void displayRowSet(RowSet rowSet) 51 throws SQLException { 52 ResultSetMetaData rsMetaData = rowSet.getMetaData(); 53 rowSet.beforeFirst(); 54 while (rowSet.next()) { 55 for (int i = 1; i <= rsMetaData.getColumnCount(); i++) 56 System.out.printf("%-12s\t", rowSet.getObject(i)); 57 System.out.println(); 58 } 59 } 60 }

If you replace JdbcRowSet with CachedRowSet in line 13, the database is not changed. To make the changes on the CachedRowSet effective in the database, you must invoke the acceptChanges() method after you make all the changes, as follows:

// Write changes back to the database ((com.sun.rowset.CachedRowSetImpl)rowSet).acceptChanges();

This method automatically reconnects to the database and writes all the changes back to the database.

M35_LIAN1878_11_GE_C35.indd 18 12/29/17 5:50 PM

35.5 RowSet, JdbcRowSet, and CachedRowSet 35-19

35.5.4 RowSetEvent A RowSet object fires a RowSetEvent whenever the object’s cursor has moved, a row has changed, or the entire row set has changed. This event can be used to synchronize a RowSet with the components that rely on the RowSet. For example, a visual component that displays the con- tents of a RowSet should be synchronized with the RowSet. The RowSetEvent can be used to achieve synchronization. The handlers in RowSetListener are cursorMoved(RowSetEvent), rowChanged(RowSetEvent), and cursorSetChanged(RowSetEvent).

Listing 35.7 gives an example that demonstrates RowSetEvent. A listener for RowSetEvent is registered in lines 14–26. When rowSet.execute() (line 33) is executed, the entire row set is changed, so the listener’s rowSetChanged handler is invoked. When rowSet.last() (line 35) is executed, the cursor is moved, so the listener’s cursorMoved handler is invoked. When rowSet.updateRow() (line 37) is executed, the row is updated, so the listener’s row- Changed handler is invoked.

Listing 35.7 TestRowSetEvent.java 1 import java.sql.*; 2 import javax.sql.*; 3 import com.sun.rowset.*; 4 5 public class TestRowSetEvent { 6 public static void main(String[] args) 7 throws SQLException, ClassNotFoundException { 8 // Load the JDBC driver 9 Class.forName("com.mysql.jdbc.Driver"); 10 System.out.println("Driver loaded"); 11 12 // Create a row set 13 RowSet rowSet = new JdbcRowSetImpl(); 14 rowSet.addRowSetListener(new RowSetListener() { 15 public void cursorMoved(RowSetEvent e) { 16 System.out.println("Cursor moved"); 17 } 18 19 public void rowChanged(RowSetEvent e) { 20 System.out.println("Row changed"); 21 } 22 23 public void rowSetChanged(RowSetEvent e) { 24 System.out.println("row set changed"); 25 } 26 }); 27 28 // Set RowSet properties 29 rowSet.setUrl("jdbc:mysql://localhost/javabook"); 30 rowSet.setUsername("scott"); 31 rowSet.setPassword("tiger"); 32 rowSet.setCommand("select * from Student"); 33 rowSet.execute(); 34 35 rowSet.last(); // Cursor moved 36 rowSet.updateString("lastName", "Yao"); // Update column 37 rowSet.updateRow(); // Row updated 38 39 // Close the connection 40 rowSet.close(); 41 } 42 }

M35_LIAN1878_11_GE_C35.indd 19 12/29/17 5:50 PM

35-20 Chapter 35 Advanced Java Database Programming

35.5.1 What are the advantages of RowSet? 35.5.2 What are JdbcRowSet and CachedRowSet? What are the differences between

them?

35.5.3 How do you create a JdbcRowSet and a CachedRowSet? 35.5.4 Can you scroll and update a RowSet? What method must be invoked to write the

changes in a CachedRowSet to the database?

35.5.5 Describe the handlers in RowSetListener.

35.6 Storing and Retrieving Images in JDBC You can store and retrieve images using JDBC.

A database can store not only numbers and strings, but also images. SQL3 introduced a new data type called BLOB (Binary Large OBject) for storing binary data, which can be used to store images. Another new SQL3 type is CLOB (Character Large OBject) for storing a large text in the character format. JDBC introduced the interfaces java.sql.Blob and java. sql.Clob to support mapping for these new SQL types. You can use getBlob, setBi- naryStream, getClob, setBlob, and setClob, to access SQL BLOB and CLOB values in the interfaces ResultSet and PreparedStatement.

To store an image into a cell in a table, the corresponding column for the cell must be of the BLOB type. For example, the following SQL statement creates a table whose type for the flag column is BLOB:

create table Country(name varchar(30), flag blob, description varchar(255));

In the preceding statement, the description column is limited to 255 characters, which is the upper limit for MySQL. For Oracle, the upper limit is 32,672 bytes. For a large character field, you can use the CLOB type for Oracle, which can store up to two GB of characters. MySQL does not support CLOB. However, you can use BLOB to store a long string and con- vert binary data into characters.

Note MS Access database does not support the BLOB and CLOB types.

To insert a record with images to a table, define a prepared statement like this one:

PreparedStatement pstmt = connection.prepareStatement( "insert into Country values(?, ?, ?)");

Images are usually stored in files. You may first get an instance of InputStream for an image file then use the setBinaryStream method to associate the input stream with a cell in the table, as follows:

// Store image to the table cell File file = new File(imageFilename); InputStream inputImage = new FileInputStream(file); pstmt.setBinaryStream(2, inputImage, (int)(file.length()));

To retrieve an image from a table, use the getBlob method, as shown below:

// Store image to the table cell Blob blob = rs.getBlob(1); ImageIcon imageIcon = new ImageIcon( blob.getBytes(1, (int)blob.length()));

Listing 35.8 gives a program that demonstrates how to store and retrieve images in JDBC. The program first creates the Country table and stores data to it. Then the program retrieves

Point Check

Point Key

M35_LIAN1878_11_GE_C35.indd 20 12/29/17 5:50 PM

35.6 Storing and Retrieving Images in JDBC 35-21

the country names from the table and adds them to a combo box. When the user selects a name from the combo box, the country's flag and description are displayed, as shown in Figure 35.4.

Figure 35.4 The program enables you to retrieve data, including images, from a table and displays them.

Listing 35.8 StoreAndRetrieveImage.java 1 import java.sql.*; 2 import java.io.*; 3 import javafx.application.Application; 4 import javafx.scene.Scene; 5 import javafx.scene.control.ComboBox; 6 import javafx.scene.control.Label; 7 import javafx.scene.image.Image; 8 import javafx.scene.image.ImageView; 9 import javafx.scene.layout.BorderPane; 10 import javafx.stage.Stage; 11 12 public class StoreAndRetrieveImage extends Application { 13 // Connection to the database 14 private Connection connection; 15 16 // Statement for static SQL statements 17 private Statement stmt; 18 19 // Prepared statement 20 private PreparedStatement pstmt = null; 21 private DescriptionPane descriptionPane 22 = new DescriptionPane(); 23 24 private ComboBox<String> cboCountry = new ComboBox<>(); 25 26 @Override // Override the start method in the Application class 27 public void start(Stage primaryStage) { 28 try { 29 connectDB(); // Connect to DB 30 storeDataToTable(); //Store data to the table (including image) 31 fillDataInComboBox(); // Fill in combo box 32 retrieveFlagInfo(cboCountry.getSelectionModel().getSelectedItem()); 33 } 34 catch (Exception ex) { 35 ex.printStackTrace(); 36 } 37 38 BorderPane paneForComboBox = new BorderPane(); 39 paneForComboBox.setLeft(new Label("Select a country: ")); 40 paneForComboBox.setCenter(cboCountry); 41 cboCountry.setPrefWidth(400); 42 BorderPane pane = new BorderPane();

M35_LIAN1878_11_GE_C35.indd 21 12/29/17 5:50 PM

35-22 Chapter 35 Advanced Java Database Programming

43 pane.setTop(paneForComboBox); 44 pane.setCenter(descriptionPane); 45 46 Scene scene = new Scene(pane, 350, 150); 47 primaryStage.setTitle("StoreAndRetrieveImage"); 48 primaryStage.setScene(scene); // Place the scene in the stage 49 primaryStage.show(); // Display the stage 50 51 cboCountry.setOnAction(e -> 52 retrieveFlagInfo(cboCountry.getValue())); 53 } 54 55 private void connectDB() throws Exception { 56 // Load the driver 57 Class.forName("com.mysql.jdbc.Driver"); 58 System.out.println("Driver loaded"); 59 60 // Establish connection 61 connection = DriverManager.getConnection 62 ("jdbc:mysql://localhost/javabook", "scott", "tiger"); 63 System.out.println("Database connected"); 64 65 // Create a statement for static SQL 66 stmt = connection.createStatement(); 67 68 // Create a prepared statement to retrieve flag and description 69 pstmt = connection.prepareStatement("select flag, description " + 70 "from Country where name = ?"); 71 } 72 73 private void storeDataToTable() { 74 String[] countries = {"Canada", "UK", "USA", "Germany", 75 "Indian", "China"}; 76 77 String[] imageFilenames = {"image/ca.gif", "image/uk.gif", 78 "image/us.gif", "image/germany.gif", "image/india.gif", 79 "image/china.gif"}; 80 81 String[] descriptions = {"A text to describe Canadian " + 82 "flag is omitted", "British flag ...", "American flag ...", 83 "German flag ...", "Indian flag ...", "Chinese flag ..."}; 84 85 try { 86 // Create a prepared statement to insert records 87 PreparedStatement pstmt = connection.prepareStatement( 88 "insert into Country values(?, ?, ?)"); 89 90 // Store all predefined records 91 for (int i = 0; i < countries.length; i++) { 92 pstmt.setString(1, countries[i]); 93 94 // Store image to the table cell 95 java.net.URL url = 96 this.getClass().getResource(imageFilenames[i]); 97 InputStream inputImage = url.openStream(); 98 pstmt.setBinaryStream(2, inputImage, 99 (int)(inputImage.available())); 100 101 pstmt.setString(3, descriptions[i]); 102 pstmt.executeUpdate();

M35_LIAN1878_11_GE_C35.indd 22 12/29/17 5:50 PM

35.6 Storing and Retrieving Images in JDBC 35-23

103 } 104 105 System.out.println("Table Country populated"); 106 } 107 catch (Exception ex) { 108 ex.printStackTrace(); 109 } 110 } 111 112 private void fillDataInComboBox() throws Exception { 113 ResultSet rs = stmt.executeQuery("select name from Country"); 114 while (rs.next()) { 115 cboCountry.getItems().add(rs.getString(1)); 116 } 117 cboCountry.getSelectionModel().selectFirst(); 118 } 119 120 private void retrieveFlagInfo(String name) { 121 try { 122 pstmt.setString(1, name); 123 ResultSet rs = pstmt.executeQuery(); 124 if (rs.next()) { 125 Blob blob = rs.getBlob(1); 126 ByteArrayInputStream in = new ByteArrayInputStream 127 (blob.getBytes(1, (int)blob.length())); 128 Image image = new Image(in); 129 ImageView imageView = new ImageView(image); 130 descriptionPane.setImageView(imageView); 131 descriptionPane.setTitle(name); 132 String description = rs.getString(2); 133 descriptionPane.setDescription(description); 134 } 135 } 136 catch (Exception ex) { 137 System.err.println(ex); 138 } 139 } 140 }

DescriptionPane (line 21) is a component for displaying a country (name, flag, and descrip- tion). This component was presented in Listing 16.6, DescriptionPane.java.

The storeDataToTable method (lines 73–110) populates the table with data. The fill- DataInComboBox method (lines 112–118) retrieves the country names and adds them to the combo box. The retrieveFlagInfo(name) method (lines 120–139) retrieves the flag and description for the specified country name.

35.6.1 How do you store images into a database? 35.6.2 How do you retrieve images from a database? 35.6.3 Does Oracle support the SQL3 BLOB type and CLOB type? What about MySQL

and Access?

Point Check

BLOB type 35-20 CLOB type 35-20 batch mode 35-2 cached row set 35-15

row set 35-15 scrollable result set 35-2 updatable result set 35-11

Key Terms

M35_LIAN1878_11_GE_C35.indd 23 12/29/17 5:50 PM

35-24 Chapter 35 Advanced Java Database Programming

ChapTer summary 1. This chapter developed a universal SQL client that can be used to access any local or

remote relational database.

2. You can use the addBatch(SQLString) method to add SQL statements to a statement for batch processing.

3. You can create a statement to specify that the result set be scrollable and updatable. By default, the result set is neither of these.

4. The RowSet can be used to simplify Java database programming. A RowSet object is scrollable and updatable. A RowSet can fire a RowSetEvent.

5. You can store and retrieve image data in JDBC using the SQL BLOB type.

Quiz Answer the quiz for this chapter online at the book Companion Website.

programming exerCises *35.1 (Batch update) Write a program that inserts a thousand records to a database, and

compare the performance with and without batch updates, as shown in Figure 35.5a. Suppose the table is defined as follows:

create table Temp(num1 double, num2 double, num3 double)

Use the Math.random() method to generate random numbers for each record. Create a dialog box that contains DBConnectionPanel, discussed in Exercise 34.3. Use this dialog box to connect to the database. When you click the Connect to Database button in Figure 35.5a, the dialog box in Figure 35.5b is displayed.

Figure 35.5 The program demonstrates the performance improvements that result from using batch updates.

(a) (b)

**35.2 (Scrollable result set) Write a program that uses the buttons First, Next, Prior, Last, Insert, Delete, and Update, and modify a single record in the Address table, as shown in Figure 35.6.

M35_LIAN1878_11_GE_C35.indd 24 12/29/17 5:50 PM

Programming Exercises 35-25

The Address table is defined as follows:

create table Address ( firstname varchar(25), mi char(1), lastname varchar(25), street varchar(40), city varchar(20), state varchar(2), zip varchar(5), telephone varchar(10), email varchar(30), primary key (firstname, mi, lastname) );

*35.3 (Display table contents) Write a program that displays the content for a given table. As shown in Figure 35.7a, you enter a table and click the Show Contents button to display the table contents in a table view.

Figure 35.6 You can use the buttons to display and modify a single record in the Address table.

Figure 35.7 (a) Enter a table name to display the table contents. (b) Select a table name from the combo box to display its contents.

(a) (b)

*35.4 (Find tables and showing their contents) Write a program that fills in table names in a combo box, as shown in Figure 35.7b. You can select a table from the combo box to display its contents in a table view.

M35_LIAN1878_11_GE_C35.indd 25 12/29/17 5:50 PM

35-26 Chapter 35 Advanced Java Database Programming

**35.5 (Revise SQLClient.java) Rewrite Listing 35.1, SQLClient.java, to display the query result in a TableView, as shown in Figure 35.8.

Figure 35.8 The query result is displayed in a TableView.

*35.5 (Populate Salary table) Rewrite Programming Exercise 34.8 using a batch mode to improve performance.

M35_LIAN1878_11_GE_C35.indd 26 12/29/17 5:50 PM

Objectives ■■ To describe Java’s internationalization features (§36.1).

■■ To construct a locale with language, country, and variant (§36.2).

■■ To display date and time based on locale (§36.3).

■■ To display numbers, currencies, and percentages based on locale (§36.4).

■■ To develop applications for international audiences using resource bundles (§36.5).

■■ To specify encoding schemes for text I/O (§36.6).

Internationalization

Chapter

M36_LIAN1878_11_GE_C36.indd 1 12/29/17 5:50 PM

36-2 Chapter 36 Internationalization

36.1 Introduction This chapter introduces writing Java code for international audience.

Many websites maintain several versions of webpages so that readers can choose one written in a language they understand. Because there are so many languages in the world, it would be highly problematic to create and maintain enough different versions to meet the needs of all cli- ents everywhere. Java comes to the rescue. Java is the first language designed from the ground up to support internationalization. In consequence, it allows your programs to be customized for any number of countries or languages without requiring cumbersome changes in the code.

Here are the major Java features that support internationalization:

■■ Java characters use Unicode, a 16-bit encoding scheme established by the Unicode Consortium to support the interchange, processing, and display of written texts in the world’s diverse languages. The use of Unicode encoding makes it easy to write Java programs that can manipulate strings in any international language. (To see all the Unicode characters, visit mindprod.com/jgloss/reuters.html.)

■■ Java provides the Locale class to encapsulate information about a specific locale. A Locale object determines how locale-sensitive information, such as date, time, and number, is displayed, and how locale-sensitive operations, such as sorting strings, are performed. The classes for formatting date, time, and numbers, and for sorting strings are grouped in the java.text package.

■■ Java uses the ResourceBundle class to separate locale-specific information, such as status messages and GUI component labels, from the program. The information is stored outside the source code and can be accessed and loaded dynamically at runtime from a ResourceBundle, rather than hard-coded into the program.

In this chapter, you will learn how to format dates, numbers, currencies, and percentages for different regions, countries, and languages. You will also learn how to use resource bundles to define which images and strings are used by a component, depending on the user’s locale and preferences.

36.2 The Locale Class The Locale class defines a locale: language and nation.

A Locale object represents a geographical, political, or cultural region in which a specific language or custom is used. For example, Americans speak English, and the Chinese speak Chinese. The conventions for formatting dates, numbers, currencies, and percentages may dif- fer from one country to another. The Chinese, for instance, use year/month/day to represent the date, while Americans use month/day/year. It is important to realize that locale is not defined only by country. For example, Canadians speak either Canadian English or Canadian French, depending on which region of Canada they reside in.

To create a Locale object, use one of the three constructors with a specified language and optional country and variant, as shown in Figure 36.1.

The language should be a valid language code—that is, one of the lowercase two-letter codes defined by ISO-639. For example, zh stands for Chinese, da for Danish, en for English, de for German, and ko for Korean. Table 36.1 lists the language codes.

The country should be a valid ISO country code—that is, one of the uppercase, two-letter codes defined by ISO-3166. For example, CA stands for Canada, CN for China, DK for Denmark, DE for Germany, and US for the United States. Table 36.2 lists the country codes.

The argument variant is rarely used and is needed only for exceptional or system-dependent situations to designate information specific to a browser or vendor. For example, the Norwe- gian language has two sets of spelling rules, a traditional one called bokmål and a new one called nynorsk. The locale for traditional spelling would be created as follows:

new Locale("no", "NO", "B");

Point Key

M36_LIAN1878_11_GE_C36.indd 2 12/29/17 5:50 PM

36.2 The Locale Class 36-3

Code Language Code Language

da Danish ja Japanese

de German ko Korean

el Greek nl Dutch

en English no Norwegian

es Spanish pt Portuguese

fi Finnish sv Swedish

fr French tr Turkish

it Italian zh Chinese

Table 31.1 Common Language Codes

java.util.Locale

+Locale(language: String) +Locale(language: String, country: String) +Locale(language: String, country: String, variant: String)

+getCountry(): String +getLanguage(): String +getVariant(): String +getDefault(): Locale +getDisplayCountry(): String +getDisplayLanguage(): String +getDisplayName(): String

+getDisplayVariant(): String +getAvailableLocales(): Locale[]

Constructs a locale from a language code. Constructs a locale from language and country codes. Constructs a locale from language, country, and variant codes.

Returns the country/region code for this locale. Returns the language code for this locale. Returns the variant code for this locale. Gets the default locale on the machine. Returns the name of the country as expressed in the current locale. Returns the name of the language as expressed in the current locale. Returns the name for the locale. For example, the name is Chinese

(China) for the locale Locale.CHINA. Returns the name for the locale’s variant if it exists. Returns the available locales in an array.

Figure 36.1 The Locale class encapsulates a locale.

Code Country Code Country

AT Austria IE Ireland

BE Belgium HK Hong Kong

CA Canada IT Italy

CH Switzerland JP Japan

CN China KR Korea

DE Germany NL Netherlands

DK Denmark NO Norway

ES Spain PT Portugal

FI Finland SE Sweden

FR France TR Turkey

GB United Kingdom TW Taiwan

GR Greece US United States

Table 31.2 Common Country Codes

For convenience, the Locale class contains many predefined locale constants. Locale.CANADA is for the country Canada and language English; Locale.CANADA_FRENCH is for the country Canada and language French. Several other common constants are:

Locale.US, Locale.UK, Locale.FRANCE, Locale.GERMANY, Locale.ITALY, Locale.CHINA, Locale.KOREA, Locale.JAPAN, and Locale.TAIWAN

M36_LIAN1878_11_GE_C36.indd 3 12/29/17 5:50 PM

36-4 Chapter 36 Internationalization

The Locale class also provides the following constants based on language:

Locale.CHINESE, Locale.ENGLISH, Locale.FRENCH, Locale.GERMAN, Locale.ITALIAN, Locale.JAPANESE, Locale.KOREAN, Locale.SIMPLIFIED_CHINESE, and Locale.TRADITIONAL_CHINESE

Tip You can invoke the static method getAvailableLocales() in the Locale class to obtain all the available locales supported in the system. For example,

Locale[] availableLocales = Calendar.getAvailableLocales();

returns all the locales in an array.

Tip Your machine has a default locale. You may override it by supplying the language and region parameters when you run the program, as follows:

java –Duser.language=zh –Duser.region=CN MainClass

An operation that requires a Locale to perform its task is called locale sensitive. Displaying a number such as a date or time, for example, is a locale-sensitive operation; the number should be formatted according to the customs and conventions of the user’s locale. The sections that follow introduce locale-sensitive operations.

36.2.1 How does Java support international characters in languages like Chinese and Arabic? 36.2.2 How do you construct a Locale object? How do you get all the available locales

from a Calendar object?

36.2.3 How do you create a locale for the French-speaking region of Canada? How do you create a locale for the Netherlands?

36.3 Displaying Date and Time The representation of date and time is dependent on locale.

Applications often need to obtain date and time. Java provides a system-independent encapsu- lation of date and time in the java.util.Date class; it also provides java.util.TimeZone for dealing with time zones, and java.util.Calendar for extracting detailed informa- tion from Date. Different locales have different conventions for displaying date and time. Should the year, month, or day be displayed first? Should slashes, periods, or colons be used to separate fields of the date? What are the names of the months in the language? The java.text.DateFormat class can be used to format date and time in a locale-sensitive way for display to the user. The Date class was introduced in Section 9.6.1, “The Date Class,” and the Calendar class and its subclass GregorianCalendar were introduced in Section 13.4, “Case Study: Calendar and GregorianCalendar.”

36.3.1 The TimeZone Class TimeZone represents a time zone offset and also figures out daylight savings. To get a TimeZone object for a specified time zone ID, use TimeZone.getTimeZone(id). To set a time zone in a Calendar object, use the setTimeZone method with a time zone ID. For exam- ple, cal.setTimeZone(TimeZone.getTimeZone("CST")) sets the time zone to Central Standard Time. To find all the available time zones supported in Java, use the static method getAvailableIDs() in the TimeZone class. In general, the international time zone ID is a string in the form of continent/city like Europe/Berlin, Asia/Taipei, and America/Washington. You can also use the static method getDefault() in the TimeZone class to obtain the default time zone on the host machine.

Point Check

Point Key

M36_LIAN1878_11_GE_C36.indd 4 12/29/17 5:50 PM

36.3 Displaying Date and Time 36-5

36.3.2 The DateFormat Class The DateFormat class can be used to format date and time in a number of styles. The DateFormat class supports several standard formatting styles. To format date and time, simply create an instance of DateFormat using one of the three static methods getDateInstance, getTimeInstance, and getDateTimeInstance and apply the format(Date) method on the instance, as shown in Figure 36.2.

java.text.DateFormat

+format(date: Date): String +getDateInstance(): DateFormat +getDateInstance(dateStyle: int): DateFormat +getDateInstance(dateStyle: int, aLocale: Locale): DateFormat

+getDateTimeInstance(): DateFormat

+getDateTimeInstance(dateStyle: int, timeStyle: int): DateFormat

+getDateTimeInstance(dateStyle: int, timeStyle: int, aLocale: Locale): DateFormat

+getInstance(): DateFormat

Formats a date into a date/time string. Gets the date formatter with the default formatting style for the default locale. Gets the date formatter with the given formatting style for the default locale. Gets the date formatter with the given formatting style for the given locale.

Gets the date and time formatter with the default formatting style for the default locale.

Gets the date and time formatter with the given date and time formatting styles for the default locale.

Gets the date and time formatter with the given formatting styles for the given locale.

Gets a default date and time formatter that uses the SHORT style for both the date and the time.

Figure 36.2 The DateFormat class formats date and time.

The dateStyle and timeStyle are one of the following constants: DateFormat.SHORT, DateFormat.MEDIUM, DateFormat.LONG, DateFormat.FULL. The exact result depends on the locale, but generally,

■■ SHORT is completely numeric, such as 7/24/98 (for date) and 4:49 PM (for time).

■■ MEDIUM is longer, such as 24-Jul-98 (for date) and 4:52:09 PM (for time).

■■ LONG is even longer, such as July 24, 1998 (for date) and 4:53:16 PM EST (for time).

■■ FULL is completely specified, such as Friday, July 24, 1998 (for date) and 4:54:13 o’clock PM EST (for time).

The statements given below display current time with a specified time zone (CST), format- ting style (full date and full time), and locale (US).

GregorianCalendar calendar = new GregorianCalendar(); DateFormat formatter = DateFormat.getDateTimeInstance( DateFormat.FULL, DateFormat.FULL, Locale.US); TimeZone timeZone = TimeZone.getTimeZone("CST"); formatter.setTimeZone(timeZone); System.out.println("The local time is " + formatter.format(calendar.getTime()));

36.3.3 The SimpleDateFormat Class The date and time formatting subclass, SimpleDateFormat, enables you to choose any user- defined pattern for date and time formatting. The constructor shown below can be used to create a SimpleDateFormat object, and the object can be used to convert a Date object into a string with the desired format.

public SimpleDateFormat(String pattern)

M36_LIAN1878_11_GE_C36.indd 5 12/29/17 5:50 PM

36-6 Chapter 36 Internationalization

The parameter pattern is a string consisting of characters with special meanings. For exam- ple, y means year, M means month, d means day of the month, G is for era designator, h means hour, m means minute of the hour, s means second of the minute, and z means time zone. Therefore, the following code will display a string like “Current time is 1997.11.12 AD at 04:10:18 PST” because the pattern is “yyyy.MM.dd G ‘at’ hh:mm:ss z”.

SimpleDateFormat formatter = new SimpleDateFormat("yyyy.MM.dd G 'at' hh:mm:ss z"); date currentTime = new Date(); String dateString = formatter.format(currentTime); System.out.println("Current time is " + dateString);

36.3.4 The DateFormatSymbols Class The DateFormatSymbols class encapsulates localizable date-time formatting data, such as the names of the months and the names of the days of the week, as shown in Figure 36.3.

For example, the following statement displays the month names and weekday names for the default locale:

DateFormatSymbols symbols = new DateFormatSymbols(); String[] monthNames = symbols.getMonths(); for (int i = 0; i < monthNames.length; i++) { System.out.println(monthNames[i]); // Display January, ... }

String[] weekdayNames = symbols.getWeekdays(); for (int i = 0; i < weekdayNames.length; i++) { System.out.println(weekdayNames[i]); // Display Sunday, Monday, ... }

java.text.DateFormatSymbols

+DateFormatSymbols() +DateFormatSymbols(Locale locale) +getAmPmStrings(): String[] +getEras(): String[] +getMonths(): String[] +setMonths(newMonths: String[]): void +getShortMonths(): String[] +setShortMonths(newShortMonths: String[]): void

+getWeekdays(): String[] +setWeekdays(newWeekdays: String[]): void +getShotWeekdays(): String[] +setShortWeekdays(newWeekdays: String[]): void

Constructs a DateFormatSymbols object for the default locale. Constructs a DateFormatSymbols object by for the given locale. Gets AM/PM strings. For example: "AM" and "PM".

Gets era strings. For example: "AD" and "BC".

Gets month strings. For example: "January", "February", etc.

Sets month strings for this locale.

Gets short month strings. For example: "Jan", "Feb", etc.

Sets short month strings for this locale.

Gets weekday strings. For example: "Sunday", "Monday", etc.

Sets weekday strings.

Gets short weekday strings. For example: "Sun", "Mon", etc.

Sets short weekday strings. For example: "Sun", "Mon", etc.

Figure 36.3 The DateFormatSymbols class encapsulates localizable date-time formatting data.

The following two examples demonstrate how to display date, time, and calendar based on locale. The first example creates a clock and displays date and time in locale-sensitive format. The second example displays several different calendars with the names of the days shown in the appropriate local language.

36.3.5 Example: Displaying an International Clock Write a program that displays a clock to show the current time based on the specified locale and time zone. The locale and time zone are selected from the combo boxes that contain the available locales and time zones in the system, as shown in Figure 36.4.

M36_LIAN1878_11_GE_C36.indd 6 12/29/17 5:50 PM

36.3 Displaying Date and Time 36-7

Figure 36.4 The program displays a clock that shows the current time with the specified locale and time zone.

Figure 36.5 WorldClockApp contains WorldClockControl, and WorldClockControl contains WorldClock.

WorldClockApp

+start(primaryStage: Stage) : void +main(args: String[]): void

javafx.application.Application

WorldClockControl

-clock: WorldClock -cboLocales: ComboBox<String> -cboTimeZones: ComboBox<String> -availableLocales: Locale[] -availableTimeZones: String[]

+WorldClockControl() -setAvailableLocales(): void -setAvailableTimeZones(): void

WorldClock

-timeZone: TimeZone -locale: Locale -clock: ClockPane -lblDigitTime: Label

+WorldClock() +setTimeZone(timeZone:

TimeZone): void +setLocale(locale: Locale): void -setCurrentTime(): void

1 11 1

javafx.scene.layout.BorderPanejavafx.scene.layout.BorderPane

Here are the major steps in the program:

1. Create a subclass of BorderPane named WorldClock (see Listing 36.1) to contain an instance of the ClockPane class (developed in Listing 14.21, ClockPane.java), and place it in the center. Create a Label to display the digit time, and place it in the bottom. Use the GregorianCalendar class to obtain the current time for a specific locale and time zone.

2. Create a subclass of BorderPanel named WorldClockControl (see Listing 36.2) to contain an instance of WorldClock and two instances of ComboBox for selecting locales and time zones.

3. Create an application named WorldClockApp (see Listing 36.3) to display an instance of WorldClockControl.

The relationship among these classes is shown in Figure 36.5.

lisTing 36.1 WorldClock.java 1 import java.util.Calendar; 2 import java.util.TimeZone; 3 import java.util.GregorianCalendar; 4 import java.text.*; 5 import java.util.Locale; 6 import javafx.animation.KeyFrame; 7 import javafx.animation.Timeline; 8 import javafx.event.ActionEvent; 9 import javafx.event.EventHandler; 10 import javafx.geometry.Pos; 11 import javafx.scene.control.Label;

M36_LIAN1878_11_GE_C36.indd 7 12/29/17 5:50 PM

36-8 Chapter 36 Internationalization

12 import javafx.scene.layout.BorderPane; 13 import javafx.util.Duration; 14 15 public class WorldClock extends BorderPane { 16 private TimeZone timeZone = TimeZone.getTimeZone("EST"); 17 private Locale locale = Locale.getDefault(); 18 private ClockPane clock = new ClockPane(); // Still clock 19 private Label lblDigitTime = new Label(); 20 21 public WorldClock() { 22 setCenter(clock); 23 setBottom(lblDigitTime); 24 BorderPane.setAlignment(lblDigitTime, Pos.CENTER); 25 26 EventHandler<ActionEvent> eventHandler = e -> { 27 setCurrentTime(); // Set a new clock time 28 }; 29 30 // Create an animation for a running clock 31 Timeline animation = new Timeline( 32 new KeyFrame(Duration.millis(1000), eventHandler)); 33 animation.setCycleCount(Timeline.INDEFINITE); 34 animation.play(); // Start animation 35 36 // Resize the clock 37 widthProperty().addListener(ov -> clock.setWidth(getWidth())); 38 heightProperty().addListener(ov -> clock.setHeight(getHeight())); 39 } 40 41 public void setTimeZone(TimeZone timeZone) { 42 this.timeZone = timeZone; 43 } 44 45 public void setLocale(Locale locale) { 46 this.locale = locale; 47 } 48 49 private void setCurrentTime() { 50 Calendar calendar = new GregorianCalendar(timeZone, locale); 51 clock.setHour(calendar.get(Calendar.HOUR)); 52 clock.setMinute(calendar.get(Calendar.MINUTE)); 53 clock.setSecond(calendar.get(Calendar.SECOND)); 54 55 // Display digit time on the label 56 DateFormat formatter = DateFormat.getDateTimeInstance 57 (DateFormat.MEDIUM, DateFormat.LONG, locale); 58 formatter.setTimeZone(timeZone); 59 lblDigitTime.setText(formatter.format(calendar.getTime())); 60 } 61 }

lisTing 36.2 WorldClockControl.java 1 import java.util.*; 2 import javafx.geometry.Pos; 3 import javafx.scene.control.ComboBox; 4 import javafx.scene.control.Label; 5 import javafx.scene.layout.BorderPane; 6 import javafx.scene.layout.GridPane; 7 8 public class WorldClockControl extends BorderPane {

M36_LIAN1878_11_GE_C36.indd 8 12/29/17 5:50 PM

36.3 Displaying Date and Time 36-9

9 // Obtain all available locales and time zone ids 10 private Locale[] availableLocales = Locale.getAvailableLocales(); 11 private String[] availableTimeZones = TimeZone.getAvailableIDs(); 12 13 // Comboxes to display available locales and time zones 14 private ComboBox<String> cboLocales = new ComboBox<>(); 15 private ComboBox<String> cboTimeZones = new ComboBox<>(); 16 17 // Create a clock 18 private WorldClock clock = new WorldClock(); 19 20 public WorldClockControl() { 21 // Initialize cboLocales with all available locales 22 setAvailableLocales(); 23 24 // Initialize cboTimeZones with all available time zones 25 setAvailableTimeZones(); 26 27 // Initialize locale and time zone 28 clock.setLocale( 29 availableLocales[cboLocales.getSelectionModel() 30 .getSelectedIndex()]); 31 clock.setTimeZone(TimeZone.getTimeZone( 32 availableTimeZones[cboTimeZones.getSelectionModel() 33 .getSelectedIndex()])); 34 35 GridPane pane = new GridPane(); 36 pane.setHgap(5); 37 pane.add(new Label("Locale"), 0, 0); 38 pane.add(new Label("Time Zone"), 0, 1); 39 pane.add(cboLocales, 1, 0); 40 pane.add(cboTimeZones, 1, 1); 41 42 setTop(pane); 43 setCenter(clock); 44 BorderPane.setAlignment(pane, Pos.CENTER); 45 BorderPane.setAlignment(clock, Pos.CENTER); 46 47 cboLocales.setOnAction(e -> 48 clock.setLocale(availableLocales[cboLocales. 49 getSelectionModel().getSelectedIndex()])); 50 cboTimeZones.setOnAction(e -> 51 clock.setTimeZone(TimeZone.getTimeZone( 52 availableTimeZones[cboTimeZones. 53 getSelectionModel().getSelectedIndex()]))); 54 } 55 56 private void setAvailableLocales() { 57 for (int i = 0; i < availableLocales.length; i++) 58 cboLocales.getItems().add(availableLocales[i] 59 .getDisplayName() + " " + availableLocales[i].toString()); 60 61 cboLocales.getSelectionModel().selectFirst(); 62 } 63 64 private void setAvailableTimeZones() { 65 // Sort time zones 66 Arrays.sort(availableTimeZones); 67 for (int i = 0; i < availableTimeZones.length; i++) { 68 cboTimeZones.getItems().add(availableTimeZones[i]); 69 }

M36_LIAN1878_11_GE_C36.indd 9 12/29/17 5:50 PM

36-10 Chapter 36 Internationalization

70 cboTimeZones.getSelectionModel().selectFirst(); 71 } 72 }

lisTing 36.3 WorldClockApp.java 1 import javafx.application.Application; 2 import javafx.scene.Scene; 3 import javafx.stage.Stage; 4 5 public class WorldClockApp extends Application { 6 @Override // Override the start method in the Application class 7 public void start(Stage primaryStage) { 8 // Create a scene and place it in the stage 9 Scene scene = new Scene(new WorldClockControl(), 450, 350); 10 primaryStage.setTitle("WorldClockApp"); // Set the stage title 11 primaryStage.setScene(scene); // Place the scene in the stage 12 primaryStage.show(); // Display the stage 13 } 14 }

The WorldClock class creates an instance of ClockPane (line 18) and places it in the center of the border pane (line 22). The setCurrentTime() method uses GregorianCalendar to obtain a Calendar object for the specified locale and time zone (line 50). The clock time is updated every one second using the current Calendar object in lines 51–53.

An instance of DateFormat is created (lines 56–57) and is used to format the date in accordance with the locale (line 59).

The WorldClockControl class contains an instance of WorldClock and two combo boxes. The combo boxes store all the available locales and time zones (lines 56–71). The newly selected locale and time zone are set in the clock (lines 47–53) and used to display a new time based on the current locale and time zone.

36.3.6 Example: Displaying a Calendar Write a program that displays a calendar based on the specified locale, as shown in Figure 36.6. The user can specify a locale from a combo box that consists of a list of all the available locales supported by the system. When the program starts, the calendar for the current month of the year is displayed. The user can use the Prior and Next buttons to browse the calendar.

Figure 36.6 The calendar program displays a calendar with a specified locale.

Here are the major steps in the program:

1. Define a subclass of BorderPane named CalendarPane (see Listing 36.4) to display the calendar for the given year and month based on the specified locale.

M36_LIAN1878_11_GE_C36.indd 10 12/29/17 5:50 PM

36.3 Displaying Date and Time 36-11

lisTing 36.4 CalendarPane.java 1 import java.text.DateFormatSymbols; 2 import java.text.SimpleDateFormat; 3 import java.util.Calendar; 4 import java.util.GregorianCalendar; 5 import java.util.Locale; 6 import javafx.geometry.Pos; 7 import javafx.scene.control.Label; 8 import javafx.scene.layout.BorderPane; 9 import javafx.scene.layout.GridPane; 10 import javafx.scene.paint.Color; 11 import javafx.scene.text.TextAlignment; 12 13 public class CalendarPane extends BorderPane { 14 // The header label 15 private Label lblHeader = new Label(); 16 17 // Maximum number of labels to display day names and days 18 private Label[] lblDay = new Label[49]; 19 20 private Calendar calendar; 21 private int month; // The specified month 22 private int year; // The specified year 23 private Locale locale = Locale.CHINA; 24 25 public CalendarPane() { 26 // Create labels for displaying days 27 for (int i = 0; i < 49; i++) { 28 lblDay[i] = new Label(); 29 lblDay[i].setTextAlignment(TextAlignment.RIGHT);

Figure 36.7 CalendarApp contains CalendarPane.

CalendarPane

-month: int -year: int -calendar: java.util.Calendar -locale: Locale

+getMonth(): int +setMonth(newMonth: int): void +getYear(): int +setYear(newYear: int): void +setLocale(newLocale: Locale): void +showHeader(): void +showDayNames(): void +showDays(): void

CalendarApp

-calendarPane: CalendarPane -cboLocale: ComboBox<String> -btPrior: Button -btNext: Button -locales: java.utilLocale[]

+start(primaryStage: Stage): void +main(args: String[]): void

javafx.scene.layout.BorderPane javafx.application.Application

1 1

2. Define an application named CalendarApp (Listing 36.5). Create a pane to hold an instance of CalendarPane in the center, two buttons, Prior and Next in the bottom, and a combo box in the top of the pane. The relationships among these classes are shown in Figure 36.7.

M36_LIAN1878_11_GE_C36.indd 11 12/29/17 5:50 PM

36-12 Chapter 36 Internationalization

30 } 31 32 showDayNames(); // Display day names for the locale 33 34 GridPane dayPane = new GridPane(); 35 dayPane.setAlignment(Pos.CENTER); 36 37 dayPane.setHgap(10); 38 dayPane.setVgap(10); 39 for (int i = 0; i < 49; i++) { 40 dayPane.add(lblDay[i], i % 7, i / 7); 41 } 42 43 // Place header and calendar body in the pane 44 this.setTop(lblHeader); 45 BorderPane.setAlignment(lblHeader, Pos.CENTER); 46 this.setCenter(dayPane); 47 48 // Set current month and year 49 calendar = new GregorianCalendar(); 50 month = calendar.get(Calendar.MONTH); 51 year = calendar.get(Calendar.YEAR); 52 updateCalendar(); 53 54 // Show calendar 55 showHeader(); 56 showDays(); 57 } 58 59 /** Update the day names based on locale */ 60 private void showDayNames() { 61 DateFormatSymbols dfs = new DateFormatSymbols(locale); 62 String dayNames[] = dfs.getWeekdays(); 63 64 // jlblDay[0], jlblDay[1], ..., jlblDay[6] for day names 65 for (int i = 0; i < 7; i++) { 66 lblDay[i].setText(dayNames[i + 1]); 67 } 68 } 69 70 /** Update the header based on locale */ 71 private void showHeader() { 72 SimpleDateFormat sdf = 73 new SimpleDateFormat("MMMM yyyy", locale); 74 String header = sdf.format(calendar.getTime()); 75 lblHeader.setText(header); 76 } 77 78 public void showDays() { 79 // Get the day of the first day in a month 80 int startingDayOfMonth = calendar.get(Calendar.DAY_OF_WEEK); 81 82 // Fill the calendar with the days before this month 83 Calendar cloneCalendar = (Calendar) calendar.clone(); 84 cloneCalendar.add(Calendar.DATE, -1); // Becomes preceding month 85 int daysInPrecedingMonth = cloneCalendar.getActualMaximum( 86 Calendar.DAY_OF_MONTH); 87 88 for (int i = 0; i < startingDayOfMonth - 1; i++) { 89 lblDay[i + 7].setTextFill(Color.LIGHTGRAY);

M36_LIAN1878_11_GE_C36.indd 12 12/29/17 5:50 PM

36.3 Displaying Date and Time 36-13

90 lblDay[i + 7].setText(daysInPrecedingMonth 91 - startingDayOfMonth + 2 + i + ""); 92 } 93 94 // Display days of this month 95 int daysInCurrentMonth = calendar.getActualMaximum( 96 Calendar.DAY_OF_MONTH); 97 for (int i = 1; i <= daysInCurrentMonth; i++) { 98 lblDay[i - 2 + startingDayOfMonth + 7].setTextFill(Color.BLACK); 99 lblDay[i - 2 + startingDayOfMonth + 7].setText(i + ""); 100 } 101 102 // Fill the calendar with the days after this month 103 int j = 1; 104 for (int i = daysInCurrentMonth - 1 + startingDayOfMonth + 7; 105 i < 49; i++) { 106 lblDay[i].setTextFill(Color.LIGHTGRAY); 107 lblDay[i].setText(j++ + ""); 108 } 109 } 110 111 /** Set the calendar to the first day of the 112 * specified month and year 113 */ 114 public void updateCalendar() { 115 calendar.set(Calendar.YEAR, year); 116 calendar.set(Calendar.MONTH, month); 117 calendar.set(Calendar.DATE, 1); 118 } 119 120 public int getMonth() { 121 return month; 122 } 123 124 public void setMonth(int newMonth) { 125 month = newMonth; 126 updateCalendar(); 127 showHeader(); 128 showDays(); 129 } 130 131 public int getYear() { 132 return year; 133 } 134 135 public void setYear(int newYear) { 136 year = newYear; 137 updateCalendar(); 138 showHeader(); 139 showDays(); 140 } 141 142 public void setLocale(Locale locale) { 143 this.locale = locale; 144 updateCalendar(); 145 showDayNames(); 146 showHeader(); 147 showDays(); 148 } 149 }

M36_LIAN1878_11_GE_C36.indd 13 12/29/17 5:50 PM

36-14 Chapter 36 Internationalization

CalendarPane is created to control and display the calendar. It displays the month and year in the header, and the day names and days in the calendar body. The header and day names are locale sensitive.

The showHeader method (lines 71–76) displays the calendar title in a form like “MMMM yyyy”. The SimpleDateFormat class used in the showHeader method is a subclass of DateFormat. SimpleDateFormat allows you to customize the date format to display the date in various nonstandard styles.

The showDayNames method (lines 60–68) displays the day names in the calendar. The DateFormatSymbols class used in the showDayNames method is a class for encapsulating localizable date-time formatting data, such as the names of the months, the names of the days of the week, and the time-zone data. The getWeekdays method is used to get an array of day names.

The showDays method (lines 60–68) displays the days for the specified month of the year. As you can see in Figure 36.6, the labels before the current month are filled with the last few days of the preceding month, and the labels after the current month are filled with the first few days of the next month.

To fill the calendar with the days before the current month, a clone of calendar, named cloneCalendar, is created to obtain the days for the preceding month (line 83). cloneCalendar is a copy of calendar with separate memory space. Thus you can change the properties of cloneCalendar without corrupting the calendar object. The clone() method is defined in the Object class, which was introduced in Section 13.7, “The Cloneable Interface.” You can clone any object as long as its defining class imple- ments the Cloneable interface. The Calendar class implements Cloneable.

The cloneCalendar.getActualMaximum(Calendar.DAY_OF_MONTH) method (lines 95–96) returns the number of days in the month for the specified calendar.

lisTing 36.5 CalendarApp.java 1 import java.util.Locale; 2 import javafx.application.Application; 3 import javafx.geometry.Pos; 4 import javafx.scene.Scene; 5 import javafx.scene.control.Button; 6 import javafx.scene.control.ComboBox; 7 import javafx.scene.control.Label; 8 import javafx.scene.layout.BorderPane; 9 import javafx.scene.layout.HBox; 10 import javafx.stage.Stage; 11 12 public class CalendarApp extends Application { 13 private CalendarPane calendarPane = new CalendarPane(); 14 private Button btPrior = new Button("Prior"); 15 private Button btNext = new Button("Next"); 16 private ComboBox<String> cboLocales = new ComboBox<>(); 17 private Locale[] availableLocales = Locale.getAvailableLocales(); 18 19 @Override // Override the start method in the Application class 20 public void start(Stage primaryStage) { 21 HBox hBox = new HBox(5); 22 hBox.getChildren().addAll(btPrior, btNext); 23 hBox.setAlignment(Pos.CENTER); 24 25 // Initialize cboLocales with all available locales 26 setAvailableLocales(); 27 HBox hBoxLocale = new HBox(5); 28 hBoxLocale.getChildren().addAll( 29 new Label("Select a locale"), cboLocales); 30 31 BorderPane pane = new BorderPane();

M36_LIAN1878_11_GE_C36.indd 14 12/29/17 5:50 PM

36.3 Displaying Date and Time 36-15

32 pane.setCenter(calendarPane); 33 pane.setTop(hBoxLocale); 34 hBoxLocale.setAlignment(Pos.CENTER); 35 pane.setBottom(hBox); 36 hBox.setAlignment(Pos.CENTER); 37 38 // Create a scene and place it in the stage 39 Scene scene = new Scene(pane, 600, 300); 40 primaryStage.setTitle("CalendarApp"); // Set the stage title 41 primaryStage.setScene(scene); // Place the scene in the stage 42 primaryStage.show(); // Display the stage 43 44 btPrior.setOnAction(e -> { 45 int currentMonth = calendarPane.getMonth(); 46 if (currentMonth == 0) { // The previous month is 11 for Dec 47 calendarPane.setYear(calendarPane.getYear() - 1); 48 calendarPane.setMonth(11); 49 } 50 else { 51 calendarPane.setMonth((currentMonth - 1) % 12); 52 } 53 }); 54 55 btNext.setOnAction(e -> { 56 int currentMonth = calendarPane.getMonth(); 57 if (currentMonth == 11) // The next month is 0 for Jan 58 calendarPane.setYear(calendarPane.getYear() + 1); 59 60 calendarPane.setMonth((currentMonth + 1) % 12); 61 }); 62 63 cboLocales.setOnAction(e -> 64 calendarPane.setLocale(availableLocales[cboLocales. 65 getSelectionModel().getSelectedIndex()])); 66 } 67 68 private void setAvailableLocales() { 69 for (int i = 0; i < availableLocales.length; i++) 70 cboLocales.getItems().add(availableLocales[i] 71 .getDisplayName() + " " + availableLocales[i].toString()); 72 73 cboLocales.getSelectionModel().selectFirst(); 74 } 75 }

CalendarApp creates the user interface and handles the button actions and combo box item selections for locales. The Locale.getAvailableLocales() method (line 17) is used to find all the available locales that have calendars. Its getDisplayName() method returns the name of each locale and adds the name to the combo box (lines 70–71). When the user selects a locale name in the combo box, a new locale is passed to calendarPane, and a new calendar is displayed based on the new locale (lines 63–65).

36.3.1 How do you set the time zone “PST” for a Calendar object? 36.3.2 How do you display current date and time in German? 36.3.3 How do you use the SimpleDateFormat class to display date and time using the

pattern “yyyy.MM.dd hh:mm:ss”?

36.3.4 In line 66 of Listing 36.2, WorldClockControl.java, Arrays. sort(availableTimeZones) is used to sort the available time zones. What happens if you attempt to sort the available locales using Arrays.sort(availa bleLocales)?

Point Check

M36_LIAN1878_11_GE_C36.indd 15 12/29/17 5:50 PM

36-16 Chapter 36 Internationalization

36.4 Formatting Numbers You can format numbers based on locales.

Formatting numbers is highly locale dependent. For example, number 5000.555 is displayed as 5,000.555 in the United States, but as 5 000,555 in France and as 5.000,555 in Germany.

Numbers are formatted using the java.text.NumberFormat class, an abstract base class that provides the methods for formatting and parsing numbers, as shown in Figure 36.8.

Point Key

java.text.NumberFormat

+getInstance(): NumberFormat +getInstance(locale: Locale): NumberFormat +getIntegerInstance(): NumberFormat +getIntegerInstance(locale: Locale): NumberFormat +getCurrencyInstance(): NumberFormat +getNumberInstance(): NumberFormat

+getNumberInstance(locale: Locale): NumberFormat +getPercentInstance(): NumberFormat

+getPercentInstance(locale: Locale): NumberFormat +format (number: double): String

+format (number: long): String

+getMaximumFractionDigits(): int

+setMaximumFractionDigits(newValue: int): void

+getMinimumFractionDigits(): int

+setMinimumFractionDigits(newValue: int): void

+getMaximumIntegerDigits(): int

+setMaximumIntegerDigits(newValue: int): void

+getMinimumIntegerDigits(): int

+setMinimumIntegerDigits(newValue: int): void

+isGroupingUsed(): boolean

+setGroupingUsed(newValue: boolean): void +parse(source: String): Number +getAvailableLocales(): Locale[]

Returns a default number format for the default locale. Returns a default number format for the specified locale. Returns an integer number format for the default locale. Returns an integer number format for the specified locale.

Returns a currency format for the current default locale. Same as getInstance().

Same as getInstance(locale).

Returns a percentage format for the default locale.

Returns a percentage format for the specified locale.

Formats a floating-point number.

Formats an integer.

Returns the maximum number of allowed fraction digits.

Sets the maximum number of allowed fraction digits.

Returns the minimum number of allowed fraction digits.

Sets the minimum number of allowed fraction digits.

Returns the maximum number of allowed integer digits in a fraction number.

Sets the maximum number of allowed integer digits in a fraction number.

Returns the minimum number of allowed integer digits in a fraction number.

Sets the minimum number of allowed integer digits in a fraction number.

Returns true if grouping is used in this format. For example, in the English locale, with grouping on, the number 1234567 is formatted as "1,234,567".

Sets whether or not grouping will be used in this format. Parses string into a number. Gets the set of locales for which NumberFormats are installed.

Figure 36.8 The NumberFormat class provides the methods for formatting and parsing numbers.

With NumberFormat, you can format and parse numbers for any locale. Your code will be completely independent of locale conventions for decimal points, thousands-separators, currency format, and percentage formats.

36.4.1 Plain Number Format You can get an instance of NumberFormat for the current locale using Number Format. get Instance() or NumberFormat.getNumberInstance and for the specified locale using NumberFormat.getInstance(Locale) or NumberFormat.getNumberInstance(Locale).

M36_LIAN1878_11_GE_C36.indd 16 12/29/17 5:50 PM

36.4 Formatting Numbers 36-17

You can then invoke format(number) on the NumberFormat instance to return a formatted number as a string.

For example, to display number 5000.555 in France, use the following code:

NumberFormat numberFormat = NumberFormat.getInstance(Locale.FRANCE); System.out.println(numberFormat.format(5000.555));

You can control the display of numbers with such methods as setMaximumFraction Digits and setMinimumFractionDigits. For example, 5000.555 will be displayed as 5000.6 if you use numberFormat.setMaximumFractionDigits(1).

36.4.2 Currency Format To format a number as a currency value, use NumberFormat.getCurrency- Instance() to get the currency number format for the current locale or NumberFormat. getCurrencyInstance(Locale) to get the currency number for the specified locale.

For example, to display number 5000.555 as currency in the United States, use the fol- lowing code:

NumberFormat currencyFormat = NumberFormat.getCurrencyInstance(Locale.US); System.out.println(currencyFormat.format(5000.555));

5000.555 is formatted into $5,000,56. If the locale is set to France, the number will be format- ted into 5,000,56 €.

36.4.3 Percent Format To format a number in a percent, use NumberFormat.getPercentInstance() or NumberFormat.getPercentInstance(Locale) to get the percent number format for the current locale or the specified locale.

For example, to display number 0.555367 as a percent in the United States, use the follow- ing code:

NumberFormat percentFormat = NumberFormat.getPercentInstance(Locale.US); System.out.println(percentFormat.format(0.555367));

0.555367 is formatted into 56%. By default, the format truncates the fraction part in a percent number. If you want to keep three digits after the decimal point, use percentFormat. setMinimumFractionDigits(3). So 0.555367 would be displayed as 55.537%.

36.4.4 Parsing Numbers You can format a number into a string using the format(numericalValue) method. You can also use the parse(String) method to convert a formatted plain number, currency value, or percent number with the conventions of a certain locale into an instance of java.lang. Number. The parse method throws a java.text.ParseException if parsing fails. For example, U.S. $5,000.56 can be parsed into a number using the following statements:

NumberFormat currencyFormat = NumberFormat.getCurrencyInstance(Locale.US); try { Number number = currencyFormat.parse("$5,000.56"); System.out.println(number.doubleValue()); } catch (java.text.ParseException ex) { System.out.println("Parse failed"); }

M36_LIAN1878_11_GE_C36.indd 17 12/29/17 5:50 PM

36-18 Chapter 36 Internationalization

36.4.5 The DecimalFormat Class If you want even more control over the format or parsing, cast the NumberFormat you get from the factory methods to a java.text.DecimalFormat, which is a subclass of NumberFormat. You can then use the applyPattern(String pattern) method of the DecimalFormat class to specify the patterns for displaying the number.

A pattern can specify the minimum number of digits before the decimal point and the maxi- mum number of digits after the decimal point. The characters '0' and '#' are used to specify a required digit and an optional digit, respectively. The optional digit is not displayed if it is zero. For example, the pattern "00.0##" indicates minimum two digits before the decimal point and maximum three digits after the decimal point. If there are more actual digits before the decimal point, all of them are displayed. If there are more than three digits after the decimal point, the number of digits is rounded. Applying the pattern "00.0##", number 111.2226 is formatted to 111.223, number 1111.2226 to 1111.223, number 1.22 to 01.22, and number 1 to 01.0. Here is the code:

NumberFormat numberFormat = NumberFormat.getInstance(Locale.US); DecimalFormat decimalFormat = (DecimalFormat)numberFormat; decimalFormat.applyPattern("00.0##"); System.out.println(decimalFormat.format(111.2226)); System.out.println(decimalFormat.format(1111.2226)); System.out.println(decimalFormat.format(1.22)); System.out.println(decimalFormat.format(1));

The character '%' can be put at the end of a pattern to indicate that a number is formatted as a percentage. This causes the number to be multiplied by 100 and appends a percent sign %.

36.4.5 Example: Formatting Numbers Create a loan calculator for computing loans. The calculator allows the user to choose locales, and displays numbers in accordance with locale-sensitive format. As shown in Figure 36.9, the user enters interest rate, number of years, and loan amount, then clicks the Compute button to display the interest rate in percentage format, the number of years in normal number format, and the loan amount, total payment, and monthly payment in currency format. Listing 36.6 gives the solution to the problem.

lisTing 36.6 NumberFormatDemo.java 1 import java.util.*; 2 import java.text.NumberFormat; 3 import javafx.application.Application; 4 import javafx.geometry.Pos; 5 import javafx.scene.Scene; 6 import javafx.scene.control.Button; 7 import javafx.scene.control.ComboBox; 8 import javafx.scene.control.Label; 9 import javafx.scene.control.TextField; 10 import javafx.scene.layout.GridPane; 11 import javafx.scene.layout.HBox; 12 import javafx.scene.layout.VBox; 13 import javafx.stage.Stage; 14 15 public class NumberFormatDemo extends Application { 16 // Combo box for selecting available locales 17 private ComboBox<String> cboLocale = new ComboBox<>(); 18 19 // Text fields for interest rate, year, and loan amount 20 private TextField tfInterestRate = new TextField("6.75"); 21 private TextField tfNumberOfYears = new TextField("15"); 22 private TextField tfLoanAmount = new TextField("107000");

M36_LIAN1878_11_GE_C36.indd 18 12/29/17 5:50 PM

36.4 Formatting Numbers 36-19

23 private TextField tfFormattedInterestRate = new TextField(); 24 private TextField tfFormattedNumberOfYears = new TextField(); 25 private TextField tfFormattedLoanAmount = new TextField(); 26 27 // Text fields for monthly payment and total payment 28 private TextField tfTotalPayment = new TextField(); 29 private TextField tfMonthlyPayment = new TextField(); 30 31 // Compute button 32 private Button btCompute = new Button("Compute"); 33 34 // Current locale 35 private Locale locale = Locale.getDefault(); 36 37 // Declare locales to store available locales 38 private Locale locales[] = Calendar.getAvailableLocales(); 39 40 /** Initialize the combo box */ 41 public void initializeComboBox() { 42 // Add locale names to the combo box 43 for (int i = 0; i < locales.length; i++) 44 cboLocale.getItems().add(locales[i].getDisplayName()); 45 } 46 47 @Override // Override the start method in the Application class 48 public void start(Stage primaryStage) { 49 initializeComboBox(); 50 51 // Pane to hold the combo box for selecting locales 52 HBox hBox = new HBox(5); 53 hBox.getChildren().addAll( 54 new Label("Choose a Locale"), cboLocale); 55 56 // Pane to hold the input 57 GridPane gridPane = new GridPane(); 58 gridPane.add(new Label("Interest Rate"), 0, 0); 59 gridPane.add(tfInterestRate, 1, 0); 60 gridPane.add(tfFormattedInterestRate, 2, 0); 61 gridPane.add(new Label("Number of Years"), 0, 1); 62 gridPane.add(tfNumberOfYears, 1, 1); 63 gridPane.add(tfFormattedNumberOfYears, 2, 1); 64 gridPane.add(new Label("Loan Amount"), 0, 2); 65 gridPane.add(tfLoanAmount, 1, 2); 66 gridPane.add(tfFormattedLoanAmount, 2, 2); 67 68 // Pane to hold the output 69 GridPane gridPaneOutput = new GridPane(); 70 gridPaneOutput.add(new Label("Monthly Payment"), 0, 0); 71 gridPaneOutput.add(tfMonthlyPayment, 1, 0); 72 gridPaneOutput.add(new Label("Total Payment"), 0, 1); 73 gridPaneOutput.add(tfTotalPayment, 1, 1); 74 75 // Set text field alignment 76 tfFormattedInterestRate.setAlignment(Pos.BASELINE_RIGHT); 77 tfFormattedNumberOfYears.setAlignment(Pos.BASELINE_RIGHT); 78 tfFormattedLoanAmount.setAlignment(Pos.BASELINE_RIGHT); 79 tfTotalPayment.setAlignment(Pos.BASELINE_RIGHT); 80 tfMonthlyPayment.setAlignment(Pos.BASELINE_RIGHT); 81 82 // Set editable false 83 tfFormattedInterestRate.setEditable(false);

M36_LIAN1878_11_GE_C36.indd 19 12/29/17 5:50 PM

36-20 Chapter 36 Internationalization

84 tfFormattedNumberOfYears.setEditable(false); 85 tfFormattedLoanAmount.setEditable(false); 86 tfTotalPayment.setEditable(false); 87 tfMonthlyPayment.setEditable(false); 88 89 VBox vBox = new VBox(5); 90 vBox.getChildren().addAll(hBox, 91 new Label("Enter Annual Interest Rate, " + 92 "Number of Years, and Loan Amount"), gridPane, 93 new Label("Payment"), gridPaneOutput, btCompute); 94 95 // Create a scene and place it in the stage 96 Scene scene = new Scene(vBox, 400, 300); 97 primaryStage.setTitle("NumberFormatDemo"); // Set the stage title 98 primaryStage.setScene(scene); // Place the scene in the stage 99 primaryStage.show(); // Display the stage 100 101 // Register listeners 102 cboLocale.setOnAction(e -> { 103 locale = locales[cboLocale 104 .getSelectionModel().getSelectedIndex()]; 105 computeLoan(); 106 }); 107 108 btCompute.setOnAction(e -> computeLoan()); 109 } 110 111 /** Compute payments and display results locale-sensitive format */ 112 private void computeLoan() { 113 // Retrieve input from user 114 double loan = new Double(tfLoanAmount.getText()).doubleValue(); 115 double interestRate = 116 new Double(tfInterestRate.getText()).doubleValue() / 1240; 117 int numberOfYears = 118 new Integer(tfNumberOfYears.getText()).intValue(); 119 120 // Calculate payments 121 double monthlyPayment = loan * interestRate/ 122 (1 - (Math.pow(1 / (1 + interestRate), numberOfYears * 12))); 123 double totalPayment = monthlyPayment * numberOfYears * 12; 124 125 // Get formatters 126 NumberFormat percentFormatter = 127 NumberFormat.getPercentInstance(locale); 128 NumberFormat currencyForm = 129 NumberFormat.getCurrencyInstance(locale); 130 NumberFormat numberForm = NumberFormat.getNumberInstance(locale); 131 percentFormatter.setMinimumFractionDigits(2); 132 133 // Display formatted input 134 tfFormattedInterestRate.setText( 135 percentFormatter.format(interestRate * 12)); 136 tfFormattedNumberOfYears.setText 137 (numberForm.format(numberOfYears)); 138 tfFormattedLoanAmount.setText(currencyForm.format(loan)); 139 140 // Display results in currency format 141 tfMonthlyPayment.setText(currencyForm.format(monthlyPayment)); 142 tfTotalPayment.setText(currencyForm.format(totalPayment)); 143 } 144 }

M36_LIAN1878_11_GE_C36.indd 20 12/29/17 5:50 PM

36.5 Resource Bundles 36-21

Figure 36.9 The locale determines the format of the numbers displayed in the loan calculator.

The computeLoan method (lines 112–143) gets the input on interest rate, number of years, and loan amount from the user, computes monthly payment and total payment, and displays annual interest rate in percentage format, number of years in normal number format, and loan amount, monthly payment, and total payment in locale-sensitive format.

The statement percentFormatter.setMinimumFractionDigits(2) (line 131) sets the minimum number of fractional parts to 2. Without this statement, 0.075 would be dis- played as 7% rather than 7.5%.

36.4.1 Write the code to format number 12345.678 in the United Kingdom locale. Keep two digits after the decimal point.

36.4.2 Write the code to format number 12345.678 in U.S. currency. 36.4.3 Write the code to format number 0.345678 as percentage with at least three digits

after the decimal point.

36.4.4 Write the code to parse 3,456.78 into a number. 36.4.5 Write the code that uses the DecimalFormat class to format number 12345.678

using the pattern “0.0000#”.

36.5 Resource Bundles You can use resource bundles to customize locale-sensitive information.

The NumberFormatDemo in the preceding example displays the numbers, currencies, and percentages in local customs, but displays all the message strings, titles, and button labels in English. In this section, you will learn how to use resource bundles to localize message strings, titles, button labels, and so on.

A resource bundle is a Java class file or text file that provides locale-specific informa- tion. This information can be accessed by Java programs dynamically. When a locale- specific resource is needed—a message string, for example—your program can load it from the resource bundle appropriate for the desired locale. In this way, you can write program code that is largely independent of the user’s locale, isolating most, if not all, of the locale-specific information in resource bundles.

With resource bundles, you can write programs that separate the locale-sensitive part of your code from the locale-independent part. The programs can easily handle multiple locales, and can easily be modified later to support even more locales.

The resources are placed inside the classes that extend the ResourceBundle class or a subclass of ResourceBundle. Resource bundles contain key/value pairs. Each key uniquely identifies a locale-specific object in the bundle. You can use the key to retrieve the object. ListResourceBundle is a convenient subclass of ResourceBundle that is often used to

Point Check

Point Key

M36_LIAN1878_11_GE_C36.indd 21 12/29/17 5:50 PM

36-22 Chapter 36 Internationalization

simplify the creation of resource bundles. Here is an example of a resource bundle that contains four keys using ListResourceBundle:

// MyResource.java: resource file public class MyResource extends java.util.ListResourceBundle { static final Object[][] contents = { {"nationalFlag", "us.gif"}, {"nationalAnthem", "us.au"}, {"nationalColor", Color.red}, {"annualGrowthRate", new Double(7.8)} }; public Object[][] getContents() { return contents; } }

Keys are case-sensitive strings. In this example, the keys are nationalFlag, national- Anthem, nationalColor, and annualGrowthRate. The values can be any type of Object.

If all the resources are strings, they can be placed in a convenient text file with the extension .properties. A typical property file would look like this:

#Wed Jul 01 07:23:24 EST 1998 nationalFlag=us.gif nationalAnthem=us.au

To retrieve values from a ResourceBundle in a program, you first need to create an instance of ResourceBundle using one of the following two static methods:

public static final ResourceBundle getBundle(String baseName) throws MissingResourceException

public static final ResourceBundle getBundle (String baseName, Locale locale) throws MissingResourceException

The first method returns a ResourceBundle for the default locale, and the second method returns a ResourceBundle for the specified locale. baseName is the base name for a set of classes, each of which describes the information for a given locale. These classes are named in Table 36.3.

For example, MyResource_en_BR.class stores resources specific to the United Kingdom, MyResource_en_US.class stores resources specific to the United States, and MyResource_ en.class stores resources specific to all the English-speaking countries.

1. BaseName_language_country_variant.class

2. BaseName_language_country.class

3. BaseName_language.class

4. BaseName.class

5. BaseName_language_country_variant.properties

6. BaseName_language_country.properties

7. BaseName_language.properties

8. BaseName.properties

Table 36.3 Resource Bundle Naming Conventions

The getBundle method attempts to load the class that matches the specified locale by language, country, and variant by searching the file names in the order shown in Table 36.3. The files searched in this order form a resource chain. If no file is found in the resource

M36_LIAN1878_11_GE_C36.indd 22 12/29/17 5:50 PM

36.5 Resource Bundles 36-23

chain, the getBundle method raises a MissingResourceException, a subclass of RuntimeException.

Once a resource bundle object is created, you can use the getObject method to retrieve the value according to the key. For example,

ResourceBundle res = ResourceBundle.getBundle("MyResource"); String flagFile = (String)res.getObject("nationalFlag"); String anthemFile = (String)res.getObject("nationalAnthem"); Color color = (Color)res.getObject("nationalColor"); double growthRate = (Double)res.getObject("annualGrowthRate");

Tip If the resource value is a string, the convenient getString method can be used to replace the getObject method. The getString method simply casts the value returned by getObject to a string.

What happens if a resource object you are looking for is not defined in the resource bundle? Java employs an intelligent look-up scheme that searches the object in the parent file along the resource chain. This search is repeated until the object is found or all the parent files in the resource chain have been searched. A MissingResourceException is raised if the search is unsuccessful.

Let us modify the NumberFormatDemo program in the preceding example so it displays messages, title, and button labels in multiple languages, as shown in Figure 36.10.

You need to provide a resource bundle for each language. Suppose the program supports three languages: English (default), Chinese, and French. The resource bundle for the English language, named MyResource.properties, is given as follows:

#MyResource.properties for English language Number_Of_Years=Years Total_Payment=French Total\ Payment Enter_Interest_Rate=Enter\ Interest\ Rate,\ Years,\ and\ Loan\ Amount Payment=Payment Compute=Compute Annual_Interest_Rate=Interest\ Rate Number_Formatting=Number\ Formatting\ Demo Loan_Amount=Loan\ Amount Choose_a_Locale=Choose\ a\ Locale Monthly_Payment=Monthly\ Payment

Figure 36.10 The program displays the strings in multiple languages.

M36_LIAN1878_11_GE_C36.indd 23 12/29/17 5:50 PM

36-24 Chapter 36 Internationalization

The resource bundle for the Chinese language, named MyResource_zh.properties, is given as follows:

#MyResource_zh.properties for Chinese language Choose_a_Locale = \u9078\u64c7\u570b\u5bb6 Enter_Interest_Rate = \u8f38\u5165\u5229\u7387,\ u5e74\u9650,\ u8cbo8\u6b3e\u7e3d\u984d Annual_Interest_Rate = \u5229\u7387 Number_Of_Years = \u5e74\u9650 Loan_Amount = \u8cbo8\u6b3e\u984d\u5ea6 Payment = \u4ed8\u606f Monthly_Payment = \u6708\u4ed8 Total_Payment = \u7e3d\u984d Compute = \u8a08\u7b97\u8cbo8\u6b3e\u5229\u606f

The resource bundle for the French language, named MyResource_fr.properties, is given as follows:

#MyResource_fr.properties for French language Number_Of_Years=annees Annual_Interest_Rate=le taux d'interet Loan_Amount=Le montant du pret Enter_Interest_Rate=inscrire le taux d'interet, les annees, et le montant du pret Payment=paiement Compute=Calculer l'hypotheque Number_Formatting=demonstration du formatting des chiffres Choose_a_Locale=Choisir la localite Monthly_Payment=versement mensuel Total_Payment=reglement total

The resource-bundle file should be placed in the class directory (e.g., c:\book for the examples in this book). The program is given in Listing 36.7.

lisTing 36.7 ResourceBundleDemo.java 1 import java.util.*; 2 import java.text.NumberFormat; 3 import javafx.application.Application; 4 import javafx.geometry.Pos; 5 import javafx.scene.Scene; 6 import javafx.scene.control.Button; 7 import javafx.scene.control.ComboBox; 8 import javafx.scene.control.Label; 9 import javafx.scene.control.TextField; 10 import javafx.scene.layout.GridPane; 11 import javafx.scene.layout.HBox; 12 import javafx.scene.layout.VBox; 13 import javafx.stage.Stage; 14 15 public class ResourceBundleDemo extends Application { 16 private ResourceBundle res 17 = ResourceBundle.getBundle("MyResource"); 18 19 // Create labels 20 private Label lblInterestRate = 21 new Label(res.getString("Annual_Interest_Rate")); 22 private Label lblNumberOfYears = 23 new Label(res.getString("Number_Of_Years")); 24 private Label lblLoanAmount =

M36_LIAN1878_11_GE_C36.indd 24 12/29/17 5:50 PM

36.5 Resource Bundles 36-25

25 new Label(res.getString("Loan_Amount")); 26 private Label lblMonthlyPayment = 27 new Label(res.getString("Monthly_Payment")); 28 private Label lblTotalPayment = 29 new Label(res.getString("Total_Payment")); 30 private Label lblPayment = 31 new Label(res.getString("Payment")); 32 private Label lblChooseALocale = 33 new Label(res.getString("Choose_a_Locale")); 34 private Label lblEnterInterestRate = 35 new Label(res.getString("Enter_Interest_Rate")); 36 37 // Combo box for selecting available locales 38 private ComboBox<String> cboLocale = new ComboBox<>(); 39 40 // Text fields for interest rate, year, and loan amount 41 private TextField tfInterestRate = new TextField("6.75"); 42 private TextField tfNumberOfYears = new TextField("15"); 43 private TextField tfLoanAmount = new TextField("107000"); 44 private TextField tfFormattedInterestRate = new TextField(); 45 private TextField tfFormattedNumberOfYears = new TextField(); 46 private TextField tfFormattedLoanAmount = new TextField(); 47 48 // Text fields for monthly payment and total payment 49 private TextField tfTotalPayment = new TextField(); 50 private TextField tfMonthlyPayment = new TextField(); 51 52 // Compute button 53 private Button btCompute = new Button("Compute"); 54 55 // Current locale 56 private Locale locale = Locale.getDefault(); 57 58 // Declare locales to store available locales 59 private Locale locales[] = Calendar.getAvailableLocales(); 60 61 /** Initialize the combo box */ 62 public void initializeComboBox() { 63 // Add locale names to the combo box 64 for (int i = 0; i < locales.length; i++) 65 cboLocale.getItems().add(locales[i].getDisplayName()); 66 } 67 68 @Override // Override the start method in the Application class 69 public void start(Stage primaryStage) { 70 initializeComboBox(); 71 72 // Pane to hold the combo box for selecting locales 73 HBox hBox = new HBox(5); 74 hBox.getChildren().addAll(lblChooseALocale, cboLocale); 75 76 // Pane to hold the input 77 GridPane gridPane = new GridPane(); 78 gridPane.add(lblInterestRate, 0, 0); 79 gridPane.add(tfInterestRate, 1, 0); 80 gridPane.add(tfFormattedInterestRate, 2, 0); 81 gridPane.add(lblNumberOfYears, 0, 1); 82 gridPane.add(tfNumberOfYears, 1, 1); 83 gridPane.add(tfFormattedNumberOfYears, 2, 1); 84 gridPane.add(lblLoanAmount, 0, 2);

M36_LIAN1878_11_GE_C36.indd 25 12/29/17 5:50 PM

36-26 Chapter 36 Internationalization

85 gridPane.add(tfLoanAmount, 1, 2); 86 gridPane.add(tfFormattedLoanAmount, 2, 2); 87 88 // Pane to hold the output 89 GridPane gridPaneOutput = new GridPane(); 90 gridPaneOutput.add(lblMonthlyPayment, 0, 0); 91 gridPaneOutput.add(tfMonthlyPayment, 1, 0); 92 gridPaneOutput.add(lblTotalPayment, 0, 1); 93 gridPaneOutput.add(tfTotalPayment, 1, 1); 94 95 // Set text field alignment 96 tfFormattedInterestRate.setAlignment(Pos.BASELINE_RIGHT); 97 tfFormattedNumberOfYears.setAlignment(Pos.BASELINE_RIGHT); 98 tfFormattedLoanAmount.setAlignment(Pos.BASELINE_RIGHT); 99 tfTotalPayment.setAlignment(Pos.BASELINE_RIGHT); 100 tfMonthlyPayment.setAlignment(Pos.BASELINE_RIGHT); 101 102 // Set editable false 103 tfFormattedInterestRate.setEditable(false); 104 tfFormattedNumberOfYears.setEditable(false); 105 tfFormattedLoanAmount.setEditable(false); 106 tfTotalPayment.setEditable(false); 107 tfMonthlyPayment.setEditable(false); 108 109 VBox vBox = new VBox(5); 110 vBox.getChildren().addAll(hBox, lblEnterInterestRate, 111 gridPane, lblPayment, gridPaneOutput, btCompute); 112 113 // Create a scene and place it in the stage 114 Scene scene = new Scene(vBox, 400, 300); 115 primaryStage.setTitle("ResourceBundleDemo"); // Set the stage title 116 primaryStage.setScene(scene); // Place the scene in the stage 117 primaryStage.show(); // Display the stage 118 119 // Register listeners 120 cboLocale.setOnAction(e -> { 121 locale = locales[cboLocale 122 .getSelectionModel().getSelectedIndex()]; 123 updateStrings(); 124 computeLoan(); 125 }); 126 127 btCompute.setOnAction(e -> computeLoan()); 128 } 129 130 /** Compute payments and display results locale-sensitive format */ 131 private void computeLoan() { 132 // Retrieve input from user 133 double loan = new Double(tfLoanAmount.getText()).doubleValue(); 134 double interestRate = 135 new Double(tfInterestRate.getText()).doubleValue() / 1240; 136 int numberOfYears = 137 new Integer(tfNumberOfYears.getText()).intValue(); 138 139 // Calculate payments 140 double monthlyPayment = loan * interestRate/ 141 (1 - (Math.pow(1 / (1 + interestRate), numberOfYears * 12))); 142 double totalPayment = monthlyPayment * numberOfYears * 12;

M36_LIAN1878_11_GE_C36.indd 26 12/29/17 5:50 PM

36.5 Resource Bundles 36-27

143 144 // Get formatters 145 NumberFormat percentFormatter = 146 NumberFormat.getPercentInstance(locale); 147 NumberFormat currencyForm = 148 NumberFormat.getCurrencyInstance(locale); 149 NumberFormat numberForm = NumberFormat.getNumberInstance(locale); 150 percentFormatter.setMinimumFractionDigits(2); 151 152 // Display formatted input 153 tfFormattedInterestRate.setText( 154 percentFormatter.format(interestRate * 12)); 155 tfFormattedNumberOfYears.setText 156 (numberForm.format(numberOfYears)); 157 tfFormattedLoanAmount.setText(currencyForm.format(loan)); 158 159 // Display results in currency format 160 tfMonthlyPayment.setText(currencyForm.format(monthlyPayment)); 161 tfTotalPayment.setText(currencyForm.format(totalPayment)); 162 } 163 164 /** Update resource strings */ 165 private void updateStrings() { 166 res = ResourceBundle.getBundle("MyResource", locale); 167 lblInterestRate.setText(res.getString("Annual_Interest_Rate")); 168 lblNumberOfYears.setText(res.getString("Number_Of_Years")); 169 lblLoanAmount.setText(res.getString("Loan_Amount")); 170 lblTotalPayment.setText(res.getString("Total_Payment")); 171 lblMonthlyPayment.setText(res.getString("Monthly_Payment")); 172 btCompute.setText(res.getString("Compute")); 173 lblChooseALocale.setText(res.getString("Choose_a_Locale")); 174 lblEnterInterestRate.setText( 175 res.getString("Enter_Interest_Rate")); 176 lblPayment.setText(res.getString("Payment")); 177 } 178 }

Property resource bundles are implemented as text files with a .properties extension, and are placed in the same location as the class files for the program. ListResourceBundles are provided as Java class files. Because they are implemented using Java source code, new and modified ListResourceBundles need to be recompiled for deployment. With Property- ResourceBundles, there is no need for recompilation when translations are modified or added to the application. Nevertheless, ListResourceBundles provide considerably better performance than PropertyResourceBundles.

If the resource bundle is not found or a resource object is not found in the resource bundle, a MissingResourceException is raised. Since MissingResourceException is a subclass of RuntimeException, you do not need to catch the exception explicitly in the code.

This example is the same as Listing 36.6, NumberFormatDemo.java, except that the program contains the code for handling resource strings. The updateString method (lines 165–177) is responsible for displaying the locale-sensitive strings. This method is invoked when a new locale is selected in the combo box.

36.5.1 How does the getBundle method locate a resource bundle? 36.5.2 How does the getObject method locate a resource? Point

Check

M36_LIAN1878_11_GE_C36.indd 27 12/29/17 5:50 PM

36-28 Chapter 36 Internationalization

36.6 Character Encoding You can specify an encoding scheme for file IO to read and write Unicode characters.

Java programs use Unicode. When you read a character using text I/O, the Unicode code of the character is returned. The encoding of the character in the file may be different from the Unicode encoding. Java automatically converts it to the Unicode. When you write a character using text I/O, Java automatically converts the Unicode of the character to the encoding speci- fied for the file. This is pictured in Figure 36.11.

Point Key

Figure 36.11 The encoding of the file may be different from the encoding used in the program.

Program

The Unicode of the character is returned

A character is converted into Unicode

The Unicode of the character is sent out

A character is converted into the code for the speci�ed encoding

A character stored in a speci�ed encoding

You can specify an encoding scheme using a constructor of Scanner/PrintWriter for text I/O, as follows:

public Scanner(File file, String encodingName) public PrintWriter(File file, String encodingName)

For a list of encoding schemes supported in Java, see http://download.oracle.com/javase/1.5.0/ docs/guide/intl/encoding.doc.html and mindprod.com/jgloss/encoding.html. For example, you may use the encoding name GB18030 for simplified Chinese characters, Big5 for traditional Chinese characters, Cp939 for Japanese characters, Cp933 for Korean characters, and Cp838 for Thai characters.

The following code in Listing 36.8 creates a file using the GB18030 encoding (line 8). You have to read the text using the same encoding (line 12). The output is shown in Figure 36.12a.

lisTing 36.8 EncodingDemo.java 1 import java.util.*; 2 import java.io.*; 3 import javafx.application.Application; 4 import javafx.scene.Scene; 5 import javafx.scene.layout.StackPane; 6 import javafx.stage.Stage; 7 import javafx.scene.text.Text; 8 9 public class EncodingDemo extends Application { 10 @Override // Override the start method in the Application class 11 public void start(Stage primaryStage) throws Exception { 12 try ( 13 PrintWriter output = new PrintWriter("temp.txt", "GB18030"); 14 ) { 15 output.print("\u6B22\u8FCE Welcome \u03b1\u03b2\u03b3"); 16 } 17

M36_LIAN1878_11_GE_C36.indd 28 12/29/17 5:50 PM

Chapter Summary 36-29

18 try ( 19 Scanner input = new Scanner(new File("temp.txt"), "GB18030"); 20 ) { 21 StackPane pane = new StackPane(); 22 pane.getChildren().add(new Text(input.nextLine())); 23 24 // Create a scene and place it in the stage 25 Scene scene = new Scene(pane, 200, 200); 26 primaryStage.setTitle("EncodingDemo"); // Set the stage title 27 primaryStage.setScene(scene); // Place the scene in the stage 28 primaryStage.show(); // Display the stage 29 } 30 } 31 }

Figure 36.12 You can specify an encoding scheme for a text file.

(a) Using GB18030 encoding (b) Using default encoding

locale 36-2 resource bundle 36-21

file encoding scheme 36-28

Key Terms

If you don’t specify an encoding in lines 13 and 19, the system’s default encoding scheme is used. The US default encoding is ASCII. ASCII code uses 8 bits. Java uses the 16-bit Unicode. If a Unicode is not an ASCII code, the character '?' is written to the file. Thus, when you write \u6B22 to an ASCII file, the ? character is written to the file. When you read it back, you will see the ? character, as shown in Figure 36.12b.

To find out the default encoding on your system, use

System.out.println(System.getProperty("file.encoding"));

The default encoding name is Cp1252 on Windows, which is a variation of ASCII.

36.6.1 How do you specify an encoding scheme for a text file? 36.6.2 What would happen if you wrote a Unicode character to an ASCII text file? 36.6.3 How do you find the default encoding name on your system?

Point Check

ChapTer summary 1. Java is the first language designed from the ground up to support internationalization.

In consequence, it allows your programs to be customized for any number of countries or languages without requiring cumbersome changes in the code.

2. Java characters use Unicode in the program. The use of Unicode encoding makes it easy to write Java programs that can manipulate strings in any international language.

M36_LIAN1878_11_GE_C36.indd 29 12/29/17 5:50 PM

36-30 Chapter 36 Internationalization

3. Java provides the Locale class to encapsulate information about a specific locale. A Locale object determines how locale-sensitive information, such as date, time, and number, is displayed, and how locale-sensitive operations, such as sorting strings, are performed. The classes for formatting date, time, and numbers, and for sorting strings are grouped in the java.text package.

4. Different locales have different conventions for displaying date and time. The java. text.DateFormat class and its subclasses can be used to format date and time in a locale-sensitive way for display to the user.

5. To format a number for the default or a specified locale, use one of the factory class methods in the NumberFormat class to get a formatter. Use getInstance or getNum- berInstance to get the normal number format. Use getCurrencyInstance to get the currency number format. Use getPercentInstance to get a format for displaying percentages.

6. Java uses the ResourceBundle class to separate locale-specific information, such as status messages and GUI component labels, from the program. The information is stored outside the source code and can be accessed and loaded dynamically at runtime from a ResourceBundle, rather than hard-coded into the program.

7. You can specify an encoding for a text file when constructing a PrintWriter or a Scanner.

Quiz Answer the quiz for this chapter online at the book Companion Website.

programming exerCises

Sections 36.1–36.2 *36.1 (Unicode viewer) Develop a GUI application that displays Unicode characters, as

shown in Figure 36.13. The user specifies a Unicode in the text field and presses the Enter key to display a sequence of Unicode characters starting with the specified Unicode. The Unicode characters are displayed in a scrollable text area of 20 lines. Each line contains 16 characters preceded by the Unicode that is the code for the first character on the line.

Figure 36.13 The program displays the Unicode characters.

M36_LIAN1878_11_GE_C36.indd 30 12/29/17 5:50 PM

Programming Exercises 36-31

**36.2 (Display date and time) Write a program that displays the current date and time as shown in Figure 36.14. The program enables the user to select a locale, time zone, date style, and time style from the combo boxes.

Figure 36.14 The program displays the current date and time.

Figure 36.15 The calendar and clock display the current date and time.

Section 36.3 36.3 (Place the calendar and clock in a panel) Write a program that displays the current

date in a calendar and current time in a clock, as shown in Figure 36.15. Enable the program to run standalone.

36.4 (Find the available locales and time zone IDs) Write two programs to display the available locales and time zone IDs using buttons, as shown in Figure 36.16.

Figure 36.16 The program displays available locales and time zones using buttons.

Section 36.4 *36.5 (Compute loan amortization schedule) Rewrite Exercise 5.22 using JavaFX, as

shown in Figure 36.17. The program allows the user to set the loan amount, loan

M36_LIAN1878_11_GE_C36.indd 31 12/29/17 5:50 PM

36-32 Chapter 36 Internationalization

period, and interest rate, and displays the corresponding interest, principal, and balance in the currency format.

36.6 (Convert dollars to other currencies) Write a program that converts U.S. dollars to Canadian dollars, German marks, and British pounds, as shown in Figure 36.18. The user enters the U.S. dollar amount and the conversion rate, and clicks the Con- vert button to display the converted amount.

Figure 36.17 The program displays the loan payment schedule.

Figure 36.18 The program converts U.S. dollars to Canadian dollars, German marks, and British pounds.

36.7 (Compute loan payments) Rewrite Listing 2.8, ComputeLoan.java, to display the monthly payment and total payment in currency.

36.8 (Use the DecimalFormat class) Rewrite Exercise 5.8 to display at most two digits after the decimal point for the temperature using the DecimalFormat class.

Section 36.5 *36.9 (Use resource bundle) Modify the example for displaying a calendar in Section 36.3.6,

“Example: Displaying a Calendar,” to localize the labels “Choose a locale” and “Calendar Demo” in French, German, Chinese, or a language of your choice.

**36.10 (Flag and anthem) Rewrite Listing 16.13, ImageAudioAnimation.java, to use the resource bundle to retrieve image and audio files.

(Hint: When a new country is selected, set an appropriate locale for it. Have your program look for the flag and audio file from the resource file for the locale.)

Section 36.6 **36.11 (Specify file encodings) Write a program named Exercise36_11Writer

that writes 1307 * 16 Chinese Unicode characters starting from \u0E00 to a file named Exercise36_11.gb using the GBK encoding scheme. Output 16

M36_LIAN1878_11_GE_C36.indd 32 12/29/17 5:50 PM

Programming Exercises 36-33

Figure 36.19 The program displays the file using the specified encoding scheme.

characters per line and separate the characters with spaces. Write a program named Exercise36_11Reader that reads all the characters from a file using a specified encoding. Figure 36.19 displays the file using the GBK encoding scheme.

M36_LIAN1878_11_GE_C36.indd 33 12/29/17 5:50 PM

Objectives ■■ To explain how a servlet works (§37.2).

■■ To create/develop/run servlets (§37.3).

■■ To deploy servlets on application servers such as Tomcat and GlassFish (§37.3).

■■ To describe the servlets API (§37.4).

■■ To create simple servlets (§37.5).

■■ To create and process HTML forms (§37.6).

■■ To develop servlets to access databases (§37.7).

■■ To use hidden fields, cookies, and HttpSession to track sessions (§37.8).

Servlets

Chapter

M37_LIAN1878_11_GE_C37.indd 1 12/29/17 5:55 PM

37-2 Chapter 37 Servlets

37.1 Introduction Java Servlets is the foundation for developing Web applications using Java.

Servlets are Java programs that run on a Web server. They can be used to process client requests or produce dynamic webpages. For example, you can write servlets to generate dynamic web- pages that display stock quotes or process client registration forms and store registration data in a database. This chapter introduces the concept of Java servlets. You will learn how to develop Java servlets using NetBeans.

Note You can develop servlets without using an IDE. However, using an IDE such as NetBeans can greatly simplify the development task. The tool can automatically create the sup- porting directories and files. We choose NetBeans because it has the best support for Java Web development. You can still use your favorite IDE or no IDE for this chapter.

Note Servlets are the foundation of Java Web technologies. JSP, JSF, and Java Web services are based on servlets. A good understanding of servlets helps you see the big picture of Java Web technology and learn JSP, JSF, and Web services.

37.2 HTML and Common Gateway Interface Java servlets are Java programs that function like CGI programs. They are executed upon request from a Web browser.

Java servlets run in the Web environment. To understand Java servlets, let us review HTML and the Common Gateway Interface (CGI).

37.2.1 Static Web Contents You create webpages using HTML. Your webpages are stored as files on the Web server. The files are usually stored in the /htdocs directory on Unix, as shown in Figure 37.1. A user types a URL for the file from a Web browser. The browser contacts the Web server and requests the file. The server finds the file and returns it to the browser. The browser then displays the file to the user. This works fine for static information that does not change regardless of who requests it or when it is requested. Static information is stored in files. The information in the files can be updated, but at any given time every request for the same document returns exactly the same result.

Point Key

Figure 37.1 A Web browser requests a static HTML page from a Web server.

Web Browser

HTML Page

http://www.webserverhost.com/index.html

Web Server

Web Server Host

Host Machine File System

/htdocs/index.html

Point Key

M37_LIAN1878_11_GE_C37.indd 2 12/29/17 5:55 PM

37.2 HTML and Common Gateway Interface 37-3

37.2.2 Dynamic Web Contents and Common Gateway Interface Not all information, however, is static in nature. Stock quotes are updated whenever a trade takes place. Election vote counts are updated constantly on Election Day. Weather reports are frequently updated. The balance in a customer’s bank account is updated whenever a transac- tion takes place. To view up-to-date information on the Web, the HTML pages for displaying this information must be generated dynamically. Dynamic Web pages are generated by Web servers. The Web server needs to run certain programs to process user requests from Web browsers in order to produce a customized response.

The Common Gateway Interface, or CGI, was proposed to generate dynamic Web content. The interface provides a standard framework for Web servers to interact with external pro- grams, known as CGI programs. As shown in Figure 37.2, the Web server receives a request from a Web browser and passes it to the CGI program. The CGI program processes the request and generates a response at runtime. CGI programs can be written in any language, but the Perl language is the most popular choice. CGI programs are typically stored in the /cgi-bin directory. Here is a pseudocode example of a CGI program for displaying a customer’s bank account balance:

1. Obtain account ID and password.

2. Verify account ID and password. If it fails, generate an HTML page to report incorrect account ID and password, and exit.

3. Retrieve account balance from the database; generate an HTML page to display the account ID and balance.

Figure 37.2 A Web browser requests a dynamic HTML page from a Web server.

Web Browser

HTML Page returned

Send a request URL Web Server

Execute CGI Program

Web Server Host

Host Machine File System

/htdocs/index.html

Get CGI Code

Generate Response

Spawn CGI Process

URL Example http://www.server.com/cgi-bin/getBalance.cgi? accountId=scott&password=tiger

37.2.3 The GET and POST Methods The two most common HTTP requests, also known as methods, are GET and POST. The Web browser issues a request using a URL or an HTML form to trigger the Web server to execute a CGI program. HTML forms will be introduced in §37.6, “HTML Forms.” When issuing a CGI request directly from a URL, the GET method is used. This URL is known as a query string. The URL query string consists of the location of the CGI program, the parameters, and their values. For example, the following URL causes the CGI program getBalance to be invoked on the server side:

http://www.webserverhost.com/cgi-bin/ getBalance.cgi?accountId=scott+smith&password=tiger

M37_LIAN1878_11_GE_C37.indd 3 12/29/17 5:55 PM

37-4 Chapter 37 Servlets

The ? symbol separates the program from the parameters. The parameter name and value are associated using the = symbol. Parameter pairs are separated using the & symbol. The + symbol denotes a space character. So, here accountId is scott smith.

When issuing a request from an HTML form, either a GET method or a POST method can be used. The form explicitly specifies one of these. If the GET method is used, the data in the form are appended to the request string as if it were submitted using a URL. If the POST method is used, the data in the form are packaged as part of the request file. The server program obtains the data by reading the file. The POST method is more secure than the GET method.

Note The GET and POST methods both send requests to the Web server. The POST method always triggers the execution of the corresponding CGI program. The GET method may not cause the CGI program to be executed, if the previous same request is cached in the Web browser. Web browsers often cache webpages so that the same request can be quickly responded to without contacting the Web server. The browser checks the request sent through the GET method as a URL query string. If the results for the exact same URL are cached on a disk, then the previous webpages for the URL may be displayed. To ensure that a new webpage is always displayed, use the POST method. For example, use a POST method if the request will actually update the database. If your request is not time sensitive, such as finding the address of a student in the database, use the GET method to speed up performance.

37.2.4 From CGI to Java Servlets CGI provides a relatively simple approach for creating dynamic Web applications that accept a user request, process it on the server side, and return responses to the Web browser. But CGI is very slow when handling a large number of requests simultaneously, because the Web server spawns a process for executing each CGI program. Each process has its own runtime environ- ment that contains and runs the CGI program. It is not difficult to imagine what will happen if many CGI programs were executed simultaneously. System resource would be quickly exhausted, potentially causing the server to crash.

Several new approaches have been developed to remedy the performance problem of CGI programs. Java servlets are one successful technology for this purpose. Java servlets are Java programs that function like CGI programs. They are executed upon request from a Web browser. All servlets run inside a servlet container, also referred to as a servlet server or a servlet engine. A servlet container is a single process that runs in a Java Virtual Machine. The JVM creates a thread to handle each servlet. Java threads have much less overhead than full- blown processes. All the threads share the same memory allocated to the JVM. Since the JVM persists beyond the life cycle of a single servlet execution, servlets can share objects already created in the JVM. For example, if multiple servlets access the same database, they can share the connection object. Servlets are much more efficient than CGI.

Servlets have other benefits that are inherent in Java. As Java programs, they are object oriented, portable, and platform independent. Since you know Java, you can develop servlets immediately with the support of Java API for accessing databases and network resources.

37.2.1 What is the common gateway interface? 37.2.2 What are the differences between the GET and POST methods in an HTML form? 37.2.3 Can you submit a GET request directly from a URL? Can you submit a POST

request directly from a URL?

37.2.4 What is wrong in the following URL for submitting a GET request to the servlet FindScore on host liang at port 8084 with parameter name?

http://liang:8084/findScore?name=“P Yates”

37.2.5 What are the differences between CGI and servlets?

Point Check

M37_LIAN1878_11_GE_C37.indd 4 12/29/17 5:55 PM

37.3 Creating and Running Servlets 37-5

37.3 Creating and Running Servlets An IDE such as NetBeans is an effective tool for creating Java servlet.

To run Java servlets, you need a servlet container. Many servlet containers are available for free. Two popular ones are Tomcat (developed by Apache, www.apache.org) and Glass- Fish (developed by Sun, glassfish.dev.java.net). Both Tomcat and GlassFish are bundled and integrated with NetBeans 7 (Java EE version). When you run a servlet from NetBeans, Tomcat or GlassFish will be automatically started. You can choose to use either of them, or any other application server. GlassFish has more features than Tomcat and it takes more system resource.

37.3.1 Creating a Servlet Before our introduction to the servlet API, let us look at a simple example to see how servlets work. A servlet to some extent resembles a JavaFX program. Every Java applet is a subclass of the Application class. You need to override appropriate methods in the Application class to implement the application. Every servlet is a subclass of the HttpServlet class. You need to override appropriate methods in the HttpServlet class to implement the servlet. Listing 37.1 is a servlet that generates a response in HTML using the doGet method.

Point Key

Listing 37.1 FirstServlet.java 1 package chapter37; 2 3 import javax.servlet.*; 4 import javax.servlet.http.*; 5 6 public class FirstServlet extends HttpServlet { 7 /** Handle the HTTP GET method. 8 * @param request servlet request 9 * @param response servlet response 10 */ 11 protected void doGet(HttpServletRequest request, 12 HttpServletResponse response) 13 throws ServletException, java.io.IOException { 14 response.setContentType("text/html"); 15 java.io.PrintWriter out = response.getWriter(); 16 // output your page here 17 out.println("<html>"); 18 out.println("<head>"); 19 out.println("<title>Servlet</title>"); 20 out.println("</head>"); 21 out.println("<body>"); 22 out.println("Hello, Java Servlets"); 23 out.println("</body>"); 24 out.println("</html>"); 25 out.close(); 26 } 27 }

The doGet method (line 11) is invoked when the Web browser issues a request using the GET method. The doGet method has two parameters: request and response. request is for obtaining data from the Web browser, and response is for sending data back to the browser. Line 14 indicates that data are sent back to the browser as text/html. Line 15 obtains an instance of PrintWriter for actually outputting data to the browser.

M37_LIAN1878_11_GE_C37.indd 5 12/29/17 5:55 PM

37-6 Chapter 37 Servlets

37.3.2 Creating Servlets in NetBeans NetBeans is updated frequently. The current version is 8 at the time of this writing. To create a servlet in NetBeans 8, you have to first create a Web project, as follows:

1. Choose File, New Project to display the New Project dialog box. Choose Java Web in the Categories section and Web Application in the Projects section, as shown in Figure 37.3a. Click Next to display the New Web Application dialog box, as shown in Figure 37.3b.

2. Enter liangweb in the Project Name field and c:\book in the Project Location field. Check Set as Main Project. Click Next to display the dialog box for specifying server and settings, as shown in Figure 37.4.

3. Select GlassFish Server 4.1 for server and Java EE 7 Web for Java EE Version. Click Finish to create the Web project, as shown in Figure 37.5.

Now you can create a servlet in the project, as follows:

1. Right-click the liangweb node in the project pane to display a context menu. Choose New, Servlet to display the New Servlet dialog box, as shown in Figure 37.6.

Figure 37.3 (a) Choose Web Application to create a Web project. (b) Specify project name and location.

(a) (b)

Figure 37.4 Choose servers and settings.

M37_LIAN1878_11_GE_C37.indd 6 12/29/17 5:55 PM

37.3 Creating and Running Servlets 37-7

Figure 37.5 A new Web project is created.

Figure 37.6 You can create a servlet in the New Servlet dialog box.

2. Enter FirstServlet in the Class Name field and chapter37 in the Package field and click Next to display the Configure Servlet Deployment dialog box, as shown in Figure 37.7.

3. Select the checkbox to add the servlet information to web.xml and click Finish to create the servlet. A servlet template is now created in the project, as shown in Figure 37.8.

M37_LIAN1878_11_GE_C37.indd 7 12/29/17 5:55 PM

37-8 Chapter 37 Servlets

Figure 37.7 You need to click the checkbox to add servlet information to web.xml.

Figure 37.8 A new servlet class is created in the project.

M37_LIAN1878_11_GE_C37.indd 8 12/29/17 5:55 PM

37.4 The Servlet API 37-9

4. Replace the code in the content pane for the servlet using the code in Listing 37.1.

5. Right-click liangweb node in the Project pane to display a context menu and choose Run to launch the Web server. In the Web browser, enter http://localhost:8084/liang- web/FirstServlet in the URL. You will now see the servlet result displayed, as shown in Figure 37.9.

Figure 37.9 Servlet result is displayed in a Web browser.

Note If the servlet is not displayed in the browser, do the following: 1. Make sure that you have added the servlet in the xml.web file. 2. Right-click liangweb and choose Clean and Build. 3. Right-click liangweb and choose Run. Reenter http://localhost:8084/ liangweb/FirstServlet in the URL. If still not working, exit NetBeans and restart it.

Note Depending on the server setup, you may have a port number other than 8084.

Tip You can deploy a Web application using a Web archive file (WAR) to a Web application server (e.g., Tomcat). To create a WAR file for the liangweb project, right-click liangweb and choose Build Project. You can now locate liangweb.war in the c:\book\liang- web\dist folder. To deploy on Tomcat, simply place liangweb.war into the webapps directory. When Tomcat starts, the .war file will be automatically installed.

Note If you wish to use NetBeans as the development tool and Tomcat as the deployment server, please see Supplement V.E, “Tomcat Tutorial.”

37.3.1 Can you display an HTML file (e.g. c:\ test.html) by typing the complete file name in the Address field of Internet Explorer? Can you run a servlet by simply typing the servlet class file name?

37.3.2 How do you create a Web project in NetBeans? 37.3.3 How do you create a servlet in NetBeans? 37.3.4 How do you run a servlet in NetBeans? 37.3.5 When you run a servlet from NetBeans, what is the port number by default? What

happens if the port number is already in use?

37.3.6 What is the .war file? How do you obtain a .war file for a Web project in NetBeans?

37.4 The Servlet API The Servlet interface defines the methods init, service, and destroy to manag- ing the life-cylce of a serlvet.

Point Check

Point Key

M37_LIAN1878_11_GE_C37.indd 9 12/29/17 5:55 PM

37-10 Chapter 37 Servlets

You have to know the servlet API in order to understand the source code in Listing3 7.1, in FirstServlet.java. The servlet API provides the interfaces and classes that support servlets. These interfaces and classes are grouped into two packages, javax.servlet and javax. servlet.http, as shown in Figure 37.10. The javax.servlet package provides basic interfaces, and the javax.servlet.http package provides classes and interfaces derived from them, which provide specific means for servicing HTTP requests.

37.4.1 The Servlet Interface The javax.servlet.Servlet interface defines the methods that all servlets must implement. The methods are listed below:

/** Invoked for every servlet constructed */ public void init() throws ServletException;

/** Invoked to respond to incoming requests */ public void service(ServletRequest request, ServletResponse response) throws ServletException, IOException;

/** Invoked to release resource by the servlet */ public void destroy();

The init, service, and destroy methods are known as life-cycle methods and are called in the following sequence (see Figure 37.11):

1. The init method is called when the servlet is first created and is not called again as long as the servlet is not destroyed. This resembles an applet’s init method, which is invoked after the applet is created and is not invoked again as long as the applet is not destroyed.

Figure 37.10 The servlet API contains interfaces and classes that you use to develop and run servlets.

ServletCon�g

ServletRequest

ServletResponse

javax.Servlet* javax.Servlet.http.*

Servlet GenericServlet HttpServlet

HttpServletRequest

HttpServletResponse

Figure 37.11 The JVM uses the init, service, and destroy methods to control the servlet.

Loaded Initialized

Invokes the init method

Served DestroyedCreated

Creates the servlet using its constructor

JVM loads the servlet class

Invokes the service method

Invokes destroy() after a timeout period has passed or the Web server is being stopped

The same servlet is invoked again, bypassing the Loaded, Created, and Initialized states, as long as it has not been destroyed

Servlet is invoked for the �rst time

Invokes the service method

M37_LIAN1878_11_GE_C37.indd 10 12/29/17 5:55 PM

37.4 The Servlet API 37-11

2. The service method is invoked each time the server receives a request for the servlet. The server spawns a new thread and invokes service.

3. The destroy method is invoked after a timeout period has passed or as the Web server is terminated. This method releases resources for the servlet.

37.4.2 The GenericServlet Class, ServletConfig Interface, and HttpServlet Class

The javax.servlet.GenericServlet class defines a generic, protocol-independent servlet. It implements javax.servlet.Servlet and javax.servlet.ServletConfig. ServletConfig is an interface that defines four methods (getInitParameter, getInit- ParameterNames, getServletContext, and getServletName) for obtaining information from a Web server during initialization. All the methods in Servlet and ServletConfig are implemented in GenericServlet except service. Therefore, GenericServlet is an abstract class.

The javax.servlet.http.HttpServlet class defines a servlet for the HTTP protocol. It extends GenericServlet and implements the service method. The service method is implemented as a dispatcher of HTTP requests. The HTTP requests are processed in the following methods:

■■ doGet is invoked to respond to a GET request.

■■ doPost is invoked to respond to a POST request.

■■ doDelete is invoked to respond to a DELETE request. Such a request is normally used to delete a file on the server.

■■ doPut is invoked to respond to a PUT request. Such a request is normally used to send a file to the server.

■■ doOptions is invoked to respond to an OPTIONS request. This returns information about the server, such as which HTTP methods it supports.

■■ doTrace is invoked to respond to a TRACE request. Such a request is normally used for debugging. This method returns an HTML page that contains appropriate trace information.

All these methods use the following signature:

protected void doXxx(HttpServletRequest req, HttpServletResponse resp) throws ServletException, java.io.IOException

The HttpServlet class provides default implementation for these methods. You need to override doGet, doPost, doDelete, and doPut if you want the servlet to process a GET, POST, DELETE, or PUT request. By default, nothing will be done. Normally, you should not override the doOptions method unless the servlet implements new HTTP methods beyond those implemented by HTTP 1.1. Nor is there any need to override the doTrace method.

Note GET and POST requests are often used, whereas DELETE, PUT, OPTIONS, and TRACE are not. For more information about these requests, please refer to the HTTP 1.1 specifica- tion from www.cis.ohio-state.edu/htbin/rfc/rfc2068.html.

Note Although the methods in HttpServlet are all nonabstract, HttpServlet is defined as an abstract class. Thus you cannot create a servlet directly from HttpServlet. Instead you have to define your servlet by extending HttpServlet.

M37_LIAN1878_11_GE_C37.indd 11 12/29/17 5:55 PM

37-12 Chapter 37 Servlets

The relationship of these interfaces and classes is shown in Figure 37.12.

Figure 37.12 HttpServlet inherits abstract class GenericServlet, which implements interfaces Servlet and ServletConfig.

«interface» javax.servlet.ServletCon�g

+getInitParameter(name: String): String

+getInitParameterNames(): Enumeration

+getServletContext(): ServletContext

+getServletName(): String

«interface» javax.servlet.Servlet

+init(con�g: ServletCon�g): void

+service(req: ServletRequest, resp: ServletResponse): void

+destroy(): void

+getServletInfo(): String

javax.servlet.GenericServlet javax.servlet.http.HttpServlet

+doGet(req: HttpServletRequest, resp: HttpServletResponse): void

+doPost(req: HttpServletRequest, resp: HttpServletResponse): void

+doDelete(req: HttpServletRequest, resp: HttpServletResponse): void

+doPut(req: HttpServletRequest, resp: HttpServletResponse): void

+doOptions(req: HttpServletRequest, resp: HttpServletResponse): void

+doTrace(req: HttpServletRequest, resp: HttpServletResponse): void

37.4.3 The ServletRequest Interface and HttpServlet Request Interface

Every doXxx method in the HttpServlet class has a parameter of the HttpServlet Request type, which is an object that contains HTTP request information, including parameter name and values, attributes, and an input stream. HttpServletRequest is a subinterface of Servlet- Request. ServletRequest defines a more general interface to provide information for all kinds of clients. The frequently used methods in these two interfaces are shown in Figure 37.13.

37.4.4 The ServletResponse Interface and HttpServlet Response Interface

Every doXxx method in the HttpServlet class has a parameter of the HttpServlet- Response type, which is an object that assists a servlet in sending a response to the client. HttpServletResponse is a subinterface of ServletResponse. ServletResponse defines a more general interface for sending output to the client.

The frequently used methods in these two interfaces are shown in Figure 37.14.

37.4.1 Describe the life cycle of a servlet. 37.4.2 Suppose you started the Web server, ran the following servlet twice by issuing an

appropriate URL from the Web browser, and finally stopped Tomcat. What was displayed on the console when the servlet was first invoked? What was displayed on the console when the servlet was invoked for the second time? What was displayed on the console when Tomcat was shut down?

import javax.servlet.*; import javax.servlet.http.*; import java.io.*; public class Test extends HttpServlet { public Test() { System.out.println("Constructor called"); }

/** Initialize variables */ public void init() throws ServletException {

Point Check

M37_LIAN1878_11_GE_C37.indd 12 12/29/17 5:55 PM

37.4 The Servlet API 37-13

Figure 37.13 HttpServletRequest is a subinterface of ServletRequest.

«interface» javax.servlet.http.HttpServletRequest

+getHeader(name: String): String

+getMethod(): String

+getQueryString(): String

+getCookies(): javax.servlet.http.Cookies[]

+getSession(create: boolean): HttpSession

«interface» javax.servlet.ServletRequest

+getParamter(name: String): String

+getParameterValues(): String[]

+getRemoteAddr(): String

+getRemoteHost(): String

Returns the value of a request parameter as a String, or null if the parameter does not exist. Request parameters are extra information sent with the request. For HTTP servlets, parameters are contained in the query string or posted from data. Only use this method when you are sure that the parameter has only one value. If it has more than one value, use getParameterValues.

Returns the Internet Protocol (IP) address of the client that sent the request.

Returns the fully quali�ed name of the client that sent the request, or the IP address of the client if the name cannot be determined.

Returns the value of the speci�ed request header as a String. If the request did not include a header of the speci�ed name, this method returns null. Since the header name is case- insensitive, you can use this method with any request header.

Returns the name of the HTTP method with which this request was made; for example, GET, POST, DELETE, PUT, OPTIONS, or TRACE.

Returns the query string that is contained in the request URL after the path. This method returns null if the URL does not have a query string.

Returns an array containing all of the Cookie objects the client sent with the request. This method returns null if no cookies were sent. Using cookies is introduced in Section 37.8.2, “Session Tracking Using Cookies.”

getSession(true) returns the current session associated with this request. If the request does not have a session, it creates one. getSession(false) returns the current session associated with the request. If the request does not have a session, it returns null. The getSession method is used in session tracking, which is introduced in Section 37.8.3, “Session Tracking Using the Servlet API.”

Figure 37.14 HttpServletResponse is a subinterface of ServletResponse.

«interface» javax.servlet.http.HttpServletResponse

+addCookie(Cookie cookie): void

«interface» javax.servlet.ServletResponse

+getWriter(): java.io.PrintWriter

+setContentType(type: String): void

Returns a PrintWriter object that can send character text to the client.

Sets the content type of the response being sent to the client before writing response to the client. When you are writing HTML to the client, the type should be set to “text/html.” For plain text, use “text/plain.” For sending a gif image to the browser, use “image/gif.”

Adds the speci�ed cookie to the response. This method can be called multiple times to set more than one cookie.

System.out.println("init called"); }

/** Process the HTTP Get request */ public void doGet(HttpServletRequest request, HttpServletResponse response) throws ServletException, IOException { System.out.println("doGet called"); }

/** Clean up resources */ public void destroy() { System.out.println("destroy called"); } }

M37_LIAN1878_11_GE_C37.indd 13 12/29/17 5:55 PM

37-14 Chapter 37 Servlets

37.5 Creating Servlets You can define a servlet class by extending the HttpServlet class and implement the doGet and doPost methods.

Servlets are the opposite of Java applets. Java applets run from a Web browser on the client side. To write Java programs, you define classes. To write a Java applet, you define a class that extends the Applet class. The Web browser runs and controls the execution of the applet through the methods defined in the Applet class. Similarly, to write a Java servlet, you define a class that extends the HttpServlet class. The servlet container runs and controls the execu- tion of the servlet through the methods defined in the HttpServlet class. Like a Java applet, a servlet does not have a main method. A servlet depends on the servlet engine to call the methods. Every servlet has a structure like the one shown below:

package chapter37;

import javax.servlet.*; import javax.servlet.http.*; import java.io.*;

public class MyServlet extends HttpServlet { /** Called by the servlet engine to initialize servlet */ public void init() throws ServletException { ... }

/** Process the HTTP Get request */ public void doGet(HttpServletRequest request, HttpServletResponse response) throws ServletException, IOException { ... }

/** Process the HTTP Post request */ public void doPost(HttpServletRequest request, HttpServletResponse response) throws ServletException, IOException { ... }

/** Called by the servlet engine to release resource */ public void destroy() { ... }

// Other methods if necessary }

The servlet engine controls the servlets using init, doGet, doPost, destroy, and other meth- ods. By default, the doGet and doPost methods do nothing. To handle a GET request, you need to override the doGet method; to handle a POST request, you need to override the doPost method.

Listing 37.2 gives a simple Java servlet that generates a dynamic webpage for displaying the current time, as shown in Figure 37.15.

Point Key

Figure 37.15 Servlet CurrentTime displays the current time.

M37_LIAN1878_11_GE_C37.indd 14 12/29/17 5:55 PM

37.6 HTML Forms 37-15

The HttpServlet class has a doGet method. The doGet method is invoked when the browser issues a request to the servlet using the GET method. Your servlet class should override the doGet method to respond to the GET request. In this case, you write the code to display the current time.

Servlets return responses to the browser through an HttpServletResponse object. Since the setContentType("text/html") method sets the MIME type to “text/html,” the browser will display the response in HTML. The getWriter method returns a PrintWriter object for sending HTML back to the client.

Note The URL query string uses the GET method to issue a request to the servlet. The current time may not be current if the webpage for displaying the current time is cached. To ensure that a new current time is displayed, refresh the page in the browser. In the next example, you will write a new servlet that uses the POST method to obtain the current time.

37.6 HTML Forms HTML forms are used to collect and submit data from a client to a Web server.

HTML forms enable you to submit data to the Web server in a convenient form. As shown in Figure 37.16, the form can contain text fields, text area, check boxes, combo boxes, lists, radio buttons, and buttons.

Point Key

Figure 37.16 An HTML form may contain text fields, radio buttons, combo boxes, lists, check boxes, text areas, and buttons.

Listing 37.2 CurrentTime.java 1 package chapter37; 2 3 import javax.servlet.*; 4 import javax.servlet.http.*; 5 import java.io.*; 6 7 public class CurrentTime extends HttpServlet { 8 /** Process the HTTP Get request */ 9 public void doGet(HttpServletRequest request, HttpServletResponse 10 response) throws ServletException, IOException { 11 response.setContentType("text/html"); 12 PrintWriter out = response.getWriter(); 13 out.println("<p>The current time is " + new java.util.Date()); 14 out.close(); // Close stream 15 } 16 }

M37_LIAN1878_11_GE_C37.indd 15 12/29/17 5:55 PM

37-16 Chapter 37 Servlets

The HTML code for creating the form in Figure 37.16 is given in Listing 37.3. (If you are unfamiliar with HTML, please see Supplement V.A, “HTML and XHTML Tutorial.”)

Listing 37.3 StudentRegistrationForm.html 1  2 <html> 3 <head> 4 <title>Student Registration Form</title> 5 </head> 6 <body> 7 <h3>Student Registration Form</h3> 8 9 <form action = "GetParameters" 10 method = "get"> 11  12 <p><label>Last Name</label> 13 <input type = "text" name = "lastName" size = "20" /> 14 <label>First Name</label> 15 <input type = "text" name = "firstName" size = "20" /> 16 <label>MI</label> 17 <input type = "text" name = "mi" size = "1" /></p> 18 19  20 <p><label>Gender:</label> 21 <input type = "radio" name = "gender" value = "M" checked /> 22 Male 23 <input type = "radio" name = "gender" value = "F" /> Female</p> 24 25  26 <p><label>Major</label> 27 <select name = "major" size = "1"> 28 <option value = "CS">Computer Science</option> 29 <option value = "Math">Mathematics</option> 30 <option>English</option> 31 <option>Chinese</option> 32 </select> 33 34  35 <label>Minor</label> 36 <select name = "minor" size = "2" multiple> 37 <option>Computer Science</option> 38 <option>Mathematics</option> 39 <option>English</option> 40 <option>Chinese</option> 41 </select></p> 42 43  44 <p><label>Hobby:</label> 45 <input type = "checkbox" name = "tennis" /> Tennis 46 <input type = "checkbox" name = "golf" /> Golf 47 <input type = "checkbox" name = "pingPong" checked />Ping Pong 48 </p> 49 50  51 <p>Remarks:</p> 52 <p><textarea name = "remarks" rows = "3" cols = "56"> 53 </textarea></p> 54

M37_LIAN1878_11_GE_C37.indd 16 12/29/17 5:55 PM

37.6 HTML Forms 37-17

The following HTML tags are used to construct HTML forms:

■■ <form> . . . </form> defines a form body. The attributes for the <form> tag are action and method. The action attribute specifies the server program to be exe- cuted on the Web server when the form is submitted. The method attribute is either get or post.

■■ <label> . . . </label> simply defines a label.

■■ <input> defines an input field. The attributes for this tag are type, name, value, checked, size, and maxlength. The type attribute specifies the input type. Possible types are text for a one-line text field, radio for a radio button, and checkbox for a check box. The name attribute gives a formal name for the attribute. This name attribute is used by the servlet program to retrieve its associated value. The names of the radio buttons in a group must be identical. The value attribute specifies a default value for a text field and text area. The checked attribute indicates whether a radio button or a check box is initially checked. The size attribute specifies the size of a text field, and the maxlength attribute specifies the maximum length of a text field.

■■ <select> . . . </select> defines a combo box or a list. The attributes for this tag are name, size, and multiple. The size attribute specifies the number of rows visible in the list. The multiple attribute specifies that multiple values can be selected from a list. Set size to 1 and do not use a multiple for a combo box.

■■ <option> . . . </option> defines a selection list within a <select> . . . </select> tag. This tag may be used with the value attribute to specify a value for the selected option (e.g., <option value = "CS">Computer Science). If no value is specified, the selected option is the value.

■■ <textarea> . . . </textarea> defines a text area. The attributes are name, rows, and cols. The rows and cols attributes specify the number of rows and columns in a text area.

Note You can create the HTML file from NetBeans. Right-click liangweb and choose New, HTML, to display the New HTML File dialog box. Enter StudentRegistrationForm as the file name and click Finish to create the file.

37.6.1 Obtaining Parameter Values from HTML Forms To demonstrate how to obtain parameter values from an HTML form, Listing 37.4 creates a servlet to obtain all the parameter values from the preceding student registration form in Figure 37.16 and display their values, as shown in Figure 37.17.

55  56 <p><input type = "submit" value = "Submit" /> 57 <input type = "reset" value = "Reset" /></p> 58 </form> 59 </body> 60 </html>

Listing 37.4 GetParameters.java 1 package chapter37; 2 3 import javax.servlet.*;

M37_LIAN1878_11_GE_C37.indd 17 12/29/17 5:55 PM

37-18 Chapter 37 Servlets

Figure 37.17 The servlet displays the parameter values entered in Figure 37.16.

The HTML form is already created in StudentRegistrationForm.html and displayed in Figure 37.16. Since the action for the form is GetParameters, clicking the Submit button invokes the GetParameters servlet.

Each GUI component in the form has a name attribute. The servlet uses the name attribute in the getParameter(attributeName) method to obtain the parameter value as a string. In case of a list with multiple values, use the getParameterValues(attributeName) method to return the parameter values in an array of strings (line 20).

4 import javax.servlet.http.*; 5 import java.io.*; 6 7 public class GetParameters extends HttpServlet { 8 /** Process the HTTP Post request */ 9 public void doGet(HttpServletRequest request, HttpServletResponse 10 response) throws ServletException, IOException { 11 response.setContentType("text/html"); 12 PrintWriter out = response.getWriter(); 13 14 // Obtain parameters from the client 15 String lastName = request.getParameter("lastName"); 16 String firstName = request.getParameter("firstName"); 17 String mi = request.getParameter("mi"); 18 String gender = request.getParameter("gender"); 19 String major = request.getParameter("major"); 20 String[] minors = request.getParameterValues("minor"); 21 String tennis = request.getParameter("tennis"); 22 String golf = request.getParameter("golf"); 23 String pingPong = request.getParameter("pingPong"); 24 String remarks = request.getParameter("remarks"); 25 26 out.println("Last Name: <b>" + lastName + "</b> First Name: <b>" 27 + firstName + "</b> MI: <b>" + mi + "</b><br>"); 28 out.println("Gender: <b>" + gender + "</b><br>"); 29 out.println("Major: <b>" + major + "</b> Minor: <b>"); 30 31 if (minors != null) 32 for (int i = 0; i < minors.length; i++) 33 out.println(minors[i] + " "); 34 35 out.println("</b><br> Tennis: <b>" + tennis + "</b> Golf: <b>" + 36 golf + "</b> PingPong: <b>" + pingPong + "</b><br>"); 37 out.println("Remarks: <b>" + remarks + "</b>"); 38 out.close(); // Close stream 39 } 40 }

M37_LIAN1878_11_GE_C37.indd 18 12/29/17 5:55 PM

37.6 HTML Forms 37-19

You may optionally specify the value attribute in a text field, text area, combo box, list, check box, or radio button in an HTML form. For text field and text area, the value attribute specifies a default value to be displayed in the text field and text area. The user can type in new values to replace it. For combo box, list, check box, and radio button, the value attribute specifies the parameter value to be returned from the getParameter and getParameter- Values methods. If the value attribute is not specified for a combo box or a list, it returns the selected string from the combo box or the list. If the value attribute is not specified for a radio button or a check box, it returns string on for a checked radio button or a checked check box, and returns null for an unchecked check box.

Note If an attribute does not exist, the getParameter(attributeName) method returns null. If an empty value of the parameter is passed to the servlet, the getParameter (attributeName) method returns a string with an empty value. In this case, the length of the string is 0.

37.6.2 Obtaining Current Time Based on Locale and Time Zone This example creates a servlet that processes the GET and POST requests. The GET request generates a form that contains a combo box for locale and a combo box for time zone, as shown in Figure 37.18a. The user can choose a locale and a time zone from this form to submit a POST request to obtain the current time based on the locale and time zone, as shown in Figure 37.18b.

Figure 37.18 The GET method in the TimeForm servlet displays a form in (a), and the POST method in the TimeForm servlet displays the time based on locale and time zone in (b).

(a) (b)

Listing 37.5 gives the servlet.

Listing 37.5 TimeForm.java 1 package chapter37; 2 3 import javax.servlet.*; 4 import javax.servlet.http.*; 5 import java.io.*; 6 import java.util.*; 7 import java.text.*; 8 9 public class TimeForm extends HttpServlet { 10 private static final String CONTENT_TYPE = "text/html"; 11 private Locale[] allLocale = Locale.getAvailableLocales(); 12 private String[] allTimeZone = TimeZone.getAvailableIDs(); 13 14 /** Process the HTTP Get request */ 15 public void doGet(HttpServletRequest request, HttpServletResponse

M37_LIAN1878_11_GE_C37.indd 19 12/29/17 5:55 PM

37-20 Chapter 37 Servlets

When you run this servlet, the servlet TimeForm’s doGet method is invoked to generate the time form dynamically. The method of the form is POST, and the action invokes the same servlet, TimeForm. When the form is submitted to the server, the doPost method is invoked to process the request.

The variables allLocale and allTimeZone (lines 11–12), respectively, hold all the avail- able locales and time zone IDs. The names of the locales are displayed in the locale list. The values for the locales are the indexes of the locales in the array allLocale. The time zone IDs

16 response) throws ServletException, IOException { 17 response.setContentType(CONTENT_TYPE); 18 PrintWriter out = response.getWriter(); 19 out.println("<h3>Choose locale and time zone</h3>"); 20 out.println("<form method=\"post\" action=" + 21 "TimeForm>"); 22 out.println("Locale <select size=\"1\" name=\"locale\">"); 23 24 // Fill in all locales 25 for (int i = 0; i < allLocale.length; i++) { 26 out.println("<option value=\"" + i +"\">" + 27 allLocale[i].getDisplayName() + "</option>"); 28 } 29 out.println("</select>"); 30 31 // Fill in all time zones 32 out.println("<p>Time Zone<select size=\"1\" name=\"timezone\">"); 33 for (int i = 0; i < allTimeZone.length; i++) { 34 out.println("<option value=\"" + allTimeZone[i] +"\">" + 35 allTimeZone[i] + "</option>"); 36 } 37 out.println("</select>"); 38 39 out.println("<p><input type=\"submit\" value=\"Submit\" >"); 40 out.println("<input type=\"reset\" value=\"Reset\"></p>"); 41 out.println("</form>"); 42 out.close(); // Close stream 43 } 44 45 /** Process the HTTP Post request */ 46 public void doPost(HttpServletRequest request, HttpServletResponse 47 response) throws ServletException, IOException { 48 response.setContentType(CONTENT_TYPE); 49 PrintWriter out = response.getWriter(); 50 out.println("<html>"); 51 int localeIndex = Integer.parseInt( 52 request.getParameter("locale")); 53 String timeZoneID = request.getParameter("timezone"); 54 out.println("<head><title>Current Time</title></head>"); 55 out.println("<body>"); 56 Calendar calendar = 57 new GregorianCalendar(allLocale[localeIndex]); 58 TimeZone timeZone = TimeZone.getTimeZone(timeZoneID); 59 DateFormat dateFormat = DateFormat.getDateTimeInstance( 60 DateFormat.FULL, DateFormat.FULL, allLocale[localeIndex]); 61 dateFormat.setTimeZone(timeZone); 62 out.println("Current time is " + 63 dateFormat.format(calendar.getTime()) + "</p>"); 64 out.println("</body></html>"); 65 out.close(); // Close stream 66 } 67 }

M37_LIAN1878_11_GE_C37.indd 20 12/29/17 5:55 PM

37.7 Database Programming in Servlets 37-21

are strings. They are displayed in the time zone list. They are also the values for the list. The indexes of the locale and the time zone are passed to the servlet as parameters. The doPost method obtains the values of the parameters (lines 51–53) and finds the current time based on the locale and time zone.

Note If you choose an Asian locale (e.g., Chinese, Korean, or Japanese), the time will not be displayed properly, because the default character encoding is UTF-8. To fix this problem, insert the following statement in line 48 to set an international character encoding:

response.setCharacterEncoding("GB18030");

For information on encoding, see Section 36.6.6, “Character Encoding.”

37.7 Database Programming in Servlets Servlets can access and manipulate databases using JDBC.

Many dynamic Web applications use databases to store and manage data. Servlets can connect to any relational database via JDBC. In Chapter 34, Java Database Programming, you learned how to create Java programs to access and manipulate relational databases via JDBC. Con- necting a servlet to a database is no different from connecting a Java application or applet to a database. If you know Java servlets and JDBC, you can combine them to develop interesting and practical Web-based interactive projects.

To demonstrate connecting to a database from a servlet, let us create a servlet that processes a registration form. The client enters data in an HTML form and submits the form to the server, as shown in Figure 37.19. The result of the submission is shown in Figure 37.20. The server collects the data from the form and stores them in a database.

Point Key

Figure 37.19 The HTML form enables the user to enter student information.

Figure 37.20 The servlet processes the form and stores data in a database.

M37_LIAN1878_11_GE_C37.indd 21 12/29/17 5:55 PM

37-22 Chapter 37 Servlets

The registration data are stored in an Address table consisting of the following fields: firstName, mi, lastName, street, city, state, zip, telephone, and email, defined in the following statement:

create table Address ( firstname varchar(25), mi char(1), lastname varchar(25), street varchar(40), city varchar(20), state varchar(2), zip varchar(5), telephone varchar(10), email varchar(30) )

MySQL, Oracle, and Access were used in Chapter 34. You can use any relational database. If the servlet uses a database driver other than the JDBC-ODBC driver (e.g., the MySQL JDBC driver and the Oracle JDBC driver), you need to add the JDBC driver (e.g., mysqljdbc.jar for MySQL and ojdbc6.jar for Oracle) into the Libraries node in the project.

Create an HTML file named SimpleRegistration.html in Listing 37.6 for collecting the data and sending them to the database using the post method.

Listing 37.6 SimpleRegistration.html 1  2 <html> 3 <head> 4 <title>Simple Registration without Confirmation</title> 5 </head> 6 <body> 7 Please register to your instructor’s student address book. 8 9 <form method = "post" action = "SimpleRegistration"> 10 <p>Last Name <font color = "#FF0000">*</font> 11 <input type = "text" name = "lastName">  12 First Name <font color = "#FF0000">*</font> 13 <input type = "text" name = "firstName">  14 MI <input type = "text" name = "mi" size = "3"> 15 </p> 16 <p>Telephone 17 <input type = "text" name = "telephone" size = "20">  18 Email 19 <input type = "text" name = "email" size = "28">  20 </p> 21 <p>Street <input type = "text" name = "street" size = "50"> 22 </p> 23 <p>City <input type = "text" name = "city" size = "23">  24 State 25 <select size = "1" name = "state"> 26 <option value = "GA">Georgia-GA</option> 27 <option value = "OK">Oklahoma-OK</option> 28 <option value = "IN">Indiana-IN</option> 29 </select>  30 Zip <input type = "text" name = "zip" size = "9"> 31 </p> 32 <p><input type = "submit" name = "Submit" value = "Submit"> 33 <input type = "reset" value = "Reset"> 34 </p> 35 </form>

M37_LIAN1878_11_GE_C37.indd 22 12/29/17 5:55 PM

37.7 Database Programming in Servlets 37-23

36 <p><font color = "#FF0000">* required fields</font></p> 37 </body> 38 </html>

Create the servlet named SimpleRegistration in Listing 37.7.

Listing 37.7 SimpleRegistration.java 1 package chapter37; 2 3 import javax.servlet.*; 4 import javax.servlet.http.*; 5 import java.io.*; 6 import java.sql.*; 7 8 public class SimpleRegistration extends HttpServlet { 9 // Use a prepared statement to store a student into the database 10 private PreparedStatement pstmt; 11 12 /** Initialize variables */ 13 public void init() throws ServletException { 14 initializeJdbc(); 15 } 16 17 /** Process the HTTP Post request */ 18 public void doPost(HttpServletRequest request, HttpServletResponse 19 response) throws ServletException, IOException { 20 response.setContentType("text/html"); 21 PrintWriter out = response.getWriter(); 22 23 // Obtain parameters from the client 24 String lastName = request.getParameter("lastName"); 25 String firstName = request.getParameter("firstName"); 26 String mi = request.getParameter("mi"); 27 String phone = request.getParameter("telephone"); 28 String email = request.getParameter("email"); 29 String address = request.getParameter("street"); 30 String city = request.getParameter("city"); 31 String state = request.getParameter("state"); 32 String zip = request.getParameter("zip"); 33 34 try { 35 if (lastName.length() == 0 || firstName.length() == 0) { 36 out.println("Last Name and First Name are required"); 37 } 38 else { 39 storeStudent(lastName, firstName, mi, phone, email, 40 address, city, state, zip); 41 42 out.println(firstName + " " + lastName + 43 " is now registered in the database"); 44 } 45 } 46 catch(Exception ex) { 47 out.println("Error: " + ex.getMessage()); 48 } 49 finally { 50 out.close(); // Close stream 51 } 52 } 53

M37_LIAN1878_11_GE_C37.indd 23 12/29/17 5:55 PM

37-24 Chapter 37 Servlets

The init method (line 13) is executed once when the servlet starts. After the servlet has started, the servlet can be invoked many times as long as it is alive in the servlet container. Load the driver and connect to the database from the servlet’s init method (line 14). If a prepared statement or a callable statement is used, it should also be created in the init method. In this example, a prepared statement is desirable, because the servlet always uses the same insert statement with different values.

A servlet can connect to any relational database via JDBC. The initializeJdbc method in this example connects to a MySQL database (line 58). Once connected, it creates a prepared statement for inserting a student record into the database. MySQL is used in this example; you can replace it with any relational database.

Last name and first name are required fields. If either of them is empty, the servlet sends an error message to the client (lines 35–36). Otherwise, the servlet stores the data in the database using the prepared statement.

37.7.1 What would be displayed if you changed the content type to html/plain in Listing 37.2, CurrentTime.java?

37.7.2 The statement out.close() is used to close the output stream to response. Why isn’t this statement enclosed in a try-catch block?

Point Check

54 /** Initialize database connection */ 55 private void initializeJdbc() { 56 try { 57 // Load the JDBC driver 58 Class.forName("com.mysql.jdbc.Driver"); 59 System.out.println("Driver loaded"); 60 61 // Establish a connection 62 Connection conn = DriverManager.getConnection 63 ("jdbc:mysql://localhost/javabook", "scott", "tiger"); 64 System.out.println("Database connected"); 65 66 // Create a Statement 67 pstmt = conn.prepareStatement("insert into Address " + 68 "(lastName, firstName, mi, telephone, email, street, city, " 69 + "state, zip) values (?, ?, ?, ?, ?, ?, ?, ?, ?)"); 70 } 71 catch (Exception ex) { 72 ex.printStackTrace(); 73 } 74 } 75 76 /** Store a student record to the database */ 77 private void storeStudent(String lastName, String firstName, 78 String mi, String phone, String email, String address, 79 String city, String state, String zip) throws SQLException { 80 pstmt.setString(1, lastName); 81 pstmt.setString(2, firstName); 82 pstmt.setString(3, mi); 83 pstmt.setString(4, phone); 84 pstmt.setString(5, email); 85 pstmt.setString(6, address); 86 pstmt.setString(7, city); 87 pstmt.setString(8, state); 88 pstmt.setString(9, zip); 89 pstmt.executeUpdate(); 90 } 91 }

M37_LIAN1878_11_GE_C37.indd 24 12/29/17 5:55 PM

37.8 Session Tracking 37-25

37.7.3 What happens when you invoke request.getParameter(paramName) if paramName does not exist?

37.7.4 How do you write a text field, combo box, check box, and text area in an HTML form?

37.7.5 How do you retrieve the parameter value for a text field, combo box, list, check box, radio button, and text area from an HTML form?

37.7.6 If the servlet uses a database driver other than the JDBC-ODBC driver, where should the driver be placed in NetBeans?

37.8 Session Tracking You can perform session tracking using hidden values in a form, using cookies, or using HttpSession.

Web servers use the Hyper-Text Transport Protocol (HTTP). HTTP is a stateless protocol. An HTTP Web server cannot associate requests from a client, and therefore treats each request independently. This protocol works fine for simple Web browsing, where each request typi- cally results in an HTML file or a text file being sent back to the client. Such simple requests are isolated. However, the requests in interactive Web applications are often related. Consider the two requests in the following scenario:

Request 1: A client sends registration data to the server; the server then returns the data to the user for confirmation.

Request 2: The client confirms the data that was submitted in Request 1.

In Request 2, the data submitted in Request 1 are confirmed. These two requests are related in a session. A session can be defined as a series of related interactions between a single client and the Web server over a period of time. Tracking data among requests in a session is known as session tracking.

This section introduces three techniques for session tracking: using hidden values, using cookies, and using the session tracking tools from servlet API.

37.8.1 Session Tracking Using Hidden Values You can track a session by passing data from the servlet to the client as hidden values in a dynamically generated HTML form by including a field like this one:

The next request will submit the data back to the servlet. The servlet retrieves this hidden value just like any other parameter value, using the getParameter method.

Let us use an example to demonstrate using hidden values in a form. The example creates a servlet that processes a registration form. The client submits the form using the GET method, as shown in Figure 37.21. The server collects the data in the form, displays them to the client, and asks the client for confirmation, as shown in Figure 37.22. The client confirms the data by submitting the request with the hidden values using the POST method. Finally, the servlet writes the data to a database.

Create an HTML form named Registration.html in Listing 37.8 for collecting the data and sending it to the database using the GET method for confirmation. This file is almost identical to Listing 37.6, SimpleRegistration.html except that the action is replaced by Registra- tion (line 9).

Point Key

M37_LIAN1878_11_GE_C37.indd 25 12/29/17 5:55 PM

37-26 Chapter 37 Servlets

Figure 37.22 The servlet asks the client for confirmation of the input.

Figure 37.21 The registration form collects user information.

Listing 37.8 Registration.html 1  2 <html> 3 <head> 4 <title>Using Hidden Data for Session Tracking</title> 5 </head> 6 <body> 7 Please register to your instructor’s student address book. 8 9 <form method = "get" action = "Registration"> 10 <p>Last Name <font color = "#FF0000">*</font> 11 <input type = "text" name = "lastName">   12 First Name <font color = "#FF0000">*</font> 13 <input type = "text" name = "firstName">   14 MI <input type = "text" name = "mi" size = "3"> 15 </p> 16 <p>Telephone 17 <input type = "text" name = "telephone" size = "20">   18 Email

M37_LIAN1878_11_GE_C37.indd 26 12/29/17 5:55 PM

37.8 Session Tracking 37-27

Create the servlet named Registration in Listing 37.9.

19 <input type = "text" name = "email" size = "28">   20 </p> 21 <p>Street <input type = "text" name = "street" size = "50"> 22 </p> 23 <p>City <input type = "text" name = "city" size = "23">   24 State 25 <select size = "1" name = "state"> 26 <option value = "GA">Georgia-GA</option> 27 <option value = "OK">Oklahoma-OK</option> 28 <option value = "IN">Indiana-IN</option> 29 </select>   30 Zip <input type = "text" name = "zip" size = "9"> 31 </p> 32 <p><input type = "submit" name = "Submit" value = "Submit"> 33 <input type = "reset" value = "Reset"> 34 </p> 35 </form> 36 <p><font color = "#FF0000">* required fields</font></p> 37 </body> 38 </html>

Listing 37.9 Registration.java 1 package chapter37; 2 3 import javax.servlet.*; 4 import javax.servlet.http.*; 5 import java.io.*; 6 import java.sql.*; 7 8 public class Registration extends HttpServlet { 9 // Use a prepared statement to store a student into the database 10 private PreparedStatement pstmt; 11 12 /** Initialize variables */ 13 public void init() throws ServletException { 14 initializeJdbc(); 15 } 16 17 /** Process the HTTP Get request */ 18 public void doGet(HttpServletRequest request, HttpServletResponse 19 response) throws ServletException, IOException { 20 response.setContentType("text/html"); 21 PrintWriter out = response.getWriter(); 22 23 // Obtain data from the form 24 String lastName = request.getParameter("lastName"); 25 String firstName = request.getParameter("firstName"); 26 String mi = request.getParameter("mi"); 27 String telephone = request.getParameter("telephone"); 28 String email = request.getParameter("email"); 29 String street = request.getParameter("street"); 30 String city = request.getParameter("city"); 31 String state = request.getParameter("state"); 32 String zip = request.getParameter("zip"); 33 34 if (lastName.length() == 0 || firstName.length() == 0) { 35 out.println("Last Name and First Name are required"); 36 }

M37_LIAN1878_11_GE_C37.indd 27 12/29/17 5:55 PM

37-28 Chapter 37 Servlets

37 else { 38 // Ask for confirmation 39 out.println("You entered the following data"); 40 out.println("<p>Last name: " + lastName); 41 out.println("<br>First name: " + firstName); 42 out.println("<br>MI: " + mi); 43 out.println("<br>Telephone: " + telephone); 44 out.println("<br>Email: " + email); 45 out.println("<br>Address: " + street); 46 out.println("<br>City: " + city); 47 out.println("<br>State: " + state); 48 out.println("<br>Zip: " + zip); 49 50 // Set the action for processing the answers 51 out.println("<p><form method=\"post\" action=" + 52 "Registration>"); 53 // Set hidden values 54 out.println("<p><input type=\"hidden\" " + 55 "value=" + lastName + " name=\"lastName\">"); 56 out.println("<p><input type=\"hidden\" " + 57 "value=" + firstName + " name=\"firstName\">"); 58 out.println("<p><input type=\"hidden\" " + 59 "value=" + mi + " name=\"mi\">"); 60 out.println("<p><input type=\"hidden\" " + 61 "value=" + telephone + " name=\"telephone\">"); 62 out.println("<p><input type=\"hidden\" " + 63 "value=" + email + " name=\"email\">"); 64 out.println("<p><input type=\"hidden\" " + 65 "value=" + street + " name=\"street\">"); 66 out.println("<p><input type=\"hidden\" " + 67 "value=" + city + " name=\"city\">"); 68 out.println("<p><input type=\"hidden\" " + 69 "value=" + state + " name=\"state\">"); 70 out.println("<p><input type=\"hidden\" " + 71 "value=" + zip + " name=\"zip\">"); 72 out.println("<p><input type=\"submit\" value=\"Confirm\" >"); 73 out.println("</form>"); 74 } 75 76 out.close(); // Close stream 77 } 78 79 /** Process the HTTP Post request */ 80 public void doPost(HttpServletRequest request, HttpServletResponse 81 response) throws ServletException, IOException { 82 response.setContentType("text/html"); 83 PrintWriter out = response.getWriter(); 84 85 try { 86 String lastName = request.getParameter("lastName"); 87 String firstName = request.getParameter("firstName"); 88 String mi = request.getParameter("mi"); 89 String telephone = request.getParameter("telephone"); 90 String email = request.getParameter("email"); 91 String street = request.getParameter("street"); 92 String city = request.getParameter("city"); 93 String state = request.getParameter("state"); 94 String zip = request.getParameter("zip"); 95 96 storeStudent(lastName, firstName, mi, telephone, email, 97 street, city, state, zip);

M37_LIAN1878_11_GE_C37.indd 28 12/29/17 5:55 PM

37.8 Session Tracking 37-29

The servlet processes the GET request by generating an HTML page that displays the client’s input and asks for the client’s confirmation. The input data consist of hidden values in the newly generated forms, so they will be sent back in the confirmation request. The confirmation request uses the POST method. The servlet retrieves the hidden values and stores them in the database.

Since the first request does not write anything to the database, it is appropriate to use the GET method. Since the second request results in an update to the database, the POST method must be used.

Note The hidden values could also be sent from the URL query string if the request used the GET method.

98 99 out.println(firstName + " " + lastName + 100 " is now registered in the database"); 101 } 102 catch(Exception ex) { 103 out.println("Error: " + ex.getMessage()); 104 } 105 } 106 107 /** Initialize database connection */ 108 private void initializeJdbc() { 109 try { 110 // Load the JDBC driver 111 Class.forName("com.mysql.jdbc.Driver"); 112 System.out.println("Driver loaded"); 113 114 // Establish a connection 115 Connection conn = DriverManager.getConnection 116 ("jdbc:mysql://localhost/javabook" , "scott", "tiger"); 117 System.out.println("Database connected"); 118 119 // Create a Statement 120 pstmt = conn.prepareStatement("insert into Address " + 121 "(lastName, firstName, mi, telephone, email, street, city, " 122 + "state, zip) values (?, ?, ?, ?, ?, ?, ?, ?, ?)"); 123 } 124 catch (Exception ex) { 125 System.out.println(ex); 126 } 127 } 128 129 /** Store a student record to the database */ 130 private void storeStudent(String lastName, String firstName, 131 String mi, String phone, String email, String address, 132 String city, String state, String zip) throws SQLException { 133 pstmt.setString(1, lastName); 134 pstmt.setString(2, firstName); 135 pstmt.setString(3, mi); 136 pstmt.setString(4, phone); 137 pstmt.setString(5, email); 138 pstmt.setString(6, address); 139 pstmt.setString(7, city); 140 pstmt.setString(8, state); 141 pstmt.setString(9, zip); 142 pstmt.executeUpdate(); 143 } 144 }

M37_LIAN1878_11_GE_C37.indd 29 12/29/17 5:55 PM

37-30 Chapter 37 Servlets

37.8.2 Session Tracking Using Cookies You can track sessions using cookies, which are small text files that store sets of name/value pairs on the disk in the client’s computer. Cookies are sent from the server through the instruc- tions in the header of the HTTP response. The instructions tell the browser to create a cookie with a given name and its associated value. If the browser already has a cookie with the key name, the value will be updated. The browser will then send the cookie with any request submitted to the same server. Cookies can have expiration dates set, after which they will not be sent to the server. The javax.servlet.http.Cookie is used to create and manipulate cookies, as shown in Figure 37.23.

Figure 37.23 Cookie stores a name/value pair and other information about the cookie.

javax.servlet.http.Cookie

+Cookie(name: String, value: String)

+getName(): String

+getValue(): String

+setValue(newValue: String): void

+getMaxAge(): int

+setMaxAge(expiration: int): void

+getSecure(): boolean

+setSecure(flag: boolean): void

+getComment(): String

+setComment(purpose: String): void

Creates a cookie with the specified name-value pair.

Returns the name of the cookie.

Returns the value of the cookie.

Assigns a new value to a cookie after the cookie is created.

Returns the maximum age of the cookie, specified in seconds.

Specifies the maximum age of the cookie. By default, this value is –1, which implies that the cookie persists until the browser exits. If you set this value to 0, the cookie is deleted.

Returns true if the browser is sending cookies only over a secure protocol.

Indicates to the browser whether the cookie should only be sent using a secure protocol, such as HTTPS or SSL.

Returns the comment describing the purpose of this cookie, or null if the cookie has no comment.

Sets the comment for this cookie.

To send a cookie to the browser, use the addCookie method in the HttpServlet- Response class, as shown below:

response.addCookie(cookie);

where response is an instance of HttpServletResponse. To obtain cookies from a browser, use

request.getCookies();

where request is an instance of HttpServletRequest. To demonstrate the use of cookies, let us create an example that accomplishes the same

task as Listing 37.9, Registration.java. Instead of using hidden values for session tracking, it uses cookies.

Create the servlet named RegistrationWithCookie in Listing 37.10. Create an HTML file named RegistrationWithCookie.html that is identical to Registration.html except that the action is replaced by RegistrationWithCookie.java.

Listing 37.10 RegistrationWithCookie.java 1 package chapter37; 2 3 import javax.servlet.*; 4 import javax.servlet.http.*; 5 import java.io.*;

M37_LIAN1878_11_GE_C37.indd 30 12/29/17 5:55 PM

37.8 Session Tracking 37-31

6 import java.sql.*; 7 8 public class RegistrationWithCookie extends HttpServlet { 9 private static final String CONTENT_TYPE = "text/html"; 10 // Use a prepared statement to store a student into the database 11 private PreparedStatement pstmt; 12 13 /** Initialize variables */ 14 public void init() throws ServletException { 15 initializeJdbc(); 16 } 17 18 /** Process the HTTP Get request */ 19 public void doGet(HttpServletRequest request, HttpServletResponse 20 response) throws ServletException, IOException { 21 response.setContentType("text/html"); 22 PrintWriter out = response.getWriter(); 23 24 // Obtain data from the form 25 String lastName = request.getParameter("lastName"); 26 String firstName = request.getParameter("firstName"); 27 String mi = request.getParameter("mi"); 28 String telephone = request.getParameter("telephone"); 29 String email = request.getParameter("email"); 30 String street = request.getParameter("street"); 31 String city = request.getParameter("city"); 32 String state = request.getParameter("state"); 33 String zip = request.getParameter("zip"); 34 35 if (lastName.length() == 0 || firstName.length() == 0) { 36 out.println("Last Name and First Name are required"); 37 } 38 else { 39 // Create cookies and send cookies to browsers 40 Cookie cookieLastName = new Cookie("lastName", lastName); 41 // cookieLastName.setMaxAge(1000); 42 response.addCookie(cookieLastName); 43 Cookie cookieFirstName = new Cookie("firstName", firstName); 44 response.addCookie(cookieFirstName); 45 // cookieFirstName.setMaxAge(0); 46 Cookie cookieMi = new Cookie("mi", mi); 47 response.addCookie(cookieMi); 48 Cookie cookieTelephone = new Cookie("telephone", telephone); 49 response.addCookie(cookieTelephone); 50 Cookie cookieEmail = new Cookie("email", email); 51 response.addCookie(cookieEmail); 52 Cookie cookieStreet = new Cookie("street", street); 53 response.addCookie(cookieStreet); 54 Cookie cookieCity = new Cookie("city", city); 55 response.addCookie(cookieCity); 56 Cookie cookieState = new Cookie("state", state); 57 response.addCookie(cookieState); 58 Cookie cookieZip = new Cookie("zip", zip); 59 response.addCookie(cookieZip); 60 61 // Ask for confirmation 62 out.println("You entered the following data"); 63 out.println("<p>Last name: " + lastName); 64 out.println("<br>First name: " + firstName); 65 out.println("<br>MI: " + mi); 66 out.println("<br>Telephone: " + telephone);

M37_LIAN1878_11_GE_C37.indd 31 12/29/17 5:55 PM

37-32 Chapter 37 Servlets

67 out.println("<br>Email: " + email); 68 out.println("<br>Street: " + street); 69 out.println("<br>City: " + city); 70 out.println("<br>State: " + state); 71 out.println("<br>Zip: " + zip); 72 73 // Set the action for processing the answers 74 out.println("<p><form method=\"post\" action=" + 75 "RegistrationWithCookie>"); 76 out.println("<p><input type=\"submit\" value=\"Confirm\" >"); 77 out.println("</form>"); 78 } 79 80 out.close(); // Close stream 81 } 82 83 /** Process the HTTP Post request */ 84 public void doPost(HttpServletRequest request, HttpServletResponse 85 response) throws ServletException, IOException { 86 response.setContentType(CONTENT_TYPE); 87 PrintWriter out = response.getWriter(); 88 89 String lastName = ""; 90 String firstName = ""; 91 String mi = ""; 92 String telephone = ""; 93 String email = ""; 94 String street = ""; 95 String city = ""; 96 String state = ""; 97 String zip = ""; 98 99 // Read the cookies 100 Cookie[] cookies = request.getCookies(); 101 102 // Get cookie values 103 for (int i = 0; i < cookies.length; i++) { 104 if (cookies[i].getName().equals("lastName")) 105 lastName = cookies[i].getValue(); 106 else if (cookies[i].getName().equals("firstName")) 107 firstName = cookies[i].getValue(); 108 else if (cookies[i].getName().equals("mi")) 109 mi = cookies[i].getValue(); 110 else if (cookies[i].getName().equals("telephone")) 111 telephone = cookies[i].getValue(); 112 else if (cookies[i].getName().equals("email")) 113 email = cookies[i].getValue(); 114 else if (cookies[i].getName().equals("street")) 115 street = cookies[i].getValue(); 116 else if (cookies[i].getName().equals("city")) 117 city = cookies[i].getValue(); 118 else if (cookies[i].getName().equals("state")) 119 state = cookies[i].getValue(); 120 else if (cookies[i].getName().equals("zip")) 121 zip = cookies[i].getValue(); 122 } 123 124 try { 125 storeStudent(lastName, firstName, mi, telephone, email, street, 126 city, state, zip); 127

M37_LIAN1878_11_GE_C37.indd 32 12/29/17 5:55 PM

37.8 Session Tracking 37-33

You have to create a cookie for each value you want to track, using the Cookie class’s only constructor, which defines a cookie’s name and value as shown below (line 40):

Cookie cookieLastName = new Cookie("lastName", lastName);

To send the cookie to the browser, use a statement like this one (line 42):

response.addCookie(cookieLastName);

If a cookie with the same name already exists in the browser, its value is updated; otherwise, a new cookie is created.

128 out.println(firstName + " " + lastName + 129 " is now registered in the database"); 130 131 out.close(); // Close stream 132 } 133 catch(Exception ex) { 134 out.println("Error: " + ex.getMessage()); 135 } 136 } 137 138 /** Initialize database connection */ 139 private void initializeJdbc() { 140 try { 141 // Load the JDBC driver 142 Class.forName("com.mysql.jdbc.Driver"); 143 System.out.println("Driver loaded"); 144 145 // Establish a connection 146 Connection conn = DriverManager.getConnection 147 ("jdbc:mysql://localhost/javabook" , "scott", "tiger"); 148 System.out.println("Database connected"); 149 150 // Create a Statement 151 pstmt = conn.prepareStatement("insert into Address " + 152 "(lastName, firstName, mi, telephone, email, street, city, " 153 + "state, zip) values (?, ?, ?, ?, ?, ?, ?, ?, ?)"); 154 } 155 catch (Exception ex) { 156 System.out.println(ex); 157 } 158 } 159 160 /** Store a student record to the database */ 161 private void storeStudent(String lastName, String firstName, 162 String mi, String telephone, String email, String street, 163 String city, String state, String zip) throws SQLException { 164 pstmt.setString(1, lastName); 165 pstmt.setString(2, firstName); 166 pstmt.setString(3, mi); 167 pstmt.setString(4, telephone); 168 pstmt.setString(5, email); 169 pstmt.setString(6, street); 170 pstmt.setString(7, city); 171 pstmt.setString(8, state); 172 pstmt.setString(9, zip); 173 pstmt.executeUpdate(); 174 } 175 }

M37_LIAN1878_11_GE_C37.indd 33 12/29/17 5:55 PM

37-34 Chapter 37 Servlets

Cookies are automatically sent to the Web server with each request from the client. The servlet retrieves all the cookies into an array using the getCookies method (line 100):

Cookie[] cookies = request.getCookies();

To obtain the name of the cookie, use the getName method (line 104):

String name = cookies[i].getName();

The cookie’s value can be obtained using the getValue method:

String value = cookies[i].getValue();

Cookies are stored as strings just like form parameters and hidden values. If a cookie represents a numeric value, you have to convert it into an integer or a double, using the parseInt method in the Integer class or the parseDouble method in the Double class.

By default, a newly created cookie persists until the browser exits. However, you can set an expiration date, using the setMaxAge method, to allow a cookie to stay in the browser for up to 2,147,483,647 seconds (approximately 24,855 days).

37.8.3 Session Tracking Using the Servlet API You have now learned both session tracking using hidden values and session tracking using cookies. These two session-tracking methods have problems. They send data to the browser either as hidden values or as cookies. The data are not secure, and anybody with knowledge of computers can obtain them. The hidden data are in HTML form, which can be viewed from the browser. Cookies are stored in the Cache directory of the browser. Because of security concerns, some browsers do not accept cookies. The client can turn the cookies off and limit their number. Another problem is that hidden data and cookies pass data as strings. You cannot pass objects using these two methods.

To address these problems, Java servlet API provides the javax.servlet.http . HttpSession interface, which provides a way to identify a user across more than one page request or visit to a website and to store information about that user. The servlet container uses this interface to create a session between an HTTP client and an HTTP server. The session persists for a specified time period, across more than one connection or page request from the user. A session usually corresponds to one user, who may visit a site many times. The session enables tracking of a large set of data. The data can be stored as objects and are secure because they are kept on the server side.

To use the Java servlet API for session tracking, first create a session object using the getSession() method in the HttpServletRequest interface:

HttpSession session = request.getSession();

This obtains the session or creates a new session if the client does not have a session on the server.

The HttpSession interface provides the methods for reading and storing data to the session, and for manipulating the session, as shown in Figure 37.24.

Note HTTP is stateless. So how does the server associate a session with multiple requests from the same client? This is handled behind the scenes by the servlet container and is transparent to the servlet programmer.

To demonstrate using HttpSession, let us rewrite Listing 37.9, Registration.java, and List- ing 37.10, RegistrationWithCookie.java. Instead of using hidden values or cookies for session tracking, it uses servlet HttpSession.

M37_LIAN1878_11_GE_C37.indd 34 12/29/17 5:55 PM

37.8 Session Tracking 37-35

Create the servlet named RegistrationWithHttpSession in Listing 37.11. Create an HTML file named RegistrationWithHttpSession.html that is identical to Regis- tration.html except that the action is replaced by RegistrationWithHttpSession.

Figure 37.24 HttpSession establishes a persistent session between a client with multiple requests and the server.

<<interface>> javax.servlet.http.HttpSession

+getAttribute(name: String): Object

+setAttribute(name: String, value: Object): void

+getId(): String

+getLastAccessedTime(): long

+invalidate(): void

+isNew(): boolean

+removeAttribute(name: String): void

+getMaxInactiveInterval(): int +setMaxInactiveInterval(interval: int): void

Returns the object bound with the specified name in this session, or null if no object is bound under the name.

Binds an object to this session, using the specified name. If an object of the same name is already bound to the session, the object is replaced.

Returns a string containing the unique identifier assigned to this session. The identifier is assigned by the servlet container and is implementation dependent.

Returns the last time the client sent a request associated with this session, as the number of milliseconds since midnight January 1, 1970 GMT, and marked by the time the container received the request.

Invalidates this session, then unbinds any objects bound to it.

Returns true if the session was just created in the current request.

Removes the object bound with the specified name from this ses- sion. If the session does not have an object bound with the specified name, this method does nothing.

Returns the time, in seconds, between client requests before the servlet container will invalidate this session. A negative time indicates that the session will never time-out. Use setMaxInac- tiveInterval to specify this value.

Listing 37.11 RegistrationWithHttpSession.java 1 package chapter37; 2 3 import javax.servlet.*; 4 import javax.servlet.http.*; 5 import java.io.*; 6 import java.sql.*; 7 8 public class RegistrationWithHttpSession extends HttpServlet { 9 // Use a prepared statement to store a student into the database 10 private PreparedStatement pstmt; 11 12 /** Initialize variables */ 13 public void init() throws ServletException { 14 initializeJdbc(); 15 } 16 17 /** Process the HTTP Get request */ 18 public void doGet(HttpServletRequest request, HttpServletResponse 19 response) throws ServletException, IOException {

M37_LIAN1878_11_GE_C37.indd 35 12/29/17 5:55 PM

37-36 Chapter 37 Servlets

20 // Set response type and output stream to the browser 21 response.setContentType("text/html"); 22 PrintWriter out = response.getWriter(); 23 24 // Obtain data from the form 25 String lastName = request.getParameter("lastName"); 26 String firstName = request.getParameter("firstName"); 27 String mi = request.getParameter("mi"); 28 String telephone = request.getParameter("telephone"); 29 String email = request.getParameter("email"); 30 String street = request.getParameter("street"); 31 String city = request.getParameter("city"); 32 String state = request.getParameter("state"); 33 String zip = request.getParameter("zip"); 34 35 if (lastName.length() == 0 || firstName.length() == 0) { 36 out.println("Last Name and First Name are required"); 37 } 38 else { 39 // Create an Address object 40 Address address = new Address(); 41 address.setLastName(lastName); 42 address.setFirstName(firstName); 43 address.setMi(mi); 44 address.setTelephone(telephone); 45 address.setEmail(email); 46 address.setStreet(street); 47 address.setCity(city); 48 address.setState(state); 49 address.setZip(zip); 50 51 // Get an HttpSession or create one if it does not exist 52 HttpSession httpSession = request.getSession(); 53 54 // Store student object to the session 55 httpSession.setAttribute("address", address); 56 57 // Ask for confirmation 58 out.println("You entered the following data"); 59 out.println("<p>Last name: " + lastName); 60 out.println("<p>First name: " + firstName); 61 out.println("<p>MI: " + mi); 62 out.println("<p>Telephone: " + telephone); 63 out.println("<p>Email: " + email); 64 out.println("<p>Address: " + street); 65 out.println("<p>City: " + city); 66 out.println("<p>State: " + state); 67 out.println("<p>Zip: " + zip); 68 69 // Set the action for processing the answers 70 out.println("<p><form method=\"post\" action=" + 71 "RegistrationWithHttpSession>"); 72 out.println("<p><input type=\"submit\" value=\"Confirm\" >"); 73 out.println("</form>"); 74 } 75 76 out.close(); // Close stream 77 } 78 79 /** Process the HTTP Post request */

M37_LIAN1878_11_GE_C37.indd 36 12/29/17 5:55 PM

37.8 Session Tracking 37-37

80 public void doPost(HttpServletRequest request, HttpServletResponse 81 response) throws ServletException, IOException { 82 // Set response type and output stream to the browser 83 response.setContentType("text/html"); 84 PrintWriter out = response.getWriter(); 85 86 // Obtain the HttpSession 87 HttpSession httpSession = request.getSession(); 88 89 // Get the Address object in the HttpSession 90 Address address = (Address)(httpSession.getAttribute("address")); 91 92 try { 93 storeStudent(address); 94 95 out.println(address.getFirstName() + " " + address.getLastName() 96 + " is now registered in the database"); 97 out.close(); // Close stream 98 } 99 catch(Exception ex) { 100 out.println("Error: " + ex.getMessage()); 101 } 102 } 103 104 /** Initialize database connection */ 105 private void initializeJdbc() { 106 try { 107 // Load the JDBC driver 108 Class.forName("com.mysql.jdbc.Driver"); 109 System.out.println("Driver loaded"); 110 111 // Establish a connection 112 Connection conn = DriverManager.getConnection 113 ("jdbc:mysql://localhost/javabook" , "scott", "tiger"); 114 System.out.println("Database connected"); 115 116 // Create a Statement 117 pstmt = conn.prepareStatement("insert into Address " + 118 "(lastName, firstName, mi, telephone, email, street, city, " 119 + "state, zip) values (?, ?, ?, ?, ?, ?, ?, ?, ?)"); 120 } 121 catch (Exception ex) { 122 System.out.println(ex); 123 } 124 } 125 126 /** Store an address to the database */ 127 private void storeStudent(Address address) throws SQLException { 128 pstmt.setString(1, address.getLastName()); 129 pstmt.setString(2, address.getFirstName()); 130 pstmt.setString(3, address.getMi()); 131 pstmt.setString(4, address.getTelephone()); 132 pstmt.setString(5, address.getEmail()); 133 pstmt.setString(6, address.getStreet()); 134 pstmt.setString(7, address.getCity()); 135 pstmt.setString(8, address.getState()); 136 pstmt.setString(9, address.getZip()); 137 pstmt.executeUpdate(); 138 } 139 }

M37_LIAN1878_11_GE_C37.indd 37 12/29/17 5:55 PM

37-38 Chapter 37 Servlets

The statement (line 52)

HttpSession httpSession = request.getSession();

obtains a session, or creates a new session if the session does not exist. Since objects can be stored in HttpSession, this program defines an Address class. An

Address object is created and is stored in the session using the setAttribute method, which binds the object with a name like the one shown below (line 55):

httpSession.setAttribute("address", address);

To retrieve the object, use the following statement (line 90):

Address address = (Address)(httpSession.getAttribute("address"));

There is only one session between a client and a servlet. You can store any number of objects in a session. By default, the maximum inactive interval on many Web servers including Tomcat and GlassFish is 1800 seconds (i.e., a half-hour), meaning that the session expires if there is no activity for 30 minutes. You can change the default using the setMaxInactive Interval method. For example, to set the maximum inactive interval to one hour, use

httpSession.setMaxInactiveInterval(3600);

If you set a negative value, the session will never expire. For this servlet program to work, you have to create the Address class in NetBeans, as

follows:

1. Choose New, Java Class from the context menu of the liangweb node in the project pane to display the New Java Class dialog box.

2. Enter Address as the Class Name and chapter37 as the package name. Click Finish to create the class.

3. Enter the code, as shown in Listing 37.12.

Listing 37.12 Address.java 1 package chapter37; 2 3 public class Address { 4 private String firstName; 5 private String mi; 6 private String lastName; 7 private String telephone; 8 private String street; 9 private String city; 10 private String state; 11 private String email; 12 private String zip; 13 14 public String getFirstName() { 15 return this.firstName; 16 } 17 18 public void setFirstName(String firstName) { 19 this.firstName = firstName; 20 } 21 22 public String getMi() { 23 return this.mi; 24 }

M37_LIAN1878_11_GE_C37.indd 38 12/29/17 5:55 PM

37.8 Session Tracking 37-39

25 26 public void setMi(String mi) { 27 this.mi = mi; 28 } 29 30 public String getLastName() { 31 return this.lastName; 32 } 33 34 public void setLastName(String lastName) { 35 this.lastName = lastName; 36 } 37 38 public String getTelephone() { 39 return this.telephone; 40 } 41 42 public void setTelephone(String telephone) { 43 this.telephone = telephone; 44 } 45 46 public String getEmail() { 47 return this.email; 48 } 49 50 public void setEmail(String email) { 51 this.email = email; 52 } 53 54 public String getStreet() { 55 return this.street; 56 } 57 58 public void setStreet(String street) { 59 this.street = street; 60 } 61 62 public String getCity() { 63 return this.city; 64 } 65 66 public void setCity(String city) { 67 this.city = city; 68 } 69 70 public String getState() { 71 return this.state; 72 } 73 74 public void setState(String state) { 75 this.state = state; 76 } 77 78 public String getZip() { 79 return this.zip; 80 } 81 82 public void setZip(String zip) { 83 this.zip = zip; 84 } 85 }

M37_LIAN1878_11_GE_C37.indd 39 12/29/17 5:55 PM

37-40 Chapter 37 Servlets

This support class will also be reused in the upcoming chapters.

37.8.1 What is session tracking? What are three techniques for session tracking? 37.8.2 How do you create a cookie, send a cookie to a browser, get cookies from a

browser, get the name of a cookie, set a new value in the cookie, and set cookie expiration time?

37.8.3 Do you have to create five Cookie objects in the servlet in order to send five cookies to the browser?

37.8.4 How do you get a session, set object value for the session, and get object value from the session?

37.8.5 Suppose you inserted the following code in line 53 in Listing 37.11:

httpSession.setMaxInactiveInterval(1);

What would happen after the user clicked the Confirm button from the browser? Test your answer by running the program.

37.8.6 Suppose you inserted the following code in line 53 in Listing 37.11:

httpSession.setMaxInactiveInterval(-1);

What would happen after the user clicked the Confirm button from the browser?

Point Check

Key Terms

Common Gateway Interface 37-3 CGI programs 37-3 cookie 37-30 GET and POST methods 37-3 GlassFish 37-5 HTML form 37-15

URL query string 37-3 servlet 37-2 servlet container (servlet engine) 37-4 servelt life-cycle methods 37-10 Tomcat 37-5

ChapTer summary 1. A servlet is a special kind of program that runs from a Web server. Tomcat and GlassFish

are Web servers that can run servlets.

2. A servlet URL is specified by the host name, port, and request string (e.g., http:// localhost:8084/liangweb/ServletClass). There are several ways to invoke a servlet: (1) by typing a servlet URL from a Web browser, (2) by placing a hyper link in an HTML page, and (3) by embedding a servlet URL in an HTML form. All the requests trigger the GET method, except that in the HTML form you can explicitly specify the POST method.

3. You develop a servlet by defining a class that extends the HttpServlet class, imple- ments the doGet(HttpServletRequest, HttpServletResponse) method to respond to the GET method, and implements the doPost(HttpServletRequest, HttpServletResponse) method to respond to the POST method.

M37_LIAN1878_11_GE_C37.indd 40 12/29/17 5:55 PM

Programming Exercises 37-41

4. The request information passed from a client to the servlet is contained in an object of HttpServletRequest. You can use the methods getParameter, getParam- eterValues, getRemoteAddr, getRemoteHost, getHeader, getQueryString, getCookies, and getSession to obtain the information from the request.

5. The content sent back to the client is contained in an object of HttpServletResponse. To send content to the client, first set the type of the content (e.g., html/plain) using the setContentType(contentType) method, then output the content through an I/O stream on the HttpServletResponse object. You can obtain a character PrintWriter stream using the getWriter() method and obtain a binary Output- Stream using the getOutputStream() method.

6. A servlet may be shared by many clients. When the servlet is first created, its init method is called. It is not called again as long as the servlet is not destroyed. The ser- vice method is invoked each time the server receives a request for the servlet. The server spawns a new thread and invokes service. The destroy method is invoked after a timeout period has passed or the Web server is stopped.

7. There are three ways to track a session. You can track a session by passing data from the servlet to the client as a hidden value in a dynamically generated HTML form by includ- ing a field such as <input type="hidden" name="lastName" value="Smith">. The next request will submit the data back to the servlet. The servlet retrieves this hidden value just like any other parameter value using the getParameter method.

8. You can track sessions using cookies. A cookie is created using the constructor new Cookie(String name, String value). Cookies are sent from the server through the object of HttpServletResponse using the addCookie(aCookie) method to tell the browser to add a cookie with a given name and its associated value. If the browser already has a cookie with the key name, the value will be updated. The browser will then send the cookie with any request submitted to the same server. Cookies can have expiration dates set, after which they will not be sent to the server.

9. Java servlet API provides a session-tracking tool that enables tracking of a large set of data. A session can be obtained using the getSession() method through an HttpServletRequest object. The data can be stored as objects and are secure because they are kept on the server side using the setAttribute(String name, Object value) method.

Quiz Answer the quiz for this chapter online at the book Companion Website.

programming exerCises

Section 37.5 *37.1 (Factorial table) Write a servlet to display a table that contains factorials for the

numbers from 0 to 10, as shown in Figure 37.25.

M37_LIAN1878_11_GE_C37.indd 41 12/29/17 5:55 PM

37-42 Chapter 37 Servlets

Figure 37.25 (a) The servlet displays factorials for the numbers from 0 to 10 in a table. (b) The servlet displays the multiplication table.

(a) (b)

*37.2 (Multiplication table) Write a servlet to display a multiplication table, as shown in Figure 37.25b.

*37.3 (Visit count) Develop a servlet that displays the number of visits on the servlet. Also display the client’s host name and IP address, as shown in Figure 37.26.

Figure 37.26 The servlet displays the number of visits and the client’s host name, IP address, and request URL.

Implement this program in three different ways:

1. Use an instance variable to store count. When the servlet is created for the first time, count is 0. count is incremented every time the servlet’s doGet method is invoked. When the Web server stops, count is lost.

2. Store the count in a file named Exercise39_3.dat, and use RandomAccess- File to read the count in the servlet’s init method. The count is incremented every time the servlet’s doGet method is invoked. When the Web server stops, store the count back to the file.

3. Instead of counting total visits from all clients, count the visits by each client identified by the client’s IP address. Use Map to store a pair of IP addresses and visit counts. For the first visit, an entry is created in the map. For subsequent visits, the visit count is updated.

Section 37.6 *37.4 (Calculate tax) Write an HTML form to prompt the user to enter taxable income

and filing status, as shown in Figure 37.27a. Clicking the Compute Tax button invokes a servlet to compute and display the tax, as shown in Figure 37.27b. Use the computeTax method introduced in Listing 3.7, ComputingTax.java, to compute tax.

M37_LIAN1878_11_GE_C37.indd 42 12/29/17 5:55 PM

Programming Exercises 37-43

*37.5 (Calculate loan) Write an HTML form that prompts the user to enter loan amount, interest rate, and number of years, as shown in Figure 37.28a. Clicking the Compute Loan Payment button invokes a servlet to compute and display the monthly and total loan payments, as shown in Figure 37.28b. Use the Loan class given in Listing 10.2, Loan.java, to compute the monthly and total payments.

Figure 37.27 The servlet computes the tax.

(a) (b)

Figure 37.28 The servlet computes the loan payment.

(a) (b)

**37.6 (Find scores from text files) Write a servlet that displays the student name and the current score, given the SSN and class ID. For each class, a text file is used to store the student name, SSN, and current score. The file is named after the class ID with .txt extension. For instance, if the class ID were csci1301, the file name would be csci1301.txt. Suppose each line consists of student name, SSN, and score. These three items are separated by the # sign. Create an HTML form that enables the user to enter the SSN and class ID, as shown in Figure 37.29a. Upon clicking the Submit button, the result is displayed, as shown in Figure 37.29b. If the SSN or the class ID does not match, report an error. Assume three courses are available: CSCI1301, CSCI1302, and CSCI3720.

Figure 37.29 The HTML form accepts the SSN and class ID from the user and sends them to the servlet to obtain the score.

(a) (b)

M37_LIAN1878_11_GE_C37.indd 43 12/29/17 5:55 PM

37-44 Chapter 37 Servlets

Section 37.7 **37.7 (Find scores from database tables) Rewrite the preceding servlet. Assume for each

class, a table is used to store the student name, ssn, and score. The table name is the same as the class ID. For instance, if the class ID were csci1301, the table name would be csci1301.

*37.8 (Change the password) Write a servlet that enables the user to change the password from an HTML form, as shown in Figure 37.30a. Suppose the user information is stored in a database table named Account with three columns: username, password, and name, where name is the real name of the user. The servlet performs the fol- lowing tasks:

a. Verify that the username and old password are in the table. If not, report the error and redisplay the HTML form.

b. Verify that the new password and the confirmed password are the same. If not, report this error and redisplay the HTML form.

c. If the user information is entered correctly, update the password and report the status of the update to the user, as shown in Figure 37.30b.

Figure 37.30 The user enters the username and the old password and sets a new password. The servlet reports the status of the update to the user.

(a) (b)

**37.9 (Display database tables) Write an HTML form that prompts the user to enter or select a JDBC driver, database URL, username, password, and table name, as shown in Figure 37.31a. Clicking the Submit button displays the table content, as shown in Figure 37.31b.

Figure 37.31 The user enters database information and specifies a table to display its content.

(a) (b)

M37_LIAN1878_11_GE_C37.indd 44 12/29/17 5:55 PM

Programming Exercises 37-45

Section 37.8 *37.10 (Store cookies) Write a servlet that stores the following cookies in a browser, and

set their max age for two days.

Cookie 1: name is “color” and value is red. Cookie 2: name is “radius” and value is 5.5. Cookie 3: name is “count” and value is 2.

*37.11 (Retrieve cookies) Write a servlet that displays all the cookies on the client. The client types the URL of the servlet from the browser to display all the cookies stored on the browser. (see Figure 37.32.)

Figure 37.32 All the cookies on the client are displayed in the browser.

Comprehensive ***37.12 (Syntax highlighting) Create an HTML form that prompts the user to enter a Java

program in a text area, as shown in Figure 37.33a. The form invokes a servlet that displays the Java source code in a syntax-highlighted HTML format, as shown in Figure 37.33b. The keywords, comments, and literals are displayed in bold navy, green, and blue, respectively.

Figure 37.33 The Java code in plain text in (a) is displayed in HTML with syntax highlighted in (b).

(a) (b)

M37_LIAN1878_11_GE_C37.indd 45 12/29/17 5:55 PM

37-46 Chapter 37 Servlets

**37.13 (Access and update a Staff table) Write a Java servlet for Exercise 33.1, as shown in Figure 37.34.

Figure 37.34 The webpage lets you view, insert, and update staff information.

***37.14 (Opinion poll) Create an HTML form that prompts the user to answer a question such as “Are you a CS major?”, as shown in Figure 37.35a. When the Submit button is clicked, the servlet increases the Yes or No count in a database and displays the current Yes and No counts, as shown in Figure 37.35b.

Figure 37.35 The HTML form prompts the user to enter Yes or No for a question in (a), and the servlet updates the Yes or No counts (b).

(a) (b)

Create a table named Poll, as follows:

create table Poll ( question varchar(40) primary key, yesCount int, noCount int);

Insert one row into the table, as follows:

insert into Poll values ('Are you a CS major? ', 0, 0);

M37_LIAN1878_11_GE_C37.indd 46 12/29/17 5:55 PM

Objectives ■■ To create a simple JSP page (§38.2).

■■ To explain how a JSP page is processed (§38.3).

■■ To use JSP constructs to code JSP script (§38.4).

■■ To use predefined variables and directives in JSP (§§38.5–38.6).

■■ To use JavaBeans components in JSP (§38.7).

■■ To get and set JavaBeans properties in JSP (§38.8).

■■ To associate JavaBeans properties with input parameters (§38.9).

■■ To forward requests from one JSP page to another (§38.10).

■■ To develop an application for browsing database tables using JSP (§38.11).

Javaserver Pages

Chapter

M38_LIAN1878_11_GE_C38.indd 1 12/29/17 5:56 PM

38-2 Chapter 38 Javaserver Pages

38.1 Introduction JavaServer Pages are the Java scripts and code embedded in an HTML file.

Servlets can be used to generate dynamic Web content. One drawback, however, is that you have to embed HTML tags and text inside the Java source code. Using servlets, you have to modify the Java source code and recompile it if changes are made to the HTML text. If you have a lot of HTML script in a servlet, the program is difficult to read and maintain, since the HTML text is part of the Java program. JavaServer Pages (JSP) was introduced to remedy this drawback. JSP enables you to write regular HTML script in the normal way and embed Java code to produce dynamic content.

38.2 Creating a Simple JSP Page An IDE such an NetBeans is an effecitve tools for creating JavaServer Pages.

JSP provides an easy way to create dynamic webpages and simplify the task of building Web applications. A JavaServer page is like a regular HTML page with special tags, known as JSP tags, which enable the Web server to generate dynamic content. You can create a webpage with HTML script and enclose the Java code for generating dynamic content in the JSP tags. Here is an example of a simple JSP page:

<html> <head> <title> CurrentTime </title> </head> <body> Current time is <%= new java.util.Date() %> </body> </html>

The dynamic content is enclosed in the tag that begins with <%= and ends with %>. The current time is returned as a string by invoking the toString method of an object of the java.util. Date class.

An IDE like NetBeans can greatly simplify the task of developing JSP. To create JSP in NetBeans, first you need to create a Web project. A Web project named liangweb was created in the preceding chapter. For convenience, this chapter will create JSP in the liangweb project.

Here are the steps to create and run CurrentTime.jsp:

1. Right-click the liangweb node in the project pane and choose New, JSP to display the New JSP dialog box, as shown in Figure 38.1.

2. Enter CurrentTime in the JSP File Name field and click Finish. You will see Current- Time.jsp appearing under the webpages node in liangweb.

3. Complete the code for CurrentTime.jsp, as shown in Figure 38.2.

4. Right-click CurrentTime.jsp in the project pane and choose Run File. You will see the JSP page displayed in a Web browser, as shown in Figure 38.3.

Note Like servlets, you can develop JSP in NetBeans, create a .war file, and then deploy the .war file in a Java Web server such as Tomcat and GlassFish.

Point Key

M38_LIAN1878_11_GE_C38.indd 2 12/29/17 5:56 PM

38.2 Creating a Simple JSP Page 38-3

Figure 38.1 You can create a JSP page using NetBeans.

Figure 38.2 A template for a JSP page is created.

Figure 38.3 The result from a JSP page is displayed in a Web browser.

M38_LIAN1878_11_GE_C38.indd 3 12/29/17 5:56 PM

38-4 Chapter 38 Javaserver Pages

38.3 How Is a JSP Page Processed? JavaServer Pages are preprocessed and compiled into Java servlets by a Java Web server.

A JSP page must first be processed by a Web server before it can be displayed in a Web browser. The Web server must support JSP, and the JSP page must be stored in a file with a .jsp extension. The Web server translates the .jsp file into a Java servlet, compiles the servlet, and executes it. The result of the execution is sent to the browser for display. Figure 38.4 shows how a JSP page is processed by a Web server.

Note A JSP page is translated into a servlet when the page is requested for the first time. It is not retranslated if the page is not modified. To ensure that the first-time real user does not encounter a delay, JSP developers should test the page after it is installed.

Point Key

Figure 38.4 A JSP page is translated into a servlet.

Web Browser

HTML Page returned

Send a request URL Web Server

Servlet Engine

Web Server Host

Host Machine File System

Get Servlet

Get JSP �le

Generate Response

Generated Servlet

Process Servlet

JSP Translator

URL Example http://www.server.com:8080/servlet/JSPFile.jsp

/servlet/JSPFile.jsp

38.2.1 What is the file-name extension of a JavaServer page? How is a JSP page processed?

38.2.2 Can you create a .war that contains JSP in NetBeans? Where should the .war be placed in a Java application server?

38.2.3 You can display an HTML file (e.g., c:\test.html) by typing the complete file name in the Address field of Internet Explorer. Why can’t you display a JSP file by sim- ply typing the file name?

38.4 JSP Scripting Constructs There are three main types of JSP constructs: scripting constructs, directives, and actions.

Scripting elements enable you to specify Java code that will become part of the resultant servlet. Directives enable you to control the overall structure of the resultant servlet. Actions enable you to control the behavior of the JSP engine. This section introduces scripting constructs.

Three types of JSP scripting constructs can be used to insert Java code into a resultant servlet: expressions, scriptlets, and declarations.

A JSP expression is used to insert a Java expression directly into the output. It has the fol- lowing form:

<%= Java expression %>

Point Check

Point Key

M38_LIAN1878_11_GE_C38.indd 4 12/29/17 5:56 PM

38.4 JSP Scripting Constructs 38-5

The expression is evaluated, converted into a string, and sent to the output stream of the servlet. A JSP scriptlet enables you to insert a Java statement into the servlet’s jspService

method, which is invoked by the service method. A JSP scriptlet has the following form:

<% Java statement %>

A JSP declaration is for declaring methods or fields into the servlet. It has the following form:

<%! Java declaration %>

HTML comments have the following form:

If you don’t want the comment to appear in the resultant HTML file, use the following com- ment in JSP:

<%-- JSP Comment --%>

Listing 38.1 creates a JavaServer page that displays factorials for numbers from 0 to 10, as shown in Figure 38.5.

Figure 38.5 The JSP page displays factorials.

Listing 38.1 Factorial.jsp 1 <html> 2 <head> 3 <title> 4 Factorial 5 </title> 6 </head> 7 <body> 8 9 <% for (int i = 0; i <= 10; i++) { %> 10 Factorial of <%= i %> is 11 <%= computeFactorial(i) %> <br /> 12 <% } %> 13 14 <%! private long computeFactorial(int n) { 15 if (n == 0) 16 return 1; 17 else 18 return n * computeFactorial(n − 1);

M38_LIAN1878_11_GE_C38.indd 5 12/29/17 5:56 PM

38-6 Chapter 38 Javaserver Pages

19 } 20 %> 21 22 </body> 23 </html>

JSP scriptlets are enclosed between <% and %>. Thus,

for (int i = 0; i <= 10; i++) {, (line 9)

is a scriptlet and as such is inserted directly into the servlet’s jspService method. JSP expressions are enclosed between <%= and %>. Thus,

<%= i %>, (line 10)

is an expression and is inserted into the output stream of the servlet. JSP declarations are enclosed between <%! and %>. Thus,

<%! private long computeFactorial(int n) { ... } %>

is a declaration that defines methods or fields in the servlet. What will be different if line 9 is replaced by the two alternatives shown below? Both work

fine, but there is an important difference. In (a), i is a local variable in the servlet, whereas in (b), i is an instance variable when translated to the servlet.

<% int i = 0; %> <% for ( ; i <= 10; i++) { %>

(a)

<%! int i; %> <% for (i = 0; i <= 10; i++) { %>

(b)

Caution: For JSP, the loop body, even though it contains a single statement, must be placed inside braces. It would be wrong to delete the opening brace ({) in line 9 and the closing brace (<% } %>) in line 12.

Caution: There is no semicolon at the end of a JSP expression. For example, <%= i; %> is incor- rect. But there must be a semicolon for each Java statement in a JSP scriptlet. For example, <% int i = 0 %> is incorrect.

Caution: JSP and Java elements are case sensitive, but HTML is not.

38.4.1 What are a JSP expression, a JSP scriptlet, and a JSP declaration? How do you write these constructs in JSP?

38.4.2 Find three syntax errors in the following JSP code: <%! int k %> <% for (int j = 1; j <= 9; j++) %> <%= j; %> <br />

38.4.3 In the following JSP, which variables are instance variables, and which are local variables when it is translated into in the servlet?

<%! int k; %> <%! int i; %>

Point Check

M38_LIAN1878_11_GE_C38.indd 6 12/29/17 5:56 PM

38.5 Predefined Variables 38-7

<% for (int j = 1; j <= 9; j++) k += 1;%> <%= k><br /> <%= i><br /> <%= getTime()><br /> <% private long getTime() { long time = System.currentTimeMillis(); return time; } %>

38.5 Predefined Variables JSP provides predefined variables that can be conviniently used in the JSP code.

You can use variables in JSP. For convenience, JSP provides eight predefined variables from the servlet environment that can be used with JSP expressions and scriptlets. These variables are also known as JSP implicit objects.

■■ request represents the client’s request, which is an instance of HttpServlet- Request. You can use it to access request parameters and HTTP headers, such as cookies and host name.

■■ response represents the servlet’s response, which is an instance of HttpServlet- Response. You can use it to set response type and send output to the client.

■■ out represents the character output stream, which is an instance of PrintWriter obtained from response.getWriter(). You can use it to send character content to the client.

■■ session represents the HttpSession object associated with the request, obtained from request.getSession().

■■ application represents the ServletContext object for storing persistent data for all clients. The difference between session and application is that session is tied to one client, but application is for all clients to share persistent data.

■■ config represents the ServletConfig object for the page.

■■ pageContext represents the PageContext object. PageContext is a new class introduced in JSP to give a central point of access to many page attributes.

■■ page is an alternative to this.

As an example, let us write an HTML page that prompts the user to enter loan amount, annual interest rate, and number of years, as shown in Figure 38.6a. Clicking the Compute Loan Payment button invokes a JSP to compute and display the monthly and total loan payments, as shown in Figure 38.6b.

The HTML file is named ComputeLoan.html (Listing 38.2). The JSP file is named ComputeLoan.jsp (Listing 38.3).

Point Key

Figure 38.6 The JSP computes the loan payments.

(a) (b)

M38_LIAN1878_11_GE_C38.indd 7 12/29/17 5:56 PM

38-8 Chapter 38 Javaserver Pages

Listing 38.2 ComputeLoan.html 1  2 <html> 3 <head> 4 <title>ComputeLoan</title> 5 </head> 6 <body> 7 <form method = "get" action = "ComputeLoan.jsp"> 8 Compute Loan Payment<br /> 9 Loan Amount 10 <input type = "text" name = "loanAmount" /><br /> 11 Annual Interest Rate 12 <input type = "text" name = "annualInterestRate" /><br /> 13 Number of Years 14 <input type = "text" name = "numberOfYears" size = "3" /><br /> 15 <p><input type = "submit" name = "Submit" 16 value = "Compute Loan Payment" /> 17 <input type = "reset" value = "Reset" /></p> 18 </form> 19 </body> 20 </html>

Listing 38.3 ComputeLoan.jsp 1  2 <html> 3 <head> 4 <title>ComputeLoan</title> 5 </head> 6 <body> 7 <% double loanAmount = Double.parseDouble( 8 request.getParameter("loanAmount")); 9 double annualInterestRate = Double.parseDouble( 10 request.getParameter("annualInterestRate")); 11 double numberOfYears = Integer.parseInt( 12 request.getParameter("numberOfYears")); 13 double monthlyInterestRate = annualInterestRate / 1200; 14 double monthlyPayment = loanAmount * monthlyInterestRate / 15 (1 − 1 / Math.pow(1 + monthlyInterestRate, numberOfYears * 12)); 16 double totalPayment = monthlyPayment * numberOfYears * 12; %> 17 Loan Amount: <%= loanAmount %><br /> 18 Annual Interest Rate: <%= annualInterestRate %><br /> 19 Number of Years: <%= numberOfYears %><br /> 20 <b>Monthly Payment: <%= monthlyPayment %><br /> 21 Total Payment: <%= totalPayment %><br /></b> 22 </body> 23 </html>

ComputeLoan.html is displayed first to prompt the user to enter the loan amount, annual interest rate, and number of years. Since this file does not contain any JSP elements, it is named with an .html extension as a regular HTML file.

ComputeLoan.jsp is invoked upon clicking the Compute Loan Payment button in the HTML form. The JSP page obtains the parameter values using the predefined variable request in lines 7–12 and computes monthly payment and total payment in lines 13–16. The formula for computing monthly payment is given in Listing 2.9, ComputeLoan.java.

What is wrong if the JSP scriptlet <% in line 7 is replaced by the JSP declaration <%!? The predefined variables (e.g., request, response, and out) correspond to local variables

M38_LIAN1878_11_GE_C38.indd 8 12/29/17 5:56 PM

38.6 JSP Directives 38-9

defined in the servlet methods doGet and doPost. They must appear in JSP scriptlets, not in JSP declarations.

Tip ComputeLoan.jsp can also be invoked using the following query string: http:// localhost:8084/liangweb/ComputeLoan.jsp?loanAmount=10000&annu alInterestRate=6&numberOfYears=15.

38.5.1 Describe the predefined variables in JSP. 38.5.2 What is wrong if the JSP scriptlet <% in line 7 in ComputeLoan.jsp (Listing 38.3)

is replaced by JSP declaration <%!?

38.5.3 Can you use predefined variables (e.g., request, response, and out) in JSP declarations?

38.6 JSP Directives You can use JSP directives to instruct JSP engine on how to process the JSP code.

A JSP directive is a statement that gives the JSP engine information about the JSP page. For example, if your JSP page uses a Java class from a package other than the java.lang pack- age, you have to use a directive to import this package. The general syntax for a JSP directive is shown below:

<%@ directive attribute = "value" %>, or <%@ directive attribute1 = "value1" attribute2 = "value2" ... attributen = "valuen" %>

The possible directives are:

■■ page lets you provide information for the page, such as importing classes and setting up content type. The page directive can appear anywhere in the JSP file.

■■ include lets you insert a file into the servlet when the page is translated to a servlet. The include directive must be placed where you want the file to be inserted.

■■ taglib lets you define custom tags.

The following are useful attributes for the page directive:

■■ import specifies one or more packages to be imported for this page. For example, the directive <%@ page import="java.util.*, java.text.*" %> imports java.util.* and java.text.*.

■■ contentType specifies the content type for the resultant JSP page. By default, the content type is text/html for JSP. The default content type for servlets is text/ plain.

■■ session specifies a boolean value to indicate whether the page is part of the session. By default, session is true.

■■ buffer specifies the output stream buffer size. By default, it is 8KB. For example, the directive <%@ page buffer="10KB" %> specifies that the output buffer size is 10KB. The directive <%@ page buffer="none" %> specifies that a buffer is not used.

Point Check

Point Key

M38_LIAN1878_11_GE_C38.indd 9 12/29/17 5:56 PM

38-10 Chapter 38 Javaserver Pages

■■ autoFlush specifies a boolean value to indicate whether the output buffer should be automatically flushed when it is full or whether an exception should be raised when the buffer overflows. By default, this attribute is true. In this case, the buffer attribute cannot be none.

■■ isThreadSafe specifies a boolean value to indicate whether the page can be accessed simultaneously without data corruption. By default, it is true. If it is set to false, the JSP page will be translated to a servlet that implements the SingleTh- readModel interface.

■■ errorPage specifies a JSP page that is processed when an exception occurs in the current page. For example, the directive <%@ page errorPage="HandleError. jsp" %> specifies that HandleError.jsp is processed when an exception occurs.

■■ isErrorPage specifies a boolean value to indicate whether the page can be used as an error page. By default, this attribute is false.

Listing 38.4 gives an example that shows how to use the page directive to import a class. The example uses the Loan class created in Listing 10.2, Loan.java, to simplify Listing 38.3, ComputeLoan.jsp. You can create an object of the Loan class and use its monthly Payment() and totalPayment() methods to compute the monthly payment and total payment.

Listing 38.4 ComputeLoan1.jsp 1  2 <html> 3 <head> 4 <title>ComputeLoan Using the Loan Class</title> 5 </head> 6 <body> 7 <%@ page import = "chapter38.Loan" %> 8 <% double loanAmount = Double.parseDouble( 9 request.getParameter("loanAmount")); 10 double annualInterestRate = Double.parseDouble( 11 request.getParameter("annualInterestRate")); 12 int numberOfYears = Integer.parseInt( 13 request.getParameter("numberOfYears")); 14 Loan loan = 15 new Loan(annualInterestRate, numberOfYears, loanAmount); 16 %> 17 Loan Amount: <%= loanAmount %><br /> 18 Annual Interest Rate: <%= annualInterestRate %><br /> 19 Number of Years: <%= numberOfYears %><br /> 20 <b>Monthly Payment: <%= loan.getMonthlyPayment() %><br /> 21 Total Payment: <%= loan.getTotalPayment() %><br /></b> 22 </body> 23 </html>

This JSP uses the Loan class. You need to create the class in the liangweb project in package chapter38 as follows:

package chapter38; public class Loan { // Same as lines 2–71 in Listing 10.2, Loan.java, so omitted

The directive <%@ page import ="chapter38.Loan" %> imports the Loan class in line 7. Line 14 creates an object of Loan for the given loan amount, annual interest rate, and number of years. Lines 20–21 invoke the Loan object's monthlyPayment() and total Payment() methods to display monthly payment and total payment.

M38_LIAN1878_11_GE_C38.indd 10 12/29/17 5:56 PM

38.7 Using JavaBeans in JSP 38-11

38.6.1 Describe the JSP directives and attributes for the page directive. 38.6.2 If a class does not have a package statement, can you import it? 38.6.3 If you use a custom class from a JSP, where should the class be placed?

38.7 Using JavaBeans in JSP You can use JavaBeans to create objects for sharing among different JSP pages.

Normally you create an instance of a class in a program and use it in that program. This method is for sharing the class, not the object. JSP allows you to share the object of a class among different pages. To enable an object to be shared, its class must be a JavaBeans component. Recall that this entails the following three features:

■■ The class is public.

■■ The class has a public constructor with no arguments.

■■ The class is serializable. (This requirement is not necessary in JSP.)

To create an instance for a JavaBeans component, use the following syntax:

<jsp:useBean id = "objectName" scope = "scopeAttribute" class = "ClassName" />

This syntax is roughly equivalent to

<% ClassName objectName = new ClassName() %>

except that the scope attribute is missing. The scope attribute specifies the scope of the object, and the object is not recreated if it is already within the scope. Listed below are four possible values for the scope attribute:

■■ application specifies that the object is bound to the application. The object can be shared by all sessions of the application.

■■ session specifies that the object is bound to the client’s session. Recall that a client’s session is automatically created between a Web browser and a Web server. When a client from the same browser accesses two servlets or two JSP pages on the same server, the session is the same.

■■ page is the default scope, which specifies that the object is bound to the page.

■■ request specifies that the object is bound to the client’s request.

When <jsp:useBean id="objectName" scope="scopeAttribute" class="ClassName" /> is processed, the JSP engine first searches for an object of the class with the same id and scope. If found, the preexisting bean is used; otherwise, a new bean is created.

Here is another syntax for creating a bean:

<jsp:useBean id = "objectName" scope = "scopeAttribute" class = "ClassName" > statements </jsp:useBean>

The statements are executed when the bean is created. If a bean with the same ID and class name already exists in the scope, the statements are not executed.

Listing 38.5 creates a JavaBeans component named Count and uses it to count the number of visits to a JSP page, as shown in Figure 38.7.

Point Check

Point Key

M38_LIAN1878_11_GE_C38.indd 11 12/29/17 5:56 PM

38-12 Chapter 38 Javaserver Pages

Listing 38.5 Count.java 1 package chapter38; 2 3 public class Count { 4 private int count = 0; 5 6 /** Return count property */ 7 public int getCount() { 8 return count; 9 } 10 11 /** Increase count */ 12 public void increaseCount() { 13 count++; 14 } 15 }

The JSP page named TestBeanScope.jsp is created in Listing 38.6.

Listing 38.6 TestBeanScope.jsp 1 <-- TestBeanScope.jsp --> 2 <%@ page import = "chapter38.Count" %> 3 <jsp:useBean id = "count" scope = "application" 4 class = "chapter38.Count"> 5 </jsp:useBean> 6 <html> 7 <head> 8 <title>TestBeanScope</title> 9 </head> 10 <body> 11 <h3>Testing Bean Scope in JSP (Application)</h3> 12 <% count.increaseCount(); %> 13 You are visitor number <%= count.getCount() %><br /> 14 From host: <%= request.getRemoteHost() %> 15 and session: <%= session.getId() %> 16 </body> 17 </html>

The scope attribute specifies the scope of the bean. scope="application" (line 3) specifies that the bean is alive in the JSP engine and available for all clients to access. The bean can be shared by any client with the directive <jsp:useBean id="count" scope="application" class="Count"> (lines 3–4). Every client accessing TestBeanScope.jsp causes the count to increase by 1. The first client causes count object to be created, and subsequent access to TestBeanScope uses the same object.

If scope="application" is changed to scope="session", the scope of the bean is limited to the session from the same browser. The count will increase only if the page is requested from the same browser. If scope="application" is changed to scope="page",

Figure 38.7 The number of visits to the page is increased when the page is visited.

M38_LIAN1878_11_GE_C38.indd 12 12/29/17 5:56 PM

38.7 Using JavaBeans in JSP 38-13

the scope of the bean is limited to the page, and any other page cannot access this bean. The page will always display count 1. If scope="application" is changed to scope="request", the scope of the bean is limited to the client’s request, and any other request on the page will always display count 1.

If the page is destroyed, the count restarts from 0. You can fix the problem by storing the count in a random access file or in a database table. Assume you store the count in the Count table in a database. The Count class can be modified in Listing 38.7.

Listing 38.7 Count.java (Revised Version) 1 package chapter38; 2 3 import java.sql.*; 4 5 public class Count { 6 private int count = 0; 7 private Statement statement = null; 8 9 public Count() { 10 initializeJdbc(); 11 } 12 13 /** Return count property */ 14 public int getCount() { 15 try { 16 ResultSet rset = statement.executeQuery 17 ("select countValue from Count"); 18 rset.next(); 19 count = rset.getInt(1); 20 } 21 catch (Exception ex) { 22 ex.printStackTrace(); 23 } 24 25 return count; 26 } 27 28 /** Increase count */ 29 public void increaseCount() { 30 count++; 31 try { 32 statement.executeUpdate( 33 "update Count set countValue = " + count); 34 } 35 catch (Exception ex) { 36 ex.printStackTrace(); 37 } 38 } 39 40 /** Initialize database connection */ 41 public void initializeJdbc() { 42 try { 43 Class.forName("com.mysql.jdbc.Driver"); 44 45 // Connect to the sample database 46 Connection connection = DriverManager.getConnection 47 ("jdbc:mysql://localhost/javabook" , "scott", "tiger"); 48 49 statement = connection.createStatement(); 50 }

M38_LIAN1878_11_GE_C38.indd 13 12/29/17 5:56 PM

38-14 Chapter 38 Javaserver Pages

38.7.1 You can create an object in a JSP scriptlet. What is the difference between an object created using the new operator and a bean created using the <jsp:useBean ... > tag?

38.7.2 What is the scope attribute for? Describe four scope attributes. 38.7.3 Describe how a <jsp:useBean ... > statement is processed by the JSP engine.

38.8 Getting and Setting Properties JSP provides convenient syntax for getting and setting JavaBeans properties.

By convention, a JavaBeans component provides the get and set methods for reading and modi- fying its private properties. You can get the property in JSP using the syntax shown below:

<jsp:getProperty name = "beanId" property = "sample" />

This is roughly equivalent to

<%= beanId.getSample() %>

You can set the property in JSP using the following syntax:

<jsp:setProperty name = "beanId" property = "sample" value = "test1" />

This is equivalent to

<% beanId.setSample("test1"); %>

38.9 Associating Properties with Input Parameters Often properties are associated with input parameters. Suppose you want to get the value of the input parameter named score and set it to the JavaBeans property named score. You could write the following code:

<% double score = Double.parseDouble( request.getParameter("score")); %> <jsp:setProperty name = "beanId" property = "score" value = "<%= score %>" />

This is cumbersome. JSP provides a convenient syntax that can be used to simplify it:

<jsp:setProperty name = "beanId" property = "score" param = "score" />

Instead of using the value attribute, you use the param attribute to name an input parameter. The value of this parameter is set to the property.

Note Simple type conversion is performed automatically when a bean property is associated with an input parameter. A string input parameter is converted to an appropriate primitive

Point Check

Point Key

51 catch (Exception ex) { 52 ex.printStackTrace(); 53 } 54 } 55 }

M38_LIAN1878_11_GE_C38.indd 14 12/29/17 5:56 PM

38.9 Associating Properties with Input Parameters 38-15

data type or a wrapper class for a primitive type. For example, if the bean property is of the int type, the value of the parameter will be converted to the int type. If the bean property is of the Integer type, the value of the parameter will be converted to the Integer type.

Often the bean property and the parameter have the same name. You can use the following convenient statement to associate all the bean properties in beanId with the parameters that match the property names:

<jsp:setProperty name = "beanId" property = "*" />

38.9.1 Example: Computing Loan Payments Using JavaBeans This example uses JavaBeans to simplify Listing 38.4, ComputeLoan1.jsp, by associat- ing the bean properties with the input parameters. The new ComputeLoan2.jsp is given in Listing 38.8.

Listing 38.8 ComputeLoan2.jsp 1  2 <html> 3 <head> 4 <title>ComputeLoan Using the Loan Class</title> 5 </head> 6 <body> 7 <%@ page import = "chapter38.Loan" %> 8 <jsp:useBean id = "loan" class = "chapter38.Loan" 9 scope = "page" ></jsp:useBean> 10 <jsp:setProperty name = "loan" property = "*" /> 11 Loan Amount: <%= loan.getLoanAmount() %><br /> 12 Annual Interest Rate: <%= loan.getAnnualInterestRate() %><br /> 13 Number of Years: <%= loan.getNumberOfYears() %><br /> 14 <b>Monthly Payment: <%= loan.monthlyPayment() %><br /> 15 Total Payment: <%= loan.totalPayment() %><br /></b> 16 </body> 17 </html>

Lines 8–9

<jsp:useBean id = "loan" class = "chapter38.Loan" scope = "page" ></jsp:useBean>

create a bean named loan for the Loan class. Line 10

<jsp:setProperty name = "loan" property = "*" />

associates the bean properties loanAmount, annualInteresteRate, and numberOfYears with the input parameter values and performs type conversion automatically.

Lines 11–13 use the accessor methods of the loan bean to get the loan amount, annual inter- est rate, and number of years.

This program acts the same as in Listings 38.3 and 38.4, ComputeLoan.jsp and ComputeLoan1.jsp, but the coding is much, more simplified.

38.9.2 Example: Computing Factorials Using JavaBeans This example creates a JavaBeans component named FactorialBean and uses it to com- pute the factorial of an input number in a JSP page named FactorialBean.jsp, as shown in Figure 38.8.

M38_LIAN1878_11_GE_C38.indd 15 12/29/17 5:56 PM

38-16 Chapter 38 Javaserver Pages

Create a JavaBeans component named FactorialBean.java (Listing 38.9). Create FactorialBean.jsp (Listing 38.10).

Figure 38.8 The factorial of an input integer is computed using a method in FactorialBean.

Listing 38.9 FactorialBean.java 1 package chapter38; 2 3 public class FactorialBean { 4 private int number; 5 6 /** Return number property */ 7 public int getNumber() { 8 return number; 9 } 10 11 /** Set number property */ 12 public void setNumber(int newValue) { 13 number = newValue; 14 } 15 16 /** Obtain factorial */ 17 public long getFactorial() { 18 long factorial = 1; 19 for (int i = 1; i <= number; i++) 20 factorial *= i; 21 return factorial; 22 } 23 }

Listing 38.10 FactorialBean.jsp 1  2 <%@ page import = "chapter38.FactorialBean" %> 3 <jsp:useBean id = "factorialBeanId" 4 class = "chapter38.FactorialBean" scope = "page" > 5 </jsp:useBean> 6 <jsp:setProperty name = "factorialBeanId" property = "*" /> 7 <html> 8 <head> 9 <title> 10 FactorialBean 11 </title> 12 </head> 13 <body> 14 <h3>Compute Factorial Using a Bean</h3>

M38_LIAN1878_11_GE_C38.indd 16 12/29/17 5:56 PM

38.9 Associating Properties with Input Parameters 38-17

The jsp:useBean tag (lines 3–4) creates a bean factorialBeanId of the FactorialBean class. Line 5 <jsp:setProperty name="factorialBeanId" property="*" /> associates all the bean properties with the input parameters that have the same name. In this case, the bean property number is associated with the input parameter number. When you click the Compute Factorial button, JSP automatically converts the input value for number from string into int and sets it to factorialBean before other statements are executed.

Lines 21–22 <jsp:getProperty name="factorialBeanId" property="number" /> tag (line 21) is equivalent to <%= factorialBeanId.getNumber() %>. The method factorialBeanId.getFactorial() (line 25) returns the factorial for the number in factorialBeanId.

Design Guide Mixing a lot of Java code with HTML in a JSP page makes the code difficult to read and to maintain. You should move the Java code to a .java file as much as you can.

Following the preceding design guide, you may improve the preceding example by moving the Java code in lines 23–25 to the FactorialBean class. The new FactorialBean.java and FactorialBean.jsp are given in Listings 38.11 and 38.12.

15 <form method = "post"> 16 Enter new value: <input name = "number" /><br /><br /> 17 <input type = "submit" name = "Submit" 18 value = "Compute Factorial" /> 19 <input type = "reset" value = "Reset" /><br /><br /> 20 Factorial of 21 <jsp:getProperty name = "factorialBeanId" 22 property = "number" /> is 23 <%@ page import = "java.text.*" %> 24 <% NumberFormat format = NumberFormat.getNumberInstance(); %> 25 <%= format.format(factorialBeanId.getFactorial()) %> 26 </form> 27 </body> 28 </html>

Listing 38.11 NewFactorialBean.java 1 package chapter38; 2 3 import java.text.*; 4 5 public class NewFactorialBean { 6 private int number; 7 8 /** Return number property */ 9 public int getNumber() { 10 return number; 11 } 12 13 /** Set number property */ 14 public void setNumber(int newValue) { 15 number = newValue; 16 } 17 18 /** Obtain factorial */ 19 public long getFactorial() { 20 long factorial = 1;

M38_LIAN1878_11_GE_C38.indd 17 12/29/17 5:56 PM

38-18 Chapter 38 Javaserver Pages

There is a problem in this page. The program cannot display large factorials. For example, if you entered value 21, the program would display an incorrect factorial. To fix this problem, all you need to do is to revise the NewFactorialBean class using BigInteger to computing factorials (see Exercise 38.18).

38.9.3 Example: Displaying International Time Listing 37.5, TimeForm.java, gives a Java servlet that uses the doGet method to generate an HTML form for the user to specify a locale and time zone (see Figure 37.18a) and uses the doPost method to display the current time for the specified time zone in the specified locale (see Figure 37.18b). This section rewrites the servlet using JSP. You have to create two JSP pages, one for displaying the form, and the other for displaying the current time.

In the TimeForm.java servlet, arrays allLocale and allTimeZone are the data fields. The doGet and doPost methods both use the arrays. Since the available locales and time zones are used in both pages, it is better to create an object that contains all available locales and time zones. This object can be shared by both pages.

21 for (int i = 1; i <= number; i++) 22 factorial *= i; 23 return factorial; 24 } 25 26 /** Format number */ 27 public static String format(long number) { 28 NumberFormat format = NumberFormat.getNumberInstance(); 29 return format.format(number); 30 } 31 }

Listing 38.12 NewFactorialBean.jsp 1  2 <%@ page import = "chapter38.NewFactorialBean" %> 3 <jsp:useBean id = "factorialBeanId" 4 class = "chapter38.NewFactorialBean" scope = "page" > 5 </jsp:useBean> 6 <jsp:setProperty name = "factorialBeanId" property = "*" /> 7 <html> 8 <head> 9 <title> 10 FactorialBean 11 </title> 12 </head> 13 <body> 14 <h3>Compute Factorial Using a Bean</h3> 15 <form method = "post"> 16 Enter new value: <input name = "number" /><br /><br /> 17 <input type = "submit" name = "Submit" 18 value = "Compute Factorial" /> 19 <input type = "reset" value = "Reset" /><br /><br /> 20 Factorial of 21 <jsp:getProperty name = "factorialBeanId" 22 property = "number" /> is 23 <%= NewFactorialBean.format(factorialBeanId.getFactorial()) %> 24 </form> 25 </body> 26 </html>

M38_LIAN1878_11_GE_C38.indd 18 12/29/17 5:56 PM

38.9 Associating Properties with Input Parameters 38-19

Let us create a JavaBeans component named TimeBean.java (Listing 38.13). This class obtains all the available locales in an array in line 7 and all time zones in an array in line 8. The bean properties localeIndex and timeZoneIndex (lines 9–10) are defined to refer to an element in the arrays. The currentTimeString() method (lines 42–52) returns a string for the current time with the specified locale and time zone.

Listing 38.13 TimeBean.java 1 package chapter38; 2 3 import java.util.*; 4 import java.text.*; 5 6 public class TimeBean { 7 private Locale[] allLocale = Locale.getAvailableLocales(); 8 private String[] allTimeZone = TimeZone.getAvailableIDs(); 9 private int localeIndex; 10 private int timeZoneIndex; 11 12 public TimeBean() { 13 Arrays.sort(allTimeZone); 14 } 15 16 public Locale[] getAllLocale() { 17 return allLocale; 18 } 19 20 public String[] getAllTimeZone() { 21 return allTimeZone; 22 } 23 24 public int getLocaleIndex() { 25 return localeIndex; 26 } 27 28 public int getTimeZoneIndex() { 29 return timeZoneIndex; 30 } 31 32 public void setLocaleIndex(int index) { 33 localeIndex = index; 34 } 35 36 public void setTimeZoneIndex(int index) { 37 timeZoneIndex = index; 38 } 39 40 /** Return a string for the current time 41 * with the specified locale and time zone */ 42 public String currentTimeString( 43 int localeIndex, int timeZoneIndex) { 44 Calendar calendar = 45 new GregorianCalendar(allLocale[localeIndex]); 46 TimeZone timeZone = 47 TimeZone.getTimeZone(allTimeZone[timeZoneIndex]); 48 DateFormat dateFormat = DateFormat.getDateTimeInstance( 49 DateFormat.FULL, DateFormat.FULL, allLocale[localeIndex]); 50 dateFormat.setTimeZone(timeZone); 51 return dateFormat.format(calendar.getTime()); 52 } 53 }

M38_LIAN1878_11_GE_C38.indd 19 12/29/17 5:56 PM

38-20 Chapter 38 Javaserver Pages

Create DisplayTimeForm.jsp (Listing 38.14). This page displays a form just like the one shown in Figure 37.18a. Line 2 imports the TimeBean class. A bean is created in lines 3–5 and is used in lines 17, 19, 24, and 26 to return all locales and time zones. The scope of the bean is application (line 4), so the bean can be shared by all sessions of the application.

Listing 38.14 DisplayTimeForm.jsp 1  2 <%@ pageimport = "chapter38.TimeBean" %> 3 <jsp:useBean id = "timeBeanId" 4 class = "chapter38.TimeBean" scope = "application" > 5 </jsp:useBean> 6 7 <html> 8 <head> 9 <title> 10 Display Time Form 11 </title> 12 </head> 13 <body> 14 <h3>Choose locale and time zone</h3> 15 <form method = "post" action = "DisplayTime.jsp"> 16 Locale <select size = "1" name = "localeIndex"> 17 <% for (int i = 0; i < timeBeanId.getAllLocale().length; i++) {%> 18 <option value = "<%= i %>"> 19 <%= timeBeanId.getAllLocale()[i] %> 20 </option> 21 <%}%> 22 </select><br /> 23 Time Zone <select size = "1" name = "timeZoneIndex"> 24 <% for (int i = 0; i < timeBeanId.getAllTimeZone().length; i++) {%> 25 <option value = "<%= i %>"> 26 <%= timeBeanId.getAllTimeZone()[i] %> 27 </option> 28 <%}%> 29 </select><br /> 30 <input type = "submit" name = "Submit" 31 value = "Get Time" /> 32 <input type = "reset" value = "Reset" /> 33 </form> 34 </body> 35 </html>

Create DisplayTime.jsp (Listing 38.15). This page is invoked from DisplayTimeForm.jsp to display the time with the specified locale and time zone, just as in Figure 37.18b.

Listing 38.15 DisplayTime.jsp 1  2 <%@page pageEncoding = "GB18030"%> 3 <%@ page import = "chapter38.TimeBean" %> 4 <jsp:useBean id = "timeBeanId" 5 class = "chapter38.TimeBean" scope = "application" > 6 </jsp:useBean> 7 <jsp:setProperty name = "timeBeanId" property = "*" /> 8 9 <html> 10 <head>

M38_LIAN1878_11_GE_C38.indd 20 12/29/17 5:56 PM

38.9 Associating Properties with Input Parameters 38-21

Line 2 sets the character encoding for the page to GB18030 for displaying international characters. By default, it is UTF-8.

Line 5 imports chapter38.TimeBean and creates a bean using the same id as in the preceding page. Since the object is already created in the preceding page, the timeBeanId in this page (lines 4–6) and in the preceding page point to the same object.

38.9.4 Example: Registering Students Listing 37.11, RegistrationWithHttpSession.java, gives a Java servlet that obtains student information from an HTML form (see Figure 37.21) and displays the information for user confirmation (see Figure 37.22). Once the user confirms it, the servlet stores the data into the database. This section rewrites the servlet using JSP. You will create two JSP pages, one named GetRegistrationData.jsp for displaying the data for user confirmation and the other named StoreData.jsp for storing the data into the database.

Since every session needs to connect to the same database, you should declare a class for connecting to the database and for storing a student to the database. This class named StoreData is given in Listing 38.16. The initializeJdbc method (lines 15–31) connects to the database and creates a prepared statement for storing a record to the Address table. The storeStudent method (lines 34–45) executes the prepared statement to store a student address. The Address class is created in Listing 37.12.

11 <title> 12 Display Time 13 </title> 14 </head> 15 <body> 16 <h3>Choose locale and time zone</h3> 17 Current time is 18 <%= timeBeanId.currentTimeString(timeBeanId.getLocaleIndex(), 19 timeBeanId.getTimeZoneIndex()) %> 20 </body> 21 <html>

Listing 38.16 StoreData.java 1 package chapter38; 2 3 import java.sql.*; 4 import chapter37.Address; 5 6 public class StoreData { 7 // Use a prepared statement to store a student into the database 8 private PreparedStatement pstmt; 9 10 public StoreData() { 11 initializeJdbc(); 12 } 13 14 /** Initialize database connection */ 15 private void initializeJdbc() { 16 try { 17 Class.forName("com.mysql.jdbc.Driver"); 18 19 // Connect to the sample database 20 Connection connection = DriverManager.getConnection 21 ("jdbc:mysql://localhost/javabook" , "scott", "tiger"); 22

M38_LIAN1878_11_GE_C38.indd 21 12/29/17 5:56 PM

38-22 Chapter 38 Javaserver Pages

The HTML file that displays the form is identical to Registration.html in Listing 37.8 except that the action is replaced by HGetRegistrationData.jsp.

GetRegistrationData.jsp, which obtains the data from the form, is shown in Listing 38.17. A bean is created in lines 3–4. Line 5 obtains the property values from the form. This is a shorthand notation. Note the parameter names and the property names must be the same to use this notation.

23 // Create a Statement 24 pstmt = connection.prepareStatement("insert into Address " + 25 "(lastName, firstName, mi, telephone, email, street, city, " 26 + "state, zip) values (?, ?, ?, ?, ?, ?, ?, ?, ?)"); 27 } 28 catch (Exception ex) { 29 System.out.println(ex); 30 } 31 } 32 33 /** Store a student record to the database */ 34 public void storeStudent(Address address) throws SQLException { 35 pstmt.setString(1, address.getLastName()); 36 pstmt.setString(2, address.getFirstName()); 37 pstmt.setString(3, address.getMi()); 38 pstmt.setString(4, address.getTelephone()); 39 pstmt.setString(5, address.getEmail()); 40 pstmt.setString(6, address.getStreet()); 41 pstmt.setString(7, address.getCity()); 42 pstmt.setString(8, address.getState()); 43 pstmt.setString(9, address.getZip()); 44 pstmt.executeUpdate(); 45 } 46 }

Listing 38.17 GetRegistrationData.jsp 1  2 <%@ page import = "chapter37.Address" %> 3 <jsp:useBean id = "addressId" 4 class = "chapter37.Address" scope = "session"></jsp:useBean> 5 <jsp:setProperty name = "addressId" property = "*" /> 6 7 <html> 8 <body> 9 <h1>Registration Using JSP</h1> 10 11 <% 12 if (addressId.getLastName() == null || 13 addressId.getFirstName() == null) { 14 out.println("Last Name and First Name are required"); 15 return; // End the method 16 } 17 %> 18 19 <p>You entered the following data</p> 20 <p>Last name: <%= addressId.getLastName() %></p> 21 <p>First name: <%= addressId.getFirstName() %></p> 22 <p>MI: <%= addressId.getMi() %></p> 23 <p>Telephone: <%= addressId.getTelephone() %></p>

M38_LIAN1878_11_GE_C38.indd 22 12/29/17 5:56 PM

38.9 Associating Properties with Input Parameters 38-23

Note The scope for addressId is session, but the scope for storeDataId is application. Why? GetRegistrationData.jsp obtains student information, and StoreData.jsp stores the information in the same session. So the session scope is appropriate for addressId. All the sessions access the same database and use the same prepared statement to store data. With the application scope for storeDataId, the bean for StoreData needs to be created just once.

Note The storeStudent method in line 11 may throw a java.sql.SQLException. In JSP, you can omit the try-block for checked exceptions. In case of an exception, JSP displays an error page.

24 <p>Email: <%= addressId.getEmail() %></p> 25 <p>Address: <%= addressId.getStreet() %></p> 26 <p>City: <%= addressId.getCity() %></p> 27 <p>State: <%= addressId.getState() %></p> 28 <p>Zip: <%= addressId.getZip() %></p> 29 30  31 <form method = "post" action = "StoreStudent.jsp"> 32 <input type = "submit" value = "Confirm"> 33 </form> 34 </body> 35 </html>

GetRegistrationData.jsp invokes StoreStudent.jsp (line 31) when the user clicks the Confirm button. In Listing 38.18, the same addressId is shared with the preceding page within the scope of the same session in lines 3–4. A bean for StoreData is created in lines 5–6 with the scope of application.

Listing 38.18 StoreStudent.jsp 1  2 <%@ page import = "chapter37.Address" %> 3 <jsp:useBean id = "addressId" class = "chapter37.Address" 4 scope = "session"></jsp:useBean> 5 <jsp:useBean id = "storeDataId" class = "chapter38.StoreData" 6 scope = "application"></jsp:useBean> 7 8 <html> 9 <body> 10 <% 11 storeDataId.storeStudent(addressId); 12 13 out.println(addressId.getFirstName() + " " + 14 addressId.getLastName() + 15 " is now registered in the database"); 16 out.close(); // Close stream 17 %> 18 </body> 19 </html>

M38_LIAN1878_11_GE_C38.indd 23 12/29/17 5:56 PM

38-24 Chapter 38 Javaserver Pages

Tip Using beans is an effective way to develop JSP. You should put Java code into a bean as much as you can. The bean not only simplifies JSP programming but also makes code reusable. The bean can also be used to implement persistent sessions.

38.10 Forwarding Requests from JavaServer Pages You can use the JSP forward tag to jump to navigate to another HTML page.

Web applications developed using JSP generally consist of many pages linked together. JSP provides a forwarding tag in the following syntax that can be used to forward a page to another page:

<jsp:forward page = "destination" />

38.10.1 How do you associate bean properties with input parameters? 38.10.2 How do you write a statement to forward requests to another JSP page?

38.11 Case Study: Browsing Database Tables This section presents a very useful JSP application for browsing tables. When you start the application, the first page prompts the user to enter the JDBC driver, URL, username, and password for a database, as shown in Figure 38.9. After you log in to the database, you can select a table to browse, as shown in Figure 38.10. Clicking the Browse Table Content button displays the table content, as shown in Figure 38.11.

Point Key

Point Check

Figure 38.9 To access a database, you need to provide the JDBC driver, URL, username, and password.

Figure 38.10 You can select a table to browse from this page.

M38_LIAN1878_11_GE_C38.indd 24 12/29/17 5:56 PM

38.11 Case Study: Browsing Database Tables 38-25

Create a JavaBeans component named DBBean.java (see Listing 38.19).

Figure 38.11 The contents of the selected table are displayed.

Listing 38.19 DBBean.java 1 package chapter38; 2 3 import java.sql.*; 4 5 public class DBBean { 6 private Connection connection = null; 7 private String username; 8 private String password; 9 private String driver; 10 private String url; 11 12 /** Initialize database connection */ 13 public void initializeJdbc() { 14 try { 15 System.out.println("Driver is " + driver); 16 Class.forName(driver); 17 18 // Connect to the sample database 19 connection = DriverManager.getConnection(url, username, 20 password); 21 } 22 catch (Exception ex) { 23 ex.printStackTrace(); 24 } 25 } 26 27 /** Get tables in the database */ 28 public String[] getTables() { 29 String[] tables = null; 30 31 try { 32 DatabaseMetaData dbMetaData = connection.getMetaData(); 33 ResultSet rsTables = dbMetaData.getTables(null, null, null, 34 new String[] {"TABLE"}); 35 36 int size = 0;

M38_LIAN1878_11_GE_C38.indd 25 12/29/17 5:56 PM

38-26 Chapter 38 Javaserver Pages

37 while (rsTables.next()) size++; 38 39 rsTables = dbMetaData.getTables(null, null, null, 40 new String[] {"TABLE"}); 41 42 tables = new String[size]; 43 int i = 0; 44 while (rsTables.next()) 45 tables[i++] = rsTables.getString("TABLE_NAME"); 46 } 47 catch (Exception ex) { 48 ex.printStackTrace(); 49 } 50 51 return tables; 52 } 53 54 /** Return connection property */ 55 public Connection getConnection() { 56 return connection; 57 } 58 59 public void setUsername(String newUsername) { 60 username = newUsername; 61 } 62 63 public String getUsername() { 64 return username; 65 } 66 67 public void setPassword(String newPassword) { 68 password = newPassword; 69 } 70 71 public String getPassword() { 72 return password; 73 } 74 75 public void setDriver(String newDriver) { 76 driver = newDriver; 77 } 78 79 public String getDriver() { 80 return driver; 81 } 82 83 public void setUrl(String newUrl) { 84 url = newUrl; 85 } 86 87 public String getUrl() { 88 return url; 89 } 90 }

Create an HTML file named DBLogin.html (see Listing 38.20) that prompts the user to enter database information and three JSP files named DBLoginInitialization.jsp (see Listing 38.21), Table.jsp (see Listing 38.22), and BrowseTable.jsp (see Listing 38.23) to process and obtain database information.

M38_LIAN1878_11_GE_C38.indd 26 12/29/17 5:56 PM

38.11 Case Study: Browsing Database Tables 38-27

Listing 38.21 DBLoginInitialization.jsp 1  2 <%@ page import = "chapter38.DBBean" %> 3 <jsp:useBean id = "dBBeanId" scope = "session" 4 class = "chapter38.DBBean"> 5 </jsp:useBean> 6 <jsp:setProperty name = "dBBeanId" property = "*" /> 7 <html> 8 <head> 9 <title>DBLoginInitialization</title> 10 </head> 11 <body> 12 13 <%-- Connect to the database --%> 14 <% dBBeanId.initializeJdbc(); %> 15 16 <% if (dBBeanId.getConnection() == null) { %> 17 Error: Login failed. Try again. 18 <% } 19 else {%> 20 <jsp:forward page = "Table.jsp" /> 21 <% } %> 22 </body> 23 </html>

Listing 38.20 DBLogin.html 1  2 <html> 3 <head> 4 <title> 5 DBLogin 6 </title> 7 </head> 8 <body> 9 <form method = "post" action = "DBLoginInitialization.jsp"> 10 JDBC URL 11 <select name = "url" size = "1"> 12 <option>jdbc:odbc:ExampleMDBDataSource</option> 13 <option>jdbc:mysql://localhost/javabook</option> 14 <option>jdbc:oracle:thin:@liang.armstrong.edu:1521:orcl</option> 15 </select><br /><br /> 16 Username <input name = "username" /><br /><br /> 17 Password <input name = "password" /><br /><br /> 18 <input type = "submit" name = "Submit" value = "Login" /> 19 <input type = "reset" value = "Reset" /> 20 </form> 21 </body> 22 </html>

Listing 38.22 Table.jsp 1  2 <%@ page import = "chapter38.DBBean" %> 3 <jsp:useBean id = "dBBeanId" scope = "session" 4 class = "chapter38.DBBean"> 5 </jsp:useBean> 6 <html> 7 <head>

M38_LIAN1878_11_GE_C38.indd 27 12/29/17 5:56 PM

38-28 Chapter 38 Javaserver Pages

8 <title>Table</title> 9 </head> 10 <body> 11 <% String[] tables = dBBeanId.getTables(); 12 if (tables == null) { %> 13 No tables 14 <% } 15 else { %> 16 <form method = "post" action = "BrowseTable.jsp"> 17 Select a table 18 <select name = "tablename" size = "1"> 19 <% for (int i = 0; i < tables.length; i++) { %> 20 <option><%= tables[i] %></option> 21 <% } 22 } %> 23 </select><br /><br /><br /> 24 <input type = "submit" name = "Submit" 25 value = "Browse Table Content"> 26 <input type = "reset" value = "Reset"> 27 </form> 28 </body> 29 </html>

Listing 38.23 BrowseTable.jsp 1  2 <%@ page import = "chapter38.DBBean" %> 3 <jsp:useBean id = "dBBeanId" scope = "session" 4 class = "chapter38.DBBean" > 5 </jsp:useBean> 6 <%@ page import = "java.sql.*" %> 7 <html> 8 <head> 9 <title>BrowseTable</title> 10 </head> 11 <body> 12 13 <% String tableName = request.getParameter("tablename"); 14 15 ResultSet rsColumns = dBBeanId.getConnection().getMetaData(). 16 getColumns(null, null, tableName, null); 17 %> 18 <table border = "1"> 19 <tr> 20 <% // Add column names to the table 21 while (rsColumns.next()) { %> 22 <td><%= rsColumns.getString("COLUMN_NAME") %></td> 23 <%}%> 24 </tr> 25 26 <% Statement statement = 27 dBBeanId.getConnection().createStatement(); 28 ResultSet rs = statement.executeQuery( 29 "select * from " + tableName); 30 31 // Get column count 32 int columnCount = rs.getMetaData().getColumnCount(); 33 34 // Store rows to rowData 35 while (rs.next()) {

M38_LIAN1878_11_GE_C38.indd 28 12/29/17 5:56 PM

38.11 Case Study: Browsing Database Tables 38-29

36 out.println("<tr>"); 37 for (int i = 0; i < columnCount; i++) { %> 38 <td><%= rs.getObject(i + 1) %></td> 39 <% } 40 out.println("</tr>"); 41 } %> 42 </table> 43 </body> 44 </html>

You start the application from DBLogin.html. This page prompts the user to enter a JDBC driver, URL, username, and password to log in to a database. A list of accessible drivers and URLs is provided in the selection list. You must make sure that these database drivers are added into the Libraries node in the project.

When you click the Login button, DBLoginInitialization.jsp is invoked. When this page is processed for the first time, an instance of DBBean named dBBeanId is created. The input parameters driver, url, username, and password are passed to the bean properties. The initializeJdbc method loads the driver and establishes a connection to the database. If login fails, the connection property is null. In this case, an error message is displayed. If login suc- ceeds, control is forwarded to Table.jsp.

Table.jsp shares dBBeanId with DBLoginInitialization.jsp in the same session, so it can access connection through dBBeanId and obtain tables in the database using the database metadata. The table names are displayed in a selection box in a form. When the user selects a table name and clicks the Browse Table Content button, BrowseTable.jsp is processed.

BrowseTable.jsp shares dBBeanId with Table.jsp and DBLoginInitialization.jsp in the same session. It retrieves the table contents for the selected table from Table.jsp.

JSP Scripting Constructs Syntax

■■ <%= Java expression %> The expression is evaluated and inserted into the page.

■■ <% Java statement %> Java statements inserted in the jspService method.

■■ <%! Java declaration %> Defines data fields and methods.

■■ <%-- JSP comment %> The JSP comments do not appear in the resultant HTML file.

■■ <%@ directive attribute="value" %> The JSP directives give the JSP engine information about the JSP page. For example, <%@ page import="java.util.*, java.text.*" %> imports java.util.* and java.text.*.

■■ <jsp:useBean id="objectName" scope="scopeAttribute" class="ClassName" /> Creates a bean if new. If a bean is already created, asso- ciates the id with the bean in the same scope.

■■ <jsp:useBean id="objectName" scope="scopeAttribute" class="ClassName" > statements </jsp:useBean> The statements are executed when the bean is created. If a bean with the same id and class name already exists, the statements are not executed.

■■ <jsp:getProperty name="beanId" property="sample" /> Gets the property value from the bean, which is the same as <%= beanId.getSample() %>.

■■ <jsp:setProperty name="beanId" property="sample" value="test1" /> Sets the property value for the bean, which is the same as <% beanId. setSample("test1"); %>.

■■ <jsp:setProperty name="beanId" property="score" param="score" /> Sets the property with an input parameter.

M38_LIAN1878_11_GE_C38.indd 29 12/29/17 5:56 PM

38-30 Chapter 38 Javaserver Pages

■■ <jsp:setProperty name="beanId" property="*" /> Associates and sets all the bean properties in beanId with the input parameters that match the property names.

■■ <jsp:forward page="destination" /> Forwards this page to a new page.

JSP Predefined Variables

■■ application represents the ServletContext object for storing persistent data for all clients.

■■ config represents the ServletConfig object for the page.

■■ out represents the character output stream, which is an instance of PrintWriter, obtained from response.getWriter().

■■ page is alternative to this.

■■ request represents the client's request, which is an instance of HttpServlet- Request in the servlet's service method.

■■ response represents the client's response, which is an instance of HttpServlet- Response in the servlet's service method.

■■ session represents the HttpSession object associated with the request, obtained from request.getSession().

ChapTer summary 1. A JavaServer page is like a regular HTML page with special tags, known as JSP tags,

which enable the Web server to generate dynamic content. You can create a webpage with static HTML and enclose the code for generating dynamic content in the JSP tags.

2. A JSP page must be stored in a file with a .jsp extension. The Web server translates the .jsp file into a Java servlet, compiles the servlet, and executes it. The result of the execu- tion is sent to the browser for display.

3. A JSP page is translated into a servlet when the page is requested for the first time. It is not retranslated if the page is not modified. To ensure that the first-time real user does not encounter a delay, JSP developers should test the page after it is installed.

4. There are three main types of JSP constructs: scripting constructs, directives, and actions. Scripting elements enable you to specify Java code that will become part of the result- ant servlet. Directives enable you to control the overall structure of the resultant servlet. Actions enable you to control the behaviors of the JSP engine.

5. Three types of scripting constructs can be used to insert Java code into the resultant servlet: expressions, scriptlets, and declarations.

6. The scope attribute (application, session, page, and request) specifies the scope of a Java- Beans object. Application specifies that the object be bound to the application. Session specifies that the object be bound to the client’s session. Page is the default scope, which specifies that the object be bound to the page. Request specifies that the object be bound to the client’s request.

7. Web applications developed using JSP generally consist of many pages linked together. JSP provides a forwarding tag in the following syntax that can be used to forward a page to another page: <jsp:forward page="destination" />.

M38_LIAN1878_11_GE_C38.indd 30 12/29/17 5:56 PM

Programming Exercises 38-31

Quiz Answer the quiz for this chapter online at the book Companion Website.

programming exerCiSeS

Note Solutions to even-numbered exercises in this chapter are in exercise\jspexercise from evennumberedexercise.zip, which can be downloaded from the Companion Website.

Section 38.4 38.1 (Factorial table in JSP) Rewrite Exercise 37.1 using JSP. 38.2 (Muliplication table in JSP) Rewrite Exercise 37.2 using JSP.

Section 38.5 *38.3 (Obtain parameters in JSP) Rewrite the servlet in Listing 37.4, GetParameters.java,

using JSP. Create an HTML form that is identical to Student_Registration_Form. html in Listing 37.3 except that the action is replaced by Exercise40_3.jsp for obtaining parameter values.

Section 38.6 38.4 (Calculate tax in JSP) Rewrite Exercise 37.4 using JSP. You need to import

ComputeTax in the JSP.

*38.5 (Find scores from text files) Rewrite Exercise 37.6 using servlets. **38.6 (Find scores from database tables) Rewrite Exercise 37.7 using servlets.

Section 38.7 **38.7 (Change the password) Rewrite Exercise 37.8 using servlets.

Comprehensive *38.8 (Store cookies in JSP) Rewrite Exercise 37.10 using JSP. Use response.

addCookie(Cookie) to add a cookie.

*38.9 (Retrieve cookies in JSP) Rewrite Exercise 37.11 using JSP. Use Cookie[] cookies = request.getCookies() to get all cookies.

38.10 (Draw images) Write a JSP program that displays a country's flag and description as shown in Figure 38.12. The country code such as us is passed as a parameter in the URL. The country's flag file is named as CountryCode.gif and the description is stored in a text file named CountryCode.txt on the server. So, for the country code us, the flag file us.gif and the text file is us.txt.

***38.11 (Syntax highlighting) Rewrite Exercise 37.12 using JSP. **38.12 (Opinion poll) Rewrite Exercise 37.13 using JSP. ***38.13 (Multiple-question opinion poll) The Poll table in Exercise 37.13 contains only

one question. Suppose you have a Poll table that contains multiple questions. Write a JSP that reads all the questions from the table and display them in a form, as shown in Figure 38.13a. When the user clicks the Submit button, another JSP page is invoked. This page updates the Yes or No counts for each question and displays the current Yes and No counts for each question in the Poll table, as shown in Figure 38.13b. Note that the table may contain many questions. The questions in the figure are just examples. Sort the questions in alphabetical order.

M38_LIAN1878_11_GE_C38.indd 31 12/29/17 5:56 PM

38-32 Chapter 38 Javaserver Pages

Figure 38.13 The form prompts the user to enter Yes or No for each question in (a), and the updated Yes or No counts are displayed in (b).

Figure 38.12 The program displays an image and the description of the image.

**38.14 (Addition quiz) Write a JSP program that generates addition quizzes randomly, as shown in Figure 38.14a. After the user answers all questions, the JSP displays the result, as shown in Figure 38.14b.

Figure 38.14 The program displays addition questions in (a) and answers in (b).

**38.15 (Subtraction quiz) Write a JSP program that generates subtraction quizzes randomly, as shown in Figure 38.14a. The first number must always be greater than or equal to the second number. After the user answers all questions, the JSP displays the result, as shown in Figure 38.14b.

M38_LIAN1878_11_GE_C38.indd 32 12/29/17 5:56 PM

Programming Exercises 38-33

Figure 38.14 The program displays subtraction questions in (a) and answers in (b).

**38.16 (Guess birthday) Listing 3.3, GuessBirthDay.java, gives a program for guessing a birthday. Write a JSP program that displays five sets of numbers, as shown in Figure 38.15a. After the user checks the appropriate boxes and clicks the Find Date button, the program displays the date, as shown in Figure 38.15b.

Figure 38.15 (a) The program displays five sets of numbers for the user to check the boxes. (b) The program displays the date.

**38.17 (Guess capitals) Write a JSP that prompts the user to enter a capital for a state, as shown in Figure 38.16a. Upon receiving the user input, the program reports whether the answer is correct, as shown in Figure 38.16b. You can click the Next button to display another question. You can use a two-dimensional array to store the states and capitals, as proposed in Exercise 9.22. Create a list from the array and apply the shuffle method to reorder the list so the questions will appear in random order.

Figure 38.16 (a) The program displays a question. (b) The program displays the answer to the question.

*38.18 (Large factorial) Rewrite Listing 38.11 to handle a large factorial. Use the BigInteger class introduced in §14.12.

M38_LIAN1878_11_GE_C38.indd 33 12/29/17 5:56 PM

38-34 Chapter 38 Javaserver Pages

Figure 38.17 The JSP page lets you view, insert, and update staff information.

**38.19 (Access and update a Staff table) Write a JSP for Exercise 33.1, as shown in Figure 38.17.

*38.20 (Guess number) Write a JSP page that generates a random number between 1 and 1000 and let the user enter a guess. When the user enters a guess, the program should tell the user whether the guess is correct, too high, or too low.

M38_LIAN1878_11_GE_C38.indd 34 12/29/17 5:56 PM

Objectives ■■ To explain what JSF is (§39.1).

■■ To create a JSF project in NetBeans (§39.2.1).

■■ To create a JSF page (§39.2.2).

■■ To create a JSF managed bean (§39.2.3).

■■ To use JSF expressions in a facelet (§39.2.4).

■■ To use JSF GUI components (§39.3).

■■ To obtain and process input from a form (§39.4).

■■ To develop a calculator using JSF (§39.5).

■■ To track sessions in application, session, view, and request scopes (§39.6).

■■ To validate input using the JSF validators (§39.7).

■■ To bind database with facelets (§39.8).

■■ To open a new JSF page from the current page (§39.9).

■■ To program using contexts and dependency injection (§39.10).

JavaServer Faces

Chapter

M39_LIAN1878_11_GE_C39.indd 1 12/29/17 6:01 PM

39-2 Chapter 39 JavaServer Faces

39.1 Introduction JavaServer Faces (JSF) is a new technology for developing server-side Web applications using Java.

JSF enables you to completely separate Java code from HTML. You can quickly build Web applications by assembling reusable UI components in a page, connecting these components to Java programs and wiring client-generated events to server-side event handlers. The applica- tion developed using JSF is easy to debug and maintain.

Note This chapter introduces JSF 2, the latest standard for JavaServer Faces. You need to know XHTML (eXtensible HyperText Markup Language) and CSS (Cascading Style Sheet) to start this chapter. For information on XHTML and CSS, see Supplements V.A and V.B.

Caution The examples and exercises in this chapter were tested using NetBeans 7.3.1, GlassFish 4, and Java EE 7. You need to use NetBeans 7.3.1 or a higher version with GlassFish 4 and Java EE 7 to develop your JSF projects.

39.2 Getting Started with JSF NetBeans is an effective tool for developing JSF applications.

We begin with a simple example that illustrates the basics of developing JSF projects using NetBeans. The example is to display the date and time on the server, as shown in Figure 39.1.

JSF

JSF 2 XHTML CSS

NetBeans 7.3.1 GlassFish 4 Java EE 7

Point Key

Figure 39.1 The application displays the date and time on the server.

Point Key

39.2.1 Creating a JSF Project Here are the steps to create the application.

Step 1: Choose File, New Project to display the New Project dialog box. In this box, choose Java Web in the Categories pane and Web Application in the Projects pane. Click Next to display the New Web Application dialog box.

In the New Web Application dialog box, enter and select the following fields, as shown in Figure 39.2a:

Project Name: jsf2demo Project Location: c:\book

Step 2: Click Next to display the dialog box for choosing servers and settings. Select the following fields as shown in Figure 39.2b. (Note: You can use any server such as Glass- Fish 4.x that supports Java EE 6.)

Server: GlassFish 4 Java EE Version: Java EE 7 Web

create a project

choose server and Java EE 7

M39_LIAN1878_11_GE_C39.indd 2 12/29/17 6:01 PM

39.2 Getting Started with JSF 39-3

Step 3: Click Next to display the dialog box for choosing frameworks, as shown in Figure 39.3. Check JavaServer Faces and JSF 2.0 as Server Library. Click Finish to create the project, as shown in Figure 39.4.

choose JavaServer Faces and JSF2.2

Figure 39.2 The New Web Application dialog box enables you to create a new Web project.

(a) (b)

Figure 39.3 Check JavaServer Faces and JSF 2.2 to create a Web project.

39.2.2 A Basic JSF Page A new project was just created with a default page named index.xhtml, as shown in Figure 39.4. This page is known as a facelet, which mixes JSF tags with XHTML tags. Listing 39.1 lists the contents of index.xhtml.

Listing 39.1 index.xhtml 1 <?xml version='1.0' encoding='UTF-8' ?> 2  3 <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" 4 "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> 5 <html xmlns="http://www.w3.org/1999/xhtml" 6 xmlns:h="http://xmlns.jcp.org/jsf/html">

facelet

xml version comment DOCTYPE

default namespace JSF namespace

M39_LIAN1878_11_GE_C39.indd 3 12/29/17 6:01 PM

39-4 Chapter 39 JavaServer Faces

7 <h:head> 8 <title>Facelet Title</title> 9 </h:head> 10 <h:body> 11 Hello from Facelets 12 </h:body> 13 </html>

Line 1 is an XML declaration to state that the document conforms to the XML version 1.0 and uses the UTF-8 encoding. The declaration is optional, but it is a good practice to use it. A docu- ment without the declaration may be assumed of a different version, which may lead to errors. If an XML declaration is present, it must be the first item to appear in the document. This is because an XML processor looks for the first line to obtain information about the document so that it can be processed correctly.

Line 2 is a comment for documenting the contents in the file. XML comment always begins with .

Lines 3 and 4 specify the version of XHTML used in the document. This can be used by the Web browser to validate the syntax of the document.

An XML document consists of elements described by tags. An element is enclosed between a start tag and an end tag. XML elements are organized in a tree-like hierarchy. Elements may con- tain subelements, but there is only one root element in an XML document. All the elements must be enclosed inside the root tag. The root element in XHTML is defined using the html tag (line 5).

Each tag in XML must be used in a pair of the start tag and the end tag. A start tag begins with < followed by the tag name and ends with >. An end tag is the same as its start tag except that it begins with </. The start tag and end tag for html are <html> and </html>.

The html element is the root element that contains all other elements in an XHTML page. The starting <html> tag (lines 5 and 6) may contain one or more xmlns (XML namespace) attributes to specify the namespace for the elements used in the document. Namespaces are like Java packages. Java packages are used to organize classes and to avoid naming conflict. XHTML namespaces are used to organize tags and resolve naming conflict. If an element with the same name is defined in two namespaces, the fully qualified tag names can be used to differentiate them.

XML declaration

XML comment

DOCTYPE

element

tag

html tag

Figure 39.4 A default JSF page is created in a new Web project.

h:head

h:body

M39_LIAN1878_11_GE_C39.indd 4 12/29/17 6:01 PM

39.2 Getting Started with JSF 39-5

Each xmlns attribute has a name and a value separated by an equal sign (=). The following declaration (line 5)

xmlns="http://www.w3.org/1999/xhtml"

specifies that any unqualified tag names are defined in the default standard XHTML namespace. The following declaration (line 6)

xmlns:h="http://xmlns.jcp.org/jsf/html"

allows the tags defined in the JSF tag library to be used in the document. These tags must have a prefix h.

An html element contains a head and a body. The h:head element (lines 7–9) defines an HTML title element. The title is usually displayed in the browser window’s title bar.

An h:body element defines the page’s content. In this simple example, it contains a string to be displayed in the Web browser.

Note The XML tag names are case sensitive, whereas HTML tags are not. So, <html> is differ- ent from <HTML> in XML. Every start tag in XML must have a matching end tag, whereas some tags in HTML do not need end tags.

You can now display the page in index.xhtml by right-clicking on index.xhtml in the projects pane and choose Run File. The page is displayed in a browser, as shown in Figure 39.5.

xmlns

h:head

h:body

Figure 39.5 The index.xhtml is displayed in the browser.

Note The JSF page is processed and converted into a regular HTML page for displaying by a browser. The Java software that runs on the server side for producing the HTML page is known as Java server container or simply container. The container is responsible for handling all server-side tasks for Java EE. GlassFish is a Java server container.

39.2.3 Managed JavaBeans for JSF JSF applications are developed using the Model-View-Controller (MVC) architecture, which separates the application’s data (contained in the model) from the graphical presentation (the view). The controller is the JSF framework that is responsible for coordinating interactions between view and the model.

In JSF, the facelets are the view for presenting data. Data are obtained from Java objects. Objects are defined using Java classes. In JSF, the objects that are accessed from a facelet are JavaBeans objects. A JavaBean class is simply a public Java class with a no-arg constructor. JavaBeans may contain properties. By convention, a property is defined with a getter and a setter method. If a property only has a getter method, the property is called a read-only prop- erty. If a property only has a setter method, the property is called a write-only property. A property does not need to be defined as a data field in the class.

container

JavaBean

M39_LIAN1878_11_GE_C39.indd 5 12/29/17 6:01 PM

39-6 Chapter 39 JavaServer Faces

Our example in this section is to develop a JSF facelet to display current time. We will cre- ate a JavaBean with a getTime() method that returns the current time as a string. The facelet will invoke this method to obtain current time.

Here are the steps to create a JavaBean named TimeBean.

Step 1. Right-click the project node jsf2demo to display a context menu as shown in Figure 39.6. Choose New, JSF Managed Bean to display the New JSF Managed Bean dialog box, as shown in Figure 39.7. (Note: if you don’t see JSF Managed Bean in the menu, choose Other to locate it in the JavaServer Faces category.)

Step 2. Enter and select the following fields, as shown in Figure 39.7:

Class Name: TimeBean Package: jsf2demo Name: timeBean Scope: request Click Finish to create TimeBean.java, as shown in Figure 39.8.

Step 3. Add the getTime() method to return the current time, as shown in Listing 39.2.

Figure 39.6 Choose JSF Managed Bean to create a JavaBean for JSF.

M39_LIAN1878_11_GE_C39.indd 6 12/29/17 6:01 PM

39.2 Getting Started with JSF 39-7

Figure 39.7 Specify the name, location, and scope for the bean.

Figure 39.8 A JavaBean for JSF was created.

M39_LIAN1878_11_GE_C39.indd 7 12/29/17 6:01 PM

39-8 Chapter 39 JavaServer Faces

Listing 39.2 TimeBean.java 1 package jsf2demo; 2 3 import javax.inject.Named; 4 import javax.enterprise.context.RequestScoped; 5 6 @Named (value = "timeBean") 7 @RequestScoped 8 public class TimeBean { 9 public TimeBean() { 10 } 11 12 public String getTime() { 13 return new java.util.Date().toString(); 14 } 15 }

TimeBean is a JavaBeans with the @Named annotation, which indicates that the JSF framework will create and manage the TimeBean objects used in the application. You have learned to use the @Override annotation in Chapter 11. The @Override annotation tells the compiler that the annotated method is required to override a method in a superclass. The @Named annotation tells the compiler to generate the code to enable the bean to be used by JSF facelets.

The @RequestScope annotation specifies that the scope of the JavaBeans object is within a request. You can also use @ViewScope, @SessionScope or @ApplicationScope to specify the scope for a session or for the entire application.

39.2.4 JSF Expressions We demonstrate JSF expressions by writing a simple application that displays the current time. You can display current time by invoking the getTime() method in a TimeBean object using a JSF expression.

To keep index.xhtml intact, we create a new JSF page named CurrentTime.xhtml as follows:

Step 1. Right-click the jsf2demo node in the project pane to display a context menu and choose New, JSF Page to display the New JSF File dialog box, as shown in Figure 39.9.

Step 2. Enter CurrentTime in the File Name field, choose Facelets and click Finish to generate CurrentTime.xhtml, as shown in Figure 39.10.

Step 3. Add a JSF expression to obtain the current time, as shown in Listing 39.3.

Step 4. Right-click on CurrentTime.xhtml in the project to display a context menu and choose Run File to display the page in a browser as shown in Figure 39.1.

Listing 39.3 CurrentTime.xhtml 1 <?xml version='1.0' encoding='UTF-8' ?> 2 <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" 3 "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> 4 <html xmlns="http://www.w3.org/1999/xhtml" 5 xmlns:h="http://xmlns.jcp.org/jsf/html"> 6 <h:head> 7 <title>Display Current Time</title> 8 <meta http-equiv="refresh" content ="60" /> 9 </h:head> 10 <h:body> 11 The current time is #{timeBean.time} 12 </h:body> 13 </html>

@RequestScope

refresh page

JSF expression

@Named @RequestScoped

time property

M39_LIAN1878_11_GE_C39.indd 8 12/29/17 6:01 PM

39.2 Getting Started with JSF 39-9

Figure 39.9 The New JSF Page dialog is used to create a JSF page.

Figure 39.10 A New JSF page CurrentTime was created.

Line 11 defines a meta tag inside the h:head tag to tell the browser to refresh every 60 seconds. This line can also be written as

An element is called an empty element if there are no contents between the start tag and the end tag. In an empty element, data are typically specified as attributes in the start tag. You can close an empty element by placing a slash immediately preceding the start tag’s right angle bracket, as shown in line 8, for brevity.

empty element

M39_LIAN1878_11_GE_C39.indd 9 12/29/17 6:01 PM

39-10 Chapter 39 JavaServer Faces

Line 8 uses a JSF expression #{timeBean.time} to obtain the current time. timeBean is an object of the TimeBean class. The object name can be changed in the @Named annotation (line 6 in Listing 39.2) using the following syntax:

@Named(name = "anyObjectName")

By default, the object name is the class name with the first letter in lowercase. Note that time is a JavaBeans property because the getTime() method is defined in Time-

Beans. The JSF expression can either use the property name or invoke the method to obtain the current time. So the following two expressions are fine:

#{timeBean.time} #{timeBean.getTime()}

The syntax of a JSF expression is

#{expression}

JSF expressions bind JavaBeans objects with facelets. You will see more use of JSF expres- sions in the upcoming examples in this chapter.

39.2.1 What is JSF? 39.2.2 How do you create a JSF project in NetBeans? 39.2.3 How do you create a JSF page in a JSF project? 39.2.4 What is a facelet? 39.2.5 What is the file extension name for a facelet? 39.2.6 What is a managed bean? 39.2.7 What is the @Named annotation for? 39.2.8 What is the @RequestScope annotation for?

39.3 JSF GUI Components JSF provides many elements for displaying GUI components.

Table 39.1 lists some of the commonly used elements. The tags with the h prefix are in the JSF HTML Tag library. The tags with the f prefix are in the JSF Core Tag library.

Listing 39.4 is an example that uses some of these elements to display a student registration form, as shown in Figure 39.11.

Listing 39.4 StudentRegistrationForm.xhtml 1 <?xml version='1.0' encoding='UTF-8' ?> 2 <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" 3 "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> 4 <html xmlns="http://www.w3.org/1999/xhtml" 5 xmlns:h="http://xmlns.jcp.org/jsf/html" 6 xmlns:f="http://xmlns.jcp.org/jsf/core"> 7 <h:head> 8 <title>Student Registration Form</title> 9 </h:head> 10 <h:body> 11 <h:form> 12  13 <h3>Student Registration Form 14 <h:graphicImage name="usIcon.gif" library="image"/> 15 </h3>

Point Check

Point Key

jsf core namespace

graphicImage

M39_LIAN1878_11_GE_C39.indd 10 12/29/17 6:01 PM

39.3 JSF GUI Components 39-11

JSF Tag Description

h:form inserts an XHTML form into a page.

h:panelGroup similar to a JavaFX FlowPane.

h:panelGrid similar to a JavaFX GridPane.

h:inputText displays a textbox for entering input.

h:outputText displays a textbox for displaying output.

h:inputTextArea displays a textarea for entering input.

h:inputSecret displays a textbox for entering password.

h:outputLabel displays a label.

h:outputLink displays a hypertext link.

h:selectOneMenu displays a combo box for selecting one item.

h:selectOneRadio displays a set of radio button.

h:selectManyCheckbox displays checkboxes.

h:selectOneListbox displays a list for selecting one item.

h:selectManyListbox displays a list for selecting multiple items.

f:selectItem specifies an item in an h:selectOneMenu, h:selectOneRadio, or h:selectManyListbox.

h:message displays a message for validating input.

h:dataTable displays a data table.

h:column specifies a column in a data table.

h:graphicImage displays an image.

tabLe 39.1 JSF GUI Form Elements

Figure 39.11 A student registration form is displayed using JSF elements.

M39_LIAN1878_11_GE_C39.indd 11 12/29/17 6:01 PM

39-12 Chapter 39 JavaServer Faces

16 17  18 <h:panelGrid columns="6" style="color:green"> 19 <h:outputLabel value="Last Name"/> 20 <h:inputText id="lastNameInputText" /> 21 <h:outputLabel value="First Name" /> 22 <h:inputText id="firstNameInputText" /> 23 <h:outputLabel value="MI" /> 24 <h:inputText id="miInputText" size="1" /> 25 </h:panelGrid> 26 27  28 <h:panelGrid columns="2"> 29 <h:outputLabel>Gender </h:outputLabel> 30 <h:selectOneRadio id="genderSelectOneRadio"> 31 <f:selectItem itemValue="Male" 32 itemLabel="Male"/> 33 <f:selectItem itemValue="Female" 34 itemLabel="Female"/> 35 </h:selectOneRadio> 36 </h:panelGrid> 37 38  39 <h:panelGrid columns="4"> 40 <h:outputLabel value="Major "/> 41 <h:selectOneMenu id="majorSelectOneMenu"> 42 <f:selectItem itemValue="Computer Science"/> 43 <f:selectItem itemValue="Mathematics"/> 44 </h:selectOneMenu> 45 <h:outputLabel value="Minor "/> 46 <h:selectManyListbox id="minorSelectManyListbox"> 47 <f:selectItem itemValue="Computer Science"/> 48 <f:selectItem itemValue="Mathematics"/> 49 <f:selectItem itemValue="English"/> 50 </h:selectManyListbox> 51 </h:panelGrid> 52 53  54 <h:panelGrid columns="4"> 55 <h:outputLabel value="Hobby: "/> 56 <h:selectManyCheckbox id="hobbySelectManyCheckbox"> 57 <f:selectItem itemValue="Tennis"/> 58 <f:selectItem itemValue="Golf"/> 59 <f:selectItem itemValue="Ping Pong"/> 60 </h:selectManyCheckbox> 61 </h:panelGrid> 62 63  64 <h:panelGrid columns="1"> 65 <h:outputLabel>Remarks:</h:outputLabel> 66 <h:inputTextarea id="remarksInputTextarea" 67 style="width:400px; height:50px;" /> 68 </h:panelGrid> 69 70  71 <h:commandButton value="Register" /> 72 </h:form> 73 </h:body> 74 </html>

h:panelGrid

h:outputLabel

h:inputText

h:selectOneRadio f:selectItem

h:selectOneMenu

h:selectManyListBox

h:selectManyCheckbox

h:inputTextarea

h:commandButton

M39_LIAN1878_11_GE_C39.indd 12 12/29/17 6:01 PM

39.3 JSF GUI Components 39-13

The tags with prefix f are in the JSF core tag library. Line 6

xmlns:f="http://xmlns.jcp.org/jsf/core">

locates the library for these tags. The h:graphicImage tag displays an image in the file usIcon.gif (line 14). The file is

located in the /resources/image folder. In JSF 2.0, all resources (image files, audio files, and CCS files) should be placed under the resources folder under the Web Pages node. You can create these folders as follows:

Step 1: Right-click the Web Pages node in the project pane to display a context menu and choose New, Folder to display the New Folder dialog box. (If Folder is not in the context menu, choose Other to locate it.)

Step 2: Enter resources as the Folder Name and click Finish to create the resources folder, as shown in Figure 39.12.

Step 3: Right-click the resources node in the project pane to create the image folder under resources. You can now place usIcon.gif under the image folder.

jsf core xmlns

h:graphicImage

Figure 39.12 The resources folder was created.

JSF provides h:panelGrid and h:panelGroup elements to contain and layout subele- ments. h:panelGrid places the elements in a grid like the JavaFX GridPane. h:panelGrid places the elements in a grid with the specified number of columns. Lines 18–25 place six elements (labels and input texts) that are in an h:panelGrid. The columns attribute specifies that each row in the grid has 6 columns. The elements are placed into a row from left to right in the order they appear in the facelet. When a row is full, a new row is created to hold the elements. We used h:panelGrid in this example. You may replace it with h:panelGroup to see how the elements would be arranged.

You may use the style attribute with a JSF html tag to specify the CSS style for the ele- ment and its subelements. The style attribute in line 18 specifies the color green for all ele- ments in this h:panelGrid element.

The h:outputLabel element is for displaying a label (line 19). The value attribute speci- fies the label’s text.

The h:inputText element is for displaying a text input box for the user to enter a text (line 20). The id attribute is useful for other elements or the server program to reference this element.

The h:selectOneRadio element is for displaying a group of radio buttons (line 30). Each radio button is defined using an f:selectItem element (lines 31–34).

h:panelGrid

the style attribute

h:outputLabel

h:inputText

h:selectOneRadio

M39_LIAN1878_11_GE_C39.indd 13 12/29/17 6:01 PM

39-14 Chapter 39 JavaServer Faces

The h:selectOneMenu element is for displaying a combo box (line 41). Each item in the combo box is defined using an f:selectItem element (lines 42 and 43).

The h:selectManyListbox element is for displaying a list for the user to choose multiple items in a list (line 46). Each item in the list is defined using an f:selectItem element (lines 47–49).

The h:selectManyCheckbox element is for displaying a group of check boxes (line 56). Each item in the check box is defined using an f:selectItem element (lines 57–59).

The h:selectTextarea element is for displaying a text area for multiple lines of input (line 66). The style attribute is used to specify the width and height of the text area (line 67).

The h:commandButton element is for displaying a button (line 71). When the button is clicked, an action is performed. The default action is to request the same page from the server. The next section shows how to process the form.

39.3.1 What is the name space for JSF tags with prefix h and prefix f? 39.3.2 Describe the use of the following tags? 1 h:form, h:panelGroup, h:panelGrid, h:inputText, h:outputText, 2 h:inputTextArea, h:inputSecret, h:outputLabel, h:outputLink, 3 h:selectOneMenu, h:selectOneRadio, h:selectBooleanCheckbox, 4 h:selectOneListbox, h:selectManyListbox, h:selectItem, 5 h:message, h:dataTable, h:columm, h:graphicImage

39.4 Processing the Form Processing forms is a common task for Web programming. JSF provides tools for pro- cessing forms.

The preceding section introduced how to display a form using common JSF elements. This section shows how to obtain and process the input.

To obtain input from the form, simply bind each input element with a property in a man- aged bean. We now define a managed bean named registration as shown in Listing 39.5.

Listing 39.5 RegistrationJSFBean.java 1 package jsf2demo; 2 3 import javax.enterprise.context.RequestScoped; 4 import javax.inject.Named; 5 6 @Named(value = "registration") 7 @RequestScoped 8 public class RegistrationJSFBean { 9 private String lastName; 10 private String firstName; 11 private String mi; 12 private String gender; 13 private String major; 14 private String[] minor; 15 private String[] hobby; 16 private String remarks; 17 18 public RegistrationJSFBean() { 19 } 20 21 public String getLastName() { 22 return lastName; 23 }

h:selectOneMenu

h:selectManyListbox

h:selectManyCheckbox

h:selectTextarea

h:commandButton

Point Check

Point Key

managed bean request scope property lastName

M39_LIAN1878_11_GE_C39.indd 14 12/29/17 6:01 PM

39.4 Processing the Form 39-15

24 25 public void setLastName(String lastName) { 26 this.lastName = lastName; 27 } 28 29 public String getFirstName() { 30 return firstName; 31 } 32 33 public void setFirstName(String firstName) { 34 this.firstName = firstName; 35 } 36 37 public String getMi() { 38 return mi; 39 } 40 41 public void setMi(String mi) { 42 this.mi = mi; 43 } 44 45 public String getGender() { 46 return gender; 47 } 48 49 public void setGender(String gender) { 50 this.gender = gender; 51 } 52 53 public String getMajor() { 54 return major; 55 } 56 57 public void setMajor(String major) { 58 this.major = major; 59 } 60 61 public String[] getMinor() { 62 return minor; 63 } 64 65 public void setMinor(String[] minor) { 66 this.minor = minor; 67 } 68 69 public String[] getHobby() { 70 return hobby; 71 } 72 73 public void setHobby(String[] hobby) { 74 this.hobby = hobby; 75 } 76 77 public String getRemarks() { 78 return remarks; 79 } 80 81 public void setRemarks(String remarks) { 82 this.remarks = remarks; 83 }

M39_LIAN1878_11_GE_C39.indd 15 12/29/17 6:01 PM

39-16 Chapter 39 JavaServer Faces

84 85 public String getResponse() { 86 if (lastName == null) 87 return ""; // Request has not been made 88 else { 89 String allMinor = ""; 90 for (String s: minor) { 91 allMinor += s + " "; 92 } 93 94 String allHobby = ""; 95 for (String s: hobby) { 96 allHobby += s + " "; 97 } 98 99 return "<p style=\"color:red\">You entered <br />" + 100 "Last Name: " + lastName + "<br />" + 101 "First Name: " + firstName + "<br />" + 102 "MI: " + mi + "<br />" + 103 "Gender: " + gender + "<br />" + 104 "Major: " + major + "<br />" + 105 "Minor: " + allMinor + "<br />" + 106 "Hobby: " + allHobby + "<br />" + 107 "Remarks: " + remarks + "</p>"; 108 } 109 } 110 }

The RegistrationJSFBean class is a managed bean that defines the properties lastName, firstName, mi, gender, major, minor, and remarks, which will be bound to the elements in the JSF registration form.

The registration form can now be revised as shown in Listing 39.6. Figure 39.13 shows that new JSF page displays the user input upon clicking the Register button.

Listing 39.6 ProcessStudentRegistrationForm.xhtml 1 <?xml version='1.0' encoding='UTF-8' ?> 2 <!DOCTYPE html PUBLIC "–//W3C//DTD XHTML 1.0 Transitional//EN" 3 "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> 4 <html xmlns="http://www.w3.org/1999/xhtml" 5 xmlns:h="http://xmlns.jcp.org/jsf/html" 6 xmlns:f="http://xmlns.jcp.org/jsf/core"> 7 <h:head> 8 <title>Student Registration Form</title> 9 </h:head> 10 <h:body> 11 <h:form> 12  13 <h3>Student Registration Form 14 <h:graphicImage name="usIcon.gif" library="image"/> 15 </h3> 16 17  18 <h:panelGrid columns="6" style="color:green"> 19 <h:outputLabel value="Last Name"/> 20 <h:inputText id="lastNameInputText" 21 value="#{registration.lastName}"/> 22 <h:outputLabel value="First Name" /> 23 <h:inputText id="firstNameInputText" 24 value="#{registration.firstName}"/>

bean properties

getResponse()

jsf core namespace

bind lastName

bind firstName

M39_LIAN1878_11_GE_C39.indd 16 12/29/17 6:01 PM

39.4 Processing the Form 39-17

25 <h:outputLabel value="MI" /> 26 <h:inputText id="miInputText" size="1" 27 value="#{registration.mi}"/> 28 </h:panelGrid> 29 30  31 <h:panelGrid columns="2"> 32 <h:outputLabel>Gender </h:outputLabel> 33 <h:selectOneRadio id="genderSelectOneRadio" 34 value="#{registration.gender}"> 35 <f:selectItem itemValue="Male" 36 itemLabel="Male"/> 37 <f:selectItem itemValue="Female" 38 itemLabel="Female"/> 39 </h:selectOneRadio> 40 </h:panelGrid> 41 42  43 <h:panelGrid columns="4"> 44 <h:outputLabel value="Major "/> 45 <h:selectOneMenu id="majorSelectOneMenu" 46 value="#{registration.major}"> 47 <f:selectItem itemValue="Computer Science"/> 48 <f:selectItem itemValue="Mathematics"/> 49 </h:selectOneMenu> 50 <h:outputLabel value="Minor "/> 51 <h:selectManyListbox id="minorSelectManyListbox" 52 value="#{registration.minor}"> 53 <f:selectItem itemValue="Computer Science"/> 54 <f:selectItem itemValue="Mathematics"/> 55 <f:selectItem itemValue="English"/> 56 </h:selectManyListbox> 57 </h:panelGrid> 58 59  60 <h:panelGrid columns="4"> 61 <h:outputLabel value="Hobby: "/> 62 <h:selectManyCheckbox id="hobbySelectManyCheckbox" 63 value="#{registration.hobby}"> 64 <f:selectItem itemValue="Tennis"/> 65 <f:selectItem itemValue="Golf"/> 66 <f:selectItem itemValue="Ping Pong"/> 67 </h:selectManyCheckbox> 68 </h:panelGrid> 69 70  71 <h:panelGrid columns="1"> 72 <h:outputLabel>Remarks:</h:outputLabel> 73 <h:inputTextarea id="remarksInputTextarea" 74 style="width:400px; height:50px;" 75 value="#{registration.remarks}"/> 76 </h:panelGrid> 77 78  79 <h:commandButton value="Register" /> 80 <br /> 81 <h:outputText escape="false" style="color:red" 82 value="#{registration.response}" /> 83 </h:form> 84 </h:body> 85 </html>

bind mi

bind gender

bind major

bind minor

bind hobby

bind remarks

bind response

M39_LIAN1878_11_GE_C39.indd 17 12/29/17 6:01 PM

39-18 Chapter 39 JavaServer Faces

The new JSF form in this listing binds the h:inputText element for last name, first name, and mi with the properties lastName, firstName, and mi in the managed bean (lines 21, 24, and 27). When the Register button is clicked, the page is sent to the server, which invokes the setter methods to set the properties in the managed bean.

The h:selectOneRadio element is bound to the gender property (line 34). Each radio button has an itemValue. The selected radio button’s itemValue is set to the gender prop- erty in the bean when the page is sent to the server.

The h:selectOneMenu element is bound to the major property (line 46). When the page is sent to the server, the selected item is returned as a string and is set to the major property.

The h:selectManyListbox element is bound to the minor property (line 52). When the page is sent to the server, the selected items are returned as an array of strings and set to the minor property.

The h:selectManyCheckbox element is bound to the hobby property (line 63). When the page is sent to the server, the checked boxes are returned as an array of itemValues and set to the hobby property.

The h:selectTextarea element is bound to the remarks property (line 75). When the page is sent to the server, the content in the text area is returned as a string and set to the remarks property.

The h:outputText element is bound to the response property (line 82). This is a read- only property in the bean. It is "" if lastName is null (lines 86 and 87 in Listing 39.5). When the page is returned to the client, the response property value is displayed in the output text element (line 82).

binding input texts

binding radio buttons

binding combo box

binding list box

binding check boxes

binding text area

binding response

Figure 39.13 The user input is collected and displayed after clicking the Register button.

M39_LIAN1878_11_GE_C39.indd 18 12/29/17 6:01 PM

39.5 Case Study: Calculator 39-19

The h:outputText element’s escape attribute is set to false (line 81) to enable the contents to be displayed in HTML. By default, the escape attribute is true, which indicates the contents are considered regular text.

39.4.1 In the h:outputText tag, what is the escape attribute for? 39.4.2 Does every GUI component tag in JSF have the style attribute?

39.5 Case Study: Calculator This section gives a case study on using GUI elements and processing forms.

This section uses JSF to develop a calculator to perform addition, subtraction, multiplication, and division, as shown in Figure 39.14.

escape attribute

Point Check

Point Key

Figure 39.14 This JSF application enables you to perform addition, subtraction, multiplication, and division.

Here are the steps to develop this project:

Step 1. Create a new managed bean named calculator with the request scope as shown in Listing 39.7, CalculatorJSFBean.java.

Step 2. Create a JSF facelet in Listing 39.8, Calculator.xhtml.

Listing 39.7 CalculatorJSFBean.java 1 package jsf2demo; 2 3 import javax.inject.Named; 4 import javax.enterprise.context.RequestScoped; 5 6 @Named(value = "calculator") 7 @RequestScoped 8 public class CalculatorJSFBean { 9 private Double number1; 10 private Double number2; 11 private Double result; 12 13 public CalculatorJSFBean() { 14 } 15 16 public Double getNumber1() { 17 return number1; 18 } 19 20 public Double getNumber2() { 21 return number2; 22 } 23

create managed bean

create JSF facelet

property number1 property number2 property result

M39_LIAN1878_11_GE_C39.indd 19 12/29/17 6:01 PM

39-20 Chapter 39 JavaServer Faces

24 public Double getResult() { 25 return result; 26 } 27 28 public void setNumber1(Double number1) { 29 this.number1 = number1; 30 } 31 32 public void setNumber2(Double number2) { 33 this.number2 = number2; 34 } 35 36 public void setResult(Double result) { 37 this.result = result; 38 } 39 40 public void add() { 41 result = number1 + number2; 42 } 43 44 public void subtract() { 45 result = number1 - number2; 46 } 47 48 public void divide() { 49 result = number1 / number2; 50 } 51 52 public void multiply() { 53 result = number1 * number2; 54 } 55 }

The managed bean has three properties number1, number2, and result (lines 9–38). The methods add(), subtract(), divide(), and multiply() add, subtract, multiply, and divide number1 with number2 and assigns the result to result (lines 40–54).

Listing 39.8 Calculator.xhtml 1 <?xml version='1.0' encoding='UTF-8' ?> 2 <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" 3 "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> 4 <html xmlns="http://www.w3.org/1999/xhtml" 5 xmlns:h="http://xmlns.jcp.org/jsf/html"> 6 <h:head> 7 <title>Calculator</title> 8 </h:head> 9 <h:body> 10 <h:form> 11 <h:panelGrid columns="6"> 12 <h:outputLabel value="Number 1"/> 13 <h:inputText id="number1InputText" size ="4" 14 style="text-align: right" 15 value="#{calculator.number1}"/> 16 <h:outputLabel value="Number 2" /> 17 <h:inputText id="number2InputText" size ="4" 18 style="text-align: right" 19 value="#{calculator.number2}"/> 20 <h:outputLabel value="Result" /> 21 <h:inputText id="resultInputText" size ="4"

add

subtract

divide

multiply

right align bind text input

M39_LIAN1878_11_GE_C39.indd 20 12/29/17 6:01 PM

39.6 Session Tracking 39-21

22 style="text-align: right" 23 value="#{calculator.result}"/> 24 </h:panelGrid> 25 26 <h:panelGrid columns="4"> 27 <h:commandButton value="Add" 28 action ="#{calculator.add}"/> 29 <h:commandButton value="Subtract" 30 action ="#{calculator.subtract}"/> 31 <h:commandButton value="Multiply" 32 action ="#{calculator.multiply}"/> 33 <h:commandButton value="Divide" 34 action ="#{calculator.divide}"/> 35 </h:panelGrid> 36 </h:form> 37 </h:body> 38 </html>

Three text input components along with their labels are placed in the grid panel (lines 11–24). Four button components are placed in the grid panel (lines 26–35).

The bean property number1 is bound to the text input for Number 1 (line 15). The CSS style text-align: right (line 14) specifies that the text is right aligned in the input box.

The action attribute for the Add button is set to the add method in the calculator bean (line 28). When the Add button is clicked, the add method in the bean is invoked to add number1 with number2 and assign the result to result. Since the result property is bound to the Result input text (line 23), the new result is now displayed in the text input field.

39.6 Session Tracking You can create a managed bean at the application scope, session scope, view scope, or request scope.

JSF supports session tracking for managed beans at the application scope, session scope, view scope, and request scope. The scope is the lifetime of a bean. A request-scoped bean is alive in a single HTTP request. After the request is processed, the bean is no longer alive. A view-scoped bean lives as long as you are in the same JSF page. A session-scoped bean is alive for the entire Web session between a client and the server. An application-scoped bean lives as long as the Web application runs. In essence, a request-scoped bean is created once for a request; a view-scoped bean is created once for the view; a session-scoped bean is created once for the entire session; and an application-scoped bean is created once for the entire application. A managed bean with a session scope must be serializable because the system may need to free resources during and session and stores the bean to a file if the bean is not used for a while. When the bean is used again, the system will restore the bean to the memory.

Consider the following example that prompts the user to guess a number. When the page starts, the program randomly generates a number between 0 and 99. This number is stored in a bean. When the user enters a guess, the program checks the guess with the random number in the bean and tells the user whether the guess is too high, too low, or just right, as shown in Figure 39.15.

Here are the steps to develop this project:

Step 1. Create a new managed bean named guessNumber with the view scope as shown in Listing 39.9, GuessNumberJSFBean.java.

Step 2. Create a JSF facelet in Listing 39.10, GuessNumber.xhtml.

Point Key

scope

request scope view scope session scope application scope

create managed bean

create JSF facelet

action

M39_LIAN1878_11_GE_C39.indd 21 12/29/17 6:01 PM

39-22 Chapter 39 JavaServer Faces

Listing 39.9 GuessNumberJSFBean.java 1 package jsf2demo; 2 3 import javax.inject.Named; 4 import javax.faces.view.ViewScoped; 5 6 @Named(value = "guessNumber") 7 @ViewScoped 8 public class GuessNumberJSFBean { 9 private int number; 10 private String guessString; 11 12 public GuessNumberJSFBean() { 13 number = (int)(Math.random() * 100); 14 } 15 16 public String getGuessString() { 17 return guessString; 18 }

Figure 39.15 The user enters a guess and the program displays the result.

view scope

random number guess by user

create random number

getter method

M39_LIAN1878_11_GE_C39.indd 22 12/29/17 6:01 PM

39.6 Session Tracking 39-23

19 20 public void setGuessString(String guessString) { 21 this.guessString = guessString; 22 } 23 24 public String getResponse() { 25 if (guessString == null) 26 return ""; // No user input yet 27 28 int guess = Integer.parseInt(guessString); 29 if (guess < number) 30 return "Too low"; 31 else if (guess == number) 32 return "You got it"; 33 else 34 return "Too high"; 35 } 36 }

The managed bean uses the @ViewScope annotation (line 7) to set up the view scope for the bean. The view scope is most appropriate for this project. The bean is alive as long as the view is not changed. The bean is created when the page is displayed for the first time. A random number between 0 and 99 is assigned to number (line 13) when the bean is created. This number will not change as long as the bean is alive in the same view.

The getResponse method converts guessString from the user input to an integer (line 28) and determines if the guess is too low (line 30), too high (line 34), and just right (line 32).

Listing 39.10 GuessNumber.xhtml 1 <?xml version='1.0' encoding='UTF-8' ?> 2 <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" 3 "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> 4 <html xmlns="http://www.w3.org/1999/xhtml" 5 xmlns:h="http://xmlns.jcp.org/jsf/html"> 6 <h:head> 7 <title>Guess a number</title> 8 </h:head> 9 <h:body> 10 <h:form> 11 <h:outputLabel value="Enter you guess: "/> 12 <h:inputText style="text-align: right; width: 50px" 13 id="guessInputText" 14 value="#{guessNumber.guessString}"/> 15 <h:commandButton style="margin-left: 60px" value="Guess" /> 16 <br /> 17 <h:outputText style="color: red" 18 value="#{guessNumber.response}" /> 19 </h:form> 20 </h:body> 21 </html>

The bean property guessString is bound to the text input (line 14). The CSS style text- align: right (line 13) specifies that the text is right aligned in the input box.

The CSS style margin-left: 60px (line 15) specifies that the command button has a left margin of 60 pixels.

The bean property response is bound to the text output (line 18). The CSS style color: red (line 17) specifies that the text is displayed in red in the output box.

The project uses the view scope. What happens if the scope is changed to the request scope? Every time the page is refreshed, JSF creates a new bean with a new random number. What happens if the scope is changed to the session scope? The bean will be alive as long as the

scope

setter method

get response

check guess

bind text input

bind text output

M39_LIAN1878_11_GE_C39.indd 23 12/29/17 6:01 PM

39-24 Chapter 39 JavaServer Faces

browser is alive. What happens if the scope is changed to the application scope? The bean will be created once when the application is launched from the server. So every client will use the same random number.

39.6.1 What is a scope? What are the available scopes in JSF? Explain request scope, view scope, session scope, and application scope. How do you set a request scope, view scope, session scope, and application scope in a managed bean?

39.6.2 What happens if the bean scope in Listing 39.9, GuessNumberJSFBean.java is changed to request?

39.6.3 What happens if the bean scope in Listing 39.9, GuessNumberJSFBean.java is changed to session?

39.6.4 What happens if the bean scope in Listing 39.9, GuessNumberJSFBean.java is changed to application?

39.7 Validating Input JSF provides tools for validating user input.

In the preceding GuessNumber page, an error would occur if you entered a noninteger in the input box before clicking the Guess button. One way to fix the problem is to check the text field before processing any event. But a better way is to user the validators. You can use the standard validator tags in the JSF Core Tag Library or create custom validators. Table 39.2 lists some JSF input validator tags.

Point Check

Point Key

JSF Tag Description

f:validateLength validates the length of the input.

f:validateDoubleRange validates whether numeric input falls within acceptable range of double values.

f:validateLongRange validates whether numeric input falls within acceptable range of long values.

f:validateRequired validates whether a field is not empty.

f:validateRegex validates whether the input matches a regualar expression.

f:validateBean invokes a custom method in a bean to perform custom validation.

tabLe 39.2 JSF Input Validator Tags

Consider the following example that displays a form for collecting user input as shown in Figure 39.16. All text fields in the form must be filled. If not, error messages are displayed. The SSN must be formatted correctly. If not, an error is displayed. If all input are correct, clicking Submit displays the result in an output text, as shown in Figure 39.17.

Here are the steps to create this project.

Step 1. Create a new page in Listing 39.11, ValidateForm.xhtml.

Step 2. Create a new managed bean named validateForm, as shown in Listing 39.12.

Listing 39.11 ValidateForm.xhtml 1 <?xml version='1.0' encoding='UTF-8' ?> 2 <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" 3 "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> 4 <html xmlns="http://www.w3.org/1999/xhtml"

M39_LIAN1878_11_GE_C39.indd 24 12/29/17 6:01 PM

39.7 Validating Input 39-25

Figure 39.16 The input fields are validated.

(a) The required messages are displayed if input is required, but empty.

(b) Error messages are displayed if input is incorrect.

Figure 39.17 The correct input values are displayed.

5 xmlns:h="http://xmlns.jcp.org/jsf/html" 6 xmlns:f="http://xmlns.jcp.org/jsf/core"> 7 <h:head>

M39_LIAN1878_11_GE_C39.indd 25 12/29/17 6:01 PM

39-26 Chapter 39 JavaServer Faces

8 <title>Validate Form</title> 9 </h:head> 10 <h:body> 11 <h:form> 12 <h:panelGrid columns="3"> 13 <h:outputLabel value="Name:"/> 14 <h:inputText id="nameInputText" required="true" 15 requiredMessage="Name is required" 16 validatorMessage="Name must have 1 to 10 chars" 17 value="#{validateForm.name}"> 18 <f:validateLength minimum="1" maximum="10" /> 19 </h:inputText> 20 <h:message for="nameInputText" style="color:red"/> 21 22 <h:outputLabel value="SSN:" /> 23 <h:inputText id="ssnInputText" required="true" 24 requiredMessage="SSN is required" 25 validatorMessage="Invalid SSN" 26 value="#{validateForm.ssn}"> 27 <f:validateRegex pattern="[\d]{3}–[\d]{2}–[\d]{4}"/> 28 </h:inputText> 29 <h:message for="ssnInputText" style="color:red"/> 30 31 <h:outputLabel value="Age:" /> 32 <h:inputText id="ageInputText" required="true" 33 requiredMessage="Age is required" 34 validatorMessage="Age must be between 16 and 120" 35 value="#{validateForm.ageString}"> 36 <f:validateLongRange minimum="16" maximum="120"/> 37 </h:inputText> 38 <h:message for="ageInputText" style="color:red"/> 39 40 <h:outputLabel value="Height:" /> 41 <h:inputText id="heightInputText" required="true" 42 requiredMessage="Height is required" 43 validatorMessage="Height must be between 3.5 and 9.5" 44 value="#{validateForm.heightString}"> 45 <f:validateDoubleRange minimum="3.5" maximum="9.5"/> 46 </h:inputText> 47 <h:message for="heightInputText" style="color:red"/> 48 </h:panelGrid> 49 50 <h:commandButton value="Submit" /> 51 52 <h:outputText style="color:red" 53 value="#{validateForm.response}" /> 54 </h:form> 55 </h:body> 56 </html>

For each input text field, set its required attribute true (lines 14, 23, 32, and 41) to indicate that an input value is required for the field. When a required input field is empty, the requiredMessage is displayed (lines 15, 24, 33, and 42).

The validatorMessage attribute specifies a message to be displayed if the input field is invalid (line 16). The f:validateLength tag specifies the minimum or maximum length of the input (line 18). JSF will determine whether the input length is valid.

The h:message element displays the validatorMessage if the input is invalid. The element’s for attribute specifies the id of the element for which the message will be displayed (line 20).

The f:validateRegex tag specifies a regular expression for validating the input (line 27). For information on regular expression, see Appendix H.

required attribute

requiredMessage

validatorMessage

f:validateLength

h:message

f:validateRegex

required input required message validator message

validate length

message element

validate regex

validate integer range

validate double range

M39_LIAN1878_11_GE_C39.indd 26 12/29/17 6:01 PM

39.7 Validating Input 39-27

The f:validateLongRange tag specifies a range for an integer input using the minimum and maximum attributes (line 45). In this project, a valid age value is between 16 and 120.

The f:validateDoubleRange tag specifies a range for a double input using the minimum and maximum attributes (line 36). In this project, a valid height value is between 3.5 and 9.5.

Listing 39.12 ValidateFormJSFBean.java 1 package jsf2demo; 2 3 import javax.enterprise.context.RequestScoped; 4 import javax.inject.Named; 5 6 @Named(value = "validateForm") 7 @RequestScoped 8 public class ValidateFormJSFBean { 9 private String name; 10 private String ssn; 11 private String ageString; 12 private String heightString; 13 14 public String getName() { 15 return name; 16 } 17 18 public void setName(String name) { 19 this.name = name; 20 } 21 22 public String getSsn() { 23 return ssn; 24 } 25 26 public void setSsn(String ssn) { 27 this.ssn = ssn; 28 } 29 30 public String getAgeString() { 31 return ageString; 32 } 33 34 public void setAgeString(String ageString) { 35 this.ageString = ageString; 36 } 37 38 public String getHeightString() { 39 return heightString; 40 } 41 42 public void setHeightString(String heightString) { 43 this.heightString = heightString; 44 } 45 46 public String getResponse() { 47 if (name == null || ssn == null || ageString == null 48 || heightString == null) { 49 return ""; 50 } 51 else { 52 return "You entered " + 53 " Name: " + name + 54 " SSN: " + ssn +

f:validateLongRange

f:validateDoubleRange

some input not set

M39_LIAN1878_11_GE_C39.indd 27 12/29/17 6:01 PM

39-28 Chapter 39 JavaServer Faces

55 " Age: " + ageString + 56 " Height: " + heightString; 57 } 58 } 59 }

If an input is invalid, its value is not set to the bean. So only when all input are correct, the getResponse() method will return all input values (lines 46–58).

39.7.1 Write a tag that validates an input text with minimal length of 2 and maximum 12. 39.7.2 Write a tag that validates an input text for SSN using a regular expression. 39.7.3 Write a tag that validates an input text for a double value with minimal 4.5 and

maximum 19.9.

39.7.4 Write a tag that validates an input text for an integer value with minimal 4 and maximum 20.

39.7.5 Write a tag that makes an input text required.

39.8 Binding Database with Facelets You can bind a database in JSF applications.

Often you need to access a database from a webpage. This section gives examples of building Web applications using databases.

Consider the following example that lets the user choose a course, as shown in Figure 39.18. After a course is selected in the combo box, the students enrolled in the course are displayed in the table, as shown in Figure 39.19. In this example, all the course titles in the Course table are bound to the combo box and the query result for the students enrolled in the course is bound to the table.

Here are the steps to create this project:

Step 1. Create a managed bean named courseName with application scope, as shown in Listing 39.13.

Step 2. Create a JSF in Listing 39.14, DisplayStudent.xhtml.

Step 3. Create a cascading style sheet for formatting the table as follows:

Point Check

Point Key

managed bean

JSF page

style sheet

Figure 39.18 You need to choose a course and display the students enrolled in the course.

M39_LIAN1878_11_GE_C39.indd 28 12/29/17 6:01 PM

39.8 Binding Database with Facelets 39-29

Step 3.1. Right-click the resources node to choose New, Others to display the New File dialog box, as shown in Figure 39.20.

Step 3.2. Choose Others in the Categories section and Cascading Style Sheet in the File Types section to display the New Cascading Style Sheet dialog box, as shown in Figure 39.21.

Step 3.3. Enter tablestyle as the File Name and click Finish to create tablestyle.css under the resources node.

Step 3.4. Define the CSS style as shown in Listing 39.15.

Figure 39.19 The table displays the students enrolled in the course.

Figure 39.20 You can create CSS files for Web project in NetBenas.

M39_LIAN1878_11_GE_C39.indd 29 12/29/17 6:01 PM

39-30 Chapter 39 JavaServer Faces

Figure 39.21 The New Cascading Style Sheet dialog box creates a new style sheet file.

Listing 39.13 CourseNameJSFBean.java 1 package jsf2demo; 2 3 import java.sql.*; 4 import java.util.ArrayList; 5 import javax.enterprise.context.ApplicationScoped; 6 import javax.inject.Named; 7 8 @Named(value = "courseName") 9 @ApplicationScoped 10 public class CourseNameJSFBean { 11 private PreparedStatement studentStatement = null; 12 private String choice; // Selected course 13 private String[] titles; // Course titles 14 15 /** Creates a new instance of CourseName */ 16 public CourseNameJSFBean() { 17 initializeJdbc(); 18 } 19 20 /** Initialize database connection */ 21 private void initializeJdbc() { 22 try { 23 Class.forName("com.mysql.jdbc.Driver"); 24 System.out.println("Driver loaded"); 25 26 // Connect to the sample database 27 Connection connection = DriverManager.getConnection( 28 "jdbc:mysql://localhost/javabook", "scott", "tiger"); 29 30 // Get course titles 31 PreparedStatement statement = connection.prepareStatement( 32 "select title from course"); 33 34 ResultSet resultSet = statement.executeQuery(); 35 36 // Store resultSet into array titles 37 ArrayList<String> list = new ArrayList<>();

application scope

initialize JDBC

connect to database

get course titles

execute SQL

M39_LIAN1878_11_GE_C39.indd 30 12/29/17 6:01 PM

39.8 Binding Database with Facelets 39-31

38 while (resultSet.next()) { 39 list.add(resultSet.getString(1)); 40 } 41 titles = new String[list.size()]; // Array for titles 42 list.toArray(titles); // Copy strings from list to array 43 44 // Define a SQL statement for getting students 45 studentStatement = connection.prepareStatement( 46 "select Student.ssn, " 47 + "student.firstName, Student.mi, Student.lastName, " 48 + "Student.phone, Student.birthDate, Student.street, " 49 + "Student.zipCode, Student.deptId " 50 + "from Student, Enrollment, Course " 51 + "where Course.title = ? " 52 + "and Student.ssn = Enrollment.ssn " 53 + "and Enrollment.courseId = Course.courseId;"); 54 } 55 catch (Exception ex) { 56 ex.printStackTrace(); 57 } 58 } 59 60 public String[] getTitles() { 61 return titles; 62 } 63 64 public String getChoice() { 65 return choice; 66 } 67 68 public void setChoice(String choice) { 69 this.choice = choice; 70 } 71 72 public ResultSet getStudents() throws SQLException { 73 if (choice == null) { 74 if (titles.length == 0) 75 return null; 76 else 77 studentStatement.setString(1, titles[0]); 78 } 79 else { 80 studentStatement.setString(1, choice); // Set course title 81 } 82 83 // Get students for the specified course 84 return studentStatement.executeQuery(); 85 } 86 }

We use the same MySQL database javabook created in Chapter 34, “Java Database Program- ming.” The scope for this managed bean is application. The bean is created when the project is launched from the server. The initializeJdbc method loads the JDBC driver for MySQL (lines 23 and 24), connects to the MySQL database (lines 27 and 28), creates statement for obtaining course titles (lines 31 and 32), and creates a statement for obtaining the student infor- mation for the specified course (lines 45–53). Lines 31–42 execute the statement for obtaining course titles and store them in array titles.

The getStudents() method returns a ResultSet that consists of all students enrolled in the specified course (lines 72–85). The choice for the title is set in the statement to obtain the

titles array

get students

set a default course

set a course

return students

M39_LIAN1878_11_GE_C39.indd 31 12/29/17 6:01 PM

39-32 Chapter 39 JavaServer Faces

student for the specified title (line 80). If choice is null, the first title in the titles array is set in the statement (line 77). If no titles in the course, getStudents() returns null (line 75).

TIP In order to use the MySQL database from this project, you have to add the MySQL JDBC driver from the Libraries node in the Project pane in NetBeans.

Listing 39.14 DisplayStudent.xhtml 1 <?xml version='1.0' encoding='UTF-8' ?> 2 <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" 3 "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> 4 <html xmlns="http://www.w3.org/1999/xhtml" 5 xmlns:h="http://xmlns.jcp.org/jsf/html" 6 xmlns:f="http://xmlns.jcp.org/jsf/core"> 7 <h:head> 8 <title>Display Student</title> 9 <h:outputStylesheet name="tablestyle.css"/> 10 </h:head> 11 <h:body> 12 <h:form> 13 <h:outputLabel value="Choose a Course: " /> 14 <h:selectOneMenu value="#{courseName.choice}"> 15 <f:selectItems value="#{courseName.titles}" /> 16 </h:selectOneMenu> 17 18 <h:commandButton style="margin-left: 20px" 19 value="Display Students" /> 20 21 <br /> <br /> 22 <h:dataTable value="#{courseName.students}" var="student" 23 rowClasses="oddTableRow, evenTableRow" 24 headerClass="tableHeader" 25 styleClass="table"> 26 <h:column> 27 <f:facet name="header">SSN</f:facet> 28 #{student.ssn} 29 </h:column> 30 31 <h:column> 32 <f:facet name="header">First Name</f:facet> 33 #{student.firstName} 34 </h:column> 35 36 <h:column> 37 <f:facet name="header">MI</f:facet> 38 #{student.mi} 39 </h:column> 40 41 <h:column> 42 <f:facet name="header">Last Name</f:facet> 43 #{student.lastName} 44 </h:column> 45 46 <h:column> 47 <f:facet name="header">Phone</f:facet> 48 #{student.phone} 49 </h:column> 50 51 <h:column>

add MySQL in the Libraries node

style sheet

bind choice titles

display button

bind result set rowClasses headerClass styleClass

ssn column

firstName column

mi column

lastName column

phone column

M39_LIAN1878_11_GE_C39.indd 32 12/29/17 6:01 PM

39.8 Binding Database with Facelets 39-33

52 <f:facet name="header">Birth Date</f:facet> 53 #{student.birthDate} 54 </h:column> 55 56 <h:column> 57 <f:facet name="header">Dept</f:facet> 58 #{student.deptId} 59 </h:column> 60 </h:dataTable> 61 </h:form> 62 </h:body> 63 </html>

Line 9 specifies that the style sheet tablestyle.css created in Step 3 is used in this XMTHL file. The rowClasses = "oddTableRow, evenTableRow" attribute specifies the style applied to the rows alternately using oddTableRow and evenTableRow (line 23). The header- Classes = "tableHeader" attribute specifies that the tableHeader class is used for header style (line 24). The styleClasses = "table" attribute specifies that the table class is used for the style of all other elements in the table (line 25).

Line 14 binds the choice property in the courseName bean with the combo box. The selection values in the combo box are bound with the titles array property (line 15).

Line 22 binds the table value with a database result set using the attribute value = "#{courseName.students}". The var="student" attribute associates a row in the result set with student. Lines 26–59 specify the column values using student.ssn (line 28), student.firstName (line 33), student.mi (line 38), student.lastName (line 33), student.phone (line 48), student.birthDate (line 53), and student.deptId (line 58).

Listing 39.15 tablestyle.css 1 /* Style for table */ 2 .tableHeader { 3 font-family:"Trebuchet MS", Arial, Helvetica, sans-serif; 4 border-collapse:collapse; 5 font-size:1.1em; 6 text-align:left; 7 padding-top:5px; 8 padding-bottom:4px; 9 background-color:#A7C942; 10 color:white; 11 border:1px solid #98bf21; 12 } 13 14 .oddTableRow { 15 border:1px solid #98bf21; 16 } 17 18 .evenTableRow { 19 background-color: #eeeeee; 20 font-size:1em; 21 22 padding:3px 7px 2px 7px; 23 24 color:#000000; 25 background-color:#EAF2D3; 26 } 27 28 .table { 29 border:1px solid green; 30 }

birthDate column

deptId column

tableHeader

oddTableRow

evenTableRow

table

M39_LIAN1878_11_GE_C39.indd 33 12/29/17 6:01 PM

39-34 Chapter 39 JavaServer Faces

The style sheet file defines the style classes tableHeader (line 2) for table header style, oddTableRow for odd table rows (line 14), evenTableRow for even table rows (line 18), and table for all other table elements (line 28).

39.9 Opening New JSF Pages You can open new JSF pages from the current JSF pages.

All the examples you have seen so far use only one JSF page in a project. Suppose you want to register student information to the database. The application first displays the page as shown in Figure 39.22 to collect student information. After the user enters the information and clicks the Submit button, a new page is displayed to ask the user to confirm the input, as shown in Figure 39.23. If the user clicks the Confirm button, the data are stored into the database and the status page is displayed, as shown in Figure 39.24. If the user clicks the Go Back button, it goes back to the first page.

Point Key

Figure 39.22 This page lets the user enter input.

For this project, you need to create three JSF pages named AddressRegistration.xhtml, ConfirmAddress.xhtml, and AddressStoredStatus.xhtml in Listings 39.16–39.18. The project starts with AddressRegistration.xhtml. When clicking the Submit button, the action for the button returns “ConfirmAddress” if the last name and first name are not empty, which causes ConfirmAddress.xhtml to be displayed. When clicking the Confirm button, the status page AddressStoredStatus.xhtml is displayed. When clicking the Go Back button, the first page AddressRegistration.xhtml is now displayed.

Listing 39.16 AddressRegistration.xhtml 1 <?xml version='1.0' encoding='UTF-8' ?> 2 <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" 3 "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> 4 <html xmlns="http://www.w3.org/1999/xhtml" 5 xmlns:h="http://xmlns.jcp.org/jsf/html" 6 xmlns:f="http://xmlns.jcp.org/jsf/core"> 7 <h:head> 8 <title>Student Registration Form</title>

jsf core namespace

M39_LIAN1878_11_GE_C39.indd 34 12/29/17 6:01 PM

39.9 Opening New JSF Pages 39-35

9 </h:head> 10 <h:body> 11 <h:form> 12  13 <h3>Student Registration Form 14 <h:graphicImage name="usIcon.gif" library="image"/> 15 </h3> 16 17 Please register to your instructor's student address book. 18  19 <h:panelGrid columns="6"> 20 <h:outputLabel value="Last Name" style="color:red"/> 21 <h:inputText id="lastNameInputText" 22 value="#{addressRegistration.lastName}"/> 23 <h:outputLabel value="First Name" style="color:red"/> 24 <h:inputText id="firstNameInputText" 25 value="#{addressRegistration.firstName}"/> 26 <h:outputLabel value="MI" /> 27 <h:inputText id="miInputText" size="1" 28 value="#{addressRegistration.mi}"/> 29 </h:panelGrid> 30

Figure 39.23 This page lets the user confirm the input.

Figure 39.24 This page displays the status of the user input.

bind lastName

bind firstName

bind mi

M39_LIAN1878_11_GE_C39.indd 35 12/29/17 6:01 PM

39-36 Chapter 39 JavaServer Faces

31 <h:panelGrid columns="4"> 32 <h:outputLabel value="Telephone"/> 33 <h:inputText id="telephoneInputText" 34 value="#{addressRegistration.telephone}"/> 35 <h:outputLabel value="Email"/> 36 <h:inputText id="emailInputText" 37 value="#{addressRegistration.email}"/> 38 </h:panelGrid> 39 40 <h:panelGrid columns="4"> 41 <h:outputLabel value="Street"/> 42 <h:inputText id="streetInputText" 43 value="#{addressRegistration.street}"/> 44 </h:panelGrid> 45 46 <h:panelGrid columns="6"> 47 <h:outputLabel value="City"/> 48 <h:inputText id="cityInputText" 49 value="#{addressRegistration.city}"/> 50 <h:outputLabel value="State"/> 51 <h:selectOneMenu id="stateSelectOneMenu" 52 value="#{addressRegistration.state}"> 53 <f:selectItem itemLabel="Georgia-GA" itemValue="GA" /> 54 <f:selectItem itemLabel="Oklahoma-OK" itemValue="OK" /> 55 <f:selectItem itemLabel="Indiana-IN" itemValue="IN"/> 56 </h:selectOneMenu> 57 <h:outputLabel value="Zip"/> 58 <h:inputText id="zipInputText" 59 value="#{addressRegistration.zip}"/> 60 </h:panelGrid> 61 62  63 <h:commandButton value="Register" 64 action="#{addressRegistration.processSubmit()}"/> 65 <br /> 66 <h:outputText escape="false" style="color:red" 67 value="#{addressRegistration.requiredFields}" /> 68 </h:form> 69 </h:body> 70 </html>

Listing 39.17 ConfirmAddress.xhtml 1 <?xml version='1.0' encoding='UTF-8' ?> 2 <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" 3 "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> 4 <html xmlns="http://www.w3.org/1999/xhtml" 5 xmlns:h="http://xmlns.jcp.org/jsf/html"> 6 <h:head> 7 <title>Confirm Student Registration</title> 8 </h:head> 9 <h:body> 10 <h:form> 11 <h:outputText escape="false" style="color:red" 12 value="#{registration1.input}" /> 13 <h:panelGrid columns="2"> 14 <h:commandButton value="Confirm" 15 action = "#{registration1.storeStudent()}"/> 16 <h:commandButton value="Go Back" 17 action = "AddressRegistration"/>

bind telephone

bind email

bind street

bind city

bind state

bind zip

process register

process confirm go to AddressRegistra- tion page

M39_LIAN1878_11_GE_C39.indd 36 12/29/17 6:01 PM

39.9 Opening New JSF Pages 39-37

18 </h:panelGrid> 19 </h:form> 20 </h:body> 21 </html>

Listing 39.18 AddressStoredStatus.xhtml 1 <?xml version='1.0' encoding='UTF-8' ?> 2 <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" 3 "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> 4 <html xmlns="http://www.w3.org/1999/xhtml" 5 xmlns:h="http://xmlns.jcp.org/jsf/html"> 6 <h:head> 7 <title>Address Stored?</title> 8 </h:head> 9 <h:body> 10 <h:form> 11 <h:outputText escape="false" style="color:green" 12 value="#{registration1.status}" /> 13 </h:form> 14 </h:body> 15 </html>

Listing 39.19 AddressRegistrationJSFBean.java 1 package jsf2demo; 2 3 import javax.inject.Named; 4 import javax.enterprise.context.SessionScoped; 5 import java.sql.*; 6 import java.io.Serializable; 7 8 @Named(value = "addressRegistration") 9 @SessionScoped 10 public class AddressRegistrationJSFBean implements Serializable { 11 private String lastName; 12 private String firstName; 13 private String mi; 14 private String telephone; 15 private String email; 16 private String street; 17 private String city; 18 private String state; 19 private String zip; 20 private String status = "Nothing stored"; 21 // Use a prepared statement to store a student into the database 22 private PreparedStatement pstmt; 23 24 public AddressRegistrationJSFBean() { 25 initializeJdbc(); 26 } 27 28 public String getLastName() { 29 return lastName; 30 } 31 32 public void setLastName(String lastName) { 33 this.lastName = lastName; 34 } 35

display status

managed bean session scope

property lastName

initialize database

M39_LIAN1878_11_GE_C39.indd 37 12/29/17 6:01 PM

39-38 Chapter 39 JavaServer Faces

36 public String getFirstName() { 37 return firstName; 38 } 39 40 public void setFirstName(String firstName) { 41 this.firstName = firstName; 42 } 43 44 public String getMi() { 45 return mi; 46 } 47 48 public void setMi(String mi) { 49 this.mi = mi; 50 } 51 52 public String getTelephone() { 53 return telephone; 54 } 55 56 public void setTelephone(String telephone) { 57 this.telephone = telephone; 58 } 59 60 public String getEmail() { 61 return email; 62 } 63 64 public void setEmail(String email) { 65 this.email = email; 66 } 67 68 public String getStreet() { 69 return street; 70 } 71 72 public void setStreet(String street) { 73 this.street = street; 74 } 75 76 public String getCity() { 77 return city; 78 } 79 80 public void setCity(String city) { 81 this.city = city; 82 } 83 84 public String getState() { 85 return state; 86 } 87 88 public void setState(String state) { 89 this.state = state; 90 } 91 92 public String getZip() { 93 return zip; 94 } 95 96 public void setZip(String zip) {

M39_LIAN1878_11_GE_C39.indd 38 12/29/17 6:01 PM

39.9 Opening New JSF Pages 39-39

97 this.zip = zip; 98 } 99 100 private boolean isRquiredFieldsFilled() { 101 return !(lastName == null || firstName == null 102 || lastName.trim().length() == 0 103 || firstName.trim().length() == 0); 104 } 105 106 public String processSubmit() { 107 if (isRquiredFieldsFilled()) 108 return "ConfirmAddress"; 109 else 110 return ""; 111 } 112 113 public String getRequiredFields() { 114 if (isRquiredFieldsFilled()) 115 return ""; 116 else 117 return "Last Name and First Name are required"; 118 } 119 120 public String getInput() { 121 return "<p style=\"color:red\">You entered <br />" 122 + "Last Name: " + lastName + "<br />" 123 + "First Name: " + firstName + "<br />" 124 + "MI: " + mi + "<br />" 125 + "Telephone: " + telephone + "<br />" 126 + "Email: " + email + "<br />" 127 + "Street: " + street + "<br />" 128 + "City: " + city + "<br />" 129 + "Street: " + street + "<br />" 130 + "City: " + city + "<br />" 131 + "State: " + state + "<br />" 132 + "Zip: " + zip + "</p>"; 133 } 134 135 /** Initialize database connection */ 136 private void initializeJdbc() { 137 try { 138 // Explicitly load a MySQL driver 139 Class.forName("com.mysql.jdbc.Driver"); 140 System.out.println("Driver loaded"); 141 142 // Establish a connection 143 Connection conn = DriverManager.getConnection( 144 "jdbc:mysql://localhost/javabook", "scott", "tiger"); 145 146 // Create a Statement 147 pstmt = conn.prepareStatement("insert into Address (lastName," 148 + " firstName, mi, telephone, email, street, city, " 149 + "state, zip) values (?, ?, ?, ?, ?, ?, ?, ?, ?)"); 150 } 151 catch (Exception ex) { 152 System.out.println(ex); 153 } 154 } 155 156 /** Store an address to the database */ 157 public String storeStudent() {

go to a new page

check required fields

get input

store address

M39_LIAN1878_11_GE_C39.indd 39 12/29/17 6:01 PM

39-40 Chapter 39 JavaServer Faces

158 try { 159 pstmt.setString(1, lastName); 160 pstmt.setString(2, firstName); 161 pstmt.setString(3, mi); 162 pstmt.setString(4, telephone); 163 pstmt.setString(5, email); 164 pstmt.setString(6, street); 165 pstmt.setString(7, city); 166 pstmt.setString(8, state); 167 pstmt.setString(9, zip); 168 pstmt.executeUpdate(); 169 status = firstName + " " + lastName 170 + " is now registered in the database."; 171 } 172 catch (Exception ex) { 173 status = ex.getMessage(); 174 } 175 176 return "AddressStoredStatus"; 177 } 178 179 public String getStatus() { 180 return status; 181 } 182 }

A session-scoped managed bean must implement the java.io.Serializable interface. So, the AddressRegistration class is defined as a subtype of java.io.Serializable.

The action for the Register button in the AddressRegistration JSF page is process- Submit() (line 64 in AddressRegistration.xhtml). This method checks if last name and first name are not empty (lines 106–111 in AddressRegistrationJSFBean.java). If so, it returns a string "ConfirmAddress", which causes the ConfirmAddress JSF page to be displayed.

The ConfirmAddress JSF page displays the data entered from the user (line 12 in Confir- mAddress.xhtml). The getInput() method (lines 120–133 in AddressRegistrationJSFBean. java) collects the input.

The action for the Confirm button in the ConfirmAddress JSF page is storeStudent() (line 15 in ConfirmAddress.xhtml). This method stores the address in the database (lines 157–177 in AddressRegistrationJSFBean.java) and returns a string "AddressStoredSta- tus", which causes the AddressStoredStatus page to be displayed. The status message is displayed in this page (line 12 in AddressStoredStatus.xhtml).

The action for the Go Back button in the ConfirmAddress page is "AddressRegistra- tion" (line 17 in ConfirmAddress.xhtml). This causes the AddressRegistration page to be displayed for the user to reenter the input.

The scope of the managed bean is session (line 9 AddressRegistrationJSFBean.java) so the multiple pages can share the same bean.

Note this program loads the database driver explicitly (line 139 AddressRegistrationJSF- Bean.java). Sometimes, an IDE such as NetBeans is not able to find a suitable driver. Loading a driver explicitly can avoid this problem.

39.10 Contexts and Dependency Injection Contexts and dependency injection enables beans to be shared in multiple applications.

Contexts and dependency injection, short for CDI, allows multiple programs to share a bean. To illustrate the need for this, consider two simple webpages and a server object named track. One page contains a button and a message that displays the number of times the button is clicked from the current IP address, as shown in Figure 39.25. When the button is clicked

Point Key

update status

go to a new page

M39_LIAN1878_11_GE_C39.indd 40 12/29/17 6:01 PM

39.10 Contexts and Dependency Injection 39-41

for the first time, the user’s IP address along with count value 1 is stored in a map with the IP address as the key. When the button is clicked again, the count value for the IP address is increased in the map. The other page simply displays the total count from each IP address, as shown in Figure 39.26. The Track class is defined as shown in Listing 39.20.

Figure 39.25 The count is updated when the Click Me button is clicked.

Figure 39.26 The count for each client IP Address is displayed.

Listing 39.20 Track.java 1 package jsf2demo; 2 3 import java.util.HashMap; 4 import java.util.Map; 5 import javax.enterprise.context.ApplicationScoped; 6 7 @ApplicationScoped 8 public class Track { 9 private Map<String, Integer> map = new HashMap<>(); 10 11 public void add(String ipAddress) { 12 map.put(ipAddress, map.containsKey(ipAddress) ? 13 map.get(ipAddress) + 1 : 1); 14 } 15 16 public int getCount(String ipAddress) { 17 return map.containsKey(ipAddress) ? map.get(ipAddress) : 0; 18 } 19 20 public String getAllCount() { 21 return "Count summary is " + map; 22 } 23 }

A Track object uses a map to store an IP address and its count with IP address as a key (line 9). The add method (lines 11–14) adds an IP Address to the map. If the IP address is not in the map, a new entry is created for the IP Address with value 1. Otherwise, the value for the IP address is incremented by 1 in the map. The getCount method (lines 16–18) returns the count for an IP address. If the IP address is not in the map, the method returns 0. The

application scope

store counts

add or update count

return count

return all counts

M39_LIAN1878_11_GE_C39.indd 41 12/29/17 6:01 PM

39-42 Chapter 39 JavaServer Faces

getAllCount method (lines 20–22) simply returns a string that describes the counts for all IP address in the map.

We now create a page named IncreaseCount.xhtml (Listing 39.21) with a button for displaying the number of times a button is clicked on the client, and create a page named Dis- playCount.xhtml (Listing 39.22) for displaying the counts from all clients.

Listing 39.21 IncreaseCount.xhtml 1 <?xml version='1.0' encoding='UTF-8' ?> 2 <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" 3 "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> 4 <html xmlns="http://www.w3.org/1999/xhtml" 5 xmlns:h="http://xmlns.jcp.org/jsf/html"> 6 <h:head> 7 <title>IncreaseCount</title> 8 </h:head> 9 <h:body> 10 <h:form> 11 <h:commandButton 12 action="#{increaseCount.click()}" value="Click Me"/> 13 <br>The current count is #{increaseCount.getCount()} and your 14 IP address is #{increaseCount.getIpAddress()}.</br> 15 </h:form> 16 </h:body> 17 </html>

Listing 39.22 DisplayCount.xhtml 1 <?xml version='1.0' encoding='UTF-8' ?> 2 <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" 3 "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> 4 <html xmlns="http://www.w3.org/1999/xhtml" 5 xmlns:h="http://xmlns.jcp.org/jsf/html"> 6 <h:head> 7 <title>DisplayCount</title> 8 </h:head> 9 <h:body> 10 #{displayCount.getAllCount()}. 11 </h:body> 12 </html>

The IncreaseCount page uses the increasCount bean to process the click action (line 12), obtain the click count (line 13), and the client’s IP address (line 14). The DisplayCount page uses the displayCount bean to obtain the count from all clients (line 10). Both increas- Count and displayCount need to access the same Track object. How can you create a Track object to be used by different objects? JSF supports context dependency injection (CDI) for injecting an object into a class using the @Inject annotation. Listing 39.23 gives the implementation for IncreaseCount.java and Listing 39.24 for DisplayCount.java.

Listing 39.23 IncreaseCount.java 1 package jsf2demo; 2 3 import javax.enterprise.context.SessionScoped; 4 import javax.inject.Named; 5 import javax.faces.context.FacesContext; 6 import javax.inject.Inject; 7 import javax.servlet.http.HttpServletRequest; 8 9 @Named(value = "increaseCount")

process a click obtain count obtain IP address

obtain all counts

M39_LIAN1878_11_GE_C39.indd 42 12/29/17 6:01 PM

39.10 Contexts and Dependency Injection 39-43

10 @SessionScoped 11 public class IncreaseCount implements java.io.Serializable { 12 @Inject private Track track; 13 private String ipAddress; 14 15 public IncreaseCount() { 16 HttpServletRequest request = (HttpServletRequest)FacesContext 17 .getCurrentInstance().getExternalContext().getRequest(); 18 this.ipAddress = request.getRemoteAddr(); 19 } 20 21 public void click() { 22 track.add(ipAddress); 23 } 24 25 public String getIpAddress() { 26 return ipAddress; 27 } 28 29 public int getCount() { 30 return track.getCount(ipAddress); 31 } 32 }

Listing 39.24 DisplayCount.java 1 package jsf2demo; 2 3 import javax.enterprise.context.ApplicationScoped; 4 import javax.inject.Named; 5 import javax.inject.Inject; 6 7 @Named(value = "displayCount") 8 @ApplicationScoped 9 public class DisplayCount { 10 @Inject private Track track; 11 12 public String getAllCount() { 13 return track.getAllCount(); 14 } 15 }

The @Inject annotation in line 12 of IncreaseCount.java and line 10 of DisplayCount.java injects a Track object. This Track object is created by the Java server container. The track data fields in both classes refer to this object.

In IncreaseCount.java, the constructor obtains the IP address of a client (lines 16 and 17) and sets it in the data field ipAddress (line 18). The click method adds the ipAddress to the map in the track object (line 22).

Note the scope for Track and DisplayCount is ApplicationScoped since these two objects are created once for the entire application. However, the scope for IncreaseCount is SessionScoped since each session has its own IP Address.

Key Terms

application scope 39-21 contexts and dependency injection

(CDI) 39-40 JavaBean 39-5

request scope 39-21 scope 39-21 session scope 39-21 view scope 39-21

session scope

inject track

increase count obtain client’s IP

add an IP address

count for an IP

application scope

inject track

obtain all counts

M39_LIAN1878_11_GE_C39.indd 43 12/29/17 6:01 PM

39-44 Chapter 39 JavaServer Faces

ChapTer summary 1. JSF enables you to completely separate Java code from HTML.

2. A facelet is an XHTML page that mixes JSF tags with XHTML tags.

3. JSF applications are developed using the Model-View-Controller (MVC) architecture, which separates the application’s data (contained in the model) from the graphical pres- entation (the view).

4. The controller is the JSF framework that is responsible for coordinating interactions between view and the model.

5. In JSF, the facelets are the view for presenting data. Data are obtained from Java objects. Objects are defined using Java classes.

6. In JSF, the objects that are accessed from a facelet are JavaBeans objects.

7. The JSF expression can either use the property name or invoke the method to obtain the current time.

8. JSF provides many elements for displaying GUI components. The tags with the h prefix are in the JSF HTML Tag library. The tags with the f prefix are in the JSF Core Tag library.

9. You can specify the JavaBeans objects at the application scope, session scope, view scope, or request scope.

10. The view scope keeps the bean alive as long as you stay on the view. The view scope is between session and request scopes.

11. JSF provides several convenient and powerful ways for input validation. You can use the standard validator tags in the JSF Core Tag Library or create custom validators.

Quiz Answer the quiz for this chapter online at the book Companion Website.

programming exerCises *39.1 (Factorial table in JSF) Write a JSF page that displays a factorial page as shown in

Figure 39.27. Display the table in an h:outputText component. Set its escape property to false to display it as HTML contents.

*39.2 (Multiplication table) Write a JSF page that displays a multiplication table as shown in Figure 39.28.

*39.3 (Calculate tax) Write a JSF page to let the user enter taxable income and filing status, as shown in Figure 39.29a. Clicking the Compute Tax button computes and displays the tax, as shown in Figure 39.29b. Use the computeTax method intro- duced in Listing 3.5, ComputeTax.java, to compute tax.

*39.4 (Calculate loan) Write a JSF page that lets the user enter loan amount, interest rate, and number of years, as shown in Figure 39.30a. Click the Compute Loan Payment button to compute and display the monthly and total loan payments, as shown in Figure 39.30b. Use the Loan class given in Listing 10.2, Loan.java, to compute the monthly and total payments.

M39_LIAN1878_11_GE_C39.indd 44 12/29/17 6:01 PM

Programming Exercises 39-45

Figure 39.27 The JSF page displays factorials for the numbers from 0 to 10 in a table.

Figure 39.28 The JSF page displays the multiplication table.

*39.5 (Addition quiz) Write a JSF program that generates addition quizzes randomly, as shown in Figure 39.31a. After the user answers all questions, it displays the result, as shown in Figure 39.31b.

*39.6 (Large factorial) Rewrite Exercise 39.1 to handle large factorial as shown in Figure 39.32. Use the BigInteger class introduced in Section 10.9.

*39.7 (Guess birthday) Listing 4.3, GuessBirthday.java, gives a program for guessing a birthday. Write a JSF program that displays five sets of numbers, as shown in Figure 39.33a. After the user checks the appropriate boxes and clicks the Guess Birthday button, the program displays the birthday, as shown in Figure 39.33b.

M39_LIAN1878_11_GE_C39.indd 45 12/29/17 6:01 PM

39-46 Chapter 39 JavaServer Faces

Figure 39.29 The JSF page computes the tax.

(a)

(b)

Figure 39.30 The JSF page computes the loan payment.

(a)

(b)

M39_LIAN1878_11_GE_C39.indd 46 12/29/17 6:01 PM

Programming Exercises 39-47

Figure 39.31 The program displays addition questions in (a) and answers in (b).

(a)

(b)

*39.8 (Guess capitals) Write a JSF that prompts the user to enter a capital for a state, as shown in Figure 39.34a. Upon receiving the user input, the program reports whether the answer is correct, as shown in Figure 39.34b. You can click the Next button to display another question. You can use a two-dimensional array to store the states and capitals, as proposed in Exercise 8.37. Create a list from the array and apply the shuffle method to reorder the list so the questions will appear in random order.

*39.9 (Access and update a Staff table) Write a JSF program that views, inserts, and updates staff information stored in a database, as shown in Figure 39.35. The view button displays a record with a specified ID. The Staff table is created as follows:

create table Staff ( id char(9) not null, lastName varchar(15), firstName varchar(15),

M39_LIAN1878_11_GE_C39.indd 47 12/29/17 6:01 PM

39-48 Chapter 39 JavaServer Faces

Figure 39.32 The JSF page displays factorials for the numbers from 10 to 20 in a table.

Figure 39.33 (a) The program displays five sets of numbers for the user to check the boxes. (b) The program displays the date.

(a)

(b)

M39_LIAN1878_11_GE_C39.indd 48 12/29/17 6:01 PM

Programming Exercises 39-49

Figure 39.34 (a) The program displays a question. (b) The program displays the answer to the question.

(a)

(b)

Figure 39.35 The webpage lets you view, insert, and update staff information.

M39_LIAN1878_11_GE_C39.indd 49 12/29/17 6:01 PM

39-50 Chapter 39 JavaServer Faces

Figure 39.36 This JSF application displays four random cards.

mi char(1), address varchar(20), city varchar(20), state char(2), telephone char(10), email varchar(40), primary key (id) );

*39.10 (Random cards) Write a JSF that displays four random cards from a deck of 52 cards, as shown in Figure 39.36. When the user clicks the Refresh button, four new random cards are displayed.

***39.11 (Game: the 24-point card game) Rewrite Exercise 20.13 using JSF, as shown in Figure 39.37. Upon clicking the Refresh button, the program displays four ran- dom cards and displays an expression if a 24-point solution exists. Otherwise, it displays “No solution”.

Figure 39.37 The JSF application solves a 24-point card game.

M39_LIAN1878_11_GE_C39.indd 50 12/29/17 6:01 PM

Programming Exercises 39-51

***39.12 (Game: the 24-point card game) Rewrite Exercise 20.17 using JSF, as shown in Figure 39.38. The program lets the user enter four card values and finds a solu- tion upon clicking the Find a Solution button.

*39.13 (Day of week) Write a program that displays the day of the week for a given day, month, and year, as shown in Figure 39.39. The program lets the user select a day, month, and year, and click the Get Day of Week button to display the day of week. The Time field displays “Future” if it is a future day or “Past” otherwise. Use the Zeller’s congruence to find the day of the week (see Programming Exercise 3.21).

Figure 39.38 The user enters four numbers and the program finds a solution.

Figure 39.39 The user enters a day, month, and year and the program displays the day of the week.

M39_LIAN1878_11_GE_C39.indd 51 12/29/17 6:01 PM

39-52 Chapter 39 JavaServer Faces

*39.14 (Display total count) Revise Listing 39.22, DisplayCount.xhtml to display the total count of the button clicks form all clients and display the client’s IP address and counts in increasing order of the counts, as shown in Figure 39.40.

Figure 39.40 The total counts and individual client counts are displayed.

M39_LIAN1878_11_GE_C39.indd 52 12/29/17 6:01 PM

Objectives ■■ To explain how RMI works (§40.2).

■■ To describe the process of developing RMI applications (§40.3).

■■ To distinguish between RMI and socket-level programming (§40.4).

■■ To develop three-tier applications using RMI (§40.5).

■■ To use callbacks to develop interactive applications (§40.6).

Remote Method Invocation

Chapter

M40_LIAN1878_11_GE_C40.indd 1 12/29/17 6:02 PM

40-2 Chapter 40 Remote Method Invocation

40.1 Introduction Remote Method Invocation is a high-level Java API for Java network programming.

Remote Method Invocation (RMI) provides a framework for building distributed Java systems. Using RMI, a Java object on one system can invoke a method in an object on another system on the network. A distributed Java system can be defined as a collection of cooperative dis- tributed objects on the network. In this chapter, you will learn how to use RMI to create useful distributed applications.

40.2 RMI Basics RMI enables you to access a remote object and invoke its methods.

RMI is the Java Distributed Object Model for facilitating communications among distributed objects. RMI is a high-level API built on top of sockets. Socket-level programming allows you to pass data through sockets among computers. RMI enables you also to invoke methods in a remote object. Remote objects can be manipulated as if they were residing on the local host. The transmission of data among different machines is handled by the JVM transparently.

In many ways, RMI is an evolution of the client/server architecture. A client is a compo- nent that issues requests for services, and a server is a component that delivers the requested services. Like the client/server architecture, RMI maintains the notion of clients and servers, but the RMI approach is more flexible.

■■ An RMI component can act as both a client and a server, depending on the scenario in question.

■■ An RMI system can pass functionality from a server to a client, and vice versa. Typi- cally a client/server system only passes data back and forth between server and client.

40.2.1 How Does RMI Work? All the objects you have used before this chapter are called local objects. Local objects are accessible only within the local host. Objects that are accessible from a remote host are called remote objects. For an object to be invoked remotely, it must be defined in a Java interface accessible to both the server and the client. Furthermore, the interface must extend the java. rmi.Remote interface. Like the java.io.Serializable interface, java.rmi.Remote is a marker interface that contains no constants or methods. It is used only to identify remote objects.

The key components of the RMI architecture are listed below (see Figure 40.1):

■■ Server object interface: A subinterface of java.rmi.Remote that defines the methods for the server object.

■■ Server class: A class that implements the remote object interface.

■■ Server object: An instance of the server class.

■■ RMI registry: A utility that registers remote objects and provides naming services for locating objects.

■■ Client program: A program that invokes the methods in the remote server object.

■■ Server stub: An object that resides on the client host and serves as a surrogate for the remote server object.

■■ Server skeleton: An object that resides on the server host and communicates with the stub and the actual server object.

Point Key

M40_LIAN1878_11_GE_C40.indd 2 12/29/17 6:02 PM

40.2 RMI Basics 40-3

RMI works as follows:

1. A server object is registered with the RMI registry.

2. A client looks through the RMI registry for the remote object.

3. Once the remote object is located, its stub is returned in the client.

4. The remote object can be used in the same way as a local object. Communication between the client and the server is handled through the stub and the skeleton.

The implementation of the RMI architecture is complex, but the good news is that RMI pro- vides a mechanism that liberates you from writing the tedious code for handling parameter passing and invoking remote methods. The basic idea is to use two helper classes known as the stub and the skeleton for handling communications between client and server.

The stub and the skeleton are automatically generated. The stub resides on the client machine. It contains all the reference information the client needs to know about the server object. When a client invokes a method on a server object, it actually invokes a method that is encapsulated in the stub. The stub is responsible for sending parameters to the server and for receiving the result from the server and returning it to the client.

The skeleton communicates with the stub on the server side. The skeleton receives param- eters from the client, passes them to the server for execution, and returns the result to the stub.

40.2.2 Passing Parameters When a client invokes a remote method with parameters, passing the parameters is handled by the stub and the skeleton. Obviously, invoking methods in a remote object on a server is very different from invoking methods in a local object on a client, since the remote object is in a different address space on a separate machine. Let us consider three types of parameters:

■■ Primitive data types, such as char, int, double, or boolean, are passed by value like a local call.

■■ Local object types, such as java.lang.String, are also passed by value, but this is completely different from passing an object parameter in a local call. In a local call, an object parameter’s reference is passed, which corresponds to the memory address of the object. In a remote call, there is no way to pass the object reference, because the address on one machine is meaningless to a different JVM. Any object can be used as

Figure 40.1 Java RMI uses a registry to provide naming services for remote objects, and uses the stub and the skeleton to facilitate communications between client and server.

Client Host

Server StubClient

Program

RMI Registry Host

RMI Registry

Server Host

Server Skeleton Server

Object

(1) Register Server Object

(2) Look for Server Object

(3) Return Server Stub

(4) Data Communication

Server Object Interface

M40_LIAN1878_11_GE_C40.indd 3 12/29/17 6:02 PM

40-4 Chapter 40 Remote Method Invocation

a parameter in a remote call as long as it is serializable. The stub serializes the object parameter and sends it in a stream across the network. The skeleton deserializes the stream into an object.

■■ Remote object types are passed differently from local objects. When a client invokes a remote method with a parameter of a remote object type, the stub of the remote object is passed. The server receives the stub and manipulates the parameter through it. Passing remote objects will be discussed in Section 40.6, RMI Callbacks.

40.2.3 RMI Registry How does a client locate the remote object? The RMI registry provides the registry services for the server to register the object and for the client to locate the object.

You can use several overloaded static getRegistry() methods in the LocateRegistry class to return a reference to a Registry, as shown in Figure 40.2. Once a Registry is obtained, you can bind an object with a unique name in the registry using the bind or rebind method, or locate an object using the lookup method, as shown in Figure 40.3.

java.rmi.registry.LocateRegistry

+getRegistry(): Registry

+getRegistry(port: int): Registry

+getRegistry(host: String): Registry

+getRegistry(host:String, port: int): Registry

Returns a reference to the remote object Registry for the local host on the default registry port of 1099.

Returns a reference to the remote object Registry for the local host on the specified port.

Returns a reference to the remote object Registry on the specified host on the default registry port of 1099.

Returns a reference to the remote object Registry on the specified host and port.

Figure 40.2 The LocateRegistry class provides the methods for obtaining a registry on a host.

java.rmi.registry.Registry

+bind(name: String, obj: Remote): void

+rebind(name: String, obj: Remote): void

+unbind(name: String): void

+list(name: String): String[]

+lookup(name: String): Remote

Binds the specified name with the remote object.

Binds the specified name with the remote object. Any existing binding for the name is replaced.

Destroys the binding for the specified name that is associated with a remote object.

Returns an array of the names bound in the registry.

Returns a reference, a stub, for the remote object associated with the specified name.

Figure 40.3 The Registry class provides the methods for binding and obtaining references to remote objects in a remote object registry.

40.3 Developing RMI Applications An RMI application consists of defining server object interface, defining a server object interface implementation class, creating and registering a server object, and developing a client program.

Now that you have a basic understanding of RMI, you are ready to write simple RMI applica- tions. The steps in developing an RMI application are shown in Figure 40.4 and listed below.

Point Key

M40_LIAN1878_11_GE_C40.indd 4 12/29/17 6:02 PM

40.3 Developing RMI Applications 40-5

1. Define a server object interface that serves as the contract between the server and its clients, as shown in the following outline:

public interface ServerInterface extends Remote { public void service1(...) throws RemoteException; // Other methods }

A server object interface must extend the java.rmi.Remote interface.

2. Define a class that implements the server object interface, as shown in the following outline:

public class ServerInterfaceImpl extends UnicastRemoteObject implements ServerInterface { public void service1(...) throws RemoteException { // Implement it } // Implement other methods }

The server implementation class must extend the java.rmi.server.UnicastRemote Object class. The UnicastRemoteObject class provides support for point-to-point active object references using TCP streams.

3. Create a server object from the server implementation class and register it with an RMI registry:

ServerInterface server = new ServerInterfaceImpl(...); Registry registry = LocateRegistry.getRegistry(); registry.rebind("RemoteObjectName", server);

4. Develop a client that locates a remote object and invokes its methods, as shown in the following outline:

Registry registry = LocateRegistry.getRegistry(host); ServerInterface server = (ServerInterfaceImpl) registry.lookup("RemoteObjectName"); server.service1(...);

The example that follows demonstrates the development of an RMI application through these steps.

40.3.1 Example: Retrieving Student Scores from an RMI Server This example creates a client that retrieves student scores from an RMI server. The client, shown in Figure 40.5, displays the score for the specified name.

1. Create a server interface named StudentServerInterface in Listing 40.1. The interface tells the client how to invoke the server’s findScore method to retrieve a student score.

Figure 40.4 The steps in developing an RMI application.

4 2

1 De�ne Server Object Interface

Develop Client Program

De�ne Server Implementation Class

3 Create and Register Server Object

M40_LIAN1878_11_GE_C40.indd 5 12/29/17 6:02 PM

40-6 Chapter 40 Remote Method Invocation

Listing 40.1 StudentServerInterface.java 1 import java.rmi.*; 2 3 public interface StudentServerInterface extends Remote { 4 /** 5 * Return the score for the specified name 6 * @param name the student name 7 * @return a double score or –1 if the student is not found 8 */ 9 public double findScore(String name) throws RemoteException; 10 }

Any object that can be used remotely must be defined in an interface that extends the java.rmi.Remote interface (line 3). StudentServerInterface, extending Remote, defines the findScore method that can be remotely invoked by a client to find a stu- dent’s score. Each method in this interface must declare that it may throw a java.rmi. RemoteException (line 9). Therefore, your client code that invokes this method must be prepared to catch this exception in a try-catch block.

2. Create a server implementation named StudentServerInterfaceImpl (Listing 40.2) that implements StudentServerInterface. The findScore method returns the score for a specified student. It returns -1 if the score is not found.

Listing 40.2 StudentServerInterfaceImpl.java 1 import java.rmi.*; 2 import java.rmi.server.*; 3 import java.util.*; 4 5 public class StudentServerInterfaceImpl 6 extends UnicastRemoteObject 7 implements StudentServerInterface { 8 // Stores scores in a map indexed by name 9 private HashMap<String, Double> scores = 10 new HashMap<String, Double>(); 11 12 public StudentServerInterfaceImpl() throws RemoteException { 13 initializeStudent(); 14 } 15 16 /** Initialize student information */ 17 protected void initializeStudent() { 18 scores.put("John", new Double(90.5)); 19 scores.put("Michael", new Double(100)); 20 scores.put("Michelle", new Double(98.5)); 21 } 22 23 /** Implement the findScore method from the 24 * Student interface */ 25 public double findScore(String name) throws RemoteException { 26 Double d = (Double)scores.get(name);

Figure 40.5 You can get the score by entering a student name and clicking the Get Score button.

M40_LIAN1878_11_GE_C40.indd 6 12/29/17 6:02 PM

40.3 Developing RMI Applications 40-7

27 28 if (d == null) { 29 System.out.println("Student " + name + " is not found "); 30 return –1; 31 } 32 else { 33 System.out.println("Student " + name + "\'s score is " 34 + d.doubleValue()); 35 return d.doubleValue(); 36 } 37 } 38 }

The StudentServerInterfaceImpl class implements StudentServerInterface. This class must also extend the java.rmi.server.RemoteServer class or its subclass. RemoteServer is an abstract class that defines the methods needed to create and export remote objects. Often its subclass java.rmi.server. UnicastRemoteObject is used (line 6). This subclass implements all the abstract methods defined in RemoteServer.

StudentServerInterfaceImpl implements the findScore method (lines 25–37) defined in StudentServerInterface. For simplicity, three students, John, Michael, and Michelle, and their corresponding scores are stored in an instance of java.util. HashMap named scores. HashMap is a concrete class of the Map interface in the Java Collections Framework, which makes it possible to search and retrieve a value using a key. Both values and keys are of Object type. The findScore method returns the score if the name is in the hash map, and returns -1 if the name is not found.

3. Create a server object from the server implementation and register it with the RMI server (see Listing 40.3).

Listing 40.3 RegisterWithRMIServer.java 1 import java.rmi.registry.*; 2 3 public class RegisterWithRMIServer { 4 /** Main method */ 5 public static void main(String[] args) { 6 try { 7 StudentServerInterface obj = 8 new StudentServerInterfaceImpl(); 9 Registry registry = LocateRegistry.getRegistry(); 10 registry.rebind("StudentServerInterfaceImpl", obj); 11 System.out.println("Student server " + obj + " registered"); 12 } 13 catch (Exception ex) { 14 ex.printStackTrace(); 15 } 16 } 17 }

RegisterWithRMIServer contains a main method, which is responsible for starting the server. It performs the following tasks: (1) create a server object (line 8); (2) obtain a reference to the RMI registry (line 9), and (3) register the object in the registry (line 10).

4. Create a client named StudentServerInterfaceClient in Listing 40.4. The client locates the server object from the RMI registry and uses it to find the scores.

Listing 40.4 StudentServerInterfaceClient.java 1 import javafx.application.Application; 2 import javafx.scene.Scene;

M40_LIAN1878_11_GE_C40.indd 7 12/29/17 6:02 PM

40-8 Chapter 40 Remote Method Invocation

3 import javafx.scene.control.Button; 4 import javafx.scene.control.Label; 5 import javafx.scene.control.TextField; 6 import javafx.scene.layout.GridPane; 7 import javafx.stage.Stage; 8 import java.rmi.registry.LocateRegistry; 9 import java.rmi.registry.Registry; 10 11 public class StudentServerInterfaceClient extends Application { 12 // Declare a Student instance 13 private StudentServerInterface student; 14 15 private Button btGetScore = new Button("Get Score"); 16 private TextField tfName = new TextField(); 17 private TextField tfScore = new TextField(); 18 19 public void start(Stage primaryStage) { 20 GridPane gridPane = new GridPane(); 21 gridPane.setHgap(5); 22 gridPane.add(new Label("Name"), 0, 0); 23 gridPane.add(new Label("Score"), 0, 1); 24 gridPane.add(tfName, 1, 0); 25 gridPane.add(tfScore, 1, 1); 26 gridPane.add(btGetScore, 1, 2); 27 28 // Create a scene and place the pane in the stage 29 Scene scene = new Scene(gridPane, 250, 250); 30 primaryStage.setTitle("StudentServerInterfaceClient"); 31 primaryStage.setScene(scene); // Place the scene in the stage 32 primaryStage.show(); // Display the stage 33 34 initializeRMI(); 35 btGetScore.setOnAction(e − > getScore()); 36 } 37 38 private void getScore() { 39 try { 40 // Get student score 41 double score = student.findScore(tfName.getText().trim()); 42 43 // Display the result 44 if (score < 0) 45 tfScore.setText("Not found"); 46 else 47 tfScore.setText(new Double(score).toString()); 48 } 49 catch(Exception ex) { 50 ex.printStackTrace(); 51 } 52 } 53 54 /** Initialize RMI */ 55 protected void initializeRMI() { 56 String host = ""; 57 58 try { 59 Registry registry = LocateRegistry.getRegistry(host); 60 student = (StudentServerInterface) 61 registry.lookup("StudentServerInterfaceImpl"); 62 System.out.println("Server object " + student + " found"); 63 } 64 catch(Exception ex) {

M40_LIAN1878_11_GE_C40.indd 8 12/29/17 6:02 PM

40.3 Developing RMI Applications 40-9

65 System.out.println(ex); 66 } 67 } 68 69 /** 70 * The main method is only needed for the IDE with limited 71 * JavaFX support. Not needed for running from the command line. 72 */ 73 public static void main(String[] args) { 74 launch(args); 75 } 76 }

StudentServerInterfaceClient invokes the findScore method on the server to find the score for a specified student. The key method in StudentServerInterface- Client is the initializeRMI method (lines 55–67), which is responsible for locating the server stub.

The lookup(String name) method (line 61) returns the remote object with the speci- fied name. Once a remote object is found, it can be used just like a local object. The stub and the skeleton are used behind the scenes to make the remote method invocation work.

5. Follow the steps below to run this example.

5.1. Start the RMI registry by typing “start rmiregistry” at a DOS prompt from the book directory. By default, the port number 1099 is used by rmiregistry. To use a different port number, simply type the command “start rmiregistry portnumber” at a DOS prompt.

5.2. Start the server RegisterWithRMIServer using the following command at C:\ book directory:

C:\ book>java RegisterWithRMIServer

5.3. Run the client StudentServerInterfaceClient as an application. A sample run of the application is shown in Figure 40.5(b).

Note: You must start rmiregistry from the directory where you will run the RMI server, as shown in Figure 40.6. Otherwise, you will receive the error ClassNotFoundException on StudentServerInterfaceImpl_Stub.

Figure 40.6 To run an RMI program, first start the rmiregistry, then register the server object with the registry. The client locates it from the registry.

RMI registry

Start RMI registry

Start RMI server

Run RMI client

M40_LIAN1878_11_GE_C40.indd 9 12/29/17 6:02 PM

40-10 Chapter 40 Remote Method Invocation

Note: Server, registry, and client can be on three different machines. If you run the client and the server on separate machines, you need to place StudentServerInterface on both machines.

Caution: If you modify the remote object implementation class, you need to restart the server class to reload the object to the RMI registry. In some old versions of rmiregistry, you may have to restart rmiregistry.

40.3.1 How do you define an interface for a remote object? 40.3.2 Describe the roles of the stub and the skeleton. 40.3.3 What is java.rmi.Remote? How do you define a server class? 40.3.4 What is an RMI registry for? How do you create an RMI registry? 40.3.5 What is the command to start an RMI registry? 40.3.6 How do you register a remote object with the RMI registry? 40.3.7 What is the command to start a custom RMI server? 40.3.8 How does a client locate a remote object stub through an RMI registry? 40.3.9 How do you obtain a registry? How do you register a remote object? How do you

locate remote object?

40.4 RMI vs. Socket-Level Programming RMI is a high-level network programming and socket-level network programming is low-low-level.

RMI enables you to program at a higher level of abstraction. It hides the details of socket server, socket, connection, and sending or receiving data. It even implements a multithreading server under the hood, whereas with socket-level programming, you have to explicitly imple- ment threads for handling multiple clients.

RMI applications are scalable and easy to maintain. You can change the RMI server or move it to another machine without modifying the client program except for resetting the URL to locate the server. (To avoid resetting the URL, you can modify the client to pass the URL as a command-line parameter.) In socket-level programming, a client operation to send data requires a server operation to read it. The implementation of client and server at the socket level is tightly synchronized.

RMI clients can directly invoke the server method, whereas socket-level programming is limited to passing values. Socket-level programming is very primitive. Avoid using it to develop client/server applications. As an analogy, socket-level programming is similar to programming in assembly language, whereas RMI programming is like programming in a high-level language.

40.10 What are the advantages of RMI over socket-level programming?

40.5 Developing Three-Tier Applications Using RMI RMI can be used in the middle between a client and a database to develop scalable and flexible business applications.

Three-tier applications have gained considerable attention in recent years, largely because of the demand for more scalable and load-balanced systems to replace traditional two-tier client/ server database systems. A centralized database system does not just handle data access, but it also processes the business rules on data. Thus, a centralized database is usually heavily

Point Check

Point Key

Point Check

Point Key

M40_LIAN1878_11_GE_C40.indd 10 12/29/17 6:02 PM

40.5 Developing Three-Tier Applications Using RMI 40-11

loaded, because it requires extensive data manipulation and processing. In some situations, data processing is handled by the client and business rules are stored on the client side. It is preferable to use a middle tier as a buffer between client and database. The middle tier can be used to apply business logic and rules, and to process data to reduce the load on the database.

A three-tier architecture does more than just reduce the processing load on the server. It also provides access to multiple network sites. This is especially useful to Java clients that need to access multiple databases on different servers, since the server may change.

To demonstrate, let us rewrite the example in Section 40.3.1, Example: Retrieving Student Scores from an RMI Server, to find scores stored in a database rather than a hash map. In addi- tion, the system is capable of blocking a client from accessing a student who has not given the university permission to publish his/her score. An RMI component is developed to serve as a middle tier between client and database; it sends a search request to the database, processes the result, and returns an appropriate value to the client.

For simplicity, this example reuses the StudentServerInterface interface and Stu- dentServerInterfaceClient class from Section 40.3.1 with no modifications. All you have to do is to provide a new implementation for the server interface and create a program to register the server with the RMI. Here are the steps to complete the program:

1. Store the scores in a database table named Score that contains three columns: name, score, and permission. The permission value is 1 or 0, which indicates whether the student has given the university permission to release his/her grade. The following is the statement to create the table and insert three records:

create table Scores (name varchar(20), score number, permission number);

insert into Scores values ('John', 90.5, 1); insert into Scores values ('Michael', 100, 1); insert into Scores values ('Michelle', 100, 0);

2. Create a new server implementation named Student3TierImpl in Listing 40.5. The server retrieves a record from the Scores table, processes the retrieved information, and sends the result back to the client.

Listing 40.5 Student3TierImpl.java 1 import java.rmi.*; 2 import java.rmi.server.*; 3 import java.sql.*; 4 5 public class Student3TierImpl extends UnicastRemoteObject 6 implements StudentServerInterface { 7 // Use prepared statement for querying DB 8 private PreparedStatement pstmt; 9 10 /** Constructs Student3TierImpl object and exports it on 11 * default port. 12 */ 13 public Student3TierImpl() throws RemoteException { 14 initializeDB(); 15 } 16 17 /** Constructs Student3TierImpl object and exports it on 18 * specified port. 19 * @param port The port for exporting 20 */ 21 public Student3TierImpl(int port) throws RemoteException { 22 super(port);

M40_LIAN1878_11_GE_C40.indd 11 12/29/17 6:02 PM

40-12 Chapter 40 Remote Method Invocation

23 initializeDB(); 24 } 25 26 /** Load JDBC driver, establish connection and 27 * create statement */ 28 protected void initializeDB() { 29 try { 30 // Load the JDBC driver 31 // Class.forName("oracle.jdbc.driver.OracleDriver"); 32 Class.forName("com.mysql.jdbc.Driver "); 33 34 System.out.println("Driver registered"); 35 36 // Establish connection 37 /*Connection conn = DriverManager.getConnection 38 ("jdbc:oracle:thin:@drake.armstrong.edu:1521:orcl", 39 "scott", "tiger"); */ 40 Connection conn = DriverManager.getConnection 41 ("jdbc:mysql://localhost/javabook", "scott", "tiger"); 42 System.out.println("Database connected"); 43 44 // Create a prepared statement for querying DB 45 pstmt = conn.prepareStatement( 46 "select * from Scores where name = ?"); 47 } 48 catch (Exception ex) { 49 System.out.println(ex); 50 } 51 } 52 53 /** Return the score for specified the name 54 * Return −1 if score is not found. 55 */ 56 public double findScore(String name) throws RemoteException { 57 double score = −1; 58 try { 59 // Set the specified name in the prepared statement 60 pstmt.setString(1, name); 61 62 // Execute the prepared statement 63 ResultSet rs = pstmt.executeQuery(); 64 65 // Retrieve the score 66 if (rs.next()) { 67 if (rs.getBoolean(3)) 68 score = rs.getDouble(2); 69 } 70 } 71 catch (SQLException ex) { 72 System.out.println(ex); 73 } 74 75 return score; 76 } 77 }

Student3TierImpl is similar to Listing 40.2, StudentServerInterfaceImpl.java in Section 40.3.1 except that the Student3TierImpl class finds the score from a JDBC data source instead from a hash map.

M40_LIAN1878_11_GE_C40.indd 12 12/29/17 6:02 PM

40.6 RMI Callbacks 40-13

The table named Scores consists of three columns, name, score, and permission, where the latter indicates whether the student has given permission to show his/her score. Since SQL does not support a boolean type, permission is defined as a number whose value of 1 indicates true and of 0 indicates false.

The initializeDB() method (lines 28–51) establishes connections with the data- base and creates a prepared statement for processing the query.

The findScore method (lines 56–76) sets the name in the prepared statement, executes the statement, processes the result, and returns the score for a student whose permission is true.

3. Write a main method in the class RegisterStudent3TierServer (see Listing 40.6) that registers the server object using StudentServerInterfaceImpl, the same name as in Listing 40.2, so you can use StudentServerInterfaceClient, created in Section 40.3.1, to test the server.

Listing 40.6 RegisterStudent3TierServer.java 1 import java.rmi.registry.*; 2 3 public class RegisterStudent3TierServer { 4 public static void main(String[] args) { 5 try { 6 StudentServerInterface obj = new Student3TierImpl(); 7 Registry registry = LocateRegistry.getRegistry(); 8 registry.rebind("StudentServerInterfaceImpl", obj); 9 System.out.println("Student server " + obj + " registered"); 10 } catch (Exception ex) { 11 ex.printStackTrace(); 12 } 13 } 14 }

4. Follow the steps below to run this example.

4.1. Start RMI registry by typing “start rmiregistry” at a DOS prompt from the book directory.

4.2. Start the server RegisterStudent3TierServer using the following command at the C:\ book directory:

C:\ book>java RegisterStudent3TierServer

4.3. Run the client StudentServerInterfaceClient. A sample run is shown in Figure 40.6.

40.5.1 Describe how parameters are passed in RMI.

40.6 RMI Callbacks RMI callbacks enable the server to invoke the methods on a client.

In a traditional client/server system, a client sends a request to a server, and the server pro- cesses the request and returns the result to the client. The server cannot invoke the methods on a client. One important benefit of RMI is that it supports callbacks, which enable the server to invoke methods on the client. With the RMI callback feature, you can develop interactive distributed applications.

Point Check

Point Key

M40_LIAN1878_11_GE_C40.indd 13 12/29/17 6:02 PM

40-14 Chapter 40 Remote Method Invocation

In Section 33.6, Case Studies: Distributed TicTacToe Games, you developed a distributed TicTacToe game using stream socket programming. The example that follows demonstrates the use of the RMI callback feature to develop an interactive TicTacToe game.

All the examples you have seen so far in this chapter have simple behaviors that are easy to model with classes. The behavior of the TicTacToe game is somewhat complex. To create the classes to model the game, you need to study and understand it and distribute the process appropriately between client and server.

Clearly the client should be responsible for handling user interactions, and the server should coordinate with the client. Specifically, the client should register with the server, and the server can take two and only two players. Once a client makes a move, it should notify the server; the server then notifies the move to the other player. The server should determine the status of the game—that is, whether it has been won or drawn—and notify the players. The server should also coordinate the turns—that is, which client has the turn at a given time. The ideal approach for notifying a player is to invoke a method in the client that sets appropri- ate properties in the client or sends messages to a player. Figure 40.7 illustrates the relation- ship between clients and server.

Figure 40.7 The server coordinates the activities with the clients.

Client 1: A client makes two kinds of calls: 1. Request to play the game. 2. Notify the server of the move.

Client 2: A client makes two kinds of calls: 1. Request to play the game. 2. Notify the server of the move.

Server: The server makes three kinds of calls: 1. Notify a client of the other client’s move. 2. Notify the game status. 3. Coordinate the turn.

All the calls a client makes can be encapsulated in one remote interface named TicTacToe (Listing 40.7), and all the calls the server invokes can be defined in another interface named CallBack (Listing 40.8). These two interfaces are defined as follows:

Listing 40.7 TicTacToeInterface.java 1 import java.rmi.*; 2 3 public interface TicTacToeInterface extends Remote { 4 /** 5 * Connect to the TicTacToe server and return the token. 6 * If the returned token is ' ', the client is not connected to 7 * the server 8 */ 9 public char connect(CallBack client) throws RemoteException; 10 11 /** A client invokes this method to notify the server of its move*/ 12 public void myMove(int row, int column, char token) 13 throws RemoteException; 14 }

M40_LIAN1878_11_GE_C40.indd 14 12/29/17 6:02 PM

40.6 RMI Callbacks 40-15

Listing 40.8 CallBack.java 1 import java.rmi.*; 2 3 public interface CallBack extends Remote { 4 /** The server notifies the client for taking a turn */ 5 public void takeTurn(boolean turn) throws RemoteException; 6 7 /** The server sends a message to be displayed by the client */ 8 public void notify(java.lang.String message) 9 throws RemoteException; 10 11 /** The server notifies a client of the other player's move */ 12 public void mark(int row, int column, char token) 13 throws RemoteException; 14 }

What does a client need to do? The client interacts with the player. Assume all the cells are initially empty, and the first player takes the X token and the second player the O token. To mark a cell, the player points the mouse to the cell and clicks it. If the cell is empty, the token (X or O) is displayed. If the cell is already filled, the player’s action is ignored.

From the preceding description, it is obvious that a cell is a GUI object that handles mouse- click events and displays tokens. The candidate for such an object could be a button or a panel. Panels are more flexible than buttons. The token (X or O) can be drawn on a panel in any size, but it can be displayed only as a label on a button.

Let Cell be a subclass of JPanel. You can declare a 3 * 3 grid to be an array Cell[] [] cell = new Cell[3][3] for modeling the game. How do you know the state of a cell (marked or not)? You can use a property named marked of the boolean type in the Cell class. How do you know whether the player has a turn? You can use a property named myTurn of boolean. This property (initially false) can be set by the server through a callback.

The Cell class is responsible for drawing the token when an empty cell is clicked, so you need to write the code for listening to the MouseEvent and for painting the shape for tokens X and O. To determine which shape to draw, introduce a variable named marker of the char type. Since this variable is shared by all the cells in a client, it is preferable to declare it in the client and to declare the Cell class as an inner class of the client so this variable will be accessible to all the cells.

Now let us turn our attention to the server side. What does the server need to do? The server needs to implement TicTacToeInterface and notify the clients of the game status. The server has to record the moves in the cells and check the status every time a player makes a move. The status information can be kept in a 3 * 3 array of char. You can imple- ment a method named isFull() to check whether the board is full and a method named isWon(token) to check whether a specific player has won.

Once a client is connected to the server, the server notifies the client which token to use— that is, X for the first client and O for the second. Once a client notifies the server of its move, the server checks the game status and notifies the clients.

Now the most critical question is how the server notifies a client. You know that a client invokes a server method by creating a server stub on the client side. A server cannot directly invoke a client, because the client is not declared as a remote object. The CallBack interface was created to facilitate the server’s callback to the client. In the implementation of CallBack, an instance of the client is passed as a parameter in the constructor of CallBack. The client creates an instance of CallBack and passes its stub to the server, using a remote method named connect() defined in the server. The server then invokes the client’s method through a CallBack instance. The triangular relationship of client, CallBack implementation, and server is shown in Figure 40.8.

M40_LIAN1878_11_GE_C40.indd 15 12/29/17 6:02 PM

40-16 Chapter 40 Remote Method Invocation

Here are the steps to complete the example.

1. Create TicTacToeImpl.java (Listing 40.9) to implement TicTacToeInterface. Add a main method in the program to register the server with the RMI.

Listing 40.9 TicTacToeImpl.java 1 import java.rmi.*; 2 import java.rmi.server.*; 3 import java.rmi.registry.*; 4 import java.rmi.registry.*; 5 6 public class TicTacToeImpl extends UnicastRemoteObject 7 implements TicTacToeInterface { 8 // Declare two players, used to call players back 9 private CallBack player1 = null; 10 private CallBack player2 = null; 11 12 // board records players' moves 13 private char[][] board = new char[3][3]; 14 15 /** Constructs TicTacToeImpl object and 16 exports it on default port. 17 */ 18 public TicTacToeImpl() throws RemoteException { 19 super(); 20 } 21 22 /** Constructs TicTacToeImpl object and exports it on specified 23 * port. 24 * @param port The port for exporting 25 */

Figure 40.8 The server receives a CallBack stub from the client and invokes the remote methods defined in the CallBack interface, which can invoke the methods defined in the client.

Client

An instance of CallBackImpl created

Client methods invoked by the methods in CallBackImpl

Server

Receive a stub of CallBack

Server invokes remote object’s methods

CallBackImpl

An instance of Client created

The remote methods in CallBack

M40_LIAN1878_11_GE_C40.indd 16 12/29/17 6:02 PM

40.6 RMI Callbacks 40-17

26 public TicTacToeImpl(int port) throws RemoteException { 27 super(port); 28 } 29 30 /** 31 * Connect to the TicTacToe server and return the token. 32 * If the returned token is ' ', the client is not connected to 33 * the server 34 */ 35 public char connect(CallBack client) throws RemoteException { 36 if (player1 == null) { 37 // player1 (first player) registered 38 player1 = client; 39 player1.notify("Wait for a second player to join"); 40 return 'X'; 41 } 42 else if (player2 == null) { 43 // player2 (second player) registered 44 player2 = client; 45 player2.notify("Wait for the first player to move"); 46 player2.takeTurn(false); 47 player1.notify("It is my turn (X token)"); 48 player1.takeTurn(true); 49 return 'O'; 50 } 51 else { 52 // Already two players 53 client.notify("Two players are already in the game"); 54 return ' '; 55 } 56 } 57 58 /** A client invokes this method to notify the 59 server of its move*/ 60 public void myMove(int row, int column, char token) 61 throws RemoteException { 62 // Set token to the specified cell 63 board[row][column] = token; 64 65 // Notify the other player of the move 66 if (token == 'X') 67 player2.mark(row, column, 'X'); 68 else 69 player1.mark(row, column, 'O'); 70 71 // Check if the player with this token wins 72 if (isWon(token)) { 73 if (token == 'X') { 74 player1.notify("I won!"); 75 player2.notify("I lost!"); 76 player1.takeTurn(false); 77 } 78 else { 79 player2.notify("I won!"); 80 player1.notify("I lost!"); 81 player2.takeTurn(false); 82 } 83 } 84 else if (isFull()) { 85 player1.notify("Draw!"); 86 player2.notify("Draw!");

M40_LIAN1878_11_GE_C40.indd 17 12/29/17 6:02 PM

40-18 Chapter 40 Remote Method Invocation

87 } 88 else if (token == 'X') { 89 player1.notify("Wait for the second player to move"); 90 player1.takeTurn(false); 91 player2.notify("It is my turn, (O token)"); 92 player2.takeTurn(true); 93 } 94 else if (token == 'O') { 95 player2.notify("Wait for the first player to move"); 96 player2.takeTurn(false); 97 player1.notify("It is my turn, (X token)"); 98 player1.takeTurn(true); 99 } 100 } 101 102 /** Check if a player with the specified token wins */ 103 public boolean isWon(char token) { 104 for (int i = 0; i < 3; i++) 105 if ((board[i][0] == token) && (board[i][1] == token) 106 && (board[i][2] == token)) 107 return true; 108 109 for (int j = 0; j < 3; j++) 110 if ((board[0][j] == token) && (board[1][j] == token) 111 && (board[2][j] == token)) 112 return true; 113 114 if ((board[0][0] == token) && (board[1][1] == token) 115 && (board[2][2] == token)) 116 return true; 117 118 if ((board[0][2] == token) && (board[1][1] == token) 119 && (board[2][0] == token)) 120 return true; 121 122 return false; 123 } 124 125 /** Check if the board is full */ 126 public boolean isFull() { 127 for (int i = 0; i < 3; i++) 128 for (int j = 0; j < 3; j++) 129 if (board[i][j] == '\u0000') 130 return false; 131 132 return true; 133 } 134 135 public static void main(String[] args) { 136 try { 137 TicTacToeInterface obj = new TicTacToeImpl(); 138 Registry registry = LocateRegistry.getRegistry(); 139 registry.rebind("TicTacToeImpl", obj); 140 System.out.println("Server " + obj + " registered"); 141 } 142 catch (Exception ex) { 143 ex.printStackTrace(); 144 } 145 } 146 }

M40_LIAN1878_11_GE_C40.indd 18 12/29/17 6:02 PM

40.6 RMI Callbacks 40-19

2. Create CallBackImpl.java (Listing 40.10) to implement the CallBack interface.

Listing 40.10 CallBackImpl.java 1 import java.rmi.*; 2 import java.rmi.server.*; 3 4 public class CallBackImpl extends UnicastRemoteObject 5 implements CallBack { 6 // The client will be called by the server through callback 7 private TicTacToeClientRMI thisClient; 8 9 /** Constructor */ 10 public CallBackImpl(Object client) throws RemoteException { 11 thisClient = (TicTacToeClientRMI)client; 12 } 13 14 /** The server notifies the client for taking a turn */ 15 public void takeTurn(boolean turn) throws RemoteException { 16 thisClient.setMyTurn(turn); 17 } 18 19 /** The server sends a message to be displayed by the client */ 20 public void notify(String message )throws RemoteException { 21 thisClient.setMessage(message); 22 } 23 24 /** The server notifies a client of the other player's move */ 25 public void mark(int row, int column, char token) 26 throws RemoteException { 27 thisClient.mark(row, column, token); 28 } 29 }

3. Create a client named TicTacToeClientRMI (Listing 40.11) for interacting with a player and communicating with the server. Enable it to run standalone.

Listing 40.11 TicTacToeClientRMI.java 1 import java.rmi.*; 2 3 import javafx.application.Application; 4 import javafx.application.Platform; 5 import javafx.stage.Stage; 6 import javafx.scene.Scene; 7 import javafx.scene.control.Label; 8 import javafx.scene.layout.BorderPane; 9 import javafx.scene.layout.GridPane; 10 import javafx.scene.layout.Pane; 11 import javafx.scene.paint.Color; 12 import javafx.scene.shape.Line; 13 import javafx.scene.shape.Ellipse; 14 15 import java.rmi.registry.Registry; 16 import java.rmi.registry.LocateRegistry; 17 18 public class TicTacToeClientRMI extends Application { 19 // marker is used to indicate the token type 20 private char marker; 21 22 // myTurn indicates whether the player can move now 23 private boolean myTurn = false;

M40_LIAN1878_11_GE_C40.indd 19 12/29/17 6:02 PM

40-20 Chapter 40 Remote Method Invocation

24 25 // Indicate which player has a turn, initially it is the X player 26 private char whoseTurn = 'X'; 27 28 // Create and initialize cell 29 private Cell[][] cell = new Cell[3][3]; 30 31 // Create and initialize a status label 32 private Label lblStatus = new Label("X's turn to play"); 33 34 // ticTacToe is the game server for coordinating 35 // with the players 36 private TicTacToeInterface ticTacToe; 37 38 private Label lblIdentification = new Label(); 39 40 @Override // Override the start method in the Application class 41 public void start(Stage primaryStage) { 42 // Pane to hold cell 43 GridPane pane = new GridPane(); 44 for (int i = 0; i < 3; i++) 45 for (int j = 0; j < 3; j++) 46 pane.add(cell[i][j] = new Cell(i, j), j, i); 47 48 BorderPane borderPane = new BorderPane(); 49 borderPane.setCenter(pane); 50 borderPane.setTop(lblStatus); 51 borderPane.setBottom(lblIdentification); 52 53 // Create a scene and place it in the stage 54 Scene scene = new Scene(borderPane, 450, 170); 55 primaryStage.setTitle("TicTacToe"); // Set the stage title 56 primaryStage.setScene(scene); // Place the scene in the stage 57 primaryStage.show(); // Display the stage 58 59 new Thread( () −> { 60 try { 61 initializeRMI(); 62 } 63 catch (Exception ex) { 64 ex.printStackTrace(); 65 }}).start(); 66 } 67 68 /** Initialize RMI */ 69 protected boolean initializeRMI() throws Exception { 70 String host = ""; 71 72 try { 73 Registry registry = LocateRegistry.getRegistry(host); 74 ticTacToe = (TicTacToeInterface) 75 registry.lookup("TicTacToeImpl"); 76 System.out.println 77 ("Server object " + ticTacToe + " found"); 78 } 79 catch (Exception ex) { 80 System.out.println(ex); 81 } 82 83 // Create callback for use by the 84 // server to control the client

M40_LIAN1878_11_GE_C40.indd 20 12/29/17 6:02 PM

40.6 RMI Callbacks 40-21

85 CallBackImpl callBackControl = new CallBackImpl(this); 86 87 if ( 88 (marker = 89 ticTacToe.connect((CallBack)callBackControl)) != ' ') 90 { 91 System.out.println("connected as " + marker + " player."); 92 Platform.runLater(() −> 93 lblIdentification.setText("You are player " + marker)); 94 return true; 95 } 96 else { 97 System.out.println("already two players connected as "); 98 return false; 99 } 100 } 101 102 /** Set variable myTurn to true or false */ 103 public void setMyTurn(boolean myTurn) { 104 this.myTurn = myTurn; 105 } 106 107 /** Set message on the status label */ 108 public void setMessage(String message) { 109 Platform.runLater(() −> lblStatus.setText(message)); 110 } 111 112 /** Mark the specified cell using the token */ 113 public void mark(int row, int column, char token) { 114 cell[row][column].setToken(token); 115 } 116 117 // An inner class for a cell 118 public class Cell extends Pane { 119 // marked indicates whether the cell has been used 120 private boolean marked = false; 121 122 // row and column indicate where the cell appears on the board 123 int row, column; 124 125 // Token used for this cell 126 private char token = ' '; 127 128 public Cell(final int row, final int column) { 129 this.row = row; 130 this.column = column; 131 setStyle("-fx-border-color: black"); 132 this.setPrefSize(2000, 2000); 133 this.setOnMouseClicked(e −> handleMouseClick()); 134 } 135 136 /** Return token */ 137 public char getToken() { 138 return token; 139 } 140 141 /** Set a new token */ 142 public void setToken(char c) { 143 token = c; 144 marked = true; 145

M40_LIAN1878_11_GE_C40.indd 21 12/29/17 6:02 PM

40-22 Chapter 40 Remote Method Invocation

146 if (token == 'X') { 147 Line line1 = new Line(10, 10, 148 this.getWidth() − 10, this.getHeight() − 10); 149 line1.endXProperty().bind(this.widthProperty().subtract(10)); 150 line1.endYProperty().bind(this.heightProperty().subtract(10)); 151 Line line2 = new Line(10, this.getHeight() − 10, 152 this.getWidth() − 10, 10); 153 line2.startYProperty().bind( 154 this.heightProperty().subtract(10)); 155 line2.endXProperty().bind(this.widthProperty().subtract(10)); 156 157 // Add the lines to the pane 158 Platform.runLater(() −> 159 this.getChildren().addAll(line1, line2)); 160 } 161 else if (token == 'O') { 162 Ellipse ellipse = new Ellipse(this.getWidth() / 2, 163 this.getHeight() / 2, this.getWidth() / 2 − 10, 164 this.getHeight() / 2 − 10); 165 ellipse.centerXProperty().bind( 166 this.widthProperty().divide(2)); 167 ellipse.centerYProperty().bind( 168 this.heightProperty().divide(2)); 169 ellipse.radiusXProperty().bind( 170 this.widthProperty().divide(2).subtract(10)); 171 ellipse.radiusYProperty().bind( 172 this.heightProperty().divide(2).subtract(10)); 173 ellipse.setStroke(Color.BLACK); 174 ellipse.setFill(Color.WHITE); 175 176 Platform.runLater(() −> 177 getChildren().add(ellipse)); // Add the ellipse to the pane 178 } 179 } 180 181 /* Handle a mouse click event */ 182 private void handleMouseClick() { 183 if (myTurn && !marked) { 184 // Mark the cell 185 setToken(marker); 186 187 // Notify the server of the move 188 try { 189 ticTacToe.myMove(row, column, marker); 190 } 191 catch (RemoteException ex) { 192 System.out.println(ex); 193 } 194 } 195 } 196 } 197 198 /** 199 * The main method is only needed for the IDE with limited 200 * JavaFX support. Not needed for running from the command line. 201 */ 202 public static void main(String[] args) { 203 launch(args); 204 } 205 }

M40_LIAN1878_11_GE_C40.indd 22 12/29/17 6:02 PM

40.6 RMI Callbacks 40-23

4. Follow the steps below to run this example.

4.1. Start RMI registry by typing “start rmiregistry” at a DOS prompt from the book directory.

4.2. Start the server TicTacToeImpl using the following command at the C:\ book directory: C:\ book>java TicTacToeImpl

4.3. Run the client TicTacToeClientRMI. A sample run is shown in Figure 40.9.

TicTacToeInterface defines two remote methods, connect(CallBack client) and myMove(int row, int column, char token). The connect method plays two roles: one is to pass a CallBack stub to the server, and the other is to let the server assign a token for the player. The myMove method notifies the server that the player has made a specific move.

The CallBack interface defines three remote methods, takeTurn(boolean turn), notify(String message), and mark(int row, int column, char token). The takeTurn method sets the client’s myTurn property to true or false. The notify method displays a message on the client’s status label. The mark method marks the client’s cell with the token at the specified location.

TicTacToeImpl is a server implementation for coordinating with the clients and managing the game. The variables player1 and player2 are instances of CallBack, each of which corresponds to a client, passed from a client when the client invokes the connect method. The variable board records the moves by the two players. This information is needed to determine the game status. When a client invokes the connect method, the server assigns a token X for the first player and O for the second player, and accepts only two players. You can modify the program to accept additional clients as observers. (See Exercise 40.7 for more details).

Once two players are in the game, the server coordinates the turns between them. When a client invokes the myMove method, the server records the move and notifies the other player by marking the other player’s cell. It then checks to see whether the player wins or whether the board is full. If neither condition applies and therefore the game continues, the server gives a turn to the other player.

The CallBackImpl implements the CallBack interface. It creates an instance of TicTacToeClientRMI through its constructor. The CallBackImpl relays the server request to the client by invoking the client’s methods. When the server invokes the takeTurn method, CallBackImpl invokes the client’s setMyTurn() method to set the property myTurn in the client. When the server invokes the notify() method, CallBackImpl invokes the client’s setMessage() method to set the message on the client’s status label. When the server invokes the mark method, CallBackImpl invokes the client’s mark method to mark the specified cell.

Figure 40.9 Two players play each other through the RMI server.

M40_LIAN1878_11_GE_C40.indd 23 12/29/17 6:02 PM

40-24 Chapter 40 Remote Method Invocation

Interestingly, obtaining the TicTacToeImpl stub for the client is different from obtain- ing the CallBack stub for the server. The TicTacToeImpl stub is obtained by invoking the lookup() method through the RMI registry, and the CallBack stub is passed to the server through the connect method in the TicTacToeImpl stub. It is a common practice to obtain the first stub with the lookup method, but to pass the subsequent stubs as parameters through remote method invocations.

Since the variables myTurn and marker are defined in TicTacToeClientRMI, the Cell class is defined as an inner class within TicTacToeClientRMI in order to enable all the cells in the client to access them. Exercise 40.8 suggests alternative approaches that implement the Cell as a noninner class.

40.6.1 What is the problem if the connect method in the TicTacToeInterface is defined as

public boolean connect(CallBack client, char token) throws RemoteException;

or as

public boolean connect(CallBack client, Character token) throws RemoteException;

40.6.2 What is callback? How does callback work in RMI?

Point Check

Key Terms

callback 40-13 RMI registry 40-3

skeleton 40-3 stub 40-3

ChapTer summary 1. RMI is a high-level Java API for building distributed applications using distributed

objects.

2. The key idea of RMI is its use of stubs and skeletons to facilitate communications between objects. The stub and skeleton are automatically generated, which relieves pro- grammers of tedious socket-level network programming.

3. For an object to be used remotely, it must be defined in an interface that extends the java.rmi.Remote interface.

4. In an RMI application, the initial remote object must be registered with the RMI registry on the server side and be obtained using the lookup method through the registry on the client side. Subsequent uses of stubs of other remote objects may be passed as parameters through remote method invocations.

5. RMI is especially useful for developing scalable and load-balanced multitier distributed applications.

Quiz Answer the quiz for this chapter online at the book Companion Website.

M40_LIAN1878_11_GE_C40.indd 24 12/29/17 6:02 PM

Programming Exercises 40-25

programming exerCises

Section 40.3 *40.1 (Limit the number of clients) Modify the example in Section 40.3.1, Example:

Retrieving Student Scores from an RMI Server, to limit the number of concurrent clients to 10.

*40.2 (Compute loan) Rewrite Programming Exercise 33.1 using RMI. You need to define a remote interface for computing monthly payment and total payment.

**40.3 (Web visit count) Rewrite Programming Exercise 33.4 using RMI. You need to define a remote interface for obtaining and increasing the count.

**40.4 (Display and add addresses) Rewrite Programming Exercise 33.6 using RMI. You need to define a remote interface for adding addresses and retrieving address information.

Section 40.5 **40.5 (Address in a database table) Rewrite Programming Exercise 40.4. Assume the

address is stored in a table.

**40.6 (Three-tier application) Use the three-tier approach to modify Programming Exercise 40.4, as follows:

■■ Create a JavaFX client to manipulate student information, as shown in Figure 33.23a.

■■ Create a remote object interface with methods for retrieving, inserting, and updating student information, and an object implementation for the interface.

Section 40.6 **40.7 (Chat) Rewrite Programming Exercise 33.13 using RMI. You need to define a

remote interface for sending and receiving a message.

**40.8 (Improve TicTacToe) Modify the TicTacToe example in Section 40.6, RMI Callbacks, as follows:

■■ Allow a client to connect to the server as an observer to watch the game. ■■ Rewrite the Cell class as a noninner class.

M40_LIAN1878_11_GE_C40.indd 25 12/29/17 6:02 PM

Objectives ■■ To describe what a Web service is (§41.1).

■■ To create a Web service class (§41.2).

■■ To publish and test a Web service (§41.3).

■■ To create a Web service client reference (§41.4).

■■ To explain the role of WSDL (§41.4).

■■ To pass arguments of object type in a Web service (§41.5).

■■ To discover how a client communicates with a Web service (§41.5).

■■ To describe what SOAP requests and SOAP responses are (§41.5).

■■ To track a session in Web services (§41.6).

Web Services

Chapter

M41_LIAN1878_11_GE_C41.indd 1 12/29/17 6:03 PM

41-2 Chapter 41 Web Services

41.1 Introduction Web services is about sharing objects on the Internet.

Web service is a technology that enables programs to communicate through HTTP on the Internet. Web services enable a program on one system to invoke a method in an object on another system. You can develop and use Web services using any languages on any platform. Web services are simple and easy to develop.

Web services run on the Web using HTTP. There are several APIs for Web services. A popular standard is the Simple Object Access Protocol (SOAP), which is based on XML. The computer on which a Web service resides is referred to as a server. The server needs to make the service available to the client, known as publishing a Web service. Using a Web service from a client is known as consuming a Web service.

A client interacts with a Web service through a proxy object. The proxy object facilitates the communication between the client and the Web service. The client passes arguments to invoke methods on the proxy object. The proxy object sends the request to the server and receives the result back from the server, as shown in Figure 41.1.

Figure 41.1 A proxy object serves as a facilitator between a client and a Web service.

Client

Web service proxy object

Server

Web serviceInternet

Point Key

41.1.1 What is a Web service? 41.1.2 Can you invoke a Web service from a language other than Java? 41.1.3 Do Web services support callback? That is, can a Web service call a method from

a client’s program?

41.1.4 What is SOAP? What is it to publish a Web service? What is it to consume a Web service? What is the role of a proxy object?

41.2 Creating Web Services An IDE such as NetBeans is an effective tool for developing and deploying Web services.

There are many tools for creating Web services. This book demonstrates creating Web services using NetBeans.

Note Apache Tomcat Server does not work well with Web services. To develop and deploy Web services using NetBeans, you need to install GlassFish. For information on how to install GlassFish on NetBeans.

We now create a Web service for obtaining student scores. A Web service is a class that contains the methods for the client to invoke. Name the class ScoreService with a method named findScore(String name) that returns the score for a student.

First, you need to create a Web project using the following steps:

1. Choose File, New Project to display the New Project dialog box. In the New Project dialog box, choose Java Web in the Categories pane and choose Web Application in the Projects pane. Click Next to display the New Web Application dialog box.

Point Check

Point Key

M41_LIAN1878_11_GE_C41.indd 2 12/29/17 6:03 PM

41.2 Creating Web Services 41-3

2. Enter WebServiceProject as the project name, specify the location where you want the project to be stored, and click Next to display the Server and Setting dialog.

3. Select GlassFish 4 as the server and Java EE 7 Web as the Java EE version. Click Finish to create the project.

Now you can create the ScoreService class in the project as follows:

1. Right-click the WebServiceProject in the Project pane to display a context menu. Choose New, Web Service to display the New Web Service dialog box. (If you don’t see Web Service, click New, Other to display the New File dialog box to choose Web Service in this dialog box.)

2. Enter ScoreService in the Web Service Name field and enter chapter41 in the Pack- age field. Click Finish to create ScoreService.

3. Complete the source code as shown in Listing 41.1.

Listing 41.1 ScoreService.java 1 package chapter41; 2 3 import java.util.HashMap; 4 import javax.jws.WebService; // For annotation @WebService 5 import javax.jws.WebMethod; // For annotation @WebMethod 6 7 @WebService(name = "ScoreService", serviceName = "ScoreWebService") 8 public class ScoreService { 9 // Stores scores in a map indexed by name 10 private HashMap<String, Double> scores = 11 new HashMap<String, Double>(); 12 13 public ScoreService() { 14 scores.put("John", 90.5); 15 scores.put("Michael", 100.0); 16 scores.put("Michelle", 98.5); 17 } 18 19 @WebMethod(operationName = "findScore") 20 public double findScore(String name) { 21 Double d = scores.get(name); 22 23 if (d == null) { 24 System.out.println("Student " + name + " is not found "); 25 return –1; 26 } 27 else { 28 System.out.println("Student " + name + "\’s score is " 29 + d.doubleValue()); 30 return d.doubleValue(); 31 } 32 } 33 }

Lines 4–5 import the annotations used in the program in lines 7 and 19. Annotation is a new feature in Java, which enables you to simplify coding. The compiler will automatically gener- ate the code for the annotated directives. So, it frees the programmer from writing the detailed boilerplate code that could be generated mechanically. The annotation (line 7)

@WebService(name = "ScoreService", serviceName = "ScoreWebService")

tells the compiler that the class ScoreService is associated with the Web service named ScoreWebService.

M41_LIAN1878_11_GE_C41.indd 3 12/29/17 6:03 PM

41-4 Chapter 41 Web Services

The annotation (line 19)

@WebMethod(operationName = "findScore")

indicates that findScore is a method that can be invoked from a client. The findScore method returns a score if the name is in the hash map. Otherwise, it returns

−1.0. You can manually type the code for the service, or create it from the Design tab, as shown

in Figure 41.2.

Figure 41.2 The services can also be created from the Design pane.

41.3 Deploying and Testing Web Services Deploying a Web service is to make it available on the Internet for other programs to use.

After a Web service is created, you need to deploy it for clients to use. Deploying Web services is also known as publishing Web services. To deploy it, right-click the WebServiceProject in the Project to display a context menu and choose Deploy. This command will first undeploy the service if it was deployed and then redeploy it.

Now you can test the Web service by entering the follow URL in a browser, as shown in Figure 41.3.

http://localhost:8080/WebServiceProject/ScoreWebService?Tester

Note ScoreWebService is the name you specified in line 7 in Listing 41.1. This Web service has only one remote method named findScore. You can define an unlimited number of remote methods in a Web service class. If so, all these methods will be displayed in the test page.

To test the findScore method, enter Michael and click the findScore button. You will see that the method returns 100.0, as shown in Figure 41.4.

Point Key

M41_LIAN1878_11_GE_C41.indd 4 12/29/17 6:03 PM

41.3 Deploying and Testing Web Services 41-5

Note If your computer is connected to the Internet, you can test Web services from another computer by entering the following URL:

http://host:8080/WebServiceProject/ScoreWebService?Tester

Where host is the host name or IP address of the server on which the Web service is run- ning. On Windows, you can find your IP address by typing the command ipconfig.

Note If you are running the server on Windows, the firewall may prevent remote clients from accessing the service. To enable it, do the following:

1. In the Windows control panel, click Windows Firewall to display the Windows Firewall dialog box.

Figure 41.3 The test page enables you to test Web services.

Figure 41.4 The method returns a test value.

M41_LIAN1878_11_GE_C41.indd 5 12/29/17 6:03 PM

41-6 Chapter 41 Web Services

2. In the Advanced tab, double-click Local Area Connection to display the Advanced Settings dialog box. Check Web Server (HTTP) to enable HTTP access to the server.

3. Click OK to close the dialog box.

41.4 Consuming Web Services Consuming a Web service is for a client to use a Web service.

After a Web service is published, you can write a client program to use it. A client can be any program (standalone application, servlet/JSP/JSF application, or another Web service) and written in any language.

We will use NetBeans to create a Web service client. Our client is a GUI application. The application simply lets the user enter a name and displays the score, as shown in Figure 41.5.

Point Key

Figure 41.5 The client uses the Web service to find scores.

Let us create a project for the client. The project named ScoreWebServiceClient Project can be created as follows:

1. Choose File, New Project to display the New Project dialog box.

2. In the New Project dialog box, choose Java in the Categories pane and choose Java Application in the Projects pane. Click Next to display the New Java Application dialog box.

3. Enter ScoreWebServiceClientProject as the project name, specify the location where you want the project to be stored, and uncheck the Create Main Class check box. Click Finish to create the project.

You need to create a Web service reference to this project. The reference will enable you to create a proxy object to interact with the Web service. Here are the steps to create a Web service reference:

1. Right-click the ScoreWebServiceClientProject in the Project pane to display a context menu. Choose New, Web Service Client to display the New Web Service Client dialog box, as shown in Figure 41.6.

2. Check the WSDL URL radio button and enter http://localhost:8080/WebServiceProject/ ScoreWebService?WSDL in the WSDL URL field.

3. Enter myWebservice in the package name field. Click Finish to generate the Web service reference.

Now you will see ScoreWebService created in the Web Service References folder in the Projects tab. The IDE has generated many supporting files for the reference. You can view all the generated .java files from the Files tab in the project pane, as shown in Figure 41.7. These files will be used by the proxy object to interact with the Web service.

M41_LIAN1878_11_GE_C41.indd 6 12/29/17 6:03 PM

41.4 Consuming Web Services 41-7

Note When you created a Web service reference, you entered a WSDL URL, as shown in Figure 41.6. This creates a .wsdl file. In this case, it is named ScoreWebService.wsdl under the Web Service References folder, as shown in Figure 41.8. So what is WSDL? WSDL stands for Web Service Description Language. A .wsdl file is an XML file that describes the available Web service to the client—i.e., the remote methods, their param- eters and return value types, and so on.

Note If the Web service is modified, you need to refresh the reference for the client. To do so, right-click the Web service node under Web Service References to display a context menu and choose Refresh Client.

Now you are ready to create a client for the Web service. Right-click the ScoreWeb Service ClientProject node in the Project pane to display a context menu, and choose New, Class to create a Java client named FindScoreApp in package chapter41, as shown in Listing 41.2.

Figure 41.6 The New Web Service Client dialog box creates a Web service reference.

Figure 41.7 You can see the automatically generated boilerplate code for Web services in the Generated Sources folder in the client’s project.

M41_LIAN1878_11_GE_C41.indd 7 12/29/17 6:03 PM

41-8 Chapter 41 Web Services

Listing 41.2 FindScoreApp.java 1 package chapter41; 2 3 import javafx.application.Application; 4 import javafx.scene.Scene; 5 import javafx.scene.control.Button; 6 import javafx.scene.control.Label; 7 import javafx.scene.control.TextField; 8 import javafx.scene.layout.GridPane; 9 import javafx.stage.Stage; 10 import myWebservice.ScoreWebService; 11 import myWebservice.ScoreService; 12 13 public class FindScoreApp extends Application { 14 // Declare a service object and a proxy object 15 private ScoreWebService scoreWebService = new ScoreWebService(); 16 private ScoreService proxy 17 = scoreWebService.getScoreServicePort(); 18 19 private Button btGetScore = new Button("Get Score"); 20 private TextField tfName = new TextField(); 21 private TextField tfScore = new TextField(); 22 23 public void start(Stage primaryStage) { 24 GridPane gridPane = new GridPane(); 25 gridPane.setHgap(5); 26 gridPane.add(new Label("Name"), 0, 0); 27 gridPane.add(new Label("Score"), 0, 1); 28 gridPane.add(tfName, 1, 0); 29 gridPane.add(tfScore, 1, 1); 30 gridPane.add(btGetScore, 1, 2); 31 32 // Create a scene and place the pane in the stage 33 Scene scene = new Scene(gridPane, 250, 250); 34 primaryStage.setTitle("FindScoreApp"); // Set the stage title 35 primaryStage.setScene(scene); // Place the scene in the stage 36 primaryStage.show(); // Display the stage 37

Figure 41.8 The .wsdl file describes Web services to clients.

M41_LIAN1878_11_GE_C41.indd 8 12/29/17 6:03 PM

41.5 Passing and Returning Arguments 41-9

38 btGetScore.setOnAction(e -> getScore()); 39 } 40 41 private void getScore() { 42 try { 43 // Get student score 44 double score = proxy.findScore(tfName.getText().trim()); 45 46 // Display the result 47 if (score < 0) 48 tfScore.setText("Not found"); 49 else 50 tfScore.setText(new Double(score).toString()); 51 } 52 catch(Exception ex) { 53 ex.printStackTrace(); 54 } 55 } 56 }

The program creates a Web service object (line 11) and creates a proxy object (line 12) to interact with the Web service.

To find a score for a student, the program invokes the remote method findScore on the proxy object (line 39).

41.5 Passing and Returning Arguments The Simple Object Access Protocol (SOAP) can be used to send and return values to and from a Web service.

In the preceding example, a Web service client you created invokes the findScore method with a string argument, and the Web service executes the method and returns a score as a dou ble value. How does this work? It is the Simple Object Access Protocol (SOAP) that facilitates communications between the client and the server.

SOAP is based on XML. The message between the client and the server is described in XML. Figure 41.9 shows the SOAP request and SOAP response for the findScore method.

When invoking the findScore method, a SOAP request is sent to the server. The request contains the information about the method and the argument. As shown in Figure 41.9, the XML text

<ns1:findScore> <arg0>Michael</arg0> </ns1:findScore>

specifies that the method findScore is called with argument Michael. Upon receiving the SOAP request, the Web service parses it. After parsing it, the Web ser-

vice invokes an appropriate method with specified arguments (if any) and sends the response back in a SOAP response. As shown in Figure 41.9, the XML text

<ns1:findScoreResponse> <return>100.0</return> </ns1:findScoreResponse>

specifies that the method returns 100.0.

Point Key

M41_LIAN1878_11_GE_C41.indd 9 12/29/17 6:03 PM

41-10 Chapter 41 Web Services

The proxy object receives the SOAP response from the Web service and parses it. This process is illustrated in Figure 41.10.

Can you pass an argument of any type between a client and a Web service? No. SOAP supports only primitive types, wrapper types, arrays, String, Date, Time, List, and several other types. It also supports certain custom classes. An object that is sent to or from a server is serialized into XML. The process of serializing/deserializing objects, called

Figure 41.9 The client request and server response are described in XML.

M41_LIAN1878_11_GE_C41.indd 10 12/29/17 6:03 PM

41.5 Passing and Returning Arguments 41-11

XML serialization/deserialization, is performed automatically. For a custom class to be used with Web methods, the class must meet the following requirements:

1. The class must have a no-arg constructor.

2. Instance variables that should be serialized must have public get and set methods. The classes of these variables must be supported by SOAP.

To demonstrate how to pass an object argument of a custom class, Listing 41.3 defines a Web service class named AddressService with two remote methods:

■■ getAddress(String firstName, String lastName) that returns an Address object for the specified firstName and lastName.

■■ storeAddress(Address address) that stores a Student object to the database.

Address information is stored in a table named Address in the database. The Address class was defined in Listing 42.12, Address.java. An Address object can be passed to or returned from a remote method, since the Address class has a no-arg constructor with get and set methods for all its properties.

Here are the steps to create a Web service named AddressService and the Address class in the project.

1. Right-click the WebServiceProject node in the project pane to display a context menu. Choose New, Web Service to display the New Web Service dialog box.

2. In the Web Service Name field, enter AddressService. In the Package field, enter chapter41. Click Finish to create the service class.

3. Right-click the WebServiceProject node in the project pane to display a context menu. Choose New, Java Class to display the New Java Class dialog box.

4. In the Class Name field, enter Address. In the Package field, enter chapter37. Click Finish to create the class.

The Address class is the same as shown in Listing 37.12. Complete the AddressService class as shown in Listing 41.3.

Listing 41.3 AddressService.java 1 package chapter41; 2 3 import chapter37.Address; 4 import java.sql.*; 5 import javax.jws.WebMethod; 6 import javax.jws.WebService; 7 8 @WebService(name = "AddressService", 9 serviceName = "AddressWebService")

Figure 41.10 A proxy object sends SOAP requests and receives SOAP responses.

Client

Web service proxy object

Server

SOAP Response

SOAP Request Web

service

M41_LIAN1878_11_GE_C41.indd 11 12/29/17 6:03 PM

41-12 Chapter 41 Web Services

10 public class AddressService { 11 // statement1 for retrieving an address and statement2 for storing 12 private PreparedStatement statement1; 13 14 // statement2 for storing an address 15 private PreparedStatement statement2; 16 17 public AddressService() { 18 initializeJdbc(); 19 } 20 21 @WebMethod(operationName = "getAddress") 22 public Address getAddress(String firstName, String lastName) { 23 try { 24 statement1.setString(1, firstName); 25 statement1.setString(2, lastName); 26 ResultSet resultSet = statement1.executeQuery(); 27 28 if (resultSet.next()) { 29 Address address = new Address(); 30 address.setFirstName(resultSet.getString("firstName")); 31 address.setLastName(resultSet.getString("lastName")); 32 address.setMi(resultSet.getString("mi")); 33 address.setTelephone(resultSet.getString("telephone")); 34 address.setFirstName(resultSet.getString("email")); 35 address.setCity(resultSet.getString("telephone")); 36 address.setState(resultSet.getString("state")); 37 address.setZip(resultSet.getString("zip")); 38 return address; 39 } 40 else 41 return null; 42 } catch (SQLException ex) { 43 ex.printStackTrace(); 44 } 45 46 return null; 47 } 48 49 @WebMethod(operationName = "storeAddress") 50 public void storeAddress(Address address) { 51 try { 52 statement2.setString(1, address.getLastName()); 53 statement2.setString(2, address.getFirstName()); 54 statement2.setString(3, address.getMi()); 55 statement2.setString(4, address.getTelephone()); 56 statement2.setString(5, address.getEmail()); 57 statement2.setString(6, address.getStreet()); 58 statement2.setString(7, address.getCity()); 59 statement2.setString(8, address.getState()); 60 statement2.setString(9, address.getZip()); 61 statement2.executeUpdate(); 62 } catch (SQLException ex) { 63 ex.printStackTrace(); 64 } 65 } 66 67 /** Initialize database connection */ 68 public void initializeJdbc() {

M41_LIAN1878_11_GE_C41.indd 12 12/29/17 6:03 PM

41.5 Passing and Returning Arguments 41-13

69 try { 70 Class.forName("com.mysql.jdbc.Driver"); 71 72 // Connect to the sample database 73 Connection connection = DriverManager.getConnection( 74 "jdbc:mysql://localhost/javabook", "scott", "tiger"); 75 76 statement1 = connection.prepareStatement( 77 "select * from Address where firstName = ? and lastName = ?"); 78 statement2 = connection.prepareStatement( 79 "insert into Address " + 80 "(lastName, firstName, mi, telephone, email, street, city, " 81 + "state, zip) values (?, ?, ?, ?, ?, ?, ?, ?, ?)"); 82 } catch (Exception ex) { 83 ex.printStackTrace(); 84 } 85 } 86 }

The new Web service is named AddressWebService (line 9) for the AddressService class.

When the service is deployed, the constructor (lines 17–19) of AddressWebService is invoked to initialize a database connection and create prepared statement1 and statement2 (lines 68–85).

The findAddress method searches the address in the Address table for the specified firstName and lastName. If found, the address information is returned in an Address object (lines 29–38). Otherwise, the method returns null (line 41).

The storeAddress method stores the address information from the Address object into the database (lines 52–61).

Note Don’t forget that you have to add the MySQL library to the WebServiceProject for this example to run.

Before you can use the service, deploy it. Right-click the WebServiceProject node in the Project to display a context menu and choose Deploy.

Now you are ready to develop a Web client that uses the AddressWebService. The client is a JSP program, as shown in Figure 41.11. The program has two functions. First, the user can enter the last name and first name and click the Search button to search for a record, as shown in Figure 41.12. Second, the user can enter the complete address infor- mation and click the Store button to store the information to the database, as shown in Figure 41.13.

Let us create a project for the client. The project named AddressWebServiceClient Project can be created as follows:

1. Choose File, New Project to display the New Web Application dialog box. In the New Web Application dialog box, choose Java Web in the Categories pane and choose Web Application in the Projects pane. Click Next to display the Name and Location dialog box.

2. Enter AddressWebServiceClientProject as the project name, specify the location where you want the project to be stored, and uncheck the Set as Main Project check box. Click Next to display the Server and Settings dialog box.

3. Choose GlassFish Server 4 in the Server field, and Java EE 7 Web as in the Java EE Version field, and click Finish to create the project.

M41_LIAN1878_11_GE_C41.indd 13 12/29/17 6:03 PM

41-14 Chapter 41 Web Services

Figure 41.12 The Search button finds and displays an address.

Figure 41.13 The Store button stores the address to the database.

Figure 41.11 The TestAddressWebService page allows the user to search and store addresses.

M41_LIAN1878_11_GE_C41.indd 14 12/29/17 6:03 PM

41.5 Passing and Returning Arguments 41-15

1. Right-click the AddressWebServiceClientProject node in the Project pane to dis- play a context menu. Choose New, Web Service Client to display the New Web Service Client dialog box.

2. Check the WSDL URL radio button and enter http://localhost:8080/WebServiceProject/ AddressWebService?WSDL in the WSDL URL field.

3. Enter myWebservice in the package name field and choose JAXWS as the JAX version. Click Finish to generate the Web service reference.

Now a reference to AddressWebService is created. Note this process also copies Address.java to the client project, as shown in Figure 41.14.

Figure 41.14 Address.java is automatically copied to the Web service client reference package.

Create a JSP named TestAddressWebService in the AddressWebServiceClient Project project, as shown in Listing 41.4.

Listing 41.4 TestAddressWebService.jsp 1  2 <%@ page import = "myWebservice.Address" %> 3 <%@ page import = "myWebservice.AddressWebService" %> 4 <%@ page import = "myWebservice.AddressService" %> 5 <jsp:useBean id = "addressId" 6 class = "myWebservice.Address" scope = "session"></jsp:useBean> 7 <jsp:setProperty name = "addressId" property = "*" /> 8

M41_LIAN1878_11_GE_C41.indd 15 12/29/17 6:03 PM

41-16 Chapter 41 Web Services

9 <html> 10 <head> 11 <title>Address Information</title> 12 </head> 13 <body> 14 <form method = "post" action = "TestAddressWebService.jsp"> 15 Last Name <font color = "#FF0000">*</font> 16 <input type = "text" name = "lastName" 17 <%if (addressId.getLastName() != null) { 18 out.print("value = \"" + addressId.getLastName() + "\"");}%> 19 size = "20" />  20 21 First Name <font color = "#FF0000">*</font> 22 <input type = "text" name = "firstName" 23 <%if (addressId.getFirstName() != null) { 24 out.print("value = \"" + addressId.getFirstName() + "\"");}%> 25 size = "20" />  26 27 MI 28 <input type = "text" name = "mi" 29 <%if (addressId.getMi() != null) { 30 out.print("value = \"" + addressId.getMi() + "\" "); } %> 31 size = "3" />  32 33 <p>Telephone 34 <input type = "text" name = "telephone" 35 <%if (addressId.getTelephone() != null) { 36 out.print("value = \"" + addressId.getTelephone() + "\" ");}%> 37 size = "20" />  38 39 Email 40 <input type = "text" name = "email" 41 <%if (addressId.getEmail() != null) { 42 out.print("value = \"" + addressId.getEmail() + "\" ");}%> 43 size = "28" />  44 </p> 45 46 <p>Street 47 <input type = "text" name = "street" 48 <%if (addressId.getStreet() != null) { 49 out.print("value = \"" + addressId.getStreet() + "\" ");}%> 50 size = "50" />  51 </p> 52 53 <p>City 54 <input type = "text" name = "city" 55 <%if (addressId.getCity() != null) { 56 out.print("value = \"" + addressId.getCity() + "\" ");}%> 57 size = "23" />  58 59 State 60 <select size = "1" name = "state"> 61 <option value = "GA">Georgia-GA</option> 62 <option value = "OK">Oklahoma-OK</option> 63 <option value = "IN">Indiana-IN</option> 64 </select>  65 66 Zip 67 <input type = "text" name = "zip" 68 <%if (addressId.getZip() != null) { 69 out.print("value = \"" + addressId.getZip() + "\" "); } %>

M41_LIAN1878_11_GE_C41.indd 16 12/29/17 6:03 PM

41.6 Web Service Session Tracking 41-17

70 size = "9" />  71 </p> 72 73 <p><input type = "submit" name = "Submit" value = "Search"> 74 <input type = "submit" name = "Submit" value = "Store"> 75 <input type = "reset" value = "Reset"> 76 </p> 77 </form> 78 <p><font color = "#FF0000">* required fields</font></p> 79 80 <% 81 if (request.getParameter("Submit") != null) { 82 AddressWebService addressWebService = new AddressWebService(); 83 AddressService proxy = addressWebService.getAddressServicePort(); 84 85 if (request.getParameter("Submit").equals("Store")) { 86 proxy.storeAddress(addressId); 87 out.println(addressId.getFirstName() + " " + 88 addressId.getLastName() + " has been added to the database"); 89 } 90 else if (request.getParameter("Submit").equals("Search")) { 91 Address address = proxy.getAddress(addressId.getFirstName(), 92 addressId.getLastName()); 93 if (address == null) 94 out.print(addressId.getFirstName() + " " + 95 addressId.getLastName() + " is not in the database"); 96 else 97 addressId = address; 98 } 99 } 100 %> 101 </body> 102 </html>

Lines 2–4 import the classes for the JSP page. The Address class (line 2) was created in the WebServiceProject and was automatically copied to the AddressWebService ClientProject project when a Web service reference for AddressWebService was created. A JavaBeans object for Address was created and associated with input parameters in lines 5–7.

The UI interface was laid in the form (lines 14–77). The action for the two buttons Search and Store invokes the same page TestAddressWebService.jsp (line 14).

When a button is clicked, a proxy object for AddressWebService is obtained (lines 82–83). For the Store button, the proxy object invokes the storeAddress method to add an address to the database (line 86). For the Search button, the proxy object invokes the getAd dress method to return an address (lines 91–92). If no address is found for the specified first and last names, the returned address is null (line 93).

41.6 Web Service Session Tracking You can use the HttpSession interface to session tracking for Web.

Section 37.8.3, Session Tracking Using the Servlet API, introduced session tracking for servlets using the javax.servlet.http.HttpSession interface. You can use Http Session to implement session tracking for Web services. To demonstrate this, consider an example that generates random True/False questions for the client and grades the answers on these ques- tions for the client.

The Web client consists of two JSP pages: DisplayQuiz.jsp and GradeQuiz.jsp. The DisplayQuiz page invokes the service method getQuestion() to display the questions, as shown in Figure 41.15. When you click the Submit button, the program invokes the service

Point Key

M41_LIAN1878_11_GE_C41.indd 17 12/29/17 6:03 PM

41-18 Chapter 41 Web Services

method gradeQuiz to grade the answers. The result is displayed in the GradeQuiz page, as shown in Figure 41.16.

Why is session tracking needed for this project? Each time a client displays a quiz, it cre- ates a randomly reorder the quiz for the client. Each client gets a different quiz every time the DisplayQuiz page is refreshed. When the client submits the answer, the Web service checks the answer against the previously generated quiz. So the quiz has to be stored in the session.

For convenience, let us create the Web service class named QuizService in the Web ServiceProject in package chapter41. Listing 41.5 gives the program.

Listing 41.5 QuizService.java 1 package chapter41; 2 3 import javax.jws.WebMethod; 4 import javax.jws.WebService; 5 import java.util.List; 6 import java.util.ArrayList; 7 import com.sun.xml.ws.developer.servlet.HttpSessionScope; 8 9 @HttpSessionScope 10 @WebService(name = "QuizService", serviceName = "QuizWebService") 11 public class QuizService { 12 private ArrayList<Object[]> quiz = new ArrayList<Object[]>(); 13 14 public QuizService() { 15 // Initialize questions and answers 16 quiz.add(new Object[]{ 17 "Is Atlanta the capital of Georgia?", true}); 18 quiz.add(new Object[]{ 19 "Is Columbia the capital of South Carolina?", true}); 20 quiz.add(new Object[]{ 21 "Is Fort Wayne the capital of Indiana?", false});

Figure 41.16 The answers are graded and displayed.

Figure 41.15 The Submit button submits the answers for grading.

M41_LIAN1878_11_GE_C41.indd 18 12/29/17 6:03 PM

41.6 Web Service Session Tracking 41-19

22 quiz.add(new Object[]{ 23 "Is New Orleans the capital of Louisiana?", false}); 24 quiz.add(new Object[]{ 25 "Is Chicago the capital of Illinois?", false}); 26 27 // Shuffle to generate a random quiz for a client 28 java.util.Collections.shuffle(quiz); 29 } 30 31 @WebMethod(operationName = "getQuestions") 32 public java.util.List<String> getQuestions() { 33 34 // Extract questions from quiz 35 List<String> questions = new ArrayList<String>(); 36 for (int i = 0; i < quiz.size(); i++) { 37 questions.add((String)(quiz.get(i)[0])); 38 } 39 40 return questions; // Return questions in the quiz 41 } 42 43 @WebMethod(operationName = "gradeQuiz") 44 public List<Boolean> gradeQuiz(List<Boolean> answers) { 45 List<Boolean> result = new ArrayList<Boolean>(); 46 for (int i = 0; i < quiz.size(); i++) 47 result.add(quiz.get(i)[1] == answers.get(i)); 48 49 return result; 50 } 51 }

The Web service class named QuizService contains two methods getQuestions and gradeQuiz. The new Web service is named QuizWebService (line 10).

The annotation @HttpSessionScope (line 9) is new in JAX-WS 2.2, which enables the Web service automatically maintains a separate instance for each client session. To use this annotation, you have add JAX-WS 2.2 into your project’s library. This can be done by clicking the Library node in the project and select Add Library.

Assume five True/False questions are available from the service. The quiz is stored in an ArrayList (lines 16–25).

Each element in the list is an array with two values. The first value is a string that describes the question and the second is a Boolean value indicating whether the answer should be true or false.

A new quiz is generated in the constructor and the quiz is shuffled using the shuffle method in the Collections class (line 28).

The getQuestions method (lines 31–40) returns questions in a list. The questions are extracted from the quiz (lines 34–37) and are returned (line 39).

The gradeQuiz method (lines 42–49) checks the answers from the client with the answers in the quiz. The client’s answers are compared with the key, and the result of the grading is stored in a list. Each element in the list is a Boolean value that indicates whether the answer is correct or incorrect (lines 44–46).

After creating and publishing the Web service, let us create a project for the client. The project named QuizWebServiceClientProject can be created as follows:

1. Choose File, New Project to display the New Web Application dialog box.

2. In the New Web Application dialog box, choose Java Web in the Categories pane and choose Web Application in the Projects pane. Click Next to display the Name and Location dialog box.

M41_LIAN1878_11_GE_C41.indd 19 12/29/17 6:03 PM

41-20 Chapter 41 Web Services

3. Enter QuizWebServiceClientProject as the project name, specify the location where you want the project to be stored, and uncheck the Set as Main Project check box. Click Next to display the Server and Settings dialog box.

4. Choose GlassFish Server 4 in the Server field, and Java EE 7 Web as in the Java EE Version field, and click Finish to create the project.

To use QuizWebService, you need to create a Web service client as follows:

1. Right-click the QuizWebServiceClientProject project in the Project pane to display a context menu. Choose New, Web Service Client to display the New Web Service Client dialog box.

2. Check the WSDL URL radio button and enter http://localhost:8080/WebServiceProject/ QuizWebService?WSDL in the WSDL URL field.

3. Enter myWebservice in the Package field. Click Finish to create the reference for QuizWebService.

Now a reference to QuizWebService is created. You can create a proxy object to access the remote methods in QuizService. Listings 41.6 and 41.7 show DisplayQuiz.jsp and Grad- eQuiz.jsp.

Listing 41.6 DisplayQuiz.jsp 1  2 <%@ page import = "myWebservice.QuizWebService" %> 3 <%@ page import = "myWebservice.QuizService" %> 4 <jsp:useBean id = "quizWebService" scope = "session" 5 class = "myWebservice.QuizWebService"> 6 </jsp:useBean> 7 8 <html> 9 <body> 10 <% 11 QuizService proxy = quizWebService.getQuizServicePort(); 12 java.util.List<String> questions = 13 (java.util.ArrayList<String>)(proxy.getQuestions()); 14 %> 15 <form method = "post" action = "GradeQuiz.jsp"> 16 <table> 17 <% for (int i = 0; i < questions.size(); i++) {%> 18 <tr> 19 <td> 20 <label><%= questions.get(i) %></label> 21 </td> 22 <td> 23 <input type = "radio" name = <%= "question" + i%> 24 value = "True" /> True 25 </td> 26 <td> 27 <input type = "radio" name = <%= "question" + i%> 28 value = "False" /> False 29 </td> 30 </tr> 31 <%}%> 32 </table> 33 <p><input type = "submit" name = "Submit" value = "Submit"> 34 <input type = "reset" value = "Reset">

M41_LIAN1878_11_GE_C41.indd 20 12/29/17 6:03 PM

41.6 Web Service Session Tracking 41-21

35 </p> 36 </form> 37 </body> 38 </html>

This page generates a quiz by invoking the getQuestions() in lines 12–13. The ques- tions are displayed in a table with radio buttons (lines 16–32). Clicking the Submit button invokes GradeQuiz.jsp.

Listing 41.7 GradeQuiz.jsp 1  2 <%@ page import = "myWebservice.QuizWebService" %> 3 <%@ page import = "myWebservice.QuizService" %> 4 <jsp:useBean id = "quizWebService" scope = "session" 5 class = "myWebservice.QuizWebService"> 6 </jsp:useBean> 7 8 <html> 9 <body> 10 <% 11 QuizService proxy = quizWebService.getQuizServicePort(); 12 java.util.List<String> quiz = proxy.getQuestions(); 13 14 // Get the answer from the DisplayQuiz page 15 java.util.List<Boolean> answers = new java.util.ArrayList<Boolean>(); 16 for (int i = 0; i < quiz.size(); i++) { 17 String trueOrFalse = request.getParameter("question" + i); 18 if (trueOrFalse.equals("True")) 19 answers.add(true); // Answered true 20 else if (trueOrFalse.equals("False")) 21 answers.add(false); // Answered false 22 } 23 24 // Grade answers 25 java.util.List<Boolean> result = proxy.gradeQuiz(answers); 26 27 // Find the correct count 28 int correctCount = 0; 29 for (int i = 0; i < result.size(); i++) { 30 if (result.get(i)) 31 correctCount++; 32 } 33 %> 34 35 Out of <%= result.size() %> questions, <%= correctCount %> correct. 36 </body> 37 </html>

This page collects the answers passed from the HTML form from the DisplayQuiz page (lines 15–21), invokes the gradeQuiz method to grade the quiz (line 25), finds the correct count (lines 28–31), and displays the result (line 35).

Note You need to answer all five questions before clicking the Submit button. A runtime error will occur if a radio button is not checked. You can fix this problem in Exercise 41.5.

M41_LIAN1878_11_GE_C41.indd 21 12/29/17 6:03 PM

41-22 Chapter 41 Web Services

41.6.1 What is the annotation to specify a Web service? What is the annotation to specify a Web method?

41.6.2 How do you deploy a Web service in NetBeans? 41.6.3 Can you test a Web service from a client? 41.6.4 How do you create a Web service reference for a client? 41.6.5 What is WSDL? What is SOAP? What is a SOAP request? What is a SOAP

response?

41.6.6 Can you pass primitive type arguments to a remote method? Can you pass any object type to a remote method? Can you pass an argument of a custom type to a remote method?

41.6.7 How do you obtain an HttpSession object for tracking a Web session? 41.6.8 Can you create two Web service references in one package in the same project in

NetBeans?

41.6.9 What happens if you don’t clone the quiz in lines 40–41 in Listing 41.5, Quiz- Service.java?

Point Check

Key Terms

@WebService 41-3 @WebMethod 41-3 consuming a Web service 41-2 proxy object 41-2

publishing a Web service 41-2 Web service 41-2 Web service client reference 41-15 WSDL 41-6

ChapTer summary 1. Web services enable a Java program on one system to invoke a method in an object on

another system.

2. Web services are platform and language independent. You can develop and use Web services using any language.

3. Web services run on the Web using HTTP. SOAP is a popular protocol for implement- ing Web services.

4. The server needs to make the service available to the client, known as publishing a Web service. Using a Web service from a client is known as consuming a Web service.

5. A client interacts with a Web service through a proxy object. The proxy object facilitates the communication between the client and the Web service.

6. You need to use Java annotation @WebService to annotate a Web service and use annotation @WebMethod to annotate a remote method.

7. A Web service class may have an unlimited number of remote methods.

8. After a Web service is published, you can write a client program to use it. You have to first create a Web client reference. From the reference, you create a proxy object for facilitating communication between a server and a client.

M41_LIAN1878_11_GE_C41.indd 22 12/29/17 6:03 PM

Programming Exercisesg 41-23

9. WSDL stands for Web Service Description Language. A .wsdl file is an XML file that describes the available Web service to the client—i.e., the remote methods, their param- eters and return value types, and so on.

10. The message between the client and the server is described in XML. A SOAP request describes the information that is sent to the Web service and a SOAP response describes the information that is received from the Web service.

11. The objects passed between client and Web service are serialized in XML. Not all object types are supported by SOAP.

12. You can track sessions in Web services using the HttpSession in the same way as in servlets.

Quiz Answer the quiz for this chapter online at the book Companion Website.

programming exerCises *41.1 (Get a score from a database table) Suppose the scores are stored in the Scores

table. The table was created as follows:

create table Scores (name varchar(20), score number, permission boolean);

insert into Scores values (‘John’, 90.5, 1); insert into Scores values (’Michael’, 100, 1); insert into Scores values (’Michelle’, 100, 0);

Revise the findScore method in Listing 41.1, ScoreService.java, to obtain a score for the specified name. Note your program does not need the permission column; ignore it. The next exercise will need the permission column.

*41.2 (Permission to find scores) Revise the preceding exercise so that the find Score method returns –1 if permission is false. Add an another method named getPermission(String name) that returns 1, 0, or –1. The method returns 1 if the student is in the Scores table and permission is true, 0 if the student is in the Scores table and permission is false, and –1 if the student is not in the Scores table.

*41.3 (Compute loan) You can compute a loan payment for a loan with the specified amount, the number of years, and the annual interest rate. Write a Web service with two remote methods for computing monthly payment and total payment. Write a client program that prompts the user to enter the loan amount, the number of years, and the annual interest rate.

*41.4 (Web service visit count) Write a Web service with a method named getCount() that returns the number of the times this method has been invoked from a client. Use a session to store the count variable.

*41.5 (Quiz) The user needs to answer all five questions before clicking the Submit button in the Quiz application in Section 41.6, Web Service Session Tracking. A runtime error will occur if a radio button is not checked. Fix this problem.

M41_LIAN1878_11_GE_C41.indd 23 12/29/17 6:03 PM

Objectives ■■ To know what a 2–4 tree is (§42.1).

■■ To design the Tree24 class that implements the Tree interface (§42.2).

■■ To search an element in a 2–4 tree (§42.3).

■■ To insert an element in a 2–4 tree and know how to split a node (§42.4).

■■ To delete an element from a 2–4 tree and know how to perform transfer and fusion operations (§42.5).

■■ To traverse elements in a 2–4 tree (§42.6).

■■ To implement and test the Tree24 class (§§42.7–42.8).

■■ To analyze the complexity of the 2–4 tree (§42.9).

■■ To use B-trees for indexing large amount of data (§42.10).

2–4 Trees and B-Trees

Chapter

M42_LIAN1878_11_GE_C42.indd 1 12/29/17 6:05 PM

42-2 Chapter 42 2–4 Trees and B-Trees

42.1 Introduction A 2–4 tree, also known as a 2–3–4 tree, is a completely balanced search tree with all leaf nodes appearing on the same level.

In a 2–4 tree, a node may have one, two, or three elements as shown in Figure 42.1. An interior 2-node contains one element and two children. An interior 3-node contains two elements and three children. An interior 4-node contains three elements and four children.

Point Key

Figure 42.1 An interior node of a 2–4 tree has two, three, or four children.

(a) 2-node (b) 3-node (c) 4-node

c1c0

c2c0 c1 c2 c3c0 c1

e0 e1 e2e0 e1

Figure 42.2 A 2–4 tree is a full complete search tree.

15 27 34

29 50 60 703 16 23 24 25

Each child is a sub 2–4 tree, possibly empty. The root node has no parent, and leaf nodes have no children. The elements in the tree are distinct. The elements in a node are ordered such that

E(c0) 6 e0 6 E(c1) 6 e1 6 E(c2) 6 e2 6 E(c3)

where E(ck) denote the elements in ck . Figure 42.2 shows an example of a 2–4 tree. ck is called the left subtree of ek and ck + 1 is called the right subtree of ek .

In a binary tree, each node contains one element. A 2–4 tree tends to be shorter than a corresponding binary search tree, since a 2–4 tree node may contain two or three elements.

Pedagogical Note Run from http://liveexample.pearsoncmg.com/dsanimation/24Tree.html to see how a 2–4 tree works, as shown in Figure 42.3.

42.2 Designing Classes for 2–4 Trees The Tree24 class defines a 2–4 tree and provides methods for searching, inserting, and deleting elements.

The Tree24 class can be designed by implementing the Tree interface, as shown in Figure 42.4. The Tree interface was defined in Listing 27.3, Tree.java. The Tree24Node class defines tree nodes. The elements in the node are stored in a list named elements and the links to the child nodes are stored in a list named child, as shown in Figure 42.5.

Point Key

M42_LIAN1878_11_GE_C42.indd 2 12/29/17 6:05 PM

42.2 Designing Classes for 2–4 Trees 42-3

Figure 42.3 The animation tool enables you to insert, delete, and search elements in a 2–4 tree visually.

Figure 42.4 The Tree24 class implements Tree.

The root of the tree.

The size of the tree.

Creates a default 2-4 tree.

Creates a 2-4 tree from an array of objects.

Returns true if the element is in the tree.

Returns true if the element is added successfully.

Returns true if the element is removed from the tree successfully. Returns true if element e is in the speci�ed node. Returns the next child node to search for e.

Inserts element along with the reference to its right child to a 2- or 3-node. Splits a 4-node u into u and v, inserts e to u or v, and returns the median element. Locates the insertion point of the element in the node.

Deletes the speci�ed element from the node.

Performs a transfer and fusion operation if node u is empty.

Returns a search path that leads to element e.

1 Link

Tree<E>

Tree24<E>

-root: Tree24Node<E>

+size: int

+Tree24()

+Tree24(objects: E[])

+search(e: E): boolean

+insert(e: E): boolean

+delete(e: E): boolean

-matched(e: E, node: TreeNode<E>): boolean

-getChildNode(e: E, node: TreeNode<E>): Tree24Node<E>

-insert23(e: E, rightChildOfe: Tree24Node<E>, node: Tree24Node<E>): void

-split(e: E, rightChildOfe: Tree24Node<E>, u: Tree24Node<E>, v: Tree24Node<E>): E

-locate(e: E, node: Tree24Node<E>): int

-delete(e: E, node: Tree24Node<E>): void

-validate(e: E, u: Tree24Node<E>, path: ArrayList<Tree24Node<E>>): void

-path(e: E): ArrayList<E>

Tree24Node<E>

elements: ArrayList<E> child: ArrayList<Tree24Node<E>>

+Tree24() +Tree24(o: E)

An array list for storing the elements.

An array list for storing the links to the child nodes.

Creates an empty tree node.

Creates a tree node with an initial element.

m 0

M42_LIAN1878_11_GE_C42.indd 3 12/29/17 6:05 PM

42-4 Chapter 42 2–4 Trees and B-Trees

42.2.1 What is a 2–4 tree? What are a 2-node, 3-node, and 4-node? 42.2.2 Describe the data fields in the Tree24 class and those in the Tree24Node class. 42.2.3 What is the minimum number of elements in a 2–4 tree of height 5? What is the

maximum number of elements in a 2–4 tree of height 5?

42.3 Searching an Element Searching an element in a 2–4 tree is similar to searching an element in a binary tree. The difference is that you have to search an element within a node in addition to searching elements along the path.

To search an element in a 2–4 tree, you start from the root and scan down. If an element is not in the node, move to an appropriate subtree. Repeat the process until a match is found or you arrive at an empty subtree. The algorithm is described in Listing 42.1.

Listing 42.1 Searching an Element in a 2–4 tree 1 boolean search(E e) { 2 current = root; // Start from the root 3 4 while (current != null) { 5 if (match(e, current)) { // Element is in the node 6 return true; // Element is found 7 } 8 else { 9 current = getChildNode(e, current); // Search in a subtree 10 } 11 } 12 return false; // Element is not in the tree 13 }

The match(e, current) method checks whether element e is in the current node. The getChildNode(e, current) method returns the root of the subtree for further search. Initially, let current point to the root (line 2). Repeat searching for the element in the current node until current is null (line 4) or the element matches an element in the current node.

42.4 Inserting an Element into a 2–4 tree Inserting an element involves locating a leaf node and inserting the element into the leaf node.

To insert an element e to a 2–4 tree, locate a leaf node in which the element will be inserted. If the leaf node is a 2-node or 3-node, simply insert the element into the node. If the node is a 4-node, inserting a new element would cause an overflow. To resolve overflow, perform a split operation as follows:

■■ Let u be the leaf 4-node in which the element will be inserted and parentOfu be the parent of u, as shown in Figure 42.6(a).

Point Check

Point Key

Figure 42.5 A 2–4 tree node stores the elements and the links to the child nodes in array lists.

child.get(0) child.get(1) child.get(2) child.get(4)child.get(3)

elements.get(0) elements.get(1) elements.get(2) elements.get(3)

M42_LIAN1878_11_GE_C42.indd 4 12/29/17 6:05 PM

42.4 Inserting an Element into a 2–4 tree 42-5

■■ Create a new node named v; move e2 to v.

■■ If e 6 e1 , insert e to u; otherwise insert e to v. Assume e0 6 e 6 e1 , e is inserted into u, as shown in Figure 42.6(b).

■■ Insert e1 along with its right child (i.e., v) to the parent node, as shown in Figure 42.6(b).

Figure 42.6 The splitting operation creates a new node and inserts the median element to its parent.

parentOfu

p0 p1

e0 e1 e2

(a) Before inserting e (b) After inserting e

New child link

u v

p0 e1 p1

e0 e e2

Figure 42.7 Insertion process continues if the parent node is a 4-node.

parentOfu

(a) The parent is a 4-node

New child link

u v

p0 p1 p2

e0 e e2

(b) Inserting e1 into the parent

parentOfu Right child of e1

u ve0 e e2

p0 p1 p2

The parent node is a 3-node in Figure 42.6. So, there is room to insert e to the parent node. What happens if it is a 4-node, as shown in Figure 42.7? This requires that the parent node be split. The process is the same as splitting a leaf 4-node, except that you must also insert the element along with its right child.

The algorithm can be modified as follows:

■■ Let u be the 4-node (leaf or nonleaf) in which the element will be inserted and parentOfu be the parent of u, as shown in Figure 42.8(a).

■■ Create a new node named v, move e2 and its children c2 and c3 to v.

■■ If e 6 e1 , insert e along with its right child link to u; otherwise insert e along with its right child link to v, as shown in Figure 42.6(b), (c), (d) for the cases e0 6 e 6 e1 , e1 6 e 6 e2 , and e2 6 e , respectively.

■■ Insert e1 along with its right child (i.e., v) to the parent node, recursively.

Listing 42.2 gives an algorithm for inserting an element.

Listing 42.2 Inserting an Element to a 2–4 tree 1 public boolean insert(E e) { 2 if (root == null) 3 root = new Tree24Node<E>(e); // Create a new root for element 4 else { 5 Locate leafNode for inserting e 6 insert(e, null, leafNode); // The right child of e is null 7 } 8 9 size++; // Increase size 10 return true; // Element inserted 11 } 12

M42_LIAN1878_11_GE_C42.indd 5 12/29/17 6:05 PM

42-6 Chapter 42 2–4 Trees and B-Trees

13 private void insert(E e, Tree24Node<E> rightChildOfe, 14 Tree24Node<E> u) { 15 if (u is a 2- or 3- node) { // u is a 2- or 3-node 16 insert23(e, rightChildOfe, u); // Insert e to node u 17 } 18 else { // Split a 4-node u 19 Tree24Node<E> v = new Tree24Node<E>(); // Create a new node 20 E median = split(e, rightChildOfe, u, v); // Split u 21 22 if (u == root) { // u is the root 23 root = new Tree24Node<E>(median); // New root 24 root.child.add(u); // u is the left child of median 25 root.child.add(v); // v is the right child of median 26 } 27 else { 28 Get the parent of u, parentOfu; 29 insert(median, v, parentOfu); // Inserting median to parent 30 } 31 } 32 }

The insert(E e, Tree24Node<E> rightChildOfe, Tree24Node<E> u) method inserts element e along with its right child to node u. When inserting e to a leaf node, the right child of e is null (line 6). If the node is a 2- or 3-node, simply insert the element to the node (lines 15–17). If the node is a 4-node, invoke the split method to split the node (line 20). The split method returns the median element. Recursively invoke the insert method to insert the median element to the parent node (line 29). Figure 42.9 shows the steps of inserting elements 34, 3, 50, 20, 15, 16, 25, 27, 29, and 24 into a 2–4 tree.

42.5 Deleting an Element from a 2–4 tree Deleting an element involves locating the node that contains the element and remov- ing the element from the node.

To delete an element from a 2–4 tree, first search the element in the tree to locate the node that contains it. If the element is not in the tree, the method returns false. Let u be the node that contains the element and parentOfu be the parent of u. Consider three cases:

Point Key

Figure 42.8 An interior node may be split to resolve overflow.

parentOfu

c0 c1 c2 c3

p0 p1

e0 e1 e2

(a) Before inserting e (b) After inserting e (e0 < e < e1)

c0 c1 c2 c3

p0 e1 p1

e0 e e2

rightChildOfe

c0 c1 c2 c3

p0 e1 p1

e0 e e2

rightChildOfe (d) After inserting e (e2 < e)

c0 c1 c2 c3

p0 e1 p1

e0 e2 e

rightChildOfe

M42_LIAN1878_11_GE_C42.indd 6 12/29/17 6:05 PM

42.5 Deleting an Element from a 2–4 tree 42-7

Case 1: u is a leaf 3-node or 4-node. Delete e from u. Case 2: u is a leaf 2-node. Delete e from u. Now u is empty. This situation is known as

underflow. To remedy an underflow, consider two subcases: Case 2.1: u’s immediate left or right sibling is a 3- or 4-node. Let the node be w, as shown

in Figure 42.10(a) (assume w is a left sibling of u). Perform a transfer operation that moves an element from parentOfu to u, as shown in Figure 42.10(b), and move an element from w to replace the moved element in parentOfu, as shown in Figure 42.10(c).

Figure 42.9 The tree changes after 34, 3, 50, 20, 15, 16, 25, 27, 29, and 24 are added into an empty tree.

root in null 34 3 34 3 34 50

3 20 50 3 15 20 50

(a) (b) (c) (d) (e) (f)

(h)

15 34

3 5016 20 25

(g)

15 34

3 5016 20

(i)

15 20 34

3 5016 25 27

(j)

15 20 34

3 5016 25 27 29

(k)

3 16

27 34

502924 25

Figure 42.10 The transfer operation fills the empty node u.

(a) u is now empty (b) Move p1 to u (c) Move e1 to replace p1

parentOfu

w u

p0 p1 p2

e0 e1

parentOfu

w up1

p0 p2

e0 e1

parentOfu

w u

p0 e1 p2

e0 p1

Case 2.2: Both u’s immediate left and right siblings are 2-node if they exist (u may have only one sibling). Let the node be w, as shown in Figure 42.11(a) (assume w is a left sibling of u). Perform a fusion operation that discards u and moves an element from parentOfu to w, as shown in Figure 42.11(b). If parentOfu becomes empty, repeat Case 2 recursively to perform a transfer or a fusion on parentOfu.

Figure 42.11 The fusion operation discards the empty node u.

(a) w is a 2-node (b) Move p1 to w

parentOfu

w u

p0 p1 p2

parentOfu

p0 p2

e0 p1

M42_LIAN1878_11_GE_C42.indd 7 12/29/17 6:05 PM

42-8 Chapter 42 2–4 Trees and B-Trees

Case 3: u is a nonleaf node. Find the rightmost leaf node in the left subtree of e. Let this node be w, as shown in Figure 42.12(a). Move the last element in w to replace e in u, as shown in Figure 42.12(b). If w becomes empty, apply a transfer or fusion operation on w.

Listing 42.3 describes the algorithm for deleting an element.

Figure 42.12 An element in the internal node is replaced by an element in a leaf node.

v0 v1w

root

….. …..…..

e0 v1 e2u

v0w

root

….. …..…..

e0 e e2

(a) e is in u (b) Replace e with v1

Listing 42.3 Deleting an Element from a 2–4 tree 1 /** Delete the specified element from the tree */ 2 public boolean delete(E e) { 3 Locate the node that contains the element e 4 if (the node is found) { 5 delete(e, node); // Delete element e from the node 6 size−−; // After one element deleted 7 return true; // Element deleted successfully 8 } 9 10 return false; // Element not in the tree 11 } 12 13 /** Delete the specified element from the node */ 14 private void delete(E e, Tree24Node<E> node) { 15 if (e is in a leaf node) { 16 // Get the path that leads to e from the root 17 ArrayList<Tree24Node<E>> path = path(e); 18 19 Remove e from the node; 20 21 // Check node for underflow along the path and fix it 22 validate(e, node, path); // Check underflow node 23 } 24 else { // e is in an internal node 25 Locate the rightmost node in the left subtree of node u; 26 Get the rightmost element from the rightmost node; 27 28 // Get the path that leads to e from the root 29 ArrayList<Tree24Node<E>> path = path(rightmostElement); 30 31 Replace the element in the node with the rightmost element 32

M42_LIAN1878_11_GE_C42.indd 8 12/29/17 6:05 PM

42.5 Deleting an Element from a 2–4 tree 42-9

33 // Check node for underflow along the path and fix it 34 validate(rightmostElement, rightmostNode, path); 35 } 36 } 37 38 /** Perform a transfer or fusion operation if necessary */ 39 private void validate(E e, Tree24Node<E> u, 40 ArrayList<Tree24Node<E>> path) { 41 for (int i = path.size() − 1; i >= 0; i−−) { 42 if (u is not empty) 43 return; // Done, no need to perform transfer or fusion 44 45 Tree24Node<E> parentOfu = path.get(i − 1); // Get parent of u 46 47 // Check two siblings 48 if (left sibling of u has more than one element) { 49 Perform a transfer on u with its left sibling 50 } 51 else if (right sibling of u has more than one element) { 52 Perform a transfer on u with its right sibling 53 } 54 else if (u has left sibling) { // Fusion with a left sibling 55 Perform a fusion on u with its left sibling 56 u = parentOfu; // Back to the loop to check the parent node 57 } 58 else { // Fusion with right sibling (right sibling must exist) 59 Perform a fusion on u with its right sibling 60 u = parentOfu; // Back to the loop to check the parent node 61 } 62 } 63 }

The delete(E e) method locates the node that contains the element e and invokes the delete(E e, Tree24Node<E> node) method (line 5) to delete the element from the node.

If the node is a leaf node, get the path that leads to e from the root (line 17), delete e from the node (line 19), and invoke validate to check and fix the empty node (line 22). The validate(E e, Tree24Node<E> u, ArrayList<Tree24Node<E>> path) method performs a transfer or fusion operation if the node is empty. Since these operations may cause the parent of node u to become empty, a path is obtained in order to obtain the parents along the path from the root to node u, as shown in Figure 42.13.

If the node is a nonleaf node, locate the rightmost element in the left subtree of the node (lines 25–26), get the path that leads to the rightmost element from the root (line 29), replace

Figure 42.13 The nodes along the path may become empty as result of a transfer and fusion operation.

root

parentOfu

M42_LIAN1878_11_GE_C42.indd 9 12/29/17 6:05 PM

42-10 Chapter 42 2–4 Trees and B-Trees

e in the node with the rightmost element (line 31), and invoke validate to fix the rightmost node if it is empty (line 34).

The validate(E e, Tree24Node<E> u, ArrayList<Tree24Node<E>> path) checks whether u is empty and performs a transfer or fusion operation to fix the empty node. The validate method exits when node is not empty (line 43). Otherwise, consider one of the following cases:

1. If u has a left sibling with more than one element, perform a transfer on u with its left sibling (line 49).

2. Otherwise, if u has a right sibling with more than one element, perform a transfer on u with its right sibling (line 52).

3. Otherwise, if u has a left sibling, perform a fusion on u with its left sibling (line 55) and reset u to parentOfu (line 56).

4. Otherwise, u must have a right sibling. Perform a fusion on u with its right sibling (line 59) and reset u to parentOfu (line 60).

Only one of the preceding cases is executed. Afterward, a new iteration starts to perform a transfer or fusion operation on a new node u if needed. Figure 42.14 shows the steps of deleting elements 20, 15, 3, 6, and 34 that are deleted from a 2–4 tree in Figure 42.9(k).

42.5.1 How do you search an element in a 2–4 tree? 42.5.2 How do you insert an element into a 2–4 tree? 42.5.3 How do you delete an element from a 2–4 tree? 42.5.4 Show the change of a 2–4 tree when inserting 1, 2, 3, 4, 10, 9, 7, 5, 8, and 6 into

it, in this order.

Point Check

(a) Delete 20

3 16

27 34

502924 25

(b) Replace 20 with 16

27 34

502924 25

3 15

27 34

502924 25

3 15

16 34

502924 25

(d) Perform a transfer

3 15

16 34

502924 25

(e) Delete 15 3

16 34

502924 25

(f) Delete 3

M42_LIAN1878_11_GE_C42.indd 10 12/29/17 6:05 PM

42.7 Implementing the Tree24 Class 42-11

42.5.5 For the tree built in the preceding question, show the change of the tree after delet- ing 1, 2, 3, 4, 10, 9, 7, 5, 8, and 6 from it, in this order.

42.5.6 Show the change of a B-tree of order 6 when inserting 1, 2, 3, 4, 10, 9, 7, 5, 8, 6, 17, 25, 18, 26, 14, 52, 63, 74, 80, 19, and 27 into it, in this order.

42.5.7 For the tree built in the preceding question, show the change of the tree after delet- ing 1, 2, 3, 4, 10, 9, 7, 5, and 8, and 6 from it, in this order.

42.6 Traversing Elements in a 2–4 tree You can perform inorder, preorder, and postorder for traversing the elements in a 2–4 tree.

Inorder, preorder, and postorder traversals are useful for 2–4 trees. Inorder traversal visits the elements in increasing order. Preorder traversal visits the elements in the root, then recursively visits the subtrees from the left to right. Postorder traversal visits the subtrees from the left to right recursively, and then the elements in the root.

For example, in the 2–4 tree in Figure 42.9(k), the inorder traversal is 3 15 16 20 24 25 27 29 34 50

The preorder traversal is 20 15 3 16 27 34 24 25 29 50

The postorder traversal is 3 16 1 24 25 29 50 27 34 20

42.7 Implementing the Tree24 Class This section gives the complete implementation for the Tree24 class.

Listing 42.4 gives the complete source code for the Tree24 class.

Point Key

Figure 42.14 The tree changes after 20, 15, 3, 6, and 34 are deleted from a 2–4 tree.

24 34

5029 25

(g) Perform a transfer

502924 25

(i) Perform a fusion

24 34

5029 25

(h) Delete 16

27 34

502924 25

(j) Perform a fusion

(k) Delete 34

502924 25

(l) Replace 34 with 16 (m) Perform a fusion

27 29

5024 25

24 25 29 50

M42_LIAN1878_11_GE_C42.indd 11 12/29/17 6:05 PM

42-12 Chapter 42 2–4 Trees and B-Trees

Listing 42.4 Tree24.java 1 import java.util.ArrayList; 2 3 public class Tree24<E extends Comparable<E>> implements Tree<E> { 4 private Tree24Node<E> root; 5 private int size; 6 7 /** Create a default 2–4 tree */ 8 public Tree24() { 9 } 10 11 /** Create a 2–4 tree from an array of objects */ 12 public Tree24(E[] elements) { 13 for (int i = 0; i < elements.length; i++) 14 insert(elements[i]); 15 } 16 17 @Override /* Search an element in the tree */ 18 public boolean search(E e) { 19 Tree24Node<E> current = root; // Start from the root 20 21 while (current != null) { 22 if (matched(e, current)) { // Element is in the node 23 return true; // Element found 24 } 25 else { 26 current = getChildNode(e, current); // Search in a subtree 27 } 28 } 29 30 return false; // Element is not in the tree 31 } 32 33 /** Return true if the element is found in this node */ 34 private boolean matched(E e, Tree24Node<E> node) { 35 for (int i = 0; i < node.elements.size(); i++) 36 if (node.elements.get(i).equals(e)) 37 return true; // Element found 38 39 return false; // No match in this node 40 } 41 42 /** Locate a child node to search element e */ 43 private Tree24Node<E> getChildNode(E e, Tree24Node<E> node) { 44 if (node.child.size() == 0) 45 return null; // node is a leaf 46 47 int i = locate(e, node); // Locate the insertion point for e 48 return node.child.get(i); // Return the child node 49 } 50 51 @Override /** Insert element e into the tree 52 * Return true if the element is inserted successfully 53 */ 54 public boolean insert(E e) { 55 if (root == null) 56 root = new Tree24Node<E>(e); // Create a new root for element 57 else { 58 // Locate the leaf node for inserting e 59 Tree24Node<E> leafNode = null;

M42_LIAN1878_11_GE_C42.indd 12 12/29/17 6:05 PM

42.7 Implementing the Tree24 Class 42-13

60 Tree24Node<E> current = root; 61 while (current != null) 62 if (matched(e, current)) { 63 return false; // Duplicate element found, nothing inserted 64 } 65 else { 66 leafNode = current; 67 current = getChildNode(e, current); 68 } 69 70 // Insert the element e into the leaf node 71 insert(e, null, leafNode); // The right child of e is null 72 } 73 74 size++; // Increase size 75 return true; // Element inserted 76 } 77 78 /** Insert element e into node u */ 79 private void insert(E e, Tree24Node<E> rightChildOfe, 80 Tree24Node<E> u) { 81 // Get the search path that leads to element e 82 ArrayList<Tree24Node<E>> path = path(e); 83 84 for (int i = path.size() − 1; i >= 0; i−−) { 85 if (u.elements.size() < 3) { // u is a 2-node or 3-node 86 insert23(e, rightChildOfe, u); // Insert e to node u 87 break; // No further insertion to u's parent needed 88 } 89 else { 90 Tree24Node<E> v = new Tree24Node<E>(); // Create a new node 91 E median = split(e, rightChildOfe, u, v); // Split u 92 93 if (u == root) { 94 root = new Tree24Node<E>(median); // New root 95 root.child.add(u); // u is the left child of median 96 root.child.add(v); // v is the right child of median 97 break; // No further insertion to u's parent needed 98 } 99 else { 100 // Use new values for the next iteration in the for loop 101 e = median; // Element to be inserted to parent 102 rightChildOfe = v; // Right child of the element 103 u = path.get(i − 1); // New node to insert element 104 } 105 } 106 } 107 } 108 109 /** Insert element to a 2- or 3- and return the insertion point */ 110 private void insert23(E e, Tree24Node<E> rightChildOfe, 111 Tree24Node<E> node) { 112 int i = this.locate(e, node); // Locate where to insert 113 node.elements.add(i, e); // Insert the element into the node 114 if (rightChildOfe != null) 115 node.child.add(i + 1, rightChildOfe); // Insert the child link 116 } 117 118 /** Split a 4-node u into u and v and insert e to u or v */ 119 private E split(E e, Tree24Node<E> rightChildOfe,

M42_LIAN1878_11_GE_C42.indd 13 12/29/17 6:05 PM

42-14 Chapter 42 2–4 Trees and B-Trees

120 Tree24Node<E> u, Tree24Node<E> v) { 121 // Move the last element in node u to node v 122 v.elements.add(u.elements.remove(2)); 123 E median = u.elements.remove(1); 124 125 // Split children for a nonleaf node 126 // Move the last two children in node u to node v 127 if (u.child.size() = 0) { 128 v.child.add(u.child.remove(2)); 129 v.child.add(u.child.remove(2)); 130 } 131 132 // Insert e into a 2- or 3- node u or v. 133 if (e.compareTo(median) < 0) 134 insert23(e, rightChildOfe, u); 135 else 136 insert23(e, rightChildOfe, v); 137 138 return median; // Return the median element 139 } 140 141 /** Return a search path that leads to element e */ 142 private ArrayList<Tree24Node<E>= path(E e) { 143 ArrayList<Tree24Node<E>= list = new ArrayList<Tree24Node<E>=(); 144 Tree24Node<E> current = root; // Start from the root 145 146 while (current != null) { 147 list.add(current); // Add the node to the list 148 if (matched(e, current)) { 149 break; // Element found 150 } 151 else { 152 current = getChildNode(e, current); 153 } 154 } 155 156 return list; // Return an array of nodes 157 } 158 159 @Override /** Delete the specified element from the tree */ 160 public boolean delete(E e) { 161 // Locate the node that contains the element e 162 Tree24Node<E> node = root; 163 while (node != null) 164 if (matched(e, node)) { 165 delete(e, node); // Delete element e from node 166 size−−; // After one element deleted 167 return true; // Element deleted successfully 168 } 169 else { 170 node = getChildNode(e, node); 171 } 172 173 return false; // Element not in the tree 174 } 175 176 /** Delete the specified element from the node */ 177 private void delete(E e, Tree24Node<E> node) { 178 if (node.child.size() == 0) { // e is in a leaf node 179 // Get the path that leads to e from the root 180 ArrayList<Tree24Node<E>> path = path(e);

M42_LIAN1878_11_GE_C42.indd 14 12/29/17 6:05 PM

42.7 Implementing the Tree24 Class 42-15

181 182 node.elements.remove(e); // Remove element e 183 184 if (node == root) { // Special case 185 if (node.elements.size() == 0) 186 root = null; // Empty tree 187 return; // Done 188 } 189 190 validate(e, node, path); // Check underflow node 191 } 192 else { // e is in an internal node 193 // Locate the rightmost node in the left subtree of the node 194 int index = locate(e, node); // Index of e in node 195 Tree24Node<E> current = node.child.get(index); 196 while (current.child.size() > 0) { 197 current = current.child.get(current.child.size() − 1); 198 } 199 E rightmostElement = 200 current.elements.get(current.elements.size() − 1); 201 202 // Get the path that leads to e from the root 203 ArrayList<Tree24Node<E>= path = path(rightmostElement); 204 205 // Replace the deleted element with the rightmost element 206 node.elements.set(index, current.elements.remove( 207 current.elements.size() − 1)); 208 209 validate(rightmostElement, current, path); // Check underflow 210 } 211 } 212 213 /** Perform transfer and confusion operations if necessary */ 214 private void validate(E e, Tree24Node<E> u, 215 ArrayList<Tree24Node<E>> path) { 216 for (int i = path.size() − 1; u.elements.size() == 0; i−−) { 217 Tree24Node<E> parentOfu = path.get(i − 1); // Get parent of u 218 int k = locate(e, parentOfu); // Index of e in the parent node 219 220 // Check two siblings 221 if (k > 0 && parentOfu.child.get(k − 1).elements.size() > 1) { 222 leftSiblingTransfer(k, u, parentOfu); 223 } 224 else if (k + 1 < parentOfu.child.size() && 225 parentOfu.child.get(k + 1).elements.size() > 1) { 226 rightSiblingTransfer(k, u, parentOfu); 227 } 228 else if (k − 1 == 0) { // Fusion with a left sibling 229 // Get left sibling of node u 230 Tree24Node<E> leftNode = parentOfu.child.get(k − 1); 231 232 // Perform a fusion with left sibling on node u 233 leftSiblingFusion(k, leftNode, u, parentOfu); 234 235 // Done when root becomes empty 236 if (parentOfu == root && parentOfu.elements.size() == 0) { 237 root = leftNode; 238 break; 239 } 240 241 u = parentOfu; // Back to the loop to check the parent node

M42_LIAN1878_11_GE_C42.indd 15 12/29/17 6:05 PM

42-16 Chapter 42 2–4 Trees and B-Trees

242 } 243 else { // Fusion with right sibling (right sibling must exist) 244 // Get left sibling of node u 245 Tree24Node<E> rightNode = parentOfu.child.get(k + 1); 246 247 // Perform a fusion with right sibling on node u 248 rightSiblingFusion(k, rightNode, u, parentOfu); 249 250 // Done when root becomes empty 251 if (parentOfu == root && parentOfu.elements.size() == 0) { 252 root = rightNode; 253 break; 254 } 255 256 u = parentOfu; // Back to the loop to check the parent node 257 } 258 } 259 } 260 261 /** Locate the insertion point of the element in the node */ 262 private int locate(E o, Tree24Node<E> node) { 263 for (int i = 0; i < node.elements.size(); i++) { 264 if (o.compareTo(node.elements.get(i)) <= 0) { 265 return i; 266 } 267 } 268 269 return node.elements.size(); 270 } 271 272 /** Perform a transfer with a left sibling */ 273 private void leftSiblingTransfer(int k, 274 Tree24Node<E> u, Tree24Node<E> parentOfu) { 275 // Move an element from the parent to u 276 u.elements.add(0, parentOfu.elements.get(k − 1)); 277 278 // Move an element from the left node to the parent 279 Tree24Node<E> leftNode = parentOfu.child.get(k − 1); 280 parentOfu.elements.set(k − 1, 281 leftNode.elements.remove(leftNode.elements.size() − 1)); 282 283 // Move the child link from left sibling to the node 284 if (leftNode.child.size() > 0) 285 u.child.add(0, leftNode.child.remove( 286 leftNode.child.size() − 1)); 287 } 288 289 /** Perform a transfer with a right sibling */ 290 private void rightSiblingTransfer(int k, 291 Tree24Node<E> u, Tree24Node<E> parentOfu) { 292 // Transfer an element from the parent to u 293 u.elements.add(parentOfu.elements.get(k)); 294 295 // Transfer an element from the right node to the parent 296 Tree24Node<E> rightNode = parentOfu.child.get(k + 1); 297 parentOfu.elements.set(k, rightNode.elements.remove(0)); 298 299 // Move the child link from right sibling to the node 300 if (rightNode.child.size() > 0) 301 u.child.add(rightNode.child.remove(0)); 302 }

M42_LIAN1878_11_GE_C42.indd 16 12/29/17 6:05 PM

42.7 Implementing the Tree24 Class 42-17

303 304 /** Perform a fusion with a left sibling */ 305 private void leftSiblingFusion(int k, Tree24Node<E> leftNode, 306 Tree24Node<E> u, Tree24Node<E> parentOfu) { 307 // Transfer an element from the parent to the left sibling 308 leftNode.elements.add(parentOfu.elements.remove(k − 1)); 309 310 // Remove the link to the empty node 311 parentOfu.child.remove(k); 312 313 // Adjust child links for nonleaf node 314 if (u.child.size() > 0) 315 leftNode.child.add(u.child.remove(0)); 316 } 317 318 /** Perform a fusion with a right sibling */ 319 private void rightSiblingFusion(int k, Tree24Node<E> rightNode, 320 Tree24Node<E> u, Tree24Node<E> parentOfu) { 321 // Transfer an element from the parent to the right sibling 322 rightNode.elements.add(0, parentOfu.elements.remove(k)); 323 324 // Remove the link to the empty node 325 parentOfu.child.remove(k); 326 327 // Adjust child links for nonleaf node 328 if (u.child.size() > 0) 329 rightNode.child.add(0, u.child.remove(0)); 330 } 331 332 /** Get the number of nodes in the tree */ 333 public int getSize() { 334 return size; 335 } 336 337 /** Preorder traversal from the root */ 338 public void preorder() { 339 preorder(root); 340 } 341 342 /** Preorder traversal from a subtree */ 343 private void preorder(Tree24Node<E> root) { 344 if (root == null)return; 345 for (int i = 0; i < root.elements.size(); i++) 346 System.out.print(root.elements.get(i) + " "); 347 348 for (int i = 0; i < root.child.size(); i++) 349 preorder(root.child.get(i)); 350 } 351 352 /** Inorder traversal from the root*/ 353 public void inorder() { 354 // Left as exercise 355 } 356 357 /** Postorder traversal from the root */ 358 public void postorder() { 359 // Left as exercise 360 } 361 362 /** Return true if the tree is empty */ 363 public boolean isEmpty() {

M42_LIAN1878_11_GE_C42.indd 17 12/29/17 6:05 PM

42-18 Chapter 42 2–4 Trees and B-Trees

364 return root == null; 365 } 366 367 @Override /** Remove all elements from the tree */ 368 public void clear() { 369 root = null; 370 size = 0; 371 } 372 373 /** Return an iterator to traverse elements in the tree */ 374 public java.util.Iterator iterator() { 375 // Left as exercise 376 return null; 377 } 378 379 /** Define a 2–4 tree node */ 380 protected static class Tree24Node<E extends Comparable<E>> { 381 // elements has maximum three values 382 ArrayList<E> elements = new ArrayList<E>(3); 383 // Each has maximum four childres 384 ArrayList<Tree24Node<E>> child 385 = new ArrayList<Tree24Node<E>>(4); 386 387 /** Create an empty Tree24 node */ 388 Tree24Node() { 389 } 390 391 /** Create a Tree24 node with an initial element */ 392 Tree24Node(E o) { 393 elements.add(o); 394 } 395 } 396 }

The Tree24 class contains the data fields root and size (lines 4–5). root references the root node and size stores the number of elements in the tree.

The Tree24 class has two constructors: a no-arg constructor (lines 8–9) that constructs an empty tree and a constructor that creates an initial Tree24 from an array of elements (lines 12–15).

The search method (lines 18–31) searches an element in the tree. It returns true (line 23) if the element is in the tree and returns false if the search arrives at an empty subtree (line 30).

The matched(e, node) method (lines 34–40) checks where the element e is in the node. The getChildNode(e, node) method (lines 43–49) returns the root of a subtree where

e should be searched. The insert(E e) method inserts an element in a tree (lines 54–76). If the tree is empty,

a new root is created (line 56). The method locates a leaf node in which the element will be inserted and invokes insert(e, null, leafNode) to insert the element (line 71).

The insert(e, rightChildOfe, u) method inserts an element into node u (lines 79–107). The method first invokes path(e) (line 82) to obtain a search path from the root to node u. Each iteration of the for loop considers u and its parent parentOfu (lines 84–106). If u is a 2-node or 3-node, invoke insert23(e, rightChildOfe, u) to insert e and its child link rightChildOfe into u (line 86). No split is needed (line 87). Otherwise, create a new node v (line 90) and invoke split(e, rightChildOfe, u, v) (line 91) to split u into u and v. The split method inserts e into either u and v and returns the median in the original u. If u is the root, create a new root to hold median, and set u and v as the left and right children for median (lines 95–96). If u is not the root, insert median to parentOfu in the next iteration (lines 101–103).

M42_LIAN1878_11_GE_C42.indd 18 12/29/17 6:05 PM

42.8 Testing the Tree24 Class 42-19

The insert23(e, rightChildOfe, node) method inserts e along with the reference to its right child into the node (lines 110–116). The method first invokes locate(e, node) (line 112) to locate an insertion point, then insert e into the node (line 113). If rightChildOfe is not null, it is inserted into the child list of the node (line 115).

The split(e, rightChildOfe, u, v) method splits a 4-node u (lines 119-139). This is accomplished as follows: (1) move the last element from u to v and remove the median element from u (lines 122–123); (2) move the last two child links from u to v (lines 127–130) if u is a nonleaf node; (3) if e < median, insert e into u; otherwise, insert e into v (lines 133–136); and (4) return median (line 138).

The path(e) method returns an ArrayList of nodes searched from the root in order to locate e (lines 142–157). If e is in the tree, the last node in the path contains e. Otherwise, the last node is where e should be inserted.

The delete(E e) method deletes an element from the tree (lines 160–174). The method first locates the node that contains e and invokes delete(e, node) to delete e from the node (line 165). If the element is not in the tree, return false (line 173).

The delete(e, node) method deletes an element from node u (lines 177–211). If the node is a leaf node, obtain the path that leads to e (line 180), delete e (line 182), set root to null if the tree becomes empty (lines 184–188), and invoke validate to apply transfer and fusion operation on empty nodes (line 190). If the node is a nonleaf node, locate the rightmost element (lines 194–200), obtain the path that leads to e (line 203), replace e with the rightmost element (lines 206–207), and invoke validate to apply transfer and fusion operations on empty nodes (line 209).

The validate(e, u, path) method ensures that the tree is a valid 2–4 tree (lines 214–259). The for loop terminates when u is not empty (line 216). The loop body is executed to fix the empty node u by performing a transfer or fusion operation. If a left sibling with more than one element exists, perform a transfer on u with the left sibling (line 222). Otherwise, if a right sibling with more than one element exists, perform a transfer on u with the left sibling (line 226). Otherwise, if a left sibling exists, perform a fusion on u with the left sibling (lines 230–239), and validate parentOfu in the next loop iteration (line 241). Otherwise, perform a fusion on u with the right sibling.

The locate(e, node) method locates the index of e in the node (lines 262–270). The leftSiblingTransfer(k, u, parentOfu) method performs a transfer on u with

its left sibling (lines 273–287). The rightSiblingTransfer(k, u, parentOfu) method performs a transfer on u with its right sibling (lines 290–302). The leftSiblingFusion(k, leftNode, u, parentOfu) method performs a fusion on u with its left sibling leftNode (lines 305–316). The rightSiblingFusion(k, rightNode, u, parentOfu) method performs a fusion on u with its right sibling rightNode (lines 319–330).

The preorder() method displays all the elements in the tree in preorder (lines 338–350). The inner class Tree24Node defines a class for a node in the tree (lines 374–389).

42.8 Testing the Tree24 Class This section writes a test program for using the Tree24 class.

Listing 42.5 gives a test program. The program creates a 2–4 tree and inserts elements in lines 6–20, and deletes elements in lines 22–56.

Listing 42.5 TestTree24.java 1 public class TestTree24 { 2 public static void main(String[] args) { 3 // Create a 2–4 tree 4 Tree24<Integer> tree = new Tree24<Integer>(); 5

Point Key

M42_LIAN1878_11_GE_C42.indd 19 12/29/17 6:05 PM

42-20 Chapter 42 2–4 Trees and B-Trees

6 tree.insert(34); 7 tree.insert(3); 8 tree.insert(50); 9 tree.insert(20); 10 tree.insert(15); 11 tree.insert(16); 12 tree.insert(25); 13 tree.insert(27); 14 tree.insert(29); 15 tree.insert(24); 16 System.out.print("\nAfter inserting 24:"); 17 printTree(tree); 18 tree.insert(23); 19 tree.insert(22); 20 tree.insert(60); 21 tree.insert(70); 22 System.out.print("\nAfter inserting 70:"); 23 printTree(tree); 24 25 tree.delete(34); 26 System.out.print("\nAfter deleting 34:"); 27 printTree(tree); 28 29 tree.delete(25); 30 System.out.print("\nAfter deleting 25:"); 31 printTree(tree); 32 33 tree.delete(50); 34 System.out.print("\nAfter deleting 50:"); 35 printTree(tree); 36 37 tree.delete(16); 38 System.out.print("\nAfter deleting 16:"); 39 printTree(tree); 40 41 tree.delete(3); 42 System.out.print("\nAfter deleting 3:"); 43 printTree(tree); 44 45 tree.delete(15); 46 System.out.print("\nAfter deleting 15:"); 47 printTree(tree); 48 } 49 50 public static <E extends Comparable<E>> 51 void printTree(Tree<E> tree) { 52 // Traverse tree 53 System.out.print("\nPreorder: "); 54 tree.preorder(); 55 System.out.print("\nThe number of nodes is " + tree.getSize()); 56 System.out.println(); 57 } 58 }

M42_LIAN1878_11_GE_C42.indd 20 12/29/17 6:05 PM

42.10 B-Tree 42-21

After inserting 24:

Preorder: 20 15 3 16 27 34 24 25 29 50 The number of nodes is 10

After inserting 70: Preorder: 20 15 3 16 24 27 34 22 23 25 29 50 60 70 The number of nodes is 14

After deleting 34: Preorder: 20 15 3 16 24 27 50 22 23 25 29 60 70 The number of nodes is 13

After deleting 25: Preorder: 20 15 3 16 23 27 50 22 24 29 60 70 The number of nodes is 12

After deleting 50: Preorder: 20 15 3 16 23 27 60 22 24 29 70 The number of nodes is 11

After deleting 16: Preorder: 23 20 3 15 22 27 60 24 29 70 The number of nodes is 10

After deleting 3: Preorder: 23 20 15 22 27 60 24 29 70 The number of nodes is 9

After deleting 15: Preorder: 27 23 20 22 24 60 29 70 The number of nodes is 8

Figure 42.15 shows how the tree evolves as elements are added. After 34, 3, 50, 20, 15, 16, 25, 27, 29, and 24 are added to the tree, it is as shown in Figure 42.15(a). After inserting 23, 22, 60, and 70, the tree is as shown in Figure 42.15(b). After deleting 34, the tree is as shown in Figure 42.15(c). After deleting 25, the tree is as shown in Figure 42.15(d). After deleting 50, the tree is as shown in Figure 42.15(e). After deleting 16, the tree is as shown in Figure 42.15(f). After deleting 3, the tree is as shown in Figure 42.15(g). After deleting 15, the tree is as shown in Figure 42.15(h).

42.9 Time-Complexity Analysis Search, insertion, and deletion operations take O(logn) time in a 2–4 tree.

Since a 2–4 tree is a completely balanced binary tree, its height is at most O(log n). The search, insert, and delete methods operate on the nodes along a path in the tree. It takes a constant time to search an element within a node. So, the search method takes O(log n) time. For the insert method, the time for splitting a node takes a constant time. So, the insert method takes O(log n) time. For the delete method, it takes a constant time to perform a transfer and fusion operation. So, the delete method takes O(log n) time.

42.10 B-Tree A B-tree is a generalization of a 2–4 tree.

So far we assume the entire data set is stored in main memory. What if the data set is too large and cannot fit in the main memory, as in the case with most databases, where data is stored on disks? Suppose you use an AVL tree to organize a million records in a database table. To find a record, the average number of nodes traversed is log 2 1 ,0 0 0 ,0 0 0 ≈ 2 0 . This is fine

Point Key

M42_LIAN1878_11_GE_C42.indd 21 12/29/17 6:05 PM

42-22 Chapter 42 2–4 Trees and B-Trees

(a) After inserting 34, 3, 50, 20, 15, 16, 25, 27, 29, and 24, in this order

(b) After inserting 23, 22, 60, and 70

16 24 25

27 34

16 22 23

24 27 34

50 60 70

25 29

16 22 23

24 27 50

60 70

25 29

(d) After deleting 25

16 22

23 27 50

60 70

24 29

(e) After deleting 50

16 22

23 27 60

24 29

if all nodes are stored in main memory. However, for nodes stored on a disk, this means 20 disk reads. Disk I/O is expensive, and it is thousands of times slower than memory access. To improve performance, we need to reduce the number of disk I/Os. An efficient data structure for performing search, insertion, and deletion for data stored on secondary storage such as hard disks is the B-tree, which is a generalization of the 2–4 tree.

M42_LIAN1878_11_GE_C42.indd 22 12/29/17 6:05 PM

42.10 B-Tree 42-23

Figure 42.16 In a B-tree of order 6, each node except the root may contain between 2 and 5 keys.

3 6

8 13

9 10 15 16

18 43

20 26 27 31 32 35 36 37 45 46 47 49 50 75 76 77 78 79

28 33 48 53 65

59 60

Figure 42.15 The tree evolves as elements are inserted and deleted.

(f) After deleting 16

3 15

27 60

24 29

(g) After deleting 3

27 60

24 29

(h) After deleting 15

20 22

A B-tree of order d is defined as follows:

1. Each node except the root contains between <d/2= - 1 and d - 1 keys. 2. The root may contain up to d - 1 keys.

3. A nonleaf node with k keys has k + 1 children.

4. All leaf nodes have the same depth.

Figure 42.16 shows a B-tree of order 6. For simplicity, we use integers to represent keys. Each key is associated with a pointer that points to the actual record in the database. For sim- plicity, the pointers to the records in the database are omitted in the figure.

M42_LIAN1878_11_GE_C42.indd 23 12/29/17 6:05 PM

42-24 Chapter 42 2–4 Trees and B-Trees

The basic unit of the IO operations on a disk is a block. When you read data from a disk, the whole block that contains the data is read. You should choose an appropriate order d so that a node can fit in a single disk block. This will minimize the number of disk IOs.

A 2–4 tree is actually a B-tree of order 4. The techniques for insertion and deletion in a 2–4 tree can be easily generalized for a B-tree.

Inserting a key to a B-tree is similar to what was done for a 2–4 tree. First, locate the leaf node in which the key will be inserted. Insert the key to the node. After the insertion, if the leaf node has d keys, an overflow occurs. To resolve overflow, perform a split operation similar to the one used in a 2–4 tree, as follows:

Let u denote the node needed to be split and let m denote the median key in the node. Create a new node and move all keys greater than m to this new node. Insert m to the parent node of u. Now u becomes the left child of m and v becomes the right child of m, as shown in Figure 42.18. If inserting m into the parent node of u causes an overflow, repeat the same split process on the parent node.

Figure 42.18 (a) After inserting a new key to node u. (b) The median key k p is inserted to parentOfu.

k1 k2 … kd

parentOfu

…

k1 … kp–1

parentOfu

… kp …

kp+1 …kd new node

Figure 42.19 The transfer operation transfers a key from the parentOfu to u and transfers a key from u ’s sibling parentOfu.

… j

... i …

w u

parentOfu

… k … j

.. . …

w u

parentOfu

i … k …

... j…

w u

parentOfu

i … k

(a) Before a transfer is performed (b) Key i moved to node u (c) Key j moved to parentOfu

Figure 42.17 The keys in the left (right) subtree of key k i are less than (greater than) k i.

k1 k2 … ki …

left subtree right subtree

Note that a B-tree is a search tree. The keys in each node are placed in increasing order. Each key in an interior node has a left subtree and a right subtree, as shown in Figure 42.17. All keys in the left subtree are less than the key in the parent node, and all keys in the right subtree are greater than the key in the parent node.

A key k can be deleted from a B-tree in the same way as in a 2–4 tree. First locate the node u that contains the key. Consider two cases:

Case 1: If u is a leaf node, remove the key from u. After the removal, if u has less than <d/2= - 1 keys, an underflow occurs. To remedy an underflow, perform a transfer with a sibling

M42_LIAN1878_11_GE_C42.indd 24 12/29/17 6:05 PM

42.10 B-Tree 42-25

Case 2: u is a nonleaf node. Find the rightmost leaf node in the left subtree of k. Let this node be w, as shown in Figure 42.21(a). Move the last key in w to replace k in u, as shown in Figure 42.21(b). If w becomes underflow, apply a transfer or fusion operation on w.

Figure 42.20 The fusion operation moves key i from the parentOfu u to w and moves all keys in u to w .

… j

... i …

w u

parentOfu

… k … j i … k

.. . …

parentOfu

(a) Before a fusion is performed (b) After a fusion is performed

Figure 42.21 A key in the internal node is replaced by an element in a leaf node.

… iw

root

….. …..…..

… i …u

…w

root

….. …..…..

(a) Key is in u (b) Replace key k with key i

… k …

The performance of a B-tree depends on the number of disk IOs (i.e., the number of nodes accessed). The number of nodes accessed for search, insertion, and deletion opera- tions depends on the height of the tree. In the worst case, each node contains <d/2= -1 keys. So, the height of the tree is log < d / 2 = n , where n is the number of keys. In the best case, each node contains d - 1 keys. So, the height of the tree is log d n . Consider a B-tree of order 12 for 10 million keys. The height of the tree is between log6 10,000,000 ≈ 7 and log 1 2 1 0 ,0 0 0 ,0 0 0 ≈ 9 . So, for search, insertion, and deletion operations, the maximum number of nodes visited is 42. If you use an AVL tree, the maximum number of nodes visited is log 2 1 0 ,0 0 0 ,0 0 0 ≈ 2 4 .

Key Terms

2–3–4 tree 42-2 2–4 tree 42-2 2-node 42-2 3-node 42-2 4-node 42-2

B-tree 42-11 fusion operation 42-7 split operation 42-4 transfer operation 42-7

w of u that has more than <d/2= - 1 keys if such sibling exists, as shown in Figure 42.19. Otherwise, perform a fusion with a sibling w of u, as shown in Figure 42.20.

M42_LIAN1878_11_GE_C42.indd 25 12/29/17 6:05 PM

42-26 Chapter 42 2–4 Trees and B-Trees

ChapTer summary 1. A 2–4 tree is a completely balanced search tree. In a 2–4 tree, a node may have one,

two, or three elements.

2. Searching an element in a 2–4 tree is similar to searching an element in a binary tree. The difference is that you have searched an element within a node.

3. To insert an element to a 2–4 tree, locate a leaf node in which the element will be inserted. If the leaf node is a 2- or 3-node, simply insert the element into the node. If the node is a 4-node, split the node.

4. The process of deleting an element from a 2–4 tree is similar to that of deleting an ele- ment from a binary tree. The difference is that you have to perform transfer or fusion operations for empty nodes.

5. The height of a 2–4 tree is O (logn). So, the time complexities for the search, insert, and delete methods are O (logn).

6. A B-tree is a generalization of the 2–4 tree. Each node in a B-tree of order d can have between <d/2= -1 and d - 1 keys except the root. 2–4 trees are flatter than AVL trees and B-trees are flatter than 2–4 trees. B-trees are efficient for creating indexes for data in database systems where large amounts of data are stored on disks.

Quiz Answer the quiz for this chapter online at the book Companion Website.

programming exerCises *42.1 (Implement inorder) The inorder method in Tree24 is left as an exercise.

Implement it.

42.2 (Implement postorder) The postorder method in Tree24 is left as an exercise. Implement it.

42.3 (Implement iterator) The iterator method in Tree24 is left as an exercise. Implement it to iterate the elements using inorder.

*42.4 (Display a 2–4 tree graphically) Write a GUI program that displays a 2–4 tree. ***42.5 (2–4 tree animation) Write a GUI program that animates the 2–4 tree insert,

delete, and search methods, as shown in Figure 42.4.

**42.6 (Parent reference for Tree24) Redefine Tree24Node to add a reference to a node’s parent, as shown below:

Tree24Node<E>

elements: ArrayList<E>

child: ArrayList<Tree24Node<E>>

parent: Tree24Node<E>

An array list for storing the elements.

An array list for storing the links to the child nodes.

Refers to the parent of this node.

+Tree24()

+Tree24(o: E)

Creates an empty tree node.

Creates a tree node with an initial element.

M42_LIAN1878_11_GE_C42.indd 26 12/29/17 6:05 PM

Programming Exercises 42-27

Add the following two new methods in Tree24:

public Tree24Node<E> getParent(Tree24Node<E> node) Returns the parent for the specified node. public ArrayList<Tree24Node<E>> getPath(Tree24Node<E> node) Returns the path from the specified node to the root in an array list.

Write a test program that adds numbers 1, 2, ..., 100 to the tree and displays the paths for all leaf nodes.

***42.7 (The BTree class) Design and implement a class for B-trees.

M42_LIAN1878_11_GE_C42.indd 27 12/29/17 6:05 PM

A01_GIPE3898_09_SE_FM.indd 3 12/20/17 5:56 PM

This page intentionally left blank

Objectives ■■ To know what a red-black tree is (§43.1).

■■ To convert a red-black tree to a 2–4 tree and vice versa (§43.2).

■■ To design the RBTree class that extends the BST class (§43.3).

■■ To insert an element in a red-black tree and resolve the double-red violation if necessary (§43.4).

■■ To delete an element from a red-black tree and resolve the double-black problem if necessary (§43.5).

■■ To implement and test the RBTree class (§§43.6–43.7).

■■ To compare the performance of AVL trees, 2–4 trees, and RBTree (§43.8).

Red-Black Trees

Chapter

M43_LIAN1878_11_GE_C43.indd 1 12/29/17 6:06 PM

43-2 Chapter 43 Red-Black Trees

43.1 Introduction A red-black tree is a balanced binary search tree derived from a 2–4 tree. A red-black tree corresponds to a 2-4 tree.

Each node in a red-black tree has a color attribute red or black, as shown in Figure 43.1(a). A node is called external if its left or right subtree is empty. Note that a leaf node is exter- nal, but an external node is not necessarily a leaf node. For example, node 25 is external, but it is not a leaf. The black depth of a node is defined as the number of black nodes in a path from the node to the root. For example, the black depth of node 25 is 2, and that of node 27 is 2.

Note The red nodes appear in blue in the text.

A red-black tree has the following properties:

1. The root is black.

2. Two adjacent nodes cannot be both red.

3. All external nodes have the same black depth.

The red-black tree in Figure 43.1(a) satisfies all three properties. A red-black tree can be converted to a 2-4 tree, and vice versa. Figure 43.1(b) shows an equivalent 2-4 tree for the red-black tree in Figure 43.1(a).

Figure 43.1 A red-black tree can be represented using a 2-4 tree, and vice versa.

15 34

3 16 25

50 15 20 34

3 25 27 5016

(a) A red-black tree (b) A 2-4 tree

Point Key

43.2 Conversion between Red-Black Trees and 2-4 Trees This section discusses the correspondence between a red-black tree and a 2-4 tree.

You can design insertion and deletion algorithms for red-black trees without having knowledge of 2-4 trees. However, the correspondence between red-black trees and 2-4 trees provides use- ful intuition about the structure of red-black trees and operations. For this reason, this section discusses the correspondence between these two types of trees.

To convert a red-black tree to a 2-4 tree, simply merge every red node with its parent to create a 3-node or a 4-node. For example, the red nodes 15 and 34 are merged to their parent to create a 4-node, and the red node 27 is merged to its parent to create a 3-node, as shown in Figure 43.1(b).

Point Key

M43_LIAN1878_11_GE_C43.indd 2 12/29/17 6:06 PM

43.2 Conversion between Red-Black Trees and 2-4 Trees 43-3

To convert a 2-4 tree to a red-black tree, perform the following transformations for each node u:

1. If u is a 2-node, color it black, as shown in Figure 43.2(a).

2. If u is a 3-node with element values e0 and e1 , there are two ways to convert it. Either make e0 the parent of e1 or make e1 the parent of e0 . In any case, color the parent black and the child red, as shown in Figure 43.2(b).

3. If u is a 4-node with element values e0 , e1 , and e2 , make e1 the parent of e0 and e2 . Color e1 black and e0 and e2 red, as shown in Figure 43.2(c).

Figure 43.2 A node in a 2-4 tree can be transformed to nodes in a red-black tree.

2-3-4 Tree Equivalent red-black tree

(a) Converting a 2-node

(b) Converting a 3-node

e0 e1

e0 e1 e2

c2 e0

Figure 43.3 The conversion from a 2-4 tree to a red-black tree is not unique.

3 25 5016

3 27 5016

(a) (b) (c)

3 25 27 5016

Let us apply the transformation for the 2-4 tree in Figure 43.1(b). After transforming the 4-node, the tree is as shown in Figure 43.3(a). After transforming the 3-node, the tree is as shown in Figure 43.3(b). Note the transformation for a 3-node is not unique. Therefore, the conversion from a 2-4 tree to a red-black tree is not unique. After transforming the 3-node, the tree could also be as shown in Figure 43.3(c).

M43_LIAN1878_11_GE_C43.indd 3 12/29/17 6:06 PM

43-4 Chapter 43 Red-Black Trees

You can prove the conversion results in a red-black tree that satisfies all three properties.

Property 1. The root is black.

Proof: If the root of a 2-4 tree is a 2-node, the root of the red-black tree is black. If the root of a 2-4 tree is a 3-node or 4-node, the transformation produces a black parent at the root.

Property 2. Two adjacent nodes cannot be both red.

Proof: Since the parent of a red node is always black, no two adjacent nodes can be both red.

Property 3. All external nodes have the same black depth.

Proof: When you covert a node in a 2-4 tree to red-black tree nodes, you get one black node and zero, one, or two red nodes as its children, depending on whether the original node is a 2-, 3-, or 4-node. Only a leaf 2-4 node may produce external red-black nodes. Since a 2-4 tree is perfectly balanced, the number of black nodes in any path from the root to an external node is the same.

43.2.1 What is a red-black tree? What is an external node? What is black depth? 43.2.2 Describe the properties of a red-black tree. 43.2.3 How do you convert a red-black tree to a 2-4 tree? Is the conversion unique? 43.2.4 How do you convert a 2-4 tree to a red-black tree? Is the conversion unique?

43.3 Designing Classes for Red-Black Trees A red-black tree designs a class for a red-black tree.

A red-black tree is a binary search tree. So, you can define the RBTree class to extend the BST class, as shown in Figure 43.4. The BST and TreeNode classes are defined in §26.2.5.

Each node in a red-black tree has a color property. Because the color is either red or black, it is efficient to use the boolean type to denote it. The RBTreeNode class can be defined to extend BST.TreeNode with the color property. For convenience, we also provide the methods for checking the color and setting a new color. Note that TreeNode is defined as a static inner class in BST. RBTreeNode will be defined as a static inner class in RBTree. Note that BSTNode contains the data fields element, left, and right, which are inherited in RBTreeNode. So, RBTreeNode contains four data fields, as pictured in Figure 43.5.

Point Check

Point Key

Figure 43.4 The RBTree class extends BST with new implementations for the insert and delete methods.

1 Link

RBTreeNode<E>

–red: boolean

+RBTreeNode()

+RBTreeNode(e: E)

+isRed(): boolean

+isBlack(): boolean

+setRed(): void

+setBlack(): void

TreeNode<E>

+RBTree()

+RBTree(objects: E[])

#createNewNode(): RBTreeNode<E>

+insert(o: E): boolean

+delete(o: E): boolean

RBTree<E>

BST<E>

m 0

Creates a default red-black tree.

Creates an RBTree from an array of objects.

Override this method to create an RBTreeNode.

Returns true if the element is added successfully.

Returns true if the element is removed from the tree successfully.

M43_LIAN1878_11_GE_C43.indd 4 12/29/17 6:06 PM

43.4 Overriding the insert Method 43-5

In the BST class, the createNewNode() method creates a TreeNode object. This method is overridden in the RBTree class to create an RBTreeNode. Note the return type of the createNewNode() method in the BST class is TreeNode, but the return type of the creat- eNewNode() method in RBTree class is RBTreeNode. This is fine, since RBTreeNode is a subtype of TreeNode.

Searching an element in a red-black tree is the same as searching in a regular binary search tree. So, the search method defined in the BST class also works for RBTree.

The insert and delete methods are overridden to insert and delete an element and per- form operations for coloring and restructuring if necessary to ensure that the three properties of the red-black tree are satisfied.

Pedagogical Note Run from http://liveexample.pearsoncmg.com/dsanimation/ RBTree.html to see how a red-black tree works, as shown in Figure 43.6.

Figure 43.6 The animation tool enables you to insert, delete, and search elements in a red-black tree visually.

43.4 Overriding the insert Method This section discusses how to insert an element to red-black tree.

A new element is always inserted as a leaf node. If the new node is the root, color it black. Otherwise, color it red. If the parent of the new node is red, it violates Property 2 of the red- black tree. We call this a double-red violation.

Let u denote the new node inserted, v the parent of u, w the parent of v, and x the sibling of v. To fix the double-red violation, consider two cases:

Case 1: x is black or x is null. There are four possible configurations for u, v, w, and x, as shown in Figures 43.7(a), 43.8(a), 43.9(a), and 43.10(a). In this case, u, v, and w form a 4-node in the corresponding 2-4 tree, as shown in Figures 43.7(c), 43.8(c), 43.9(c), and 43.10(c), but are represented incorrectly in the red-black tree. To correct this error, restructure and recolor three nodes u, v, and w, as shown in Figures 43.7(b), 43.8(b), 43.9(b), and 43.10(b). Note x, y1 , y2 , and y3 may be null.

Point Key

Figure 43.5 An RBTreeNode contains data fields element, red, left, and right.

node: RBTreeNode<E>

#element: E -red: boolean #left: TreeNode #right: TreeNode

M43_LIAN1878_11_GE_C43.indd 5 12/29/17 6:06 PM

43-6 Chapter 43 Red-Black Trees

Figure 43.8 Case 1.2: v 6 u 6 w

(c)(b)(a)

y1 y2

y3 u

4020

y3 y1 y2 xy3 y1 y2

20 30 40

Figure 43.9 Case 1.3: w 6 v 6 u

(a) (b) (c)

4020

x y3 y1

y1 y2

y2x y3 y1

20 30 40

Figure 43.10 Case 1.4: w 6 u 6 v

(a) (b) (c)

4020

x y1 y2

y1 y2

y3x y1 y2

20 30 40

Figure 43.7 Case 1.1: u 6 v 6 w .

xy1 y2 y3

(a) (b) (c)

4020

y1 y2 y3

y1 y2

y3 20 30 40

Case 2: x is red. There are four possible configurations for u, v, w, w, and x, as shown in Fig- ures 43.11(a), 43.11(b), 43.11(c), and 43.11(d). All of these configurations correspond to an overflow situation in the corresponding 4-node in a 2-4 tree, as shown in Figure 43.12(a). A splitting operation is performed to fix the overflow problem in a 2-4 tree, as shown in Figure 43.12(b). We perform an equivalent recoloring operation to fix the problem in a red-black tree. Color w and u red and color two children of w black. Assume u is a left child of v, as shown

M43_LIAN1878_11_GE_C43.indd 6 12/29/17 6:06 PM

43.4 Overriding the insert Method 43-7

in Figure 43.11(a). After recoloring, the nodes are shown in Figure 43.12(c). Now w is red, if w’s parent is black, the double-red violation is fixed. Otherwise, a new double-red violation occurs at node w. We need to continue the same process to eliminate the double-red violation at w, recursively.

A more detailed algorithm for inserting an element is described in Listing 43.1.

Listing 43.1 Inserting an Element to a Red-Black Tree 1 public boolean insert(E e) { 2 boolean successful = super.insert(e); 3 if (!successful) 4 return false; // e is already in the tree 5 else { 6 ensureRBTree(e); 7 } 8 9 return true; // e is inserted 10 } 11 12 /** Ensure that the tree is a red-black tree */ 13 private void ensureRBTree(E e) { 14 Get the path that leads to element e from the root.

Figure 43.11 Case 2 has four possible configurations.

x 30

(a) (b)

x 30

Figure 43.12 Splitting a 4-node corresponds to recoloring the nodes in the red-black tree.

(a) A 4-node (b) Splitting a 4-node (c) Recoloring nodes

50 x

30 40 50

Insert a new element

Insert to its parent

25 30

M43_LIAN1878_11_GE_C43.indd 7 12/29/17 6:06 PM

43-8 Chapter 43 Red-Black Trees

15 int i = path.size() – 1; // Index to the current node in the path 16 Get u, v from the path. u is the node that contains e and v 17 is the parent of u. 18 Color u red; 19 20 if (u == root) // If e is inserted as the root, set root black 21 u.setBlack(); 22 else if (v.isRed()) 23 fixDoubleRed(u, v, path, i); // Fix double-red violation at u 24 } 25 26 /** Fix double-red violation at node u */ 27 private void fixDoubleRed(RBTreeNode<E> u, RBTreeNode<E> v, 28 ArrayList<TreeNode<E>> path, int i) { 29 Get w from the path. w is the grandparent of u. 30 31 // Get v’s sibling named x 32 RBTreeNode<E> x = (w.left == v) ? 33 (RBTreeNode<E>)(w.right) : (RBTreeNode<E>)(w.left); 34 35 if (x == null || x.isBlack()) { 36 // Case 1: v's sibling x is black 37 if (w.left == v && v.left == u) { 38 // Case 1.1: u < v < w, Restructure and recolor nodes 39 } 40 else if (w.left == v && v.right == u) { 41 // Case 1.2: v < u < w, Restructure and recolor nodes 42 } 43 else if (w.right == v && v.right == u) { 44 // Case 1.3: w < v < u, Restructure and recolor nodes 45 } 46 else { 47 // Case 1.4: w < u < v, Restructure and recolor nodes 48 } 49 } 50 else { // Case 2: v's sibling x is red 51 Color w and u red 52 Color two children of w black. 53 54 if (w is root) { 55 Set w black; 56 } 57 else if (the parent of w is red) { 58 // Propagate along the path to fix new double-red violation 59 u = w; 60 v = parent of w; 61 fixDoubleRed(u, v, path, i – 2); // i – 2 propagates upward 62 } 63 } 64 }

The insert(E e) method (lines 1–10) invokes the insert method in the BST class to create a new leaf node for the element (line 2). If the element is already in the tree, return false (line 4). Otherwise, invoke ensureRBTree(e) (line 6) to ensure that the tree satisfies the color and black depth property of the red-black tree.

The ensureRBTree(E e) method (lines 13–24) obtains the path that leads to e from the root (line 14), as shown in Figure 43.13. This path plays an important role to implement the algorithm. From this path, you get nodes u and v (lines 16–17). If u is the root, color u black (lines 20–21). If v is red, a double-red violation occurs at node u. Invoke fixDoubleRed to fix the problem.

M43_LIAN1878_11_GE_C43.indd 8 12/29/17 6:06 PM

43.5 Overriding the delete Method 43-9

The fixDoubleRed method (lines 27–63) fixes the double-red violation. It first obtains w (the parent of v) from the path (line 29) and x (the sibling of v) (lines 32–33). If x is empty or a black node, restructure and recolor three nodes u, v, and w to eliminate the problem (lines 35–49). If x is a red node, recolor the nodes u, v, w, and x (lines 51–52). If w is the root, color w black (lines 54–56). If the parent of w is red, the double-red violation reappears at w. Invoke fixDoubleRed with new u and v to fix the problem (line 61). Note that now i – 2 points to the new u in the path. This adjustment is necessary to locate the new nodes w and parent of w along the path.

Figure 43.14 shows the steps of inserting 34, 3, 50, 20, 15, 16, 25, and 27 into an empty red-black tree. When inserting 20 into the tree in (d), Case 2 applies to recolor 3 and 50 to black. When inserting 15 into the tree in (g), Case 1.4 applies to restructure and recolor nodes 15, 20, and 3. When inserting 16 into the tree in (i), Case 2 applies to recolor nodes 3 and 20 to black and nodes 15 and 16 to red. When inserting 27 into the tree in (l), Case 2 applies to recolor nodes 16 and 25 to black and nodes 20 and 27 to red. Now a new double- red problem occurs at node 20. Apply Case 1.2 to restructure and recolor nodes. The new tree is shown in (n).

43.5 Overriding the delete Method This section discusses how to delete an element to red-black tree.

To delete an element from a red-black tree, first search the element in the tree to locate the node that contains the element. If the element is not in the tree, the method returns false. Let u be the node that contains the element. If u is an internal node with both left and right children, find the rightmost node in the left subtree of u. Replace the element in u with the element in the rightmost node. Now we will only consider deleting external nodes.

Let u be an external node to be deleted. Since u is an external node, it has at most one child, denoted by childOfu. childOfu may be null. Let parentOfu denote the parent of u, as shown in Figure 43.15(a). Delete u by connecting childOfu with parentOfu, as shown in Figure 43.15(b).

Consider the following case:

■■ If u is red, we are done.

■■ If u is black and childOfu is red, color childOfu black to maintain the black height for childOfu.

■■ Otherwise, assign childOfu a fictitious double black, as shown in Figure 43.16(a). We call this a double-black problem, which indicates that the black depth is short by 1, caused by deleting a black node u.

Point Key

Figure 43.13 The path consists of the nodes from u to the root.

root

path

If path.get(i) is u, path.get(i – 1) is v and path.get(i – 2) is w.

M43_LIAN1878_11_GE_C43.indd 9 12/29/17 6:06 PM

43-10 Chapter 43 Red-Black Trees

Figure 43.14 Inserting into a red-black tree: (a) initial empty tree; (b) inserting 34; (c) inserting 3; (d) inserting 50; (e) inserting 20 causes a double red; (f) after recolor- ing (Case 2); (g) inserting 15 causes a double red; (h) after restructuring and recoloring (Case 1.4); (i) inserting 16 causes a double red; (j) after recoloring (Case 2); (k) insert- ing 25; (l) inserting 27 causes a double red at 27; (m) a double red at 20 reappears after recoloring (Case 2); and (n) after restructuring and recoloring (Case 1.2).

34root in null

(a) (b) (c) (d) (e)

(g) (h)

(j) (k)

(m) (n)

203

(f)

203

16 25

(l)

203

16 25

203

16 25

163 25

(i)

203

M43_LIAN1878_11_GE_C43.indd 10 12/29/17 6:06 PM

43.5 Overriding the delete Method 43-11

A double black in a red-black tree corresponds to an empty node for u (i.e., underflow situation) in the corresponding 2-4 tree, as shown in Figure 43.16(b). To fix the double-black problem, we will perform equivalent transfer and fusion operations. Consider three cases:

Case 1: The sibling y of childOfu is black and has a red child. This case has four possible con- figurations, as shown in Figures 43.17(a), 43.18(a), 43.19(a), and 43.20(a). The dashed circle denotes that the node is either red or black. To eliminate the double-black problem, restructure and recolor the nodes, as shown in Figures 43.17(b), 43.18(b), 43.19(b), and 43.20(b).

Figure 43.16 (a) childOfu is denoted double black. (b) u corresponds to an empty node in a 2-4 tree.

(a)

u is black

parentOfu

childOfu is black or null

parentOfu

(b)

Figure 43.15 u is an external node and childOfu may be null.

(a) Before deleting u (b) After deleting u

parentOfu

childOfu

parentOfu

childOfu

Figure 43.17 Case 1.1: The sibling y of childOfu is black and y1 is red.

(a) (b)

parent

childOfu

parent childOfu is double blacky

y1 y2

Figure 43.18 Case 1.2: The sibling y of childOfu is black and y2 is red.

parent childOfu is doubleblack y

y1 y2

parent

childOfu

y2.left y2.right

(a) (b)

M43_LIAN1878_11_GE_C43.indd 11 12/29/17 6:06 PM

43-12 Chapter 43 Red-Black Trees

Note Case 1 corresponds to a transfer operation in the 2-4 tree. For example, the correspond- ing 2-4 tree for Figure 43.17(a) is shown in Figure 43.21(a), and it is transformed into Figure 43.21(b) through a transfer operation.

Figure 43.19 Case 1.3: The sibling y of childOfu is black and y1 is red.

(a) (b)

parent

childOfu

y2 y1.left y1.right

parent

y1 y2childOfu is double black

Figure 43.20 Case 1.4: the sibling y of childOfu is black and y2 is red.

(a) (b)

parent

y1 y2childOfu is double black

parent

childOfu

Figure 43.21 Case 1 corresponds to a transfer operation in the corresponding 2-4 tree.

(a) (b)

parent

childOfu

y1 u

parent

childOfu

y1 y

Case 2: The sibling y of childOfu is black and its children are black or null. In this case, change y’s color to red. If parent is red, change it to black, and we are done, as shown in Figure 43.22. If parent is black, we denote parent double black, as shown in Figure 43.23. The double-black problem propagates to the parent node.

Figure 43.22 Case 2: Recoloring eliminates the double-black problem if parent is red.

(a) (b)

parent

y1 y2

childOfu is double black

y1 y2

childOfu

parent

M43_LIAN1878_11_GE_C43.indd 12 12/29/17 6:06 PM

43.5 Overriding the delete Method 43-13

Note Figures 43.22 and 43.23 show that childOfu is a right child of parent. If childOfu is a left child of parent, recoloring is performed identically.

Note Case 2 corresponds to a fusion operation in the 2-4 tree. For example, the corresponding 2-4 tree for Figure 43.22(a) is shown in Figure 43.24(a), and it is transformed into Figure 43.24(b) through a fusion operation.

Case 3: The sibling y of childOfu is red. In this case, perform an adjustment operation. If y is a left child of parent, let y1 and y2 be the left and right children of y, as shown in Figure 43.25. If y is a right children of parent, let y1 and y2 be the left and right child of y, as shown in Figure 43.26. In both cases, color y black and parent red. childOfu is still a fictitious double-black node. After the adjustment, the sibling of childOfu is now black, and either Case 1 or Case 2 applies. If Case 1 applies, a one-time restructuring and recoloring operation eliminates the double-black problem. If Case 2 applies, the double-black problem cannot reappear, since parent is now red. Therefore, one-time application of Case 1 or Case 2 will complete Case 3.

Note Case 3 results from the fact that a 3-node may be transformed in two ways to a red-black tree, as shown in Figure 43.27.

Figure 43.23 Case 2: Recoloring propagates the double-black problem if parent is black.

parent parent

y y

y1 y1y2 y2

childOfu

(a) (b)

childOfu is double black

Figure 43.24 Case 2 corresponds to a fusion operation in the corresponding 2-4 tree.

(a) (b)

parent

childOfu

y parent

childOfu

...

Figure 43.25 Case 3.1: y is a left red child of parent.

(a) (b)

parent

y1 y2

childOfu is double black parenty1

childOfu is double black

M43_LIAN1878_11_GE_C43.indd 13 12/29/17 6:06 PM

43-14 Chapter 43 Red-Black Trees

Based on the foregoing discussion, Listing 43.2 presents a more detailed algorithm for delet- ing an element.

Listing 43.2 Deleting an Element from a Red-Black Tree 1 public boolean delete(E e) { 2 Locate the node to be deleted 3 if (the node is not found) 4 return false; 5 6 if (the node is an internal node) { 7 Find the rightmost node in the subtree of the node; 8 Replace the element in the node with the one in rightmost; 9 The rightmost node is the node to be deleted now; 10 } 11 12 Obtain the path from the root to the node to be deleted; 13 14 // Delete the last node in the path and propagate if needed 15 deleteLastNodeInPath(path); 16

Figure 43.26 Case 3.2: y is a right red child of parent.

(a) (b)

parent

y1 y2childOfu is double black y1

childOfu is double black

parent

Figure 43.27 A 3-node may be transformed in two ways to red-black tree nodes.

y parent

childOfu

y1 y2 or

parent

childOfu is double black

y1 y2

childOfu

(a)

(b)

(c)

y2 u

childOfu

parent

M43_LIAN1878_11_GE_C43.indd 14 12/29/17 6:06 PM

43.5 Overriding the delete Method 43-15

17 size--; // After one element deleted 18 return true; // Element deleted 19 } 20 21 /** Delete the last node from the path. */ 22 public void deleteLastNodeInPath(ArrayList<TreeNode<e>> path) { 23 Get the last node u in the path; 24 Get parentOfu and grandparentOfu in the path; 25 Get childOfu from u; 26 Delete node u. Connect childOfu with parentOfu 27 28 // Recolor the nodes and fix double black if needed 29 if (childOfu == root || u.isRed()) 30 return; // Done if childOfu is root or if u is red 31 else if (childOfu != null && childOfu.isRed()) 32 childOfu.setBlack(); // Set it black, done 33 else // u is black, childOfu is null or black 34 // Fix double black on parentOfu 35 fixDoubleBlack(grandparentOfu, parentOfu, childOfu, path, i); 36 } 37 38 /** Fix the double black problem at node parent */ 39 private void fixDoubleBlack( 40 RBTreeNode<E> grandparent, RBTreeNode<E> parent, 41 RBTreeNode<E> db, ArrayList<TreeNode<E>> path, int i) { 42 Obtain y, y1, and y2 43 44 if (y.isBlack() && y1 != null && y1.isRed()) { 45 if (parent.right == db) { 46 // Case 1.1: y is a left black sibling and y1 is red 47 Restructure and recolor parent, y, and y1 to fix the problem; 48 } 49 else { 50 // Case 1.3: y is a right black sibling and y1 is red 51 Restructure and recolor parent, y1, and y to fix the problem; 52 } 53 } 54 else if (y.isBlack() && y2 != null && y2.isRed()) { 55 if (parent.right == db) { 56 // Case 1.2: y is a left black sibling and y2 is red 57 Restructure and recolor parent, y2, and y to fix the problem; 58 } 59 else { 60 // Case 1.4: y is a right black sibling and y2 is red 61 Restructure and recolor parent, y, and y2 to fix the problem; 62 } 63 } 64 else if (y.isBlack()) { 65 // Case 2: y is black and y's children are black or null 66 Recolor y to red; 67 68 if (parent.isRed()) 69 parent.setBlack(); // Done 70 else if (parent != root) { 71 // Propagate double black to the parent node 72 // Fix new appearance of double black recursively 73 db = parent; 74 parent = grandparent; 75 grandparent = 76 (i >= 3) ? (RBTreeNode<E>)(path.get(i − 3)) : null;

M43_LIAN1878_11_GE_C43.indd 15 12/29/17 6:06 PM

43-16 Chapter 43 Red-Black Trees

77 fixDoubleBlack(grandparent, parent, db, path, i − 1); 78 } 79 } 80 else if (y.isRed()) { 81 if (parent.right == db) { 82 // Case 3.1: y is a left red child of parent 83 parent.left = y2; 84 y.right = parent; 85 } 86 else { 87 // Case 3.2: y is a right red child of parent 88 parent.right = y.left; 89 y.left = parent; 90 } 91 92 parent.setRed(); // Color parent red 93 y.setBlack(); // Color y black 94 connectNewParent(grandparent, parent, y); // y is new parent 95 fixDoubleBlack(y, parent, db, path, i − 1); 96 } 97 }

The delete(E e) method (lines 1–19) locates the node that contains e (line 2). If the node does not exist, return false (lines 3–4). If the node is an internal node, find the right most node in its left subtree and replace the element in the node with the element in the right most node (lines 6–9). Now the node to be deleted is an external node. Obtain the path from the root to the node (line 12). Invoke deleteLastNodeInPath(path) to delete the last node in the path and ensure that the tree is still a red-black tree (line 15).

The deleteLastNodeInPath method (lines 22–36) obtains the last node u, parentOfu, grandparendOfu, and childOfu (lines 23–26). If childOfu is the root or u is red, the tree is fine (lines 29–30). If childOfu is red, color it black (lines 31–32). We are done. Otherwise, u is black and childOfu is null or black. Invoke fixDoubleBlack to eliminate the double- black problem (line 35).

The fixDoubleBlack method (lines 39–97) eliminates the double-black problem. Obtain y, y1, and y2 (line 42). y is the sibling of the double-black node. y1 and y2 are the left and right children of y. Consider three cases:

1. If y is black and one of its children is red, the double-black problem can be fixed by one-time restructuring and recoloring in Case 1 (lines 44–63).

2. If y is black and its children are null or black, change y to red. If parent of y is black, denote parent to be the new double-black node and invoke fixDoubleBlack recur- sively (line 77).

3. If y is red, adjust the nodes to make parent a child of y (lines 84, 89) and color parent red and y black (lines 92–93). Make y the new parent (line 94). Recursively invoke fixDoubleBlack on the same double-black node with a different color for parent (line 95).

Figure 43.28 shows the steps of deleting elements. To delete 50 from the tree in Figure 43.28(a), apply Case 1.2, as shown in Figure 43.28(b). After restructuring and recoloring, the new tree is as shown in Figure 43.28(c).

When deleting 20 in Figure 43.28(c), 20 is an internal node, and it is replaced by 16, as shown in Figure 43.28(d). Now Case 2 applies to deleting the rightmost node, as shown in Figure 43.28(e). Recolor the nodes results in a new tree, as shown in Figure 43.28(f).

When deleting 15, connect node 3 with node 20 and color node 3 black, as shown in Figure 43.28(g). We are done.

M43_LIAN1878_11_GE_C43.indd 16 12/29/17 6:06 PM

43.5 Overriding the delete Method 43-17

Figure 43.28 Delete elements from a red-black tree.

163 25

null

163 25 34

(a) Delete 50 (b) Case 1.2 (c) Delete 20

163 25 34

null3 25 34

3 25 34

(d) Copy 16 to replace 20 (e) Case 2 (f) Delete 15

25 34

null

25 34

25null

(g) Delete 3 (h) Case 3 (i) Case 2

34null

(j) Delete 25 (k) Delete 16 (l) Case 2

27 nullroot:

(m) Delete 34 (n) Delete 27 (o) Empty tree

After deleting 25, the new tree is as shown in Figure 43.28(j). Now delete 16. Apply Case 2, as shown in Figure 43.28(k). The new tree is shown in Figure 43.28(l).

After deleting 34, the new tree is as shown in Figure 43.28(m). After deleting 27, the new tree is as shown in Figure 43.28(n).

M43_LIAN1878_11_GE_C43.indd 17 12/29/17 6:06 PM

43-18 Chapter 43 Red-Black Trees

43.5.1 What are the data fields in RBTreeNode? 43.5.2 How do you insert an element into a red-black tree and how do you fix the double-

red violation?

43.5.3 How do you delete an element from a red-black tree and how do you fix the double-black problem?

43.5.4 Show the change of the tree when inserting 1, 2, 3, 4, 10, 9, 7, 5, 8, and 6 into it, in this order.

43.5.5 For the tree built in the preceding question, show the change of the tree after delet- ing 1, 2, 3, 4, 10, 9, 7, 5, 8, and 6 from it in this order.

43.6 Implementing RBTree Class This section implements the RBTree class.

Listing 43.3 gives a complete implementation for the RBTree class.

Listing 43.3 RBTree.java 1 import java.util.ArrayList; 2 3 public class RBTree<E extends Comparable<E>> extends BST<E> { 4 /** Create a default RB tree */ 5 public RBTree() { 6 } 7 8 /** Create an RB tree from an array of elements */ 9 public RBTree(E[] elements) { 10 super(elements); 11 } 12 13 @Override /** Override createNewNode to create an RBTreeNode */ 14 protected RBTreeNode<E> createNewNode(E e) { 15 return new RBTreeNode<E>(e); 16 } 17 18 @Override /** Override the insert method to 19 balance the tree if necessary */ 20 public boolean insert(E e) { 21 boolean successful = super.insert(e); 22 if (!successful) 23 return false; // e is already in the tree 24 else { 25 ensureRBTree(e); 26 } 27 28 return true; // e is inserted 29 } 30 31 /** Ensure that the tree is a red-black tree */ 32 private void ensureRBTree(E e) { 33 // Get the path that leads to element e from the root 34 ArrayList<TreeNode<E>> path = path(e); 35 36 int i = path.size() − 1; // Index to the current node in the path 37 38 // u is the last node in the path. u contains element e 39 RBTreeNode<E> u = (RBTreeNode<E>)(path.get(i));

Point Check

Point Key

M43_LIAN1878_11_GE_C43.indd 18 12/29/17 6:06 PM

43.6 Implementing RBTree Class 43-19

40 41 // v is the parent of of u, if exists 42 RBTreeNode<E> v = (u == root) ? null : 43 (RBTreeNode<E>)(path.get(i − 1)); 44 45 u.setRed(); // It is OK to set u red 46 47 if (u == root) // If e is inserted as the root, set root black 48 u.setBlack(); 49 else if (v.isRed()) 50 fixDoubleRed(u, v, path, i); // Fix double-red violation at u 51 } 52 53 /** Fix double-red violation at node u */ 54 private void fixDoubleRed(RBTreeNode<E> u, RBTreeNode<E> v, 55 ArrayList<TreeNode<E>> path, int i) { 56 // w is the grandparent of u 57 RBTreeNode<E> w = (RBTreeNode<E>)(path.get(i − 2)); 58 RBTreeNode<E> parentOfw = (w == root) ? null : 59 (RBTreeNode<E>)path.get(i – 3); 60 61 // Get v's sibling named x 62 RBTreeNode<E> x = (w.left == v) ? 63 (RBTreeNode<E>)(w.right) : (RBTreeNode<E>)(w.left); 64 65 if (x == null || x.isBlack()) { 66 // Case 1: v's sibling x is black 67 if (w.left == v && v.left == u) { 68 // Case 1.1: u < v < w, Restructure and recolor nodes 69 restructureRecolor(u, v, w, w, parentOfw); 70 71 w.left = v.right; // v.right is y3 in Figure 43.6 72 v.right = w; 73 } 74 else if (w.left == v && v.right == u) { 75 // Case 1.2: v < u < w, Restructure and recolor nodes 76 restructureRecolor(v, u, w, w, parentOfw); 77 v.right = u.left; 78 w.left = u.right; 79 u.left = v; 80 u.right = w; 81 } 82 else if (w.right == v && v.right == u) { 83 // Case 1.3: w < v < u, Restructure and recolor nodes 84 restructureRecolor(w, v, u, w, parentOfw); 85 w.right = v.left; 86 v.left = w; 87 } 88 else { 89 // Case 1.4: w < u < v, Restructure and recolor nodes 90 restructureRecolor(w, u, v, w, parentOfw); 91 w.right = u.left; 92 v.left = u.right; 93 u.left = w; 94 u.right = v; 95 } 96 } 97 else { // Case 2: v's sibling x is red 98 // Recolor nodes 99 w.setRed(); 100 u.setRed();

M43_LIAN1878_11_GE_C43.indd 19 12/29/17 6:06 PM

43-20 Chapter 43 Red-Black Trees

101 ((RBTreeNode<E>)(w.left)).setBlack(); 102 ((RBTreeNode<E>)(w.right)).setBlack(); 103 104 if (w == root) { 105 w.setBlack(); 106 } 107 else if (((RBTreeNode<E>)parentOfw).isRed()) { 108 // Propagate along the path to fix new double-red violation 109 u = w; 110 v = (RBTreeNode<E>)parentOfw; 111 fixDoubleRed(u, v, path, i − 2); // i – 2 propagates upward 112 } 113 } 114 } 115 116 /** Connect b with parentOfw and recolor a, b, c for a < b < c */ 117 private void restructureRecolor(RBTreeNode<E> a, RBTreeNode<E> b, 118 RBTreeNode<E> c, RBTreeNode<E> w, RBTreeNode<E> parentOfw) { 119 if (parentOfw == null) 120 root = b; 121 else if (parentOfw.left == w) 122 parentOfw.left = b; 123 else 124 parentOfw.right = b; 125 126 b.setBlack(); // b becomes the root in the subtree 127 a.setRed(); // a becomes the left child of b 128 c.setRed(); // c becomes the right child of b 129 } 130 131 @Override /** Delete an element from the RBTree. 132 * Return true if the element is deleted successfully 133 * Return false if the element is not in the tree */ 134 public boolean delete(E e) { 135 // Locate the node to be deleted 136 TreeNode<E> current = root; 137 while (current != null) { 138 if (e.compareTo(current.element) < 0) { 139 current = current.left; 140 } 141 else if (e.compareTo(current.element) > 0) { 142 current = current.right; 143 } 144 else 145 break; // Element is in the tree pointed by current 146 } 147 148 if (current == null) 149 return false; // Element is not in the tree 150 151 java.util.ArrayList<TreeNode<E>> path; 152 153 // current node is an internal node 154 if (current.left != null && current.right != null) { 155 // Locate the rightmost node in the left subtree of current 156 TreeNode<E> rightMost = current.left; 157 while (rightMost.right != null) { 158 rightMost = rightMost.right; // Keep going to the right 159 } 160

M43_LIAN1878_11_GE_C43.indd 20 12/29/17 6:06 PM

43.6 Implementing RBTree Class 43-21

161 path = path(rightMost.element); // Get path before replacement 162 163 // Replace the element in current by the element in rightMost 164 current.element = rightMost.element; 165 } 166 else 167 path = path(e); // Get path to current node 168 169 // Delete the last node in the path and propagate if needed 170 deleteLastNodeInPath(path); 171 172 size--; // After one element deleted 173 return true; // Element deleted 174 } 175 176 /** Delete the last node from the path. */ 177 public void deleteLastNodeInPath(ArrayList<TreeNode<E>> path) { 178 int i = path.size() − 1; // Index to the node in the path 179 // u is the last node in the path 180 RBTreeNode<E> u = (RBTreeNode<E>)(path.get(i)); 181 RBTreeNode<E> parentOfu = (u == root) ? null : 182 (RBTreeNode<E>)(path.get(i − 1)); 183 RBTreeNode<E> grandparentOfu = (parentOfu == null || 184 parentOfu == root) ? null : 185 (RBTreeNode<E>)(path.get(i − 2)); 186 RBTreeNode<E> childOfu = (u.left == null) ? 187 (RBTreeNode<E>)(u.right) : (RBTreeNode<E>)(u.left); 188 189 // Delete node u. Connect childOfu with parentOfu 190 connectNewParent(parentOfu, u, childOfu); 191 192 // Recolor the nodes and fix double black if needed 193 if (childOfu == root || u.isRed()) 194 return; // Done if childOfu is root or if u is red 195 else if (childOfu != null && childOfu.isRed()) 196 childOfu.setBlack(); // Set it black, done 197 else // u is black, childOfu is null or black 198 // Fix double black on parentOfu 199 fixDoubleBlack(grandparentOfu, parentOfu, childOfu, path, i); 200 } 201 202 /** Fix the double-black problem at node parent */ 203 private void fixDoubleBlack( 204 RBTreeNode<E> grandparent, RBTreeNode<E> parent, 205 RBTreeNode<E> db, ArrayList<TreeNode<E>> path, int i) { 206 // Obtain y, y1, and y2 207 RBTreeNode<E> y = (parent.right == db) ? 208 (RBTreeNode<E>)(parent.left) : (RBTreeNode<E>)(parent.right); 209 RBTreeNode<E> y1 = (RBTreeNode<E>)(y.left); 210 RBTreeNode<E> y2 = (RBTreeNode<E>)(y.right); 211 212 if (y.isBlack() && y1 != null && y1.isRed()) { 213 if (parent.right == db) { 214 // Case 1.1: y is a left black sibling and y1 is red 215 connectNewParent(grandparent, parent, y); 216 recolor(parent, y, y1); // Adjust colors 217 218 // Adjust child links 219 parent.left = y.right;

M43_LIAN1878_11_GE_C43.indd 21 12/29/17 6:06 PM

43-22 Chapter 43 Red-Black Trees

220 y.right = parent; 221 } 222 else { 223 // Case 1.3: y is a right black sibling and y1 is red 224 connectNewParent(grandparent, parent, y1); 225 recolor(parent, y1, y); // Adjust colors 226 227 // Adjust child links 228 parent.right = y1.left; 229 y.left = y1.right; 230 y1.left = parent; 231 y1.right = y; 232 } 233 } 234 else if (y.isBlack() && y2 != null && y2.isRed()) { 235 if (parent.right == db) { 236 // Case 1.2: y is a left black sibling and y2 is red 237 connectNewParent(grandparent, parent, y2); 238 recolor(parent, y2, y); // Adjust colors 239 240 // Adjust child links 241 y.right = y2.left; 242 parent.left = y2.right; 243 y2.left = y; 244 y2.right = parent; 245 } 246 else { 247 // Case 1.4: y is a right black sibling and y2 is red 248 connectNewParent(grandparent, parent, y); 249 recolor(parent, y, y2); // Adjust colors 250 251 // Adjust child links 252 y.left = parent; 253 parent.right = y1; 254 } 255 } 256 else if (y.isBlack()) { 257 // Case 2: y is black and y’s children are black or null 258 y.setRed(); // Change y to red 259 if (parent.isRed()) 260 parent.setBlack(); // Done 261 else if (parent != root) { 262 // Propagate double black to the parent node 263 // Fix new appearance of double black recursively 264 db = parent; 265 parent = grandparent; 266 grandparent = 267 (i >= 3) ? (RBTreeNode<E>)(path.get(i − 3)) : null; 268 fixDoubleBlack(grandparent, parent, db, path, i − 1); 269 } 270 } 271 else { // y.isRed() 272 if (parent.right == db) { 273 // Case 3.1: y is a left red child of parent 274 parent.left = y2; 275 y.right = parent; 276 } 277 else { 278 // Case 3.2: y is a right red child of parent 279 parent.right = y.left; 280 y.left = parent;

M43_LIAN1878_11_GE_C43.indd 22 12/29/17 6:06 PM

43.6 Implementing RBTree Class 43-23

281 } 282 283 parent.setRed(); // Color parent red 284 y.setBlack(); // Color y black 285 connectNewParent(grandparent, parent, y); // y is new parent 286 fixDoubleBlack(y, parent, db, path, i − 1); 287 } 288 } 289 290 /** Recolor parent, newParent, and c. Case 1 removal */ 291 private void recolor(RBTreeNode<E> parent, 292 RBTreeNode<E> newParent, RBTreeNode<E> c) { 293 // Retain the parent’s color for newParent 294 if (parent.isRed()) 295 newParent.setRed(); 296 else 297 newParent.setBlack(); 298 299 // c and parent become the children of newParent; set them black 300 parent.setBlack(); 301 c.setBlack(); 302 } 303 304 /** Connect newParent with grandParent */ 305 private void connectNewParent(RBTreeNode<E> grandparent, 306 RBTreeNode<E> parent, RBTreeNode<E> newParent) { 307 if (parent == root) { 308 root = newParent; 309 if (root != null) 310 newParent.setBlack(); 311 } 312 else if (grandparent.left == parent) 313 grandparent.left = newParent; 314 else 315 grandparent.right = newParent; 316 } 317 318 @Override /** Preorder traversal from a subtree */ 319 protected void preorder(TreeNode<E> root) { 320 if (root == null) return; 321 System.out.print(root.element + 322 (((RBTreeNode<E>)root).isRed() ? " (red) " : " (black) ")); 323 preorder(root.left); 324 preorder(root.right); 325 } 326 327 /** RBTreeNode is TreeNode plus color indicator */ 328 protected static class RBTreeNode<E extends Comparable<E>> extends 329 BST.TreeNode<E> { 330 private boolean red = true; // Indicate node color 331 332 public RBTreeNode(E e) { 333 super(e); 334 } 335 336 public boolean isRed() { 337 return red; 338 } 339 340 public boolean isBlack() { 341 return !red;

M43_LIAN1878_11_GE_C43.indd 23 12/29/17 6:06 PM

43-24 Chapter 43 Red-Black Trees

342 } 343 344 public void setBlack() { 345 red = false; 346 } 347 348 public void setRed() { 349 red = true; 350 } 351 352 int blackHeight; 353 } 354 }

The RBTree class extends BST. Like the BST class, the RBTree class has a no-arg constructor that constructs an empty RBTree (lines 5–6) and a constructor that creates an initial RBTree from an array of elements (lines 9–11).

The createNewNode() method defined in the BST class creates a TreeNode. This method is overridden to return an RBTreeNode (lines 14–16). This method is invoked in the insert method in BST to create a node.

The insert method in RBTree is overridden in lines 20–29. The method first invokes the insert method in BST, then invokes ensureRBTree(e) (line 25) to ensure that tree is still a red-black tree after inserting a new element.

The ensureRBTree(E e) method first obtains the path of nodes that lead to element e from the root (line 34). It obtains u and v (the parent of u) from the path. If u is the root, color u black (lines 47–48). If v is red, invoke fixDoubleRed to fix the double red on both u and v (lines 49–50).

The fixDoubleRed(u, v, path, i) method fixes the double-red violation at node u. The method first obtains w (the grandparent of u from the path) (line 57), parentOfw if exists (lines 58–59), and x (the sibling of v) (lines 62–63). If x is null or black, consider four sub- cases to fix the double-red violation (lines 67–96). If x is red, color w and u red and color w’s two children black (lines 101–104). If w is the root, color w black (lines 104–106). Otherwise, propagate along the path to fix the new double-red violation (lines 109–111).

The delete(E e) method in RBTree is overridden in lines 134–174. The method locates the node that contains e (lines 136–146). If the node is null, no element is found (lines 148–149). The method considers two cases:

■■ If the node is internal, find the rightmost node in its left subtree (lines 156–159). Obtain a path from the root to the rightmost node (line 161), and replace the element in the node with the element in the rightmost node (line 164).

■■ If the node is external, obtain the path from the root to the node (line 167).

The last node in the path is the node to be deleted. Invoke deleteLastNodeInPath(path) to delete it and ensure the tree is a red-black after the node is deleted (line 170).

The deleteLastNodeInPath(path) method first obtains u, parentOfu, grand- parendOfu, and childOfu (lines 180–187). u is the last node in the path. Connect childOfu as a child of parentOfu (line 190). This in effect deletes u from the tree. Consider three cases:

■■ If childOfu is the root or childOfu is red, we are done (lines 193–194).

■■ Otherwise, if childOfu is red, color it black (lines 195–196).

■■ Otherwise, invoke fixDoubleBlack to fix the double-black problem on childOfu (line 199).

M43_LIAN1878_11_GE_C43.indd 24 12/29/17 6:06 PM

43.7 Testing the RBTree Class 43-25

The fixDoubleBlack method first obtains y, y1, and y2 (lines 207–210). y is the sibling of the first double-black node, and y1 and y2 are the left and right children of y. Consider three cases:

■■ If y is black and y1 or y2 is red, fix the double-black problem for Case 1 (lines 213–255).

■■ Otherwise, if y is black, fix the double-black problem for Case 2 by recoloring the nodes. If parent is black and not a root, propagate double black to parent and recur- sively invoke fixDoubleBlack (lines 264–268).

■■ Otherwise, y is red. In this case, adjust the nodes to make parent the child of y (lines 272–281). Invoke fixDoubleBlack with the adjusted nodes (line 286) to fix the double-black problem.

The preorder(TreeNode<E> root) method is overridden to display the node colors (lines 319–325).

43.7 Testing the RBTree Class This section gives a test program that uses the RBTree class.

Listing 43.4 gives a test program. The program creates an RBTree initialized with an array of integers 34, 3, and 50 (lines 4–5), inserts elements in lines 10–22, and deletes elements in lines 25–46.

Listing 43.4 TestRBTree.java 1 public class TestRBTree { 2 public static void main(String[] args) { 3 // Create an RB tree 4 RBTree<Integer> tree = 5 new RBTree<Integer>(new Integer[]{34, 3, 50}); 6 printTree(tree); 7 8 tree.insert(20); 9 printTree(tree); 10 11 tree.insert(15); 12 printTree(tree); 13 14 tree.insert(16); 15 printTree(tree); 16 17 tree.insert(25); 18 printTree(tree); 19 20 tree.insert(27); 21 printTree(tree); 22 23 tree.delete(50); 24 printTree(tree); 25 26 tree.delete(20); 27 printTree(tree); 28 29 tree.delete(15); 30 printTree(tree); 31

Point Key

M43_LIAN1878_11_GE_C43.indd 25 12/29/17 6:06 PM

43-26 Chapter 43 Red-Black Trees

32 tree.delete(3); 33 printTree(tree); 34 35 tree.delete(25); 36 printTree(tree); 37 38 tree.delete(16); 39 printTree(tree); 40 41 tree.delete(34); 42 printTree(tree); 43 44 tree.delete(27); 45 printTree(tree); 46 } 47 48 public static <E extends Comparable<E>> 49 void printTree(BST <E> tree) { 50 // Traverse tree 51 System.out.print("\nInorder (sorted): "); 52 tree.inorder(); 53 System.out.print("\nPostorder: "); 54 tree.postorder(); 55 System.out.print("\nPreorder: "); 56 tree.preorder(); 57 System.out.print("\nThe number of nodes is " + tree.getSize()); 58 System.out.println(); 59 } 60 }

Inorder (sorted): 3 34 50

Postorder: 3 50 34 Preorder: 34 (black) 3 (red) 50 (red) The number of nodes is 3

Inorder (sorted): 3 20 34 50 Postorder: 20 3 50 34 Preorder: 34 (black) 3 (black) 20 (red) 50 (black) The number of nodes is 4

Inorder (sorted): 3 15 20 34 50 Postorder: 3 20 15 50 34 Preorder: 34 (black) 15 (black) 3 (red) 20 (red) 50 (black) The number of nodes is 5

Inorder (sorted): 3 15 16 20 34 50 Postorder: 3 16 20 15 50 34 Preorder: 34 (black) 15 (red) 3 (black) 20 (black) 16 (red) 50 (black) The number of nodes is 6

Inorder (sorted): 3 15 16 20 25 34 50 Postorder: 3 16 25 20 15 50 34 Preorder: 34 (black) 15 (red) 3 (black) 20 (black) 16 (red) 25 (red) 50 (black) The number of nodes is 7

Inorder (sorted): 3 15 16 20 25 27 34 50 Postorder: 3 16 15 27 25 50 34 20

M43_LIAN1878_11_GE_C43.indd 26 12/29/17 6:06 PM

43.8 Performance of the RBTree Class 43-27

Figure 43.14 shows how the tree evolves as elements are added to it, and Figure 43.28 shows how the tree evolves as elements are deleted from it.

43.8 Performance of the RBTree Class This search, insertion, and deletion operations take O(logn) time in a red-black tree.

The search, insertion, and deletion times in a red-black tree depend on the height of the tree. A red-black tree corresponds to a 2–4 tree. When you convert a node in a 2–4 tree to red-black tree nodes, you get one black node and zero, one, or two red nodes as its children, depending on whether the original node is a 2-node, 3-node, or 4-node. So, the height of a red-black tree

Point Key

Preorder: 20 (black) 15 (red) 3 (black) 16 (black) 34 (red) 25 (black) 27 (red) 50 (black) The number of nodes is 8

Inorder (sorted): 3 15 16 20 25 27 34 Postorder: 3 16 15 25 34 27 20 Preorder: 20 (black) 15 (red) 3 (black) 16 (black) 27 (red) 25 (black) 34 (black) The number of nodes is 7

Inorder (sorted): 3 15 16 25 27 34 Postorder: 3 15 25 34 27 16 Preorder: 16 (black) 15 (black) 3 (red) 27 (red) 25 (black) 34 (black) The number of nodes is 6

Inorder (sorted): 3 16 25 27 34 Postorder: 3 25 34 27 16 Preorder: 16 (black) 3 (black) 27 (red) 25 (black) 34 (black) The number of nodes is 5

Inorder (sorted): 16 25 27 34 Postorder: 25 16 34 27 Preorder: 27 (black) 16 (black) 25 (red) 34 (black) The number of nodes is 4

Inorder (sorted): 16 27 34 Postorder: 16 34 27 Preorder: 27 (black) 16 (black) 34 (black) The number of nodes is 3

Inorder (sorted): 27 34 Postorder: 34 27 Preorder: 27 (black) 34 (red) The number of nodes is 2

Inorder (sorted): 27 Postorder: 27 Preorder: 27 (black) The number of nodes is 1

Inorder (sorted): Postorder: Preorder: The number of nodes is 0

M43_LIAN1878_11_GE_C43.indd 27 12/29/17 6:06 PM

43-28 Chapter 43 Red-Black Trees

is at most as twice that of its corresponding 2–4 tree. Since the height of a 2–4 tree is log n, the height of a red-black tree is 2log n.

A red-black tree has the same time complexity as an AVL tree, as shown in Table 43.1. In general, a red-black is more efficient than an AVL tree, because a red-black tree requires only one-time restructuring of the nodes for insert and delete operations.

A red-black tree has the same time complexity as a 2–4 tree, as shown in Table 43.1. In general, a red-black is more efficient than a 2–4 tree for two reasons:

1. A red-black tree requires only one-time restructuring of the nodes for insert and delete operations. However, a 2–4 tree may require many splits for an insert operation and fusion for a delete operation.

2. A red-black tree is a binary search tree. A binary tree can be implemented more space efficiently than a 2–4 tree, because a node in a 2–4 tree has at most three elements and four children. Space is wasted for 2-nodes and 3-nodes in a 2–4 tree.

Listing 43.5 gives an empirical test of the performance of AVL trees, 2–4 trees, and red-black trees.

Listing 43.5 TreePerformanceTest.java 1 public class TreePerformanceTest { 2 public static void main(String[] args) { 3 final int TEST_SIZE = 500000; // Tree size used in the test 4 5 // Create an AVL tree 6 Tree<Integer> tree1 = new AVLTree<Integer>(); 7 System.out.println("AVL tree time: " + 8 getTime(tree1, TEST_SIZE) + " milliseconds"); 9 10 // Create a 2-4 tree 11 Tree<Integer> tree2 = new Tree24<Integer>(); 12 System.out.println("2-4 tree time: " 13 + getTime(tree2, TEST_SIZE) + " milliseconds"); 14 15 // Create a red-black tree 16 Tree<Integer> tree3 = new RBTree<Integer>(); 17 System.out.println("RB tree time: " 18 + getTime(tree3, TEST_SIZE) + " milliseconds"); 19 } 20 21 public static long getTime(Tree<Integer> tree, int testSize) { 22 long startTime = System.currentTimeMillis(); // Start time 23 24 // Create a list to store distinct integers 25 java.util.List<Integer> list = new java.util.ArrayList<Integer>();

Methods Red-Black Tree AVL Tree 2-4 Tree

search (e: E) O(logn) O(logn) O(logn)

insert (e: E) O(logn) O(logn) O(logn)

delete (e: E) O(logn) O(logn) O(logn)

getSize() O(l) O(l) O(l)

isEmpty() O(l) O(l) O(l)

tabLe 43.1 Time Complexities for Methods in RBTree, AVLTree, and Tree234

M43_LIAN1878_11_GE_C43.indd 28 12/29/17 6:06 PM

43.8 Performance of the RBTree Class 43-29

26 for (int i = 0; i < testSize; i++) 27 list.add(i); 28 29 java.util.Collections.shuffle(list); // Shuffle the list 30 31 // Insert elements in the list to the tree 32 for (int i = 0; i < testSize; i++) 33 tree.insert(list.get(i)); 34 35 java.util.Collections.shuffle(list); // Shuffle the list 36 37 // Delete elements in the list from the tree 38 for (int i = 0; i < testSize; i++) 39 tree.delete(list.get(i)); 40 41 // Return elapse time 42 return System.currentTimeMillis() - startTime; 43 } 44 }

AVL tree time: 7609 milliseconds

2–4 tree time: 8594 milliseconds RB tree time: 5515 milliseconds

The getTestTime method creates a list of distinct integers from 0 to testSize – 1 (lines 25–27), shuffles the list (line 29), adds the elements from the list to a tree (lines 32–33), shuf- fles the list again (line 35), removes the elements from the tree (lines 38–39), and finally returns the execution time (line 42).

The program creates an AVL (line 6), a 2-4 tree (line 11), and a red-black tree (line 16). The program obtains the execution time for adding and removing 500000 elements in the three trees.

As you see, the red-black tree performs the best, followed by the AVL tree.

Note The java.util.TreeSet class in the Java API is implemented using a red-black tree. Each entry in the set is stored in the tree. Since the search, insert, and delete methods in a red-black tree take O(log n) time, the get, add, remove, and contains methods in java.util.TreeSet take O(log n) time.

Note The java.util.TreeMap class in the Java API is implemented using a red-black tree. Each entry in the map is stored in the tree. The order of the entries is determined by their keys. Since the search, insert, and delete methods in a red-black tree take O(log n) time, the get, put, remove, and containsKey methods in java.util. TreeMap take O(log n) time.

Key Terms

black depth 43-2 double-black violation 43-11 double-red violation 43-7

external node 43-9 red-black tree 43-2

M43_LIAN1878_11_GE_C43.indd 29 12/29/17 6:06 PM

43-30 Chapter 43 Red-Black Trees

ChapTer summary 1. A red-black tree is a binary search tree, derived from a 2-4 tree. A red-black tree cor-

responds to a 2-4 tree. You can convert a red-black tree to a 2-4 tree or vice versa.

2. In a red-black tree, each node is colored red or black. The root is always black. Two adjacent nodes cannot be both red. All external nodes have the same black depth.

3. Since a red-black tree is a binary search tree, the RBTree class extends the BST class.

4. Searching an element in a red-black tree is the same as in binary search tree, since a red-black tree is a binary search tree.

5. A new element is always inserted as a leaf node. If the new node is the root, color it black. Otherwise, color it red. If the parent of the new node is red, we have to fix the double-red violation by reassigning the color and/or restructuring the tree.

6. If a node to be deleted is internal, find the rightmost node in its left subtree. Replace the element in the node with the element in the rightmost node. Delete the rightmost node.

7. If the external node to be deleted is red, simply reconnect the parent node of the external node with the child node of the external node.

8. If the external node to be deleted is black, you need to consider several cases to ensure that black height for external nodes in the tree is maintained correctly.

9. The height of a red-black tree is O(logn). So, the time complexities for the search, insert, and delete methods are O(logn).

Quiz Answer the quiz for this chapter online at the book Companion Website.

programming exerCises *43.1 (red-black tree to 2-4 tree) Write a program that converts a red-black tree to a 2-4

tree.

*43.2 (2-4 tree to red-black tree) Write a program that converts a red-black tree to a 2-4 tree.

***43.3 (red-black tree animation) Write a GUI program that animates the red-black tree insert, delete, and search methods, as shown in Figure 43.6.

**43.4 (Parent reference for RBTree) Suppose the TreeNode class defined in BST con- tains a reference to the node’s parent, as shown in Exercise 26.17. Implement the RBTree class to support this change. Write a test program that adds numbers 1, 2, . . . , 100 to the tree and displays the paths for all leaf nodes.

M43_LIAN1878_11_GE_C43.indd 30 12/29/17 6:06 PM

Objectives ■■ To know what JUnit is and how JUnit works (§44.2).

■■ To create and run a JUnit test class from the command window (§44.2).

■■ To create and run a JUnit test class from NetBeans (§44.3).

■■ To create and run a JUnit test class from Eclipse (§44.4).

Testing Using JUnit

Chapter

M44_LIAN1878_11_GE_C44.indd 1 12/29/17 6:07 PM

44-2 Chapter 44 Testing Using JUnit

44.1 Introduction JUnit is a tool for testing Java programs.

At the very beginning of this book in Section 2.16, we introduced software development pro- cess that includes requirements specification, analysis, design, implementation, testing, deploy- ment, and maintenance. Testing is an important part of this process. This chapter introduces how to test Java classes using JUnit.

44.2 JUnit Basics To test a class, you need to write a test class and run it through JUnit to generate a report for the class.

JUnit is the de facto framework for testing Java programs. JUnit is a third-party open-source library packed in a jar file. The jar file contains a tool called test runner, which is used to run test programs. Suppose you have a class named A. To test this class, you write a test class named ATest. This test class, called a test runner, contains the methods you write for testing class A. The test runner executes ATest to generate a test report, as shown in Figure 44.1.

Point Key

Figure 44.1 JUnit test runner executes the test class to generate a test report.

Test Runner

Test ReportTest Class File

… e.g., ATest.class

A.class …

Figure 44.2 The JUnit test runner displays the JUnit version.

You will see how JUnit works from an example. To create the example, first you need to download JUnit from http://sourceforge.net/projects/junit/files/. At present, the latest version is junit-4.10.jar. Download this file to c:\book\lib and add it to the classpath environment variable as follows:

set classpath=.;%classpath%;c:\book\lib\junit-4.10.jar

To test if this environment variable is set correctly, open a new command window, and type the following command:

java org.junit.runner.JUnitCore

You should see the message displayed as shown in Figure 44.2.

To use JUnit, create a test class. By convention, if the class to be tested is named A, the test class should be named ATest. A simple template of a test class may look like this:

M44_LIAN1878_11_GE_C44.indd 2 12/29/17 6:07 PM

44.2 JUnit Basics 44-3

1 package mytest; 2 3 import org.junit.*; 4 import static org.junit.Assert.*; 5 6 public class ATest { 7 @Test 8 public void m1() { 9 // Write a test method 10 } 11 12 @Test 13 public void m2() { 14 // Write another test method 15 } 16 17 @Before 18 public void setUp() throws Exception { 19 // Common objects used by test methods may be set up here 20 } 21 }

This class should be placed in a directory under mytest. Suppose the class is placed under c:\book\mytest. You need to compile it from the mytest directory and run it from c:\book as shown in the following screen shot.

Note the command to run the test from the console is:

java org.junit.runner.JUnitCore mytest.ATest

When this command is executed, JUnitCore controls the execution of ATest. It first executes the setUp() method to set up the common objects used for the test, and then executes test methods m1 and m2 in this order. You may define multiple test methods if desirable.

The following methods can be used to implement a test method:

assertTrue(booleanExpression) The method reports success if the booleanExpression evaluates true.

assertEquals(Object, Object) The method reports success if the two objects are the same using the equals method.

assertNull(Object) The method reports success if the object reference passed is null.

fail(String) The method causes the test to fail and prints out the string.

M44_LIAN1878_11_GE_C44.indd 3 12/29/17 6:07 PM

44-4 Chapter 44 Testing Using JUnit

Listing 44.1 is an example of a test class for testing java.util.ArrayList.

Listing 44.1 ArrayListTest.java 1 package mytest; 2 3 import org.junit.*; 4 import static org.junit.Assert.*; 5 import java.util.*; 6 7 public class ArrayListTest { 8 private ArrayList<String> list = new ArrayList<String>(); 9 10 @Before 11 public void setUp() throws Exception { 12 } 13 14 @Test 15 public void testInsertion() { 16 list.add("Beijing"); 17 assertEquals("Beijing", list.get(0)); 18 list.add("Shanghai"); 19 list.add("Hongkong"); 20 assertEquals("Hongkong", list.get(list.size() – 1)); 21 } 22 23 @Test 24 public void testDeletion() { 25 list.clear(); 26 assertTrue(list.isEmpty()); 27 28 list.add("A"); 29 list.add("B"); 30 list.add("C"); 31 list.remove("B"); 32 assertEquals(2, list.size()); 33 } 34 }

A test run of the program is shown in Figure 44.3. Note that you have to first compile Array- ListTest.java. The ArrayListTest class is placed in the mytest package. So you should place ArrayListTest.java in the directory named mytest.

Figure 44.3 The test report is displayed from running ArrayListTest.

M44_LIAN1878_11_GE_C44.indd 4 12/29/17 6:07 PM

44.2 JUnit Basics 44-5

No errors are reported in this JUnit run. If you mistakenly change

assertEquals(2, list.size());

in line 32 to

assertEquals(3, list.size());

Run ArrayListTest now. You will see an error reported as shown in Figure 44.4.

Figure 44.4 The test report reports an error.

You can define any number of test methods. In this example, the two test methods test- Insertion and testDeletion are defined. JUnit executes testInsertion and test- Deletion in this order.

Note The test class must be placed in a named package such as mytest in this example. The JUnit will not work if the test class is placed a default package.

Listing 44.2 gives a test class for testing the Loan class in Listing 10.2. For convenience, we create Loan.java in the same directory with LoanTest.java. The Loan class is shown in Listing 44.3.

Listing 44.2 LoanTest.java 1 package mytest; 2 3 import org.junit.*; 4 import static org.junit.Assert.*; 5 6 public class LoanTest { 7 @Before 8 public void setUp() throws Exception { 9 } 10 11 @Test 12 public void testPaymentMethods() { 13 double annualInterestRate = 2.5; 14 int numberOfYears = 5; 15 double loanAmount = 1000; 16 Loan loan = new Loan(annualInterestRate, numberOfYears, 17 loanAmount); 18

M44_LIAN1878_11_GE_C44.indd 5 12/29/17 6:07 PM

44-6 Chapter 44 Testing Using JUnit

19 assertTrue(loan.getMonthlyPayment() == 20 getMonthlyPayment(annualInterestRate, numberOfYears, 21 loanAmount)); 22 assertTrue(loan.getTotalPayment() == 23 getTotalPayment(annualInterestRate, numberOfYears, 24 loanAmount)); 25 } 26 27 /** Find monthly payment */ 28 private double getMonthlyPayment(double annualInterestRate, 29 int numberOfYears, double loanAmount) { 30 double monthlyInterestRate = annualInterestRate / 1200; 31 double monthlyPayment = loanAmount * monthlyInterestRate / (1 – 32 (1 / Math.pow(1 + monthlyInterestRate, numberOfYears * 12))); 33 return monthlyPayment; 34 } 35 36 /** Find total payment */ 37 public double getTotalPayment(double annualInterestRate, 38 int numberOfYears, double loanAmount) { 39 return getMonthlyPayment(annualInterestRate, numberOfYears, 40 loanAmount) * numberOfYears * 12; 41 } 42 }

Listing 44.3 Loan.java 1 package mytest; 2 3 public class Loan { 4 private double annualInterestRate; 5 private int numberOfYears; 6 private double loanAmount; 7 private java.util.Date loanDate; 8 9 /** Default constructor */ 10 public Loan() { 11 this(2.5, 1, 1000); 12 } 13 14 /** Construct a loan with specified annual interest rate, 15 number of years, and loan amount 16 */ 17 public Loan(double annualInterestRate, int numberOfYears, 18 double loanAmount) { 19 this.annualInterestRate = annualInterestRate; 20 this.numberOfYears = numberOfYears; 21 this.loanAmount = loanAmount; 22 loanDate = new java.util.Date(); 23 } 24 25 /** Return annualInterestRate */ 26 public double getAnnualInterestRate() { 27 return annualInterestRate; 28 } 29 30 /** Set a new annualInterestRate */ 31 public void setAnnualInterestRate(double annualInterestRate) { 32 this.annualInterestRate = annualInterestRate; 33 }

M44_LIAN1878_11_GE_C44.indd 6 12/29/17 6:07 PM

44.2 JUnit Basics 44-7

34 35 /** Return numberOfYears */ 36 public int getNumberOfYears() { 37 return numberOfYears; 38 } 39 40 /** Set a new numberOfYears */ 41 public void setNumberOfYears(int numberOfYears) { 42 this.numberOfYears = numberOfYears; 43 } 44 45 /** Return loanAmount */ 46 public double getLoanAmount() { 47 return loanAmount; 48 } 49 50 /** Set a newloanAmount */ 51 public void setLoanAmount(double loanAmount) { 52 this.loanAmount = loanAmount; 53 } 54 55 /** Find monthly payment */ 56 public double getMonthlyPayment() { 57 double monthlyInterestRate = annualInterestRate / 1200; 58 double monthlyPayment = loanAmount * monthlyInterestRate / (1 – 59 (1 / Math.pow(1 + monthlyInterestRate, numberOfYears * 12))); 60 return monthlyPayment; 61 } 62 63 /** Find total payment */ 64 public double getTotalPayment() { 65 double totalPayment = getMonthlyPayment() * numberOfYears * 12; 66 return totalPayment; 67 } 68 69 /** Return loan date */ 70 public java.util.Date getLoanDate() { 71 return loanDate; 72 } 73 }

The testPaymentMethods() in LoanTest creates an instance of Loan (line 16–17) and tests whether loan.getMonthlyPayment() returns the same value as getMonthlyPayment (annualInterestRate, numberOfYears, loanAmount). The latter method is defined in the LoanTest class (lines 28–34).

Figure 44.5 The JUnit test runner executes LoanTest and reports no errors.

M44_LIAN1878_11_GE_C44.indd 7 12/29/17 6:07 PM

44-8 Chapter 44 Testing Using JUnit

The testPaymentMethods() also tests whether loan.getTotalPayment() returns the same value as getTotalPayment(annualInterestRate, numberOfYears, loan- Amount). The latter method is defined in the LoanTest class (lines 37–41).

A sample run of the program is shown in Figure 44.5.

44.2.1 What is JUnit? 44.2.2 What is a JUnit test runner? 44.2.3 What is a test class? How do you create a test class? 44.2.4 How do you use the assertTrue method? 44.2.5 How do you use the assertEquals method?

44.3 Using JUnit from NetBeans JUnit is intergrated with NetBeans. Using NetBeans, the test program can be auto- matically generated and the test process can be automated.

An IDE such as NetBeans and Eclipse can greatly simplify the process for creating and run- ning test classes. This section introduces using JUnit from NetBeans, and the next section will introduce using JUnit from Eclipse.

If you are not familiar with NetBeans, see Supplement II.B. Assume you have installed NetBeans 8 or higher. Create a project named chapter44 as follows:

Step 1: Choose File, New Project to display the New Project dialog box.

Step 2: Choose Java in the Categories section and Java Application in the Projects section. Click Next to display the New Java Application dialog box.

Step 3: Enter chapter44 as the Project Name and c:\book as Project Location. Click Finish to create the project as shown in Figure 44.6.

To demonstrate how to create a test class, we first create a class to be tested. Let the class be Loan from Listing 10.2. Here are the steps to create the Loan class under chapter44.

Point Check

Point Key

Figure 44.6 A new project named chapter44 is created.

M44_LIAN1878_11_GE_C44.indd 8 12/29/17 6:07 PM

44.3 Using JUnit from NetBeans 44-9

Step 1: Right-click the project node chapter44 and choose New, Java Class to display the New Java Class dialog box.

Step 2: Enter Loan as Class Name and chapter44 in the Package field and click Finish to create the class.

Step 3: Copy the code in Listing 10.2 to the Loan class and make sure the first line is package chapter44, as shown in Figure 44.7.

Figure 44.7 The Loan class is created.

Figure 44.8 The Create Tests dialog box creates a Test class.

Now you can create a test class to test the Loan class as follows:

Step 1: Right-click Loan.java in the project to display a context menu and choose Tools, Create/Update Test to display the Create Test dialog box, as shown in Figure 44.8.

M44_LIAN1878_11_GE_C44.indd 9 12/29/17 6:07 PM

44-10 Chapter 44 Testing Using JUnit

Figure 44.9 You should select JUnit 4.x framework to create test classes.

Figure 44.10 The LoanTest class is automatically generated.

Step 2: Click OK. You will see the Select JUnit version dialog box displayed as shown in Figure 44.9. Choose Junit 4.x. Click OK to generate a test class named LoanTest as shown in Figure 44.10. Note that LoanTest.java is placed under the Test Packages node in the project.

M44_LIAN1878_11_GE_C44.indd 10 12/29/17 6:07 PM

44.4 Using JUnit from Eclipse 44-11

44.4 Using JUnit from Eclipse JUnit is intergrated with Eclipse. Using Eclipse, the test program can be automati- cally generated and the test process can be automated.

This section introduces using JUnit from Eclipse. If you are not familiar with Eclipse, see Supplement II.D. Assume you have installed Eclipse 4.5 or higher. Create a project named chapter50 as follows:

Step 1: Choose File, New Java Project to display the New Java Project dialog box, as shown in Figure 44.12.

Step 2: Enter chapter50 in the project name field and click Finish to create the project.

To demonstrate how to create a test class, we first create a class to be tested. Let the class be Loan from Listing 10.2. Here are the steps to create the Loan class under chapter44.

Point Key

Figure 44.11 The test report is displayed after the LoanTest class is executed.

You can now modify LoanTest by copying the code from Listing 44.2. Run LoanTest.java. You will see the test report as shown in Figure 44.11.

M44_LIAN1878_11_GE_C44.indd 11 12/29/17 6:07 PM

44-12 Chapter 44 Testing Using JUnit

Figure 44.12 The New Java Project dialog creates a new project.

Step 1: Right-click the project node chapter44 and choose New, Class to display the New Java Class dialog box, as shown in Figure 44.13.

Step 2: Enter mytest in the Package field and click Finish to create the class.

Step 3: Copy the code in Listing 10.2 to the Loan class and make sure the first line is package mytest, as shown in Figure 44.14.

Now you can create a test class to test the Loan class as follows:

Step 1: Right-click Loan.java in the project to display a context menu and choose New, JUnit Test Case to display the New JUnit Test Case dialog box, as shown in Figure 44.15.

Step 2: Click Finish. You will see a dialog prompting you to add JUnit 4 to the project build path. Click OK to add it. Now a test class named LoanTest is created as shown in Figure 44.16.

You can now modify LoanTest by copying the code from Listing 44.2. Run LoanTest.java. You will see the test report as shown in Figure 44.17.

JUnit 44-2 JUnitCore 44-2

test class 44-2 test runner 44-2

Key Terms

M44_LIAN1878_11_GE_C44.indd 12 12/29/17 6:07 PM

44.4 Using JUnit from Eclipse 44-13

Figure 44.13 The New Java Class dialog creates a new Java class.

Figure 44.14 The Loan class is created.

M44_LIAN1878_11_GE_C44.indd 13 12/29/17 6:07 PM

44-14 Chapter 44 Testing Using JUnit

Figure 44.15 The New JUnit Test Case dialog box creates a Test class.

Figure 44.16 The LoanTest class is automatically generated.

M44_LIAN1878_11_GE_C44.indd 14 12/29/17 6:07 PM

Programming Exercises 44-15

ChapTer summary 1. JUnit is an open-source framework for testing Java programs.

2. To test a Java class, you create a test class for the class to be tested and use JUnit’s test runner to execute the test class to generate a test report.

3. You can create and run a test class from the command window or use a tool such as NetBeans and Eclipse.

Quiz Answer the quiz for this chapter online at the book Companion Website.

programming exerCises 44.1 Write a test class to test the methods length, charAt, substring, and indexOf

in the java.lang.String class.

44.2 Write a test class to test the methods add, remove, addAll, removeAll, size, isEmpty, and contains in the java.util.HashSet class.

44.3 Write a test class to test the method isPrime in Listing 6.7, PrimeNumberMethod. java.

44.4 Write a test class to test the methods getBMI and getStatus in the BMI class in Listing 10.4.

Figure 44.17 The test report is displayed after the LoanTest class is executed.

M44_LIAN1878_11_GE_C44.indd 15 12/29/17 6:07 PM

A01_GIPE3898_09_SE_FM.indd 3 12/20/17 5:56 PM

This page intentionally left blank

1177

aPPendIXeS appendix a Java Keywords

appendix B The ASCII Character Set

appendix c Operator Precedence Chart

appendix d Java Modifiers

appendix e Special Floating-Point Values

appendix f Number Systems

appendix g Bitwise Operations

appendix h Regular Expressions

appendix I Enumerated Types

Z01_LIAN1878_11_GE_APPA.indd 1177 12/29/17 4:54 PM

A01_GIPE3898_09_SE_FM.indd 3 12/20/17 5:56 PM

This page intentionally left blank

1179

Java Keywords The following 50 keywords are reserved for use by the Java language:

Appendix A

abstract

assert

boolean

break

byte

case

catch

char

class

const

continue

default

double

else

enum

extends

final

finally

float

for

goto

implements

import

instanceof

int

interface

long

native

new

package

private

protected

public

return

short

static

strictfp*

super

switch

synchronized

this

throw

throws

transient

try

void

volatile

while

*The strictfp keyword is a modifier for a method or class that enables it to use strict floating-point calculations. Floating-point arithmetic can be executed in one of two modes: strict or nonstrict. The strict mode guarantees that the evaluation result is the same on all Java Virtual Machine implementations. The non- strict mode allows intermediate results from calculations to be stored in an extended format different from the standard IEEE floating-point number format. The extended format is machine dependent and enables code to be executed faster. However, when you execute the code using the nonstrict mode on different JVMs, you may not always get precisely the same results. By default, the nonstrict mode is used for floating-point calculations. To use the strict mode in a method or a class, add the strictfp keyword in the method or the class declaration. Strict floating-point may give you slightly better precision than nonstrict floating-point, but the distinction will only affect some applications. Strictness is not inherited; that is, the presence of strictfp on a class or interface declaration does not cause extended classes or interfaces to be strict.

The keywords goto and const are C+ + keywords reserved, but not currently used in Java. This enables Java compilers to identify them and to produce better error messages if they appear in Java programs.

The literal values true, false, and null are not keywords, just like literal value 100. However, you cannot use them as identifiers, just as you cannot use 100 as an identifier.

In the code listing, we use the keyword color for true, false, and null to be consistent with their coloring in Java IDEs.

Z01_LIAN1878_11_GE_APPA.indd 1179 12/29/17 4:54 PM

1180

The ASCII Character Set Tables B.1 and B.2 show ASCII characters and their respective decimal and hexadecimal codes. The decimal or hexadecimal code of a character is a combination of its row index and column index. For example, in Table B.1, the letter A is at row 6 and column 5, so its decimal equivalent is 65; in Table B.2, letter A is at row 4 and column 1, so its hexadecimal equivalent is 41.

Appendix B

Z02_LIAN1878_11_GE_APPB.indd 1180 12/29/17 4:56 PM

T a

bl e

b. 1

A SC

II C

ha ra

ct er

S et

in th

e D

ec im

al In

de x

0 1

2 3

4 5

6 7

8 9

0 nu

l so

h st

x et

x eo

t en

q ac

k be

l bs

1 nl

vt ff

cr so

si dl

e dc

l dc

2 dc

4 na

k sy

n et

b ca

n em

su b

es c

fs gs

3 rs

us sp

! ”

# $

% &

’

4 (

) *

+ ,

- .

/ 0

5 2

3 4

5 6

7 8

9 :

;

6 6

= 7

? @

A B

C D

7 F

G H

I J

K L

M N

8 P

Q R

S T

U V

W X

9 Z

[ \

] ¿

- ’

a b

10 d

e f

g h

i j

k l

11 n

o p

q r

s t

u v

12 x

y z

{ |

} ∼

de l

T a

bl e

b. 2

A SC

II C

ha ra

ct er

S et

in th

e H

ex ad

ec im

al In

de x

0 1

2 3

4 5

6 7

8 9

A B

C D

E F

0 nu

l so

h st

x et

x eo

t en

q ac

k be

l bs

ht nl

vt ff

cr so

1 dl

e dc

l dc

2 dc

3 dc

4 na

k sy

n et

b ca

n em

su b

es c

fs gs

rs us

2 sp

! ”

# $

% &

’ (

) *

+ ,

- .

3 0

1 2

3 4

5 6

7 8

9 :

; 6

= 7

4 @

A B

C D

E F

G H

I J

K L

M N

5 P

Q R

S T

U V

W X

Y Z

[ \

] ¿

6 ’

a b

c d

e f

g h

i j

k l

m n

7 p

q r

s t

u v

w x

y z

{ |

} ∼

de l

1181

Z02_LIAN1878_11_GE_APPB.indd 1181 12/29/17 4:56 PM

1182

Operator Precedence Chart The operators are shown in decreasing order of precedence from top to bottom. Operators in the same group have the same precedence, and their associativity is shown in the table.

Appendix C

Operator Name Associativity

() Parentheses Left to right

() Function call Left to right

[] Array subscript Left to right

. Object member access Left to right

++ Postincrement Left to right

–– Postdecrement Left to right

++ Preincrement Right to left

–– Predecrement Right to left

+ Unary plus Right to left

– Unary minus Right to left

! Unary logical negation Right to left

(type) Unary casting Right to left

new Creating object Right to left

* Multiplication Left to right

/ Division Left to right

% Remainder Left to right

+ Addition Left to right

– Subtraction Left to right

<< Left shift Left to right

>> Right shift with sign extension Left to right

>>> Right shift with zero extension Left to right

< Less than Left to right

<= Less than or equal to Left to right

> Greater than Left to right

>= Greater than or equal to Left to right

instanceof Checking object type Left to right

Z03_LIAN1878_11_GE_APPC.indd 1182 12/29/17 4:57 PM

Operator Name Associativity

== Equal comparison Left to right

!= Not equal Left to right

& (Unconditional AND) Left to right

^ (Exclusive OR) Left to right

| (Unconditional OR) Left to right

&& Conditional AND Left to right

|| Conditional OR Left to right

?: Ternary condition Right to left

= Assignment Right to left

+= Addition assignment Right to left

–= Subtraction assignment Right to left

*= Multiplication assignment Right to left

/= Division assignment Right to left

%= Remainder assignment Right to left

Appendix C 1183

Z03_LIAN1878_11_GE_APPC.indd 1183 12/29/17 4:57 PM

1184

Java Modifiers Modifiers are used on classes and class members (constructors, methods, data, and class-level blocks), but the final modifier can also be used on local variables in a method. A modifier that can be applied to a class is called a class modifier. A modifier that can be applied to a method is called a method modifier. A modifier that can be applied to a data field is called a data modifier. A modifier that can be applied to a class-level block is called a block modifier. The following table gives a summary of the Java modifiers.

Appendix D

Modifier Class Constructor Method Data Block Explanation

(default)* √ √ √ √ √ A class, constructor, method, or data field is visible in this package.

public √ √ √ √ A class, constructor, method, or data field is visible to all the programs in any package.

private √ √ √ A constructor, method, or data field is only visible in this class.

protected √ √ √ A constructor, method, or data field is visible in this package and in subclasses of this class in any package.

static √ √ √ Define a class method, a class data field, or a static initializa- tion block.

final √ √ √ A final class cannot be extended. A final method cannot be modified in a subclass. A final data field is a constant.

abstract √ √ An abstract class must be extended. An abstract method must be implemented in a con- crete subclass.

native √ A native method indicates that the method is implemented using a language other than Java.

* Default access doesn’t have a modifier associated with it. For example: class Test {}

Z04_LIAN1878_11_GE_APPD.indd 1184 1/2/18 10:49 PM

Modifier Class Constructor Method Data Block Explanation

synchronized √ √ Only one thread at a time can execute this method.

strictfp √ √ Use strict floating-point calculations to guarantee that the evaluation result is the same on all JVMs.

transient √ Mark a nonserializable instance data field.

Appendix D 1185

The modifiers default (no modifier), public, private, and protected are known as visibility or accessibility modifiers because they specify how classes and class members are accessed.

The modifiers public, private, protected, static, final, and abstract can also be applied to inner classes.

Z04_LIAN1878_11_GE_APPD.indd 1185 1/2/18 10:49 PM

1186

Special Floating-Point Values Dividing an integer by zero is invalid and throws ArithmeticException, but dividing a floating-point value by zero does not cause an exception. Floating-point arithmetic can over- flow to infinity if the result of the operation is too large for a double or a float, or under- flow to zero if the result is too small for a double or a float. Java provides the special floating-point values POSITIVE_INFINITY, NEGATIVE_INFINITY, and NaN (Not a Number) to denote these results. These values are defined as special constants in the Float class and the Double class.

If a positive floating-point number is divided by zero, the result is POSITIVE_INFINITY. If a negative floating-point number is divided by zero, the result is NEGATIVE_INFINITY. If a floating-point zero is divided by zero, the result is NaN, which means that the result is undefined mathematically. The string representations of these three values are Infinity, -Infinity, and NaN. For example,

System.out.print(1.0 / 0); // Print Infinity System.out.print(–1.0 / 0); // Print –Infinity System.out.print(0.0 / 0); // Print NaN

These special values can also be used as operands in computations. For example, a number divided by POSITIVE_INFINITY yields a positive zero. Table E.1 summarizes various com- binations of the /, *, %, +, and – operators.

Appendix E

Table e.1 Special Floating-Point Values

x y x/y x*y x%y x + y x - y

Finite { 0.0 { infinity { 0.0 NaN Finite Finite Finite { infinity { 0.0 { 0.0 x { infinity infinity

{ 0.0 { 0.0 NaN { 0.0 NaN { 0.0 { 0.0

{ infinity Finite { infinity { 0.0 NaN { infinity { infinity

{ infinity { infinity NaN { 0.0 NaN { infinity infinity

{ 0.0 { infinity { 0.0 NaN { 0.0 { infinity { 0.0

NaN Any NaN NaN NaN NaN NaN

Any NaN NaN NaN NaN NaN NaN

Note If one of the operands is NaN, the result is NaN.

Z05_LIAN1878_11_GE_APPE.indd 1186 12/29/17 4:59 PM

1187

Number Systems F.1 Introduction Computers use binary numbers internally, because computers are made naturally to store and process 0s and 1s. The binary number system has two digits, 0 and 1. A number or character is stored as a sequence of 0s and 1s. Each 0 or 1 is called a bit (binary digit).

In our daily life, we use decimal numbers. When we write a number such as 20 in a program, it is assumed to be a decimal number. Internally, computer software is used to convert decimal numbers into binary numbers, and vice versa.

We write computer programs using decimal numbers. However, to deal with an operating system, we need to reach down to the “machine level” by using binary numbers. Binary num- bers tend to be very long and cumbersome. Often hexadecimal numbers are used to abbreviate them, with each hexadecimal digit representing four binary digits. The hexadecimal number system has 16 digits: 0–9 and A–F. The letters A, B, C, D, E, and F correspond to the decimal numbers 10, 11, 12, 13, 14, and 15.

The digits in the decimal number system are 0, 1, 2, 3, 4, 5, 6, 7, 8, and 9. A decimal number is represented by a sequence of one or more of these digits. The value that each digit represents depends on its position, which denotes an integral power of 10. For example, the digits 7, 4, 2, and 3 in decimal number 7423 represent 7000, 400, 20, and 3, respectively, as shown below:

= 7 * 103 + 4 * 102 + 2 * 101 + 3 * 100

103 102 101 100 = 7000 + 400 + 20 + 3 = 7423

The decimal number system has 10 digits, and the position values are integral powers of 10. We say that 10 is the base or radix of the decimal number system. Similarly, since the binary number system has two digits, its base is 2, and since the hex number system has 16 digits, its base is 16.

If 1101 is a binary number, the digits 1, 1, 0, and 1 represent 1 * 23, 1 * 22, 0 * 21, and 1 * 20, respectively:

= 1 * 23 + 1 * 22 + 0 * 21 + 1 * 20

23 22 21 20 = 8 + 4 + 0 + 1 = 13

If 7423 is a hex number, the digits 7, 4, 2, and 3 represent 7 * 163, 4 * 162, 2 * 161, and 3 * 160, respectively:

7 4 2 3 = 7 * 163 + 4 * 162 + 2 * 161 + 3 * 160

163 162 161 160 = 28672 + 1024 + 32 + 3 = 29731

base radix

hexadecimal number

decimal numbers

7 4 2 3

binary numbers

1 1 0 1

Appendix F

Z06_LIAN1878_11_GE_APPF.indd 1187 12/29/17 5:00 PM

1188 Appendix F

F.2 Conversions between Binary and Decimal Numbers Given a binary number bnbn - 1bn - 2 c b2b1b0, the equivalent decimal value is

bn * 2n + bn - 1 * 2n - 1 + bn - 2 * 2n - 2 + c + b2 * 22 + b1 * 21 + b0 * 20

Here are some examples of converting binary numbers to decimals:

binary to decimal

2 1

2 3

2 7

Quotient

Remainder

122

12330

Figure F.1 You can perform number conversions using the Windows Calculator.

BinaryDecimal

Hex

Binary Conversion Formula Decimal

10 1 * 21 + 0 * 20 2

1000 1 * 23 + 0 * 22 + 0 * 21 + 0 * 20 8

10101011 1 * 27 + 0 * 26 + 1 * 25 + 0 * 24 + 1 * 23 + 0 * 22 + 1 * 21 + 1 * 20

171

To convert a decimal number d to a binary number is to find the bits bn, bn - 1, bn - 2, c , b2, b1 and b0 such that

d = bn * 2n + bn - 1 * 2n - 1 + bn - 2 * 2n - 2 + c + b2 * 22 + b1 * 21 + b0 * 20

These bits can be found by successively dividing d by 2 until the quotient is 0. The remain- ders are b0, b1, b2, c , bn - 2, bn - 1, and bn.

For example, the decimal number 123 is 1111011 in binary. The conversion is done as follows:

decimal to binary

Tip The Windows Calculator, as shown in Figure F.1, is a useful tool for performing number conversions. To run it, search for Calculator from the Start button and launch Calculator, then under View select Scientific.

Z06_LIAN1878_11_GE_APPF.indd 1188 12/29/17 5:00 PM

Appendix F 1189

F.3 Conversions between Hexadecimal and Decimal Numbers Given a hexadecimal number hnhn - 1hn - 2 c h2h1h0, the equivalent decimal value is

hn * 16n + hn - 1 * 16n - 1 + hn - 2 * 16n - 2 + c + h2 * 162 + h1 * 161 + h0 * 160

Here are some examples of converting hexadecimal numbers to decimals:

hex to decimal

Quotient

Remainder

16 7

16 123

112

1 1 1 0 0 0 1 1 0 1

D83

F.4 Conversions between Binary and Hexadecimal Numbers To convert a hexadecimal number to a binary number, simply convert each digit in the hexa- decimal number into a four-digit binary number, using Table F.1.

For example, the hexadecimal number 7B is 1111011, where 7 is 111 in binary and B is 1011 in binary.

To convert a binary number to a hexadecimal number, convert every four binary digits from right to left in the binary number into a hexadecimal number.

For example, the binary number 1110001101 is 38D, since 1101 is D, 1000 is 8, and 11 is 3, as shown below.

hex to binary

binary to hex

Hexadecimal Conversion Formula Decimal

7F 7 * 161 + 15 * 160 127

FFFF 15 * 163 + 15 * 162 + 15 * 161 + 15 * 160 65535

431 4 * 162 + 3 * 161 + 1 * 160 1073

To convert a decimal number d to a hexadecimal number is to find the hexadecimal digits hn, hn - 1, hn - 2, c , h2, h1, and h0 such that

d = hn * 16n + hn - 1 * 16n - 1 + hn - 2 * 16n - 2 + c + h2 * 162

+ h1 * 161 + h0 * 160

These numbers can be found by successively dividing d by 16 until the quotient is 0. The remainders are h0, h1, h2, c , hn - 2, hn - 1, and hn.

For example, the decimal number 123 is 7B in hexadecimal. The conversion is done as follows:

decimal to hex

Z06_LIAN1878_11_GE_APPF.indd 1189 12/29/17 5:00 PM

1190 Appendix F

Note Octal numbers are also useful. The octal number system has eight digits, 0 to 7. A deci- mal number 8 is represented in the octal system as 10.

Here are some good online resources for practicing number conversions:

■■ http://forums.cisco.com/CertCom/game/binary_game_page.htm

■■ http://people.sinclair.edu/nickreeder/Flash/binDec.htm

■■ http://people.sinclair.edu/nickreeder/Flash/binHex.htm

F.1 Convert the following decimal numbers into hexadecimal and binary numbers: 100; 4340; 2000

F.2 Convert the following binary numbers into hexadecimal and decimal numbers: 1000011001; 100000000; 100111

F.3 Convert the following hexadecimal numbers into binary and decimal numbers: FEFA9; 93; 2000

Point Check

Hexadecimal Binary Decimal

0 0000 0

1 0001 1

2 0010 2

3 0011 3

4 0100 4

5 0101 5

6 0110 6

7 0111 7

8 1000 8

9 1001 9

A 1010 10

B 1011 11

C 1100 12

D 1101 13

E 1110 14

F 1111 15

Table F.1 Converting Hexadecimal to Binary

Z06_LIAN1878_11_GE_APPF.indd 1190 12/29/17 5:00 PM

1191

Bitwise Operations To write programs at the machine-level, often you need to deal with binary numbers directly and perform operations at the bit level. Java provides the bitwise operators and shift operators defined in Table G.1.

The bit operators apply only to integer types (byte, short, int, and long). A character involved in a bit operation is converted to an integer. All bitwise operators can form bitwise assignment operators, such as =, |=, <<=, >>=, and >>>=.

Appendix G

Table g.1

Operator Name Example

(using bytes in the example) Description

& Bitwise AND 10101110 & 10010010 yields 10000010

The AND of two corresponding bits yields a 1 if both bits are 1.

| Bitwise inclusive OR

10101110 | 10010010 yields 10111110

The OR of two corresponding bits yields a 1 if either bit is 1.

^ Bitwise exclusive OR

10101110 ^ 10010010 yields 00111100

The XOR of two corresponding bits yields a 1 only if two bits are different.

~ One's complement

~10101110 yields 01010001

The operator toggles each bit from 0 to 1 and from 1 to 0.

<< Left shift 10101110 << 2 yields 10111000

The operator shifts bits in the first operand left by the number of bits specified in the second operand, filling with 0s on the right.

>> Right shift with sign extension

10101110 >> 2 yields 11101011

00101110 >> 2 yields 00001011

The operator shifts bit in the first operand right by the number of bits specified in the second operand, filling with the highest (sign) bit on the left.

>>> Unsigned right shift with zero extension

10101110 >>> 2 yields 00101011

00101110 >>> 2 yields 00001011

The operator shifts bit in the first operand right by the number of bits specified in the second operand, filling with 0s on the left.

Z07_LIAN1878_11_GE_APPG.indd 1191 1/2/18 10:50 PM

1192

Regular Expressions Often, you need to write the code to validate user input such as to check whether the input is a number, a string with all lowercase letters, or a Social Security number. How do you write this type of code? A simple and effective way to accomplish this task is to use the regular expression.

A regular expression (abbreviated regex) is a string that describes a pattern for matching a set of strings. Regular expression is a powerful tool for string manipulations. You can use regular expressions for matching, replacing, and splitting strings.

H.1 Matching Strings Let us begin with the matches method in the String class. At first glance, the matches method is very similar to the equals method. For example, the following two statements both evaluate to true:

"Java".matches("Java"); "Java".equals("Java");

However, the matches method is more powerful. It can match not only a fixed string, but also a set of strings that follow a pattern. For example, the following statements all evaluate to true:

"Java is fun".matches("Java.*") "Java is cool".matches("Java.*") "Java is powerful".matches("Java.*")

"Java.*" in the preceding statements is a regular expression. It describes a string pattern that begins with Java followed by any zero or more characters. Here, the substring .* matches any zero or more characters.

H.2 Regular Expression Syntax A regular expression consists of literal characters and special symbols. Table H.1 lists some frequently used syntax for regular expressions.

Note Backslash is a special character that starts an escape sequence in a string. So you need to use \\ to represent a literal character \.

Note Recall that a whitespace character is ' ', '\t', '\n', '\r', or '\f'. So \s is the same as [\t\n\r\f], and \S is the same as [^ \t\n\r\f].

regular expression

matches

Appendix H

Z08_LIAN1878_11_GE_APPH.indd 1192 1/2/18 10:51 PM

Table H.1 Frequently Used Regular Expressions

Regular Expression Matches Example

x a specified character x Java matches Java

. any single character Java matches J..a

(ab|cd) ab or cd ten matches t(en|im)

[abc] a, b, or c Java matches Ja[uvwx]a

[^abc] any character except a, b, or c

Java matches Ja[^ars]a

[a−z] a through z Java matches [A-M]av[a-d]

[^a−z] any character except a through z

Java matches Jav[^b-d]

[a-e[m−p]] a through e or m through p Java matches [A-G[I-M]]av[a-d]

[a-e&&[c−p]] intersection of a-e with c-p

Java matches [A-P&&[I-M]]av[a-d]

\d a digit, same as [0−9] Java2 matches "Java[\\d]"

\D a non-digit $Java matches "[\\D][\\ D]ava"

\w a word character Java1 matches "[\\w]ava[\\w]"

\W a non-word character $Java matches "[\\W][\\ w]ava"

\s a whitespace character "Java 2" matches "Java\\s2"

\S a non-whitespace char Java matches "[\\S]ava"

p* zero or more occurrences of pattern p

aaaa matches "a*" abab matches "(ab)*"

p+ one or more occurrences of pattern p

Java matches "a+" bbb matches "a+"

p? zero or one occurrence of pattern p

Java matches "J?Java" ava matches "J?ava"

p{n} exactly n occurrences of pattern p

Java matches "a{1}" Java does not match "a{2}"

p{n,} at least n occurrences of pattern p

Java matches "a{1,}" Java does not match "a{2,}"

p{n,m} between n and m occurrences (inclusive)

Java matches "a{1,9}" Java does not match "a{2,9}"

\p{P} a punctuation character !"#$%&’()*+, -./:;<=>?@ [\]^_‘{|}~

J?a matches "J\p{P}a" J?a. does not match "J\p{P}a"

Appendix H 1193

Note A word character is any letter, digit, or the underscore character. So \w is the same as [a−z[A−Z][0−9]_] or simply [a−zA−Z0−9_], and \W is the same as [^a−zA−Z0−9_].

Note The last six entries *, +, ?, {n}, {n,}, and {n, m} in Table H.1 are called quantifiers that specify how many times the pattern before a quantifier may repeat. For example, A* matches zero or more A’s, A+ matches one or more A’s, A? matches zero or one A, A{3} matches exactly AAA, A{3,} matches at least three A’s, and A{3,6} matches between 3 and 6 A’s. * is the same as {0,}, + is the same as {1,}, and ? is the same as {0,1}.

quantifier

Z08_LIAN1878_11_GE_APPH.indd 1193 1/2/18 10:51 PM

1194 Appendix H

Caution Do not use spaces in the repeat quantifiers. For example, A{3,6} cannot be written as A{3, 6} with a space after the comma.

Note You may use parentheses to group patterns. For example, (ab){3} matches ababab, but ab{3} matches abbb.

Let us use several examples to demonstrate how to construct regular expressions.

Example 1 The pattern for Social Security numbers is xxx−xx−xxxx, where x is a digit. A regular expression for Social Security numbers can be described as

[\\d]{3}-[\\d]{2}-[\\d]{4}

For example,

"111−22−3333".matches("[\\d]{3}-[\\d]{2}−[\\d]{4}") returns true. "11−22−3333".matches("[\\d]{3}-[\\d]{2}−[\\d]{4}") returns false.

Example 2 An even number ends with digits 0, 2, 4, 6, or 8. The pattern for even numbers can be described as

[\\d]*[02468]

For example,

"123".matches("[\\d]*[02468]") returns false. "122".matches("[\\d]*[02468]") returns true.

Example 3 The pattern for telephone numbers is (xxx) xxx-xxxx, where x is a digit and the first digit cannot be zero. A regular expression for telephone numbers can be described as

\\ ([1−9][\\ d]{2}\\) [\\d]{3}−[\\d]{4}

Note the parentheses symbols ( and ) are special characters in a regular expression for grouping patterns. To represent a literal ( or ) in a regular expression, you have to use \$ and \$. For example,

"(912) 921-2728".matches("\\ ([1−9][\\ d]{2}\\) [\\d]{3}−[\\d]{4}") returns true. "921-2728".matches("\\ ([1−9][\\ d]{2}\\) [\\d]{3}−[\\d]{4}") returns false.

Example 4 Suppose the last name consists of at most 25 letters, and the first letter is in uppercase. The pattern for a last name can be described as

[A−Z][a−zA−Z]{1,24}

Note you cannot have arbitrary whitespace in a regular expression. For example, [A−Z] [a-zA-Z]{1, 24} would be wrong.

Z08_LIAN1878_11_GE_APPH.indd 1194 1/2/18 10:51 PM

For example,

"Smith".matches("[A-Z][a-zA-Z]{1,24}") returns true. "Jones123".matches("[A-Z][a-zA-Z]{1,24}") returns false.

Example 5 Java identifiers are defined in Section 2.3, Identifiers.

■■ An identifier must start with a letter, an underscore (_), or a dollar sign ($). It cannot start with a digit.

■■ An identifier is a sequence of characters that consists of letters, digits, underscores (_), and dollar signs ($).

The pattern for identifiers can be described as

[a−zA−Z_$][\\w$]*

Example 6 What strings are matched by the regular expression "Welcome to (Java|HTML)"? The answer is Welcome to Java or Welcome to HTML.

Example 7 What strings are matched by the regular expression ".*"? The answer is any string.

H.3 Replacing and Splitting Strings The matches method in the String class returns true if the string matches the regular expression. The String class also contains the replaceAll, replaceFirst, and split methods for replacing and splitting strings, as shown in Figure H.1.

Appendix H 1195

Figure H.1 The String class contains the methods for matching, replacing, and splitting strings using regular expressions.

java.lang.String

+matches(regex: String): boolean +replaceAll(regex: String, replacement: String): String

+replaceFirst(regex: String, replacement: String): String

+split(regex: String): String[]

+split(regex: String, limit: int): String[]

Returns true if this string matches the pattern. Returns a new string that replaces all matching substrings with the replacement.

Returns a new string that replaces the �rst matching substring with the replacement.

Returns an array of strings consisting of the substrings split by the matches.

Same as the preceding split method except that the limit parameter controls the number of times the pattern is applied.

The replaceAll method replaces all matching substring, and the replaceFirst method replaces the first matching substring. For example, the code

System.out.println("Java Java Java".replaceAll("v\\w", "wi"));

displays

Jawi Jawi Jawi

Z08_LIAN1878_11_GE_APPH.indd 1195 1/2/18 10:51 PM

1196 Appendix H

and this code

System.out.println("Java Java Java".replaceFirst("v\\w", "wi"));

displays

Jawi Java Java

There are two overloaded split methods. The split(regex) method splits a string into substrings delimited by the matches. For example, the statement

String[] tokens = "Java1HTML2Perl".split("\\d");

splits string "Java1HTML2Perl" into Java, HTML, and Perl and saves in tokens[0], tokens[1], and tokens[2].

In the split(regex, limit) method, the limit parameter determines how many times the pattern is matched. If limit <= 0, split(regex, limit) is same as split(regex). If limit > 0, the pattern is matched at most limit −1 times. Here are some examples:

"Java1HTML2Perl".split("\\d", 0); splits into Java, HTML, Perl "Java1HTML2Perl".split("\\d", 1); splits into Java1HTML2Perl "Java1HTML2Perl".split("\\d", 2); splits into Java, HTML2Perl "Java1HTML2Perl".split("\\d", 3); splits into Java, HTML, Perl "Java1HTML2Perl".split("\\d", 4); splits into Java, HTML, Perl "Java1HTML2Perl".split("\\d", 5); splits into Java, HTML, Perl

Note By default, all the quantifiers are greedy. This means that they will match as many occurrences as possible. For example, the following statement displays JRvaa, since the first match is aaa:

System.out.println("Jaaavaa".replaceFirst("a+", "R"));

You can change a qualifier’s default behavior by appending a question mark (?) after it. The quantifier becomes reluctant or lazy, which means that it will match as few occur- rences as possible. For example, the following statement displays JRaavaa, since the first match is a:

System.out.println("Jaaavaa".replaceFirst("a+?", "R"));

Z08_LIAN1878_11_GE_APPH.indd 1196 1/2/18 10:51 PM

1197

Enumerated Types I.1 Simple Enumerated Types An enumerated type defines a list of enumerated values. Each value is an identifier. For exam- ple, the following statement declares a type, named MyFavoriteColor, with values RED, BLUE, GREEN, and YELLOW in this order:

enum MyFavoriteColor {RED, BLUE, GREEN, YELLOW};

A value of an enumerated type is like a constant and so, by convention, is spelled with all uppercase letters. So, the preceding declaration uses RED, not red. By convention, an enumer- ated type is named like a class with first letter of each word capitalized.

Once a type is defined, you can declare a variable of that type:

MyFavoriteColor color;

The variable color can hold one of the values defined in the enumerated type MyFavoriteColor or null, but nothing else. Java enumerated type is type-safe, meaning that an attempt to assign a value other than one of the enumerated values or null will result in a compile error.

The enumerated values can be accessed using the syntax

EnumeratedTypeName.valueName

For example, the following statement assigns enumerated value BLUE to variable color:

color = MyFavoriteColor.BLUE;

Note you have to use the enumerated type name as a qualifier to reference a value such as BLUE.

As with any other type, you can declare and initialize a variable in one statement:

MyFavoriteColor color = MyFavoriteColor.BLUE;

An enumerated type is treated as a special class. An enumerated type variable is therefore a reference variable. An enumerated type is a subtype of the Object class and the Comparable interface. Therefore, an enumerated type inherits all the methods in the Object class and the compareTo method in the Comparable interface. Additionally, you can use the following methods on an enumerated object:

■■ public String name(); Returns a name of the value for the object.

■■ public int ordinal(); Returns the ordinal value associated with the enumerated value. The first value in an enumerated type has an ordinal value of 0, the second has an ordinal value of 1, the third one 3, and so on.

Appendix I

Z09_LIAN1878_11_GE_APPI.indd 1197 1/2/18 10:53 PM

Listing I.1 gives a program that demonstrates the use of enumerated types.

lisTing I.1 EnumeratedTypeDemo.java 1 public class EnumeratedTypeDemo { 2 static enum Day {SUNDAY, MONDAY, TUESDAY, WEDNESDAY, THURSDAY, 3 FRIDAY, SATURDAY}; 4 5 public static void main(String[] args) { 6 Day day1 = Day.FRIDAY; 7 Day day2 = Day.THURSDAY; 8 9 System.out.println("day1’s name is " + day1.name()); 10 System.out.println("day2’s name is " + day2.name()); 11 System.out.println("day1’s ordinal is " + day1.ordinal()); 12 System.out.println("day2’s ordinal is " + day2.ordinal()); 13 14 System.out.println("day1.equals(day2) returns " + 15 day1.equals(day2)); 16 System.out.println("day1.toString() returns " + 17 day1.toString()); 18 System.out.println("day1.compareTo(day2) returns " + 19 day1.compareTo(day2)); 20 } 21 }

day1’s name is FRIDAY day2’s name is THURSDAY day1’s ordinal is 5 day2’s ordinal is 4 day1.equals(day2) returns false day1.toString() returns FRIDAY day1.compareTo(day2) returns 1

An enumerated type Day is defined in lines 2 and 3. Variables day1 and day2 are declared as the Day type and assigned enumerated values in lines 6 and 7. Since day1’s value is FRIDAY, its ordinal value is 5 (line 11). Since day2’s value is THURSDAY, its ordinal value is 4 (line 12).

Since an enumerated type is a subclass of the Object class and the Comparable interface, you can invoke the methods equals, toString, and compareTo from an enumerated object reference variable (lines 14–19). day1.equals(day2) returns true if day1 and day2 have the same ordinal value. day1.compareTo(day2) returns the difference between day1’s ordinal value and day2’s.

Alternatively, you can rewrite the code in Listing I.1 into Listing I.2.

lisTing I.2 StandaloneEnumTypeDemo.java 1 public class StandaloneEnumTypeDemo { 2 public static void main(String[] args) { 3 Day day1 = Day.FRIDAY; 4 Day day2 = Day.THURSDAY; 5 6 System.out.println("day1’s name is " + day1.name()); 7 System.out.println("day2’s name is " + day2.name()); 8 System.out.println("day1’s ordinal is " + day1.ordinal());

1198 Appendix I

define an enum type

declare an enum variable

get enum name

get enum ordinal

compare enum values

Z09_LIAN1878_11_GE_APPI.indd 1198 1/2/18 10:53 PM

9 System.out.println("day2’s ordinal is " + day2.ordinal()); 10 11 System.out.println("day1.equals(day2) returns " + 12 day1.equals(day2)); 13 System.out.println("day1.toString() returns " + 14 day1.toString()); 15 System.out.println("day1.compareTo(day2) returns " + 16 day1.compareTo(day2)); 17 } 18 } 19 20 enum Day {SUNDAY, MONDAY, TUESDAY, WEDNESDAY, THURSDAY, 21 FRIDAY, SATURDAY}

An enumerated type can be defined inside a class, as shown in lines 2 and 3 in Listing I.1, or standalone as shown in lines 20 and 21 in Listing I.2. In the former case, the type is treated as an inner class. After the program is compiled, a class named EnumeratedTypeDemo$Day. class is created. In the latter case, the type is treated as a stand-alone class. After the program is compiled, a class named Day.class is created.

Note When an enumerated type is declared inside a class, the type must be declared as a member of the class and cannot be declared inside a method. Furthermore, the type is always static. For this reason, the static keyword in line 2 in Listing I.1 may be omitted. The visibility modifiers on inner class can be also be applied to enumerated types defined inside a class.

Tip Using enumerated values (e.g., Day.MONDAY, Day.TUESDAY, and so on) rather than literal integer values (e.g., 0, 1, and so on) can make the program easier to read and maintain.

I.2 Using if or switch Statements with an Enumerated Variable An enumerated variable holds a value. Often, your program needs to perform a specific action depending on the value. For example, if the value is Day.MONDAY, play soccer; if the value is Day.TUESDAY, take piano lesson, and so on. You can use an if statement or a switch state- ment to test the value in the variable, as shown in (a) and (b).

if (day.equals(Day.MONDAY)) { // process Monday } else if (day.equals(Day.TUESDAY)) { // process Tuesday } else ...

switch (day) { case MONDAY: // process Monday break; case TUESDAY: // process Tuesday break; ... }

Equivalent

(a) (b)

In the switch statement in (b), the case label is an unqualified enumerated value (e.g., MONDAY, but not Day.MONDAY).

Appendix I 1199

Z09_LIAN1878_11_GE_APPI.indd 1199 1/2/18 10:53 PM

I.3 Processing Enumerated Values Using a Foreach Loop Each enumerated type has a static method values() that returns all enumerated values for the type in an array. For example,

Day[] days = Day.values();

You can use a regular for loop in (a) or an enhanced for loop in (b) to process all the values in the array.

I.4 Enumerated Types with Data Fields, Constructors, and Methods The simple enumerated types introduced in the preceding section define a type with a list of enumerated values. You can also define an enumerate type with data fields, constructors, and methods, as shown in Listing I.3.

Listing I.3 TrafficLight.java 1 public enum TrafficLight { 2 RED ("Please stop"), GREEN ("Please go"), 3 YELLOW ("Please caution"); 4 5 private String description; 6 7 private TrafficLight(String description) { 8 this.description = description; 9 } 10 11 public String getDescription() { 12 return description; 13 } 14 }

The enumerated values are defined in lines 2 and 3. The value declaration must be the first statement in the type declaration. A data field named description is declared in line 5 to describe an enumerated value. The constructor TrafficLight is declared in lines 7−9. The constructor is invoked whenever an enumerated value is accessed. The enumerated value’s argument is passed to the constructor, which is then assigned to description.

Listing I.4 gives a test program to use TrafficLight.

Listing I.4 TestTrafficLight.java 1 public class TestTrafficLight { 2 public static void main(String[] args) { 3 TrafficLight light = TrafficLight.RED; 4 System.out.println(light.getDescription()); 5 } 6 }

for (int i = 0; i < days.length; i++) System.out.println(days[i]);

for (Day day: days) System.out.println(day);

Equivalent

(a) (b)

1200 Appendix I

Z09_LIAN1878_11_GE_APPI.indd 1200 1/2/18 10:53 PM

An enumerated value TrafficLight.red is assigned to variable light (line 3). Accessing TrafficLight.RED causes the JVM to invoke the constructor with argument “please stop”. The methods in enumerated type are invoked in the same way as the methods in a class. light.getDescription() returns the description for the enumerated value (line 4).

Note The Java syntax requires that the constructor for enumerated types be private to prevent it from being invoked directly. The private modifier may be omitted. In this case, it is considered private by default.

Appendix I 1201

Z09_LIAN1878_11_GE_APPI.indd 1201 1/2/18 10:53 PM

A01_GIPE3898_09_SE_FM.indd 3 12/20/17 5:56 PM

This page intentionally left blank

Console Input

Scanner input = new Scanner(System.in); int intValue = input.nextInt(); long longValue = input.nextLong(); double doubleValue = input.nextDouble(); f loat f loatValue = input.nextFloat(); String string = input.next(); String line = input.nextLine();

Console Output

System.out.println(anyValue);

Primitive Data Types

byte 8 bits short 16 bits int 32 bits long 64 bits f loat 32 bits double 64 bits char 16 bits boolean true/false

Relational Operators

< less than <= less than or equal to > greater than >= greater than or equal to == equal to != not equal

Logical Operators

&& short circuit AND || short circuit OR ! NOT ^ exclusive OR

Arithmetic Operators

+ addition - subtraction * multiplication / division % remainder ++var preincrement --var predecrement var++ postincrement var-- postdecrement

switch Statements

switch (intExpression) { case value1: statements; break; ... case valuen: statements; break; default: statements; }

Companion Website: www.pearsonglobaleditions.com/liang

Assignment Operators

= assignment += addition assignment -= subtraction assignment *= multiplication assignment /= division assignment %= remainder assignment

if Statements

if (condition) { statements; }

if (condition) { statements; } else { statements; }

if (condition1) { statements; } else if (condition2) { statements; } else { statements; }

loop Statements

while (condition) { statements; }

do { statements; } while (condition);

for (init; condition; adjustment) { statements; }

Java Quick Reference

Conditional Expression

boolean-expression ? expression1 : expression2

y = (x > 0) ? 1 : -1

System.out.println(number % 2 == 0 ? "number is even" : "number is odd");

Z10_LIAN1878_11_GE_REF.indd 1203 12/29/17 5:04 PM

Frequently Used Static Constants/Methods

Math.PI Math.random() Math.pow(a, b) Math.abs(a) Math.max(a, b) Math.min(a, b) Math.sqrt(a) Math.sin(radians) Math.asin(a) Math.toRadians(degrees) Math.toDegress(radians) System.currentTimeMillis() Integer.parseInt(string) Integer.parseInt(string, radix) Double.parseDouble(string) Arrays.sort(type[] list) Arrays.binarySearch(type[] list, type key)

Array/Length/Initializer

int[] list = new int[10]; list.length; int[] list = {1, 2, 3, 4};

Multidimensional Array/Length/Initializer

int[][] list = new int[10][10]; list.length; list[0].length; int[][] list = {{1, 2}, {3, 4}};

Ragged Array

int[][] m = {{1, 2, 3, 4}, {1, 2, 3}, {1, 2}, {1}};

String Class

String s = "Welcome"; String s = new String(char[]); int length = s.length(); char ch = s.charAt(index); int d = s.compareTo(s1); boolean b = s.equals(s1); boolean b = s.startsWith(s1); boolean b = s.endsWith(s1); boolean b = s.contains(s1); String s1 = s.trim(); String s1 = s.toUpperCase(); String s1 = s.toLowerCase(); int index = s.indexOf(ch); int index = s.lastIndexOf(ch); String s1 = s.substring(ch); String s1 = s.substring(i,j); char[] chs = s.toCharArray(); boolean b = s.matches(regex); String s1 = s.replaceAll(regex,repl); String[] tokens = s.split(regex);

Companion Website: www.pearsonglobaleditions.com/liang

Object Class

Object o = new Object(); o.toString(); o.equals(o1);

File Class

File f ile = new File(f ilename); f ile.exists() f ile.renameTo(File) f ile.delete()

Text File Output

PrintWriter output = new PrintWriter(�lename); output.print(...); output.println(...); output.printf(...);

Text File Input Scanner input = new Scanner( new File(f ilename));

ArrayList Class

ArrayList<E> list = new ArrayList<>(); list.add(object); list.add(index, object); list.clear(); Object o = list.get(index); boolean b = list.isEmpty(); boolean b = list.contains(object); int i = list.size(); list.remove(index); list.set(index, object); int i = list.indexOf(object); int i = list.lastIndexOf(object);

printf Method

System.out.printf("%b %c %d %f %e %s", true, 'A', 45, 45.5, 45.5, "Welcome"); System.out.printf("%-5d %10.2f %10.2e %8s", 45, 45.5, 45.5, "Welcome");

Comparable Interface

c.compareTo(Comparable) c is a Comparable object

Java Quick Reference

Z10_LIAN1878_11_GE_REF.indd 1204 12/29/17 5:04 PM

1205

Index Symbols –– (decrement operator), 77–78 - (subtraction operator), 68, 72–73 . (dot operator), 354 . (object member access operator), 354, 453 / (division operator), 68, 72 //, in line comment syntax, 40 /*, in block comment syntax, 40 /**.*/ (Javadoc comment syntax), 40 /= (division assignment operator), 76–77 ; (semicolons), common errors, 106 \ (backslash character), as directory separator, 500 \ (escape characters), 148 || (or logical operator), 115–119 + (addition operator), 68, 72 + (string concatenation operator), 58, 153 ++ (increment operator), 77–78 += (addition assignment operator), augmented, 76–77 = (assignment operator), 64–65, 76–77 = (equals operator), 98 -= (subtraction assignment operator), 76–77 == (comparison operator), 98, 456 = (equal to operator), 98 ! (not logical operator), 115–119 != (not equal to comparison operator), 98 $ (dollar sign character), use in source code, 62 % (remainder or modulo operator), 68, 72 %= (remainder assignment operator), 76–77 && (and logical operator), 115–119 ( ) (parentheses), 36, 249 * (multiplication operator), 37, 68, 72 *= (multiplication assignment operator), 76 ^ (exclusive or logical operator), 115–119 {} (curly braces), 35, 101, 105 < (less than comparison operator), 98 <= (less than or equal to comparison operator), 98 > (greater than comparison operator), 98 >= (greater than or equal to comparison operator), 98

Numbers 24-point game, 832–833

A abs method, Math class, 144, 552 Absolute file name, 499 Abstract classes

AbstractCollection class, 798, 799 AbstractMap class, 852 AbstractSet class, 838 case study: abstract number class, 527–529 case study: Calendar and GregorianCalendar classes, 529–532 characteristics of, 526–527 Circle.java and Rectangle.java examples, 524 compared with interfaces, 545–548 GeometricObject.java example, 522–524 InputStream and OutputStream classes, 716–717 interfaces compared to, 532 key terms, 556 overview of, 390–391, 522–523

questions and excercises, 557–562 Rational.java example, 550–553 reasons for using abstract methods, 524 summary, 556–557 TestCalendar.java example, 530–532 TestGeometricObject.java example, 524–525 TestRationalClass.java example, 549–550 using as interface, 940

Abstract data type (ADT), 390 Abstract methods

characteristics of, 526 GenericMatrix.java example, 789–791 GeometricObject class, 523–524 implementing in subclasses, 523 in interfaces, 532 key terms, 556 in Number class, 552 overview of, 249–250 questions and exercises, 557–562 reasons for using, 524 summary, 556–557

abstract modifier, for denoting abstract methods, 522 Abstract number class

LargestNumbers.java, 528–529 overview of, 527–529

Abstract Windows Toolkit. see AWT (Abstract Windows Toolkit) AbstractCollection class, 799 AbstractMap class, 852 AbstractSet class, 838 Accessibility modifiers, 1185 Accessor methods. see Getter (accessor) methods acos method, trigonometry, 142–143 ActionEvent, 617–618 Actions (behaviors), object, 346 Activation records, invoking methods and, 232 Actual concrete types, 774 Actual parameters, defining methods and, 229 Ada, high-level languages, 30 add method

implementing linked lists, 951 List interface, 805

Addition (+=) assignment operator, 76–77 Addition (+) operator, 68, 72 Adelson-Velsky, G. M., 1012 Adjacency lists, representing edges, 1068–1070 Adjacency matrices

representing edges, 1068–1070 weighted, 1106

Adjacent edges, overview of, 1064 ADT (Abstract data type), 390 Aggregate operations, for collection streams

AnalyzeNumbersUsingStream.java example, 1166–1167 case study: analyzing numbers, 1166–1168 case study: counting keywords, 1171–1172 case study: counting occurrences of each letter, 1167–1168 case study: counting occurrences of each letter in string, 1168–1169 case study: finding directory size, 1170–1171 case study: occurrences of words, 1172–1174 case study: processing all elements in two-dimensional array,

1169–1170

Z11_LIAN1878_11_GE_IDX.indd 1205 1/2/18 10:54 PM

1206 Index

Algorithms, spanning tree Dijkstra’s single-source shortest-path algorithm, 1122–1127 MST algorithm, 1117–1118 Prim’s minimum spanning tree algorithm, 1115–1117

allMatch method, 1148, 1150 Ambiguous invocation, of methods, 245 American Standard Code for Information Interchange (ASCII). see

ASCII (American Standard Code for Information Interchange) And (&&) logical operator, 115–119 Animation

case study: bouncing ball, 648–651 case study: US map, 652–655 ClockAnimation.java, 647–648 FadeTransition, 644–645 key terms, 655 PathTransition, 641–644 programming exercises, 656–663 quiz, 656 summary, 655–656 Timeline, 646–647

Anonymous arrays, 282 Anonymous objects, 355 AnonymousHandlerDemo.java, 625–627 anyMatch method, 1148, 1150 Application Program Interfaces (APIs), 33 Apps, developing on Web servers, 33 Arc

overview, 597 ShowArc.java, 597–599

Arguments defining methods and, 229 passing by values, 236–239 receiving string arguments from command line, 296–299 variable-length argument lists, 288–289

ArithmeticException class, 479 Arithmetic/logic units, CPU components, 25 Array elements, 272 Array initializers, 272–273 arraycopy method, System class, 280 ArrayIndexOutOfBoundsException, 275 ArrayList class

animation of array lists, 941 case study: custom stack class, 463–464 cloning arrays, 541 compared with LinkedList, 806–808 creating and adding numbers to array lists, 460–462 creating array lists, 800–802 defined under List interface, 804 DistinctNumbers.java example, 460–462 as example of generic class, 774–775 heap sorts, 917 implementing array lists, 944–951 implementing bucket sorts, 923–924 implementing buckets, 1039 implementing stacks using array lists. see Stacks MyArrayList, 940–941, 963 MyArrayList compared with MyLinkedList, 963 MyArrayList.java example, 945–949 MyList.java example, 942–944 representing edges in graphs, 1069 SetListPerformanceTest.java example, 847–848 storing edge objects in, 1067 for storing elements in a list, 798 storing heaps in, 917 storing list of objects in, 456–457 TestArrayAndLinkedList.java, 807–808 TestArrayList.java example, 457–460

Aggregate operations, for collection streams (Continued) CollectDemo.java example, 1161–1163 CollectGroupDemo.java example, 1164–1166 CountKeywordStream.java example, 1171–1172 CountLettersUsingStream.java example, 1167–1168 CountOccurrenceOfLettersInAString.java example,

1168–1169 CountOccurrenceOfWordsStream.java example,

1173–1174 DirectorySizeStream.java example, 1170–1171 DoubleStream, 1152–1155 grouping elements using groupingby collector, 1163–1166 IntStream, 1152–1155 IntStreamDemo.java example, 1152–1155 LongStream, 1152–1155 overview of, 1146 parallel streams, 1155–1157 ParallelStreamDemo.java example, 1155–1157 quiz and exercises, 1174–1175 Stream class, 1147 stream pipelines, 1146–1152 stream reduction using collect method, 1160–1163 stream reduction using reduce method, 1157–1160 StreamDemo.java example, 1148–1149 StreamReductionDemo.java example, 1158–1160 summary, 1174 TwoDimensionalArrayStream.java example, 1169–1170

Aggregating classes, 398 Aggregating objects, 398 Aggregation relationships, objects, 398–399 AIFF audio files, 698 Algorithms, 56

analyzing Towers of Hanoi problem, 868–869 Big O notation for measuring efficiency of, 862–863 binary search, 868 bubble sort, 906–908 comparing growth functions, 869–870 comparing prime numbers, 883 determining Big O for repetition, sequence, and selection statements,

864–867 EfficientPrimeNumbers.java example, 879–882 external sorts. see External sorts finding closest pair of points, 883–886 finding convex hull for a set of points, 889–891 finding Fibonacci numbers, 871–873 finding greatest common denominator, 873–877 finding prime numbers, 877–883 GCDEuclid.java example, 875–877 GCD.java example, 874–875 gift-wrapping algorithm, 889–890 Graham’s algorithm, 890–891 graph algorithms, 1062–1063 greedy, 1001 heap sort. see Heap sorts insertion sorts, 904–906 key terms, 891 merge sorts, 909–912 overview of, 862 PrimeNumbers.java example, 878–879 quick sort, 912–916 quiz and exercises, 892–901 recurrence relations and, 869 selection sort and insertion sort, 868 SieveOfEratosthenes.java example, 882–883 solving Eight Queens problem, 886–889 for sort method, 781 summary, 891–892

Z11_LIAN1878_11_GE_IDX.indd 1206 1/2/18 10:54 PM

Index 1207

text I/O vs. binary I/O, 715–716 asin method, trigonometry, 142–143 asList method, 808 Assemblers, 29 Assembly language, 29 Assignment operator (=), 1191

augmented, 76–77 overview of, 64–65

Assignment statements (assignment expressions) assigning value to variables, 58 overview of, 64–65

Associative array, 1032 Associative arrays. see Maps Associativity, of operators, 127, 1182–1183 atan method, trigonometry, 142–143 Attributes, object, 346 Audio files

case study: national flags and anthems, 701–703 MediaDemo.java, 699–701

Autoboxing/Autounboxing, 408–409, 775–776 Average-case analysis

measuring algorithm efficiency, 862, 876 quick sort and, 916

AVL trees AVLTree.java example, 1018–1023 balancing nodes on a path, 1016–1017 deleting elements, 1018 designing classes for, 1015–1016 key terms, 1028 overriding the insert method, 1016–1017 overview of, 1012 questions and exercises, 1028–1029 rebalancing, 1012–1014 rotations for balancing, 1017 summary, 1028 TestAVLTree.java example, 1024–1027 time complexity of, 1027

AVLTree class delete method, 1023, 1027 overview of, 1018–1023 as subclass of BST class, 1015 testing, 1024–1027

AWT (Abstract Windows Toolkit) Color class, 575–576 Date class, 358–359, 529–530 Error class, 482, 484 event classes in, 618 EventObject class, 618–619 exceptions. see Exception class File class, 499–501, 714 Font class, 576–577 GeometricObject class, 523–524 GuessDate class, 328–329 IllegalArgumentException class, 485 InputMismatchException class, 480, 505 KeyEvent class, 635 MalformedURLException class, 509 MouseEvent class, 633–634 Polygon class, 599–600 String class, 410 Swing vs., 564

B Babylonian method, 263 Backslash character (\), as directory separator, 500 Backtracking algorithm, 886–889

TestMyArrayList.java example, 949–951 Vector class compared with, 803

Arrays class, 294–296 Arrays, in general

edge arrays, 1067 as fixed-size data structure, 944 implementing binary heaps using, 917 ragged arrays, 1068 sorting using Heap class, 921–922 storing lists in. see ArrayList class storing vertices in, 1066

Arrays, multi-dimensional case study: daily temperature and humidity, 326–328 case study: guessing birthdays, 328–329 overview of, 325–326 questions and exercises, 329–343 summary, 329

Arrays, single-dimensional accessing elements, 272 ArrayList class, 459–460 Arrays class, 294–296 case study: analyzing numbers, 277–278 case study: counting occurrences of letters, 285–288 case study: deck of cards, 278–280 case study: generic method for sorting, 780–781 constructing strings from, 410 converting strings to/from, 413 copying, 280–281 creating, 271–272, 536–538 declaring, 270 foreach loops, 275–277 initializers, 272–273 key terms, 299 of objects, 375–377 overview of, 270 passing to methods, 281–284 processing, 273–275 questions and exercises, 300–309 returning from methods, 284–285 searching, 289–293 serializing, 731–733 size and default values, 272 sorting, 293–294, 536–538 summary, 299–300 treating as objects in Java, 354 variable-length argument lists, 288–289

Arrays, two-dimensional case study: finding closest pair of points, 320–322 case study: grading multiple-choice test, 318–320 case study: processing all elements in a two-dimensional array,

1169–1170 case study: Sudoku, 322–325 declaring variables and creating two-dimensional arrays, 312–313 obtaining length of two-dimensional arrays, 313–314 passing to methods to two-dimensional arrays, 317–318 processing two-dimensional arrays, 315–317 questions and exercises, 329–343 ragged arrays, 314–315 representing graph edges with, 1067 representing weighted graphs, 1105–1106 summary, 329

Arrows keys, on keyboards, 28 ASCII (American Standard Code for Information Interchange)

character (char data type), 147–148 decimal and hexadecimal equivalents, 1181 encoding scheme, 25–26 text encoding, 714

Z11_LIAN1878_11_GE_IDX.indd 1207 1/2/18 10:54 PM

1208 Index

TestBSTDelete.java example, 992–995 TestBST.java example, 988–989 TestBSTWithIterator.java example, 999–1000 tree traversal, 979–980 tree visualization and MVC, 995–998 Tree.java example, 981–983

Binary trees, 976 binarySearch method

applying to lists, 814 Arrays class, 295

Binding properties BindingDemo.java, 572–573 ShowCircleCentered.java, 570–572

Bit operators, 1191 Bits (binary digits), 25 Bitwise operators, 1191 Block comments, in Welcome.java, 35 Block modifiers, 1184–1185 Block style, programming style, 41 Blocks, in Welcome.java, 35 BMI (Body Mass Index), 111–112, 394–397 Boolean accessor method, 369 boolean data type

java.util.Random, 359 overview of, 98–100

Boolean expressions case study: determining leap year, 119–120 conditional operators, 125–126 defined, 98 if statements and, 100–101 if-else statements, 102–103 writing, 108–109

Boolean literals, 99 Boolean values

defined, 98 as format specifier, 168 logical operators and, 115 redundancy in testing, 106

Boolean variables assigning, 108 overview of, 98–99 redundancy in testing, 106

BorderPane overview of, 585 ShowBorderPane.java, 585–586

Bottom-up implementation, 251–253 Bounded generic types

erasing, 786–787 GenericMatrix.java example, 789–791 MaxUsingGenericType.java example, 782–783 overview of, 779

Bounded wildcards, 784 Boxing, converting wrapper object to primitive value, 407 Braces. see Curly braces ({}) Breadth-first searches (BFS)

applications of, 1093 finding BFS trees, 1064 implementing, 1091–1092 overview of, 1090 TestBFS.java, 1092 traversing graphs, 1083

Breadth-first traversal, tree traversal, 980 break statements

controlling loops, 208–211 using with switch statements, 122, 123

Breakpoints, setting for debugging, 128 Brute-force algorithm, 873

Backward pointer, in doubly linked lists, 964 Balance factor, for AVL nodes, 1012, 1020 Balanced nodes

in AVL trees, 1012 AVLTree class, 1018–1019, 1022–1023

Base cases, in recursion, 748 BaseStream interface, 1146 BASIC, high-level languages, 30 Bean machine game, 308–309, 662 beginIndex method, for obtaining substrings from strings, 158 Behaviors (actions), object, 346 Behind the scene evaluation, expressions, 127 Best-case input

measuring algorithm efficiency, 862, 876 quick sort and, 916

BFS (breadth-first searches). see Breadth-first searches (BFS) Big O

determining for repetition, sequence, and selection statements, 864–867

for measuring algorithm efficiency, 862–863 BigDecimal class, 408–409, 527 Binary

files, 714 machine language as binary code, 29 operator, 70 searches, 290–293, 752

Binary digits (Bits), 25 Binary heaps (binary trees), 916. see also Heap sorts

complete, 916, 917, 922 Binary I/O

BufferedInputStream and BufferedOutputStream classes, 723–726

characters and strings in, 720–721 classes, 716–726 DataInputStream and DataOutputStream classes, 720–723 DetectEndOfFile.java, 723 FileInputStream and FileOutputStream classes, 717–720 FilterInputStream and FilterOutputStream classes, 720 overview of, 714 TestDataStream.java, 721–722 TestFileStream.java, 718–720 vs. text I/O, 715–716

Binary numbers converting to/from decimal, 767, 1188 converting to/from hexadecimal, 1189–1190 overview of, 1187

Binary search algorithm, 897 analyzing, 868 recurrence relations and, 869

Binary search trees (BST) BST class, 980–989 BSTAnimation.java example, 995–996 BST.java example, 983–988 BTView.java example, 996–998 case study: data compression, 1000–1005 deleting elements, 989–995 displaying/visualizing binary trees, 995–998 HuffmanCode.java example, 1002–1005 implementing using linked structure, 976–977 inserting elements, 978–979 iterators, 998–1000 key terms, 1005 overview of, 976 quiz and exercises, 1005–1009 representation of, 977–978 searching for elements, 978 summary, 1005

Z11_LIAN1878_11_GE_IDX.indd 1208 1/2/18 10:54 PM

Index 1209

Celsius, converting Fahrenheit to/from, 259–260 Chained exceptions, 495–496 char data type. see Characters (char data type) Characters (char data type)

applying numeric operators to, 248 in binary I/O, 720–721 case study: ignoring nonalphanumeric characters when checking palin-

dromes, 420–422 casting to/from numeric types, 149 comparing, 98 comparing and testing, 150–151 constructing strings from arrays of, 410 converting to strings, 413–414 decimal and hexadecimal equivalents of ASCII character set, 1181 escape characters, 148 finding, 158–159 generic method for sorting array of Comparable objects, 780 hash codes for primitive types, 1033 overview of, 147 RandomCharacter.java, 248 retrieving in strings, 153 TestRandomCharacter.java, 248–249 Unicode and ASCII and, 147–148

charAt (index) method retrieving characters in strings, 153 StringBuilder class, 419

CheckBox, 670–673 CheckBoxDemo.java, 671–673 Checked exceptions, 483 checkIndex method, 951 Checkpoint Questions, recurrence relations and, 869 Child, in BST, 977–978 Choice lists. see ComboBox Circle and Ellipse

overview, 594–595 ShowEllipse.java, 595–596

Circle class, 346, 347 Circular

doubly linked lists, 964 singly linked lists, 964

Clarity, class design guidelines, 554 Class diagrams, UML, 347 Class loaders, 40 Class modifiers, Java modifiers, 1184–1185 ClassCastException, 452 Classes

abstract. see Abstract classes abstraction and encapsulation in, 390–394 benefits of generics, 774–776 case study: designing class for matrix using generic types, 788–793 case study: designing class for stacks, 402–404 case study: designing Course class, 400–402 in Circle.java (for CircleWithPrivateDataFields), 369–370 in Circle.java (for CircleWithStaticMembers) example,

362–363 clients of, 349 commenting, 41 in ComputeExpression.java, 36–37 data field encapsulation for maintaining, 368–369 defining custom exception classes, 496–499 defining for objects, 346–348 defining generic, 776–778 design guidelines, 553–556 identifiers, 62 inner (nested) classes. see Inner (nested) classes from Java Library, 358–361 names/naming conventions, 35, 66

BST (binary search trees). see Binary search trees (BST) BST class

AVLTree class as subclass of, 1015 BST.java example, 983–988 overview of, 980–981 TestBSTDelete.java example, 992–995 TestBST.java example, 988–989 time complexity of, 993 Tree.java example, 981–983

Bubble sorts, 303 algorithms, 907 bubble sort algorithms, 907 BubbleSort.java example, 908 overview of, 906–907 time complexity of, 908

Buckets bucket sorts, 923–925 separate chaining and, 1039, 1057

BufferedInputStream and BufferedOutputStream classes, 723–726 Bugs (logic errors), 43, 128 Bus, function of, 24–25 Button, 668–670 Button, ButtonDemo.java, 669–670 ButtonBase, 668–669 ButtonDemo.java, 669–670 byte type, numeric types

hash codes for primitive types, 1033 overview of, 67

Bytecode translating Java source file into, 37 verifier, 40

Bytes defined, 25 measuring storage capacity in, 26

C C++, high-level languages, 30 C, high-level languages, 30 Cable modems, 28 Calendar class, 529–530 Call stacks

displaying in debugging, 128 invoking methods and, 232

Calling methods, 230–232 objects, 355

canRead method, File class, 500–501 canWrite method, File class, 500–501 capacity() method, StringBuilder class, 419 Case sensitivity

identifiers and, 62 in Welcome.java, 35

Casting. see Type casting Casting objects

CastingDemo.java example, 452–453 overview of, 451–452

Catching exceptions. see also try-catch blocks catch block omitted when finally clause is used, 493 CircleWithException.java example, 489 InputMismatchExceptionDemo.java example, 480–481 overview of, 485–487 QuotientWithException.java example, 478–480

CDs (compact discs), as storage device, 27 Cells

in Sudoku grid, 322 in tic-tac-toe case study, 693–698

Z11_LIAN1878_11_GE_IDX.indd 1209 1/2/18 10:54 PM

1210 Index

forEach method, 803–804 iterators for traversing collections, 802–803 key terms, 829 List interface, 804–806 Map interface, 1032 methods of List interface, 804–806 overview of, 798 PriorityQueue class, 823–824 Queue interface, 822 queues and priority queues, 821–824 quiz and exercises, 830–836 static methods for lists and collections, 813–816 summary, 829–830 TestCollection.java example, 800–802 TestIterator.java example, 802–803 Vector and Stack classes, 820–821

Collisions, in hashing double hashing, 1037–1039 handling using open addressing, 1035–1039 handling using separate chaining, 1039 linear probing, 1035–1036 overview of, 1033 quadratic probing, 1036–1037

ComboBox ComboBoxDemo.java, 682–684 overview of, 681–682

Command-line arguments, 296–299 Comments

code maintainability and, 123 programming style and, 40 in Welcome.java, 35

Common denominator, finding greatest common denominator. see Gcd (greatest common denominator)

Communication devices, computers and, 28 Compact discs (CDs), as storage device, 27 Comparable interface

ComparableRectangle.java example, 537–538 Comparator interface vs., 810 enumerated types, 1197 as example of generic interface, 774–775 generic method for sorting array of Comparable objects, 780 overview of, 535–537 PriorityQueue class and, 823 Rational class implementing, 550 SortComparableObjects.java example, 537 SortRectangles.java example, 538–539 TreeMap class and, 853

Comparator interface Comparable vs., 810 GeometricObjectComparator.java, 809 methods of, 809 PriorityQueue class and, 823, 824 sorted method, 1149–1150 SortStringByLength.java, 810–811 SortStringIgnoreCase.java, 811–812 TestComparator.java, 809–810 TestTreeSetWithComparator.java example, 843–845 TreeMap class and, 853

compare method, 809–810 compareTo method

Cloneable interface and, 540 Comparable interface defining, 535–536 ComparableRectangle.java example, 537–538 comparing strings, generic method for sorting array of Comparable

objects, 781 implementing in Rational class, 550

Classes (Continued) Point2D, 360–361 preventing extension of, 467 raw types and backward compatibility, 782–783 static variables, constants, and methods, 361–366 in TestCircleWithPrivateDataFields.java example, 370–371 in TestCircleWithStaticMembers.java example, 363–366 in UML diagram, 347, 348 variable scope and, 361–362 visibility modifiers, 366–368 in Welcome.java, 34 in WelcomeWithThreeMessages.java, 36

Classes, binary I/O BufferedInputStream and BufferedOutputStream classes,

723–726 DataInputStream and DataOutputStream classes, 720–723 DetectEndOfFile.java, 723 FileInputStream and FileOutputStream classes, 717–720 FilterInputStream and FilterOutputStream classes, 720 overview of, 716–717 TestDataStream.java, 721–722 TestFileStream.java, 718–720

Classifier, 1163 Class’s contract, 390 Clock speed, CPUs, 25 ClockPane Class

ClockPane.java, 604–606 DisplayClock.java, 603–604 paintClock method, 605–606

clone method, shallow and deep copies, 542–543 Cloneable interface

House.java example, 541–545 overview, 540–541

Closest pair problem, two-dimensional array applied to, 320–322 Closest-pair animation, 897 Cluster, 1036 COBOL, high-level languages, 30 Code

arrays for simplifying, 274–275 comments and, 123 incremental development, 186 programming. see Programs/programming reuse. see Reusable code sharing. see Sharing code in software development process, 83–84

Coding trees, 1000–1005. see also Huffman coding trees Coherent purpose, class design guidelines, 553–554 collect method, stream reduction using, 1160–1163 Collections

Collection interface, 798–802 forEach method, 803–804 iterators for traversing collections, 802–803 singleton and unmodifiable, 857 static methods for, 813–816 TestCollection.java example, 800–802

Collections class singleton and unmodifiable collections, 857 static methods, 813

Collections Framework hierarchy ArrayList and LinkedList class, 806–808 case study: displaying bouncing balls, 816–820 case study: stacks used to evaluate expressions, 825–829 Collection interface, 798–800 Comparator interface, 809–813 Dequeue interface, 822–824 designing complex data structures, 1070

Z11_LIAN1878_11_GE_IDX.indd 1210 1/2/18 10:54 PM

Index 1211

input, 34 output, 34 reading input, 59–61

Constant time, comparing growth functions, 870 Constants

class, 361–362 declaring, 362 identifiers, 62 KeyCode constants, 635 named constants, 65–66 naming conventions, 66 wrapper classes and, 405

Constructor chaining, 441–442 Constructor modifiers, 1184–1185 Constructors, 382

in abstract classes, 524 for AVLTree class, 1018–1019 for BMI class, 395–396 calling subclass constructors, 440 creating Date class, 359 creating objects with, 353 creating Random objects, 359 for DataInputStream and DataOutputStream classes, 721 generic classes and, 778 interfaces vs. abstract classes, 545 invoking with this reference, 382 for Loan class, 392–394 object methods and, 346 private, 368 for String class, 410 for StringBuilder class, 417 in TestCircle.java example, 349, 350 in TV.java example, 351 UML diagram of, 348 for UnweightedGraph class, 1074–1075 for WeightedGraph class, 1108–1109 wrapper classes and, 555

Containers creating data structures, 798 maps as, 850 removing elements from, 848 storing objects in, 799 types supported by Java Collections Framework, 798

contains method, 848, 849 continue statements, for controlling loops, 208–211 Contract, object class as, 346 Control, 567–570 Control units, CPUs, 25 Control variables, in for loops, 195–196 ControlCircle.java, 622–623 Conversion methods, for wrapper classes, 405 Convex hull

finding for set of points, 889–891 gift-wrapping algorithm applied to, 889–890 Graham’s algorithm applied to, 890–891

Copying arrays, 280–281 files, 726–728

Core, of CPU, 25 cos method, trigonometry, 142–143 Cosine function, 611 count method, 1150–1151 Counter-controlled loops, 183 Coupon collector’s problem, 303–304 Course class, 400–401 CPUs (central processing units), 25

wrapper classes and, 405–406 compareToIgnoreCase method, 156, 811 Comparison operators (==), 98, 456 Compatibility, raw types and backward compatibility, 782–783 Compile errors (Syntax errors)

common errors, 35–36 debugging, 128 programming errors, 42

Compile time error detection at, 774 restrictions on generic types, 787 Xlint:unchecked error, 782

Compilers ambiguous invocation and, 245 reporting syntax errors, 42 translating Java source file into bytecode file, 37–38 translating source program into machine code, 30, 31

Complete binary tree, 916, 917, 922 Complete graphs, 1064 Completeness, class design guidelines, 555 Complex numbers, Math class, 560 Components

ListView, 684–686 ListViewDemo.java, 686–687 quiz and exercises, 704–711 ScrollBar, 687–689 ScrollBarDemo.java, 688–689 SliderDemo.java, 691–692 sliders, 691 summary, 703–704 TextArea, 677–680 TextAreaDemo.java, 680 TextFieldDemo.java, 676–677

Composition, in designing stacks and queues, 966 Composition relationships

aggregation and, 398–399 between ArrayList and MyStack, 463–464

Compound expressions case study: stacks used to evaluate, 825 EvaluateExpression.java example, 826–829

Compression data compression using Huffman coding, 1000–1005 of hash codes, 1034–1035 HuffmanCode.java example, 1002–1005

Compute expression, 36 Computers

communication devices, 28 CPUs, 25 input/output devices, 27–28 memory, 26 OSs (operating systems), 31–32 overview of, 24–25 programming languages, 29–31 storage devices, 26–27

concat method, 153 Concatenate strings, 58, 153–154 Conditional operators, 125–126 Connect four game, 337 Connected circles problem

ConnectedCircles.java, 1089 overview of, 1088–1089

Connected graphs, 1064 Consistency, class design guidelines, 554 Consoles

defined, 34 formatting output, 167–171

Z11_LIAN1878_11_GE_IDX.indd 1211 1/2/18 10:54 PM

1212 Index

reference types. see Reference types specifying, 57 strings, 152 types of, 63 using abstract class as, 526

DataInputStream/DataOutputStream classes DetectEndOfFile.java, 723 external sorts and, 925 overview of, 720 TestDataStream.java, 721–722

Date class case study: Calendar and GregorianCalendar classes, 529–530 java.util, 358

De Morgan’s law, 117 Debugging

benefits of stepwise refinement, 256 code modularization and, 239 selections, 128

Decimal numbers BigDecimal class, 408–409 converting to hexadecimals, 206–208, 241–243, 767 converting to/from binary, 767, 1188 converting to/from hexadecimal, 1189 equivalents of ASCII character set, 1181 overview of, 1187

Declaring constants, 65, 362 Declaring exceptions

CircleWithException.java example, 489 ReadData.java example, 504–505 TestCircleWithCustomException.java example, 497–498 throws keyword for, 484, 489

Declaring methods generic methods, 779 static methods, 361

Declaring variables array variables, 270 overview of, 62–63 specifying data types and, 57–58 two-dimensional array variables, 312–313

Decrement (––) operator, 77–78 Deep copies, 543 Default field values, for data fields, 355 Degree of vertex, 1064 Delete key, on keyboards, 28 delete method, AVLTree class, 1023, 1027 Delimiters, token reading methods and, 505 Denominator. see Gcd (greatest common denominator) Denominators, in rational numbers, 548 Depth-first searches (DFS)

applications, 1087–1088 case study: connected circles problem, 1088–1089 finding DFS trees, 1064 implementing, 1085–1087 traversing graphs, 1083–1084

Depth-first traversal, tree traversal, 980 Dequeue interface, LinkedList class, 822–824 dequeue method, 968 DescriptionPane class, 679–680 Descriptive names

benefits of, 62 for variables, 57

Deserialization, of objects, 731 Design guidelines, classes, 553–556 Determining Big O

for repetition statements, 864–867 for selection statements, 864–867

Cubic time, comparing growth functions, 870 Curly braces {}

in block syntax, 35 dangers of omitting, 196 forgetting to use, 105–106

Cursor, mouse, 28 Cycle, connected graphs, 1064

D .dat files (binary), 716 Data, arrays for referencing, 270 Data compression

Huffman coding for, 1000–1005 HuffmanCode.java example, 1002–1005

Data fields accessing object data, 354–355 encapsulating, 368–371, 554 in interfaces, 534 object state represented by, 346 referencing, 355, 380–382 in TestCircle.java example, 349 in TV.java example, 351 UML diagram of, 348

Data modifiers, 1184–1185 Data streams. see DataInputStream/DataOutputStream classes Data structures. see also Collections Framework hierarchy

array lists. see ArrayList class choosing, 798 collections. see Collections first-in, first-out, 821 linked lists. see LinkedList class lists. see Lists queues. see Queues stacks. see Stacks

Data structures, implementing array lists, 944–951 GenericQueue.java example, 967 implementing MyLinkedList class, 955–963 linked lists, 951–965 lists, 940–944 MyArrayList compared with MyLinkedList, 963 MyArrayList.java example, 945–949 MyLinkedList class, 940–941, 953, 963 MyPriorityQueue.java example, 969–970 overview of, 940 priority queues, 969–970 quiz and exercises, 971–973 stacks and queues, 965–969 summary, 971 TestMyArrayList.java example, 949–951 TestMyLinkedList.java example, 954–955 TestPriorityQueue.java example, 970 TestStackQueue.java example, 967–969 variations on linked lists, 964–965

Data types ADT (Abstract data type), 390 boolean, 98–100, 359 char. see Characters (char data type) double. see double (double precision), numeric types float. see Floating-point numbers (float data type) fundamental. see Primitive types (fundamental types) generic. see Generics int. see Integers (int data type) long. see Long, numeric types numeric, 67–70

Z11_LIAN1878_11_GE_IDX.indd 1212 1/2/18 10:54 PM

Index 1213

Edges adjacency lists, 1068–1070 adjacency matrices, 1068 adjacent and incident, 1064 defining as objects, 1067 Graph.java example, 1072 on graphs, 1063 Prim’s algorithm and, 1115 representing edge arrays, 1067 TestGraph.java example, 1072 TestMinimumSpanningTree.java, 1118–1119 TestWeightedGraph.java, 1112–1113 weighted adjacency matrices, 1106 weighted edges using edge array, 1105–1106 weighted graphs, 1104 WeightedGraph class, 1107–1108

Edge-weighted graphs overview of, 1104 WeightedGraph class, 1106

Eight Queens puzzle EightQueens.java example, 887–889 recursion, 770 single-dimensional arrays, 308 solving, 886–889

Element type, specifying for arrays, 270 Emirp, 264 Encapsulation

in Circle.java (for CircleWithPrivateDataFields) example, 369–370

class design guidelines, 552 of classes, 390–394 of data fields, 368–371 information hiding with, 249 of Rational class, 552

Encoding schemes defined, 25–26 mapping characters to binary equivalents, 147

End of file exception (EOFException), 723 End-of-line style, block styles, 41 enqueue method, 968 entrySet method, Map interface, 852 Enumerated types

with data fields, constructors, and methods, 1200–1201

EnumeratedTypeDemo.java example, 1198 if statements with, 1199 simple, 1197–1199 StandaloneEnumTypeDemo.java example,

1198–1199 switch statements with, 1199 TestTrafficLight.java example, 1200–1201 TrafficLight.java example, 1200 values method, 1200

Equal (=) operator, for assignment, 98 == (equal to operator), 98 Equal to (==) operator, for comparison, 98 equalArea method, for comparing areas of geometric

objects, 525 Equals method

Arrays class, 295 Comparator interface, 809 Object class, 446

Erasure and restrictions, on generics, 786–788 Error class, 482, 484 Errors, programming. see Programming errors

(escape characters), 148

for sequence statements, 864–867 DFS (depth-first searches). see Depth-first searches (DFS) Dial-up modems, 28 Dictionaries, 1032. see also Maps Digital subscriber lines (DSLs), 28 Digital versatile disc (DVDs), 27 Digits, matching, 120 Dijkstra’s single-source shortest-path algorithm, 1122–1127 Direct recursion, 745 Directed graphs, 1063 Directories

case study: determining directory size, 753–754 DirectorySize.java, 753–754 File class and, 500 file paths, 499

disjoint method, 815 Disks, as storage device, 27 Display message

in Welcome.java, 34 in WelcomeWithThreeMessages.java, 36

distinct method, 1148, 1150–1151 Divide-and-conquer algorithm, 883–886 Divide-and-conquer strategy. see Stepwise refinement Division (/=) assignment operator, 64 Division operator (/), 68, 72 Documentation, programming and, 40 Dot operator (.), 354 Dot pitch, measuring sharpness of displays, 28 double (double precision), numeric types

converting characters and numeric values to strings, 413–414 declaring variables and, 63 generic method for sorting array of Comparable objects, 780 hash codes for primitive types, 1033 java.util.Random, 359 overview of numeric types, 67 precision of, 203

Double hashing, collision handling, 1037–1039 Double.parseDouble method, 159 DoubleStream, 1152–1155 Doubly linked lists, 964

deciding when to use, 198–200 do-while loops, 192–195 overview of, 190–192

do-while loops, 192–195 Downcasting objects, 451 drawArc method, 597–599 Drives, 27 Drop-down lists. see ComboBox DSLs (digital subscriber lines), 28 DVDs (Digital versatile disc), 27 Dynamic binding, inheritance and, 447–451 Dynamic programming

computing Fibonacci numbers, 871–873 definition, 872 Dijkstra’s algorithm, 1126 ImprovedFibonacci.java example, 872–873

E Eclipse

built in debugging, 128 creating/editing Java source code, 37

Edge arrays representing edges, 1067 weighted edges using, 1105–1106

Edge class, 1067

Z11_LIAN1878_11_GE_IDX.indd 1213 1/2/18 10:54 PM

1214 Index

Boolean. see Boolean expressions case study: stacks used to evaluate, 825 EvaluateExpression.java example, 826–829

extends keyword, interface inheritance and, 546 External sorts

complexity, 932 CreateLargeFile.java example, 925–926 implementation phases, 926–931 overview of, 925

F Factorials

case study: computing factorials, 742–745 ComputeFactorial.java, 743–745 ComputeFactorialTailRecusion.java, 763 tail recursion and, 762–763

FadeTransition, 644–645 Fahrenheit, converting Celsius to/from, 259–260 Fall-through behavior, switch statements, 123 Feet, converting to/from meters, 260 fib method, 746–748 Fibonacci, Leonardo, 746 Fibonacci numbers

algorithm for finding, 871–873 case study: computing, 745–748 ComputeFibonacci.java, 746–748 computing recursively, 765 ImprovedFibonacci.java example, 872–873 recurrence relations and, 869

File class, 499–501, 714 File I/O. see Input; Output File pointers, random-access files and, 734 FileInputStream/FileOutputStream classes

overview of, 717–718 TestFileStream.java, 718–720

Files case study: copying files, 726–728 case study: replacing text in, 506–508 File class, 499–501, 714 input/output, 502–508 key terms, 512 quiz and exercises, 514–519 reading data from, 504–505 reading data from Web, 508–510 summary, 513 TestFileClass.java example, 501 writing data to, 502–503

fill method, 815 filter method, 1148, 1150 FilterInputStream/FilterOutputStream classes, 720 final keyword, for declaring constants, 65 final modifier, for preventing classes from being extended, 467 finally clause, in exception handling, 492–493 findAny method, 1148, 1151–1152 findFirst method, 1148, 1151–1152 First-in, first-out data structures, 821 float data type. see Floating-point numbers (float data type) Floating-point numbers (float data type)

approximation, 88 converting to integers, 79 hash codes for primitive types, 1033 java.util.Random, 359 minimizing numeric errors related to loops, 202–203 numeric types for, 67 overview of numeric types, 67

Euclid’s algorithm finding greatest common divisors using, 873–877 GCDEuclid.java example, 875–877

Euler, 1062–1063 Event delegation, 618 Event handlers/event handling, 616–617, 627–629, 635 Exception class

exceptions in, 482 extending, 496 in java.lang, 496 subclasses of, 482–483

Exception handling. see also Programming errors catching exceptions, 485–487, 489–492 chained exceptions, 495–496 ChainedExceptionDemo.java example, 495–496 checked and unchecked, 483 CircleWithException.java example, 489 ClassCastException, 452 declaring exceptions (throws), 484, 489 defined, 476 defining custom exception classes, 496–499 EOFException, 723 in Exception class, 482 exception classes cannot be generic, 788 FileNotFoundException, 717 files input/output, 502–508 finally clause in, 492–493 getting information about exceptions, 487–489 in House.java example, 542 InputMismatchExceptionDemo.java example, 480–481 InvalidRadiusException.java example, 496–497 IOException, 482 key terms, 512 NotSerializableException, 731 overview of, 61, 476–481 quiz and exercises, 514–519 Quotient.java example, 476 QuotientWithException.java example, 478–480 QuotientWithIf.java example, 477 QuotientWithMethod.java example, 477–478 ReadData.java example, 504–505 ReadFileFromURL.java example, 509–510 ReplaceText.java example, 506–508 rethrowing exceptions, 494–495 summary, 513 TestCircleWithCustomException.java example, 497–499 TestCircleWithException.java example, 490–492 TestException.java example, 488–489 TestFileClass.java example, 501 throwing exceptions, 484–485, 489–492 types of exceptions, 481–483 unsupported operations of Collection interface, 800 WebCrawler.java example, 511–512 when to use exceptions, 493–494 WriteData.java example, 502–503 WriteDataWithAutoClose.java example, 503–504

Exception propagation, 485 Exclusive or (^) logical operator, 115–119 Execution stacks. see Call stacks exists method, for checking file instances, 500 Explicit casting, 79, 80, 451 Exponent method, Math class, 143 Exponential algorithms, 869, 892 Expressions

assignment statements and, 64–65 behind the scene evaluation, 127

Z11_LIAN1878_11_GE_IDX.indd 1214 1/2/18 10:54 PM

Index 1215

GenericStack class, 777–778 key terms, 793 methods, 778–780 motivation for using, 774–776 overview of, 774 questions and exercises, 794–795 raw types and backward compatibility and, 782–783 restrictions on generic types, 786–788 summary, 793–794 wildcards for specifying range of generic types, 783–786

GeometricObject class Circle.java and Rectangle.java, 524 overview of, 523 TestGeometricObject.java example, 524–525

getAbsolutePath(), File class, 501 getArea method, Circle example, 349, 350 getArray method, 317–318 getBMI method, BMI class, 395, 396 getCharacterFrequency method, 1003 getChars method, converting strings into arrays, 413 getDateCreated method, Date class, 378 getIndex method, ArrayList class, 459 getMinimumSpanningTree method, WeightedGraph class, 1119–1120 getPerimeter method, Circle example, 349 getRadius method, CircleWithPrivateDataFields.java

example, 370 getRandomLowerCaseLetter method, 285, 287 getSize method, finding directory size, 464, 1170–1171 getSource method, events, 618 getStackTrace method, for getting information about exceptions, 488 getStatus method, BMI class, 395, 396 Getter (accessor) methods

ArrayList class and, 460 encapsulation of data fields and, 369–370 implementing linked lists, 951

Gift-wrapping algorithm, 889–890 Gigabytes (GBs), of storage, 26 Gigahertz (GHz), clock speed, 25 GMT (Greenwich Mean Time), 74 Gosling, James, 32 Graham’s algorithm, 890–891 Graph interface, 1070 Graph theory, 1063 Graphical user interface (GUI), 666 Graphs

breadth-first searches (BFS), 1090–1091 case study: connected circles problem, 1088–1089 case study: nine tails problem, 1093–1099 ConnectedCircles.java, 1089 DisplayUSMap.java example, 1081–1083 Graph.java example, 1073–1074 GraphView.java example, 1080–1081 key terms, 1099 modeling, 1070–1078 overview of, 1092 questions and exercises, 1099–1105 representing edges, 1067–1070 representing vertices, 1064–1066 summary, 1099 terminology regarding, 1063–1064 TestGraph.java example, 1073–1074 traversing, 1083 UnweightedGraph.java example, 1074–1080 visualization of, 1080–1083

Greater than (>) comparison operator, 98 Greater than or equal to (>=) comparison operator, 98

special values, 1186 specifying data types, 57 specifying precision, 169

Flowcharts do-while loop, 192–193 if statements and, 100–101 if-else statements, 102 for loops, 195, 196 switch statements, 122 while loops, 182, 183

FlowPane HBox and VBox, 586–587 overview, 581 ShowFlowPane.java, 581–583

Folding, hash codes and, 1033 Font, FontDemo.java, 576–577 for loops

deciding when to use, 198–200 nesting, 200–202, 315 overview of, 195–198 processing arrays with, 273 variable scope and, 246–247

foreach (enhanced) loops implicit use of iterator by, 808 overview of, 275–277 for traversing collections, 803

forEach method, 1149 Formal generic type, 774 Formal parameters. see Parameters Format specifiers, 168–170 FORTRAN, high-level languages, 30 Forward pointer, in doubly linked lists, 964 Fractals

case study, 758–761 H-tree fractals, 771 Koch snowflake fractal, 769 SierpinskiTriangle.java, 758–761

Frames (windows) ScrollBarDemo.java, 688–689 SliderDemo.java, 691–692

Free cells, in Sudoku grid, 322 frequency method, 816 Function keys, on keyboards, 27 Functions, 229. see also Methods Fundamental types (Primitive types). see Primitive types

(fundamental types)

G Galton box, 308 Garbage collection, JVM and, 357 GBs (gigabytes), of storage, 26 Gcd (greatest common denominator)

algorithm for finding, 873–877 case study: finding greatest common denominator, 204–205 computing recursively, 765 gcd method, 239–240 GCDEuclid.java example, 875–877 GCD.java example, 874–875 Rational class and, 548–549

Generic instantiation, 774 Generics

case study: designing class for matrix using generic types, 788–793 case study: generic method for sorting array, 780–781 defining generic classes and interfaces, 776–777 erasing generic types, 786–788

Z11_LIAN1878_11_GE_IDX.indd 1215 1/2/18 10:54 PM

1216 Index

summary, 1057 TestMyHashMap.java example of map implementation, 1048–1049 TestMyHashSet.java example of set implementation, 1056–1057 what it is, 1032–1033

HashMap class concrete implementation of Map class, 850–852 implementation of Map class, 1032 load factor thresholds, 1040 overview of, 853 TestMap.java example, 853–855 types of maps, 850–851

HashSet class case study: counting keywords, 849–850 implementation of Set class, 1050 overview of, 838–842 TestHashSet.java example, 839–840 TestMethodsInCollection.java example, 840–841 types of sets, 838

Hashtable, 853 HBox and VBox

definition, 588 overview, 586 ShowHBoxVBox.java, 587–588

Heap class Heap.java example, 920–921 operations for manipulating heaps in, 920 sorting arrays with, 921–922

Heap sorts, 916–923 adding nodes to heaps, 918 arrays using heaps, 921–922 Heap class, 920–921 Heap.java example, 920–921 HeapSort.java example, 922 overview of, 916–917 removing root from heap, 918–919 storing heaps, 917 time complexity of, 922–923

Heaps adding nodes to, 918 arrays using, 921–922 binary heaps (binary trees), 916 dynamic memory allocation and, 283 height of, 922 implementing priority queues with, 969–970 removing root from, 918–919 storing, 917

Height, 976 Height of a heap, 922 Helper method, recursive

overview of, 750 RecursivePalindrome.java, 750–751

Hertz (Hz), clock speed in, 25 Hexadecimal numbers

converting to/from binary, 1189–1190 converting to/from decimal, 164–166, 206–208, 241–243, 767, 1189 equivalents of ASCII character set, 1181 overview of, 1187

Hidden data fields, 380, 382 High-level languages, 30–31 Hilbert curve, 771 Horizontal scroll bars, 688 Horizontal sliders, 690, 691 H-trees

fractals, 771 recursive approach to, 742

Huffman coding, 1000–1005

Greatest common denominator. see Gcd (greatest common denominator) Greedy algorithms

Dijkstra’s algorithm, 1126 overview of, 1001

Greenwich Mean Time (GMT), 74 GregorianCalendar class

Cloneable interface and, 540–541 in java.util package, 385 overview of, 529–530 TestCalendar.java example, 530–532

GridPane overview, 583–584 ShowGridPane.java, 584–585

Grids, representing using two-dimensional array, 322 Group classifier, 1163 Group processor, 1163 groupingby collector, grouping elements using, 1163–1166 Growth rates

algorithm for comparing, 869–870 comparing algorithms based on, 862

H Hamiltonian path/cycle, 1103 HandleEvent.java, 617–618 Hand-traces, for debugging, 128 Hangman game, 307, 516, 611, 831, 832 Hard disks, as storage device, 27 Hardware, 24 Has-a relationships

in aggregation models, 398–399 composition and, 464

Hash codes, 1033 compressing, 1034–1035 vs. hash functions, 1033–1035 for primitive types, 1033 for strings, 1033–1034

Hash functions, 1032 vs. hash codes, 1033–1035 as index to hash table, 1032

Hash map, 853, 1050 Hash set, 838 Hash tables, 1032, 1051. see also Maps

measuring fullness using load factor, 1039 parameters, 1047

hashCode method, 838, 1033 Hashing

collisions handling using open addressing, 1035–1039 collisions handling using separate chaining, 1039 compressing hash codes, 1034–1035 double hashing open addressing, 1037–1039 function, 1032–1035 hash codes for primitive types, 1033 hash codes for strings, 1033–1034 hash functions vs. hash codes, 1033–1035 key terms, 1057 linear probing open addressing, 1035–1036 load factor and rehashing, 1039–1041 map implementation with, 1041–1049 MyHashMap.java example of map implementation, 1043–1048 MyHashSet.java example of set implementation, 1050–1056 MyMap.java example of map implementation, 1042–1043 overview of, 1032 quadratic probing open addressing, 1036–1037 quiz and exercises, 1057–1059 set implementation with, 1050–1057

Z11_LIAN1878_11_GE_IDX.indd 1216 1/2/18 10:54 PM

Index 1217

hiding (encapsulation), 249 Inheritance

ArrayList object, 456–457 calling subclass constructors, 440 calling superclass methods, 442–443 case study: custom stack class, 463–464 casting objects and, 451–455 CastingDemo.java example, 452–453 Circle.java example, 436–438 constructor chaining and, 441–442 in designing stacks and queues, 966 DistinctNumbers.java example, 460–462 dynamic binding and, 447–451 equals method of Object class, 455–456 generic classes, 778 GeometricObject.java example, 435–436 interface inheritance, 532–533, 546 is-a relationships and, 464 key terms, 467 Object class and, 446 overriding methods and, 443–444 overview of, 434 preventing classes from being extended or overridden, 467 protected data and methods, 464–466 quiz and exercises, 469–473 Rectangle.java example, 438–439 summary, 468 superclasses and subclasses and, 434–440 TestArrayList.java example, 457–460 TestCircleRectangle.java example, 439–440 using super keyword, 440

Initializing variables AnalyzeNumbers.java, 277 arrays, 273 declaring variables and, 63 multidimensional arrays, 313 two-dimensional arrays, 315

Inner (nested) classes anonymous, 624–625 AnonymousHandlerDemo.java, 625–627 for defining listener classes, 623–624 KeyEventDemo.java, 636–637 ShortestPathTree class as inner class of WeightedGraph class,

1126–1127 TicTacToe.java, 694–698

Inorder traversal time complexity of, 993 tree traversal, 979

Input reading from console, 59–61 redirecting using while loops, 192 runtime errors, 43 streams. see InputStream classes

Input, process, output (IPO), 61 InputMismatchException class, 480, 505 Input/output devices, computers and, 27–28 InputStream classes

BufferedInputStream, 723–726 case study: copying files, 727 DataInputStream, 720–723 deserialization and, 731 DetectEndOfFile.java, 723 FileInputStream, 717–718 FilterInputStream, 720 ObjectInputStream, 728–729 overview of, 716–717

Huffman coding trees data compression using, 1000–1005 HuffmanCode.java example, 1002–1005

Hz (Hertz), clock speed in, 25

I Identifiers, 62 IDEs (integrated development environments) for creating/editing Java

source code, 33, 34, 37–38 IEEE (Institute of Electrical and Electronics Engineers), floating point

standard (IEEE 754), 67 if statements

common errors, 105–109 in computing body mass index, 111–112 in computing taxes, 112–115 conditional operator used with, 126 with enumerated types, 1199 nesting, 103 overview of, 100–102 SimpleIfDemo.java example, 101–102

if-else statements conditional expressions and, 126 dangling else ambiguity, 106–107 multi-way, 103–105 overview of, 103–105 recursion and, 748

IllegalArgumentException class, 485 Image, 578–580 Image class, 578 Image icons, ComboBoxDemo.java, 682 Images, ShowImage.java, 579–580 ImageView, 578–580 Immutable

BigInteger and BigDecimal classes, 408–409 class, 377 objects, 377–378 Rational class, 552 String object, 410–411 wrapper classes, 405

Implementation (coding), in software development process, 83–84 Implementation methods, 253–256 Implicit casting, 149, 451 Importing, types of import statements, 60 Increment method, in Increment.java example, 236 Increment (++) operator, 77–78 Incremental development

benefits of stepwise refinement, 256 coding incrementally, 186

Indentation, programming style, 41 Indexed variables elements, 272 Indexes

accessing elements in arrays, 270, 272 finding characters/substrings in a string, 158–159 List interface and, 804, 805 MyList.java example, 942–944 string index range, 153

indexOf method, 158–159 implementing MyLinkedList, 962 List interface, 804 MyArrayList.java example, 942, 947, 949

Indirect recursion, 745 Infinite loops, 184 Infinite recursion, 745 Information

getting information about exceptions, 487–489

Z11_LIAN1878_11_GE_IDX.indd 1217 1/2/18 10:54 PM

1218 Index

Cloneable interface, 540–541 Comparable interface, 535–536 ComparableRectangle.java example, 537–538 for defining common class behaviors, 588 defining generic, 776–778 House.java example, 541–545 key terms, 556 overview of, 523 questions and exercises, 557–562 raw types and backward compartibility, 782 SortComparableObjects.java example, 537 SortRectangles.java example, 538–539 summary, 556–557 TestEdible.java example, 532–535

Intermediate method, Stream interface, 1146 Interned strings, 411 Internet, 33 Interpreters, translating source program into machine code, 30, 31 IntStream, 1152–1155 Invoking methods, 230, 231, 355, 779

binary I/O classes, 716–717 BufferedInputStream and BufferedOutputStream classes,

723–726 case study: copying files, 727 case study: replacing text, 506–508 Copy.java, 727–728 DataInputStream and DataOutputStream classes, 720–723 DetectEndOfFile.java, 723 FileInputStream and FileOutputStream classes, 717–720 FilterInputStream and FilterOutputStream classes, 720 handling text I/O in Java, 714–715 key terms, 736 object I/O, 728–729 overview of, 503, 714 questions and exercises, 737–740 random-access files, 733–736 reading data from file using Scanner class, 504–506 reading data from the Web, 508–510 serializable interface, 730–731 serializing arrays, 731–733 summary, 737 TestDataStream.java, 721–722 TestFileStream.java, 718–720 TestObjectInputStream.java, 730 TestObjectOutputStream.java, 729–730 TestRandomAccessFile.java, 735–736 text I/O vs. binary I/O, 715–716 types of I/O devices, 27–28 writing data to file using PrintWriter class, 502–503

IOException, 717, 718 IPO (input, process, output), 61 Is-a relationships

design guide for when to use interfaces vs. classes, 546 inheritance and, 464

isAbsolute method, for checking file instances, 500–501 isDigit method, Character class, 166 isDirectory method, for checking file instances, 500–501 isFile method, for checking file instances, 500–501 isHidden method, for checking file instances, 500–501 Is-kind-of relationships, 546 isPalindrome method

RecursivePalindrome.java, 750–751 as tail-recursive method, 762

isPrime method, prime numbers, 241 isValid method, applying to Sudoku grid, 324 Iterable interface, 802

InputStream classes (Continued) TestDataStream.java, 721–722 TestFileStream.java, 718–720 TestObjectInputStream.java, 730

Insert key, on keyboards, 28 insert method

AVLTree class, 1027 overriding, 1016–1017

Insertion order, LinkedHashMap class, 853 Insertion sort algorithms

analyzing, 868 recurrence relations and, 869

Insertion sorts algorithms, 904–906 InsertionSort.java example, 905–906 time complexity of, 906

Instance methods accessing object data and methods, 355 in Circle.java (for CircleWithStaticMembers), 362–363 class design guidelines, 348–349 invoking, 392, 395 when to use instance methods vs. static, 363–366

Instance variables accessing object data and methods, 355 class design guidelines, 365 static variables compared with, 361–363 in TestCircleWithStaticMembers.java example, 363 when to use instance variables vs. static, 363–366

Instances. see also Objects checking file instantiation, 500 checking object instantiation, 346, 452 generic instantiation, 774

Institute of Electrical and Electronics Engineers (IEEE), floating point standard (IEEE 754), 67

int data type. see Integers (int data type) Integer.parseInt method, 159 Integers (int data type)

ArrayList for, 461 BigInteger class, 408–409 bit operators and, 1191 case study: designing class for matrix using generic types, 788, 789 casting to/from char types, 149 converting characters and numeric values to strings, 413–414 declaring variables and, 63 division of, 476–480 finding larger between two, 229 floating-point numbers converted to, 78–79 generic method for sorting array of Comparable objects, 780 greatest common denominator of, 873 hash codes for primitive types, 1033 IntegerMatrix.java example, 791 java.util.Random, 359 numeric types for, 67 sorting, 923 sorting int values, 929 specifying data types, 57 TestIntegerMatrix.java example, 792

Integrated development environments (IDEs), 33, 34 for creating/editing Java source code, 37–38 overview of, 33

Intelligent guesses, 186 Interfaces

abstract classes compared with, 543–545 benefits of, 538 benefits of generics, 774 case study: Rational class, 548–549

Z11_LIAN1878_11_GE_IDX.indd 1218 1/2/18 10:54 PM

Index 1219

button, 668–670 ButtonDemo.java, 669–670 case study: developing tic-tac-toe game, 693–698 case study: national flags and anthems, 701–703 CheckBox, 670–673 CheckBoxDemo.java, 671–673 ComboBox, 681–684 ComboBoxDemo.java, 682–684 DescriptionPane.java, 679–680 Labeled and Label, 666–668 LabelWithGraphic.java, 666–668 ListView, 684–686 ListViewDemo.java, 686–687 MediaDemo.java, 699–701 programming exercises, 704–711 quiz, 704 RadioButton, 673–675 RadioButtonDemo.java, 674–675 ScrollBar, 687–689 ScrollBarDemo.java, 688–689 Slider, 690–693 SliderDemo.java, 691–692 TextArea, 677–680 TextAreaDemo.java, 680 Textfield, 676–677 TextFieldDemo.java, 676–677 TicTacToe.java, 694–698 video and audio, 698–701

java.io File class, 499–501 PrintWriter class, 502 RandomAccessFile class, 734

java.lang Comparable interface, 536 Exception class, 496 Number class, 527 packages, 84 Throwable class, 481–483, 487

java.net MalformedURLException class, 509 URL class, 508

java.util Arrays class, 294–296 Calendar class, 529–530 creating stacks, 828 Date class, 358 EventObject class, 618–619 GregorianCalendar class, 385, 529–530 Java Collections Framework and, 798 Random class, 359 Scanner class, 60, 504–506

jdb debugger, 128 JDK (Java Development Toolkit)

jdb debugger in, 128 overview of, 34

JVM (Java Virtual Machine) defined, 38 detecting runtime errors, 476 garbage collection, 280 heap as storage area in, 283 interned string and, 411

K KBs (kilobytes), 26 Key constants, 635

Iteration/iterators advantages and variations of, 1000 binary search trees and, 998–1000 implementing MyLinkedList, 960, 962 Iterable interface, 999 Iterator object, 802 lists and, 802–803 loops and, 182 MyArrayList.java example, 948, 949 recursion compared with, 762 TestIterator.java example, 802–803 TestMyArrayList.java example, 950 traversing collections, 802–803

J Java Collections Framework. see Collections Framework hierarchy java command, for executing Java program, 39 Java Development Toolkit (JDK)

jdb debugger in, 128 overview of, 33, 34

Java EE (Java Enterprise Edition), 34 Java language specification, 33–34 Java Library, 358–361 Java ME (Java Micro Edition), 34 Java programming

creating, compiling, and executing programs, 37–40 displaying text in message dialog box, 45 high-level languages, 30 introduction to, 34 simple examples, 34–37 using Eclipse, 47–50 using NetBeans, 45–47

Java SE (Java Standard Edition), 34 Java Virtual Machine. see JVM (Java Virtual Machine) javac command, for compiling Java program, 39 Javadoc comments (/**.*/), 40 JavaFX

Arc, 597–599 binding properties, 570–572 BorderPane, 585–586 case study: ClockPane Class, 602–606 Circle and Ellipse, 594–596 Color class, 575–576 FlowPane, 581–583 Font class, 576–577 GridPane, 583–585 HBox and VBox, 586–587 Image and ImageView Classes, 578–580 key terms, 607 Layout panes, 580 Line, 591–592 nodes, 573–574 panes, 567–570 Polygon and Polyline, 599–602 quiz and exercises, 608–613 Rectangle, 592–594 shapes, 589 structure, 564–567 summary, 607–608 vs. Swing and AWT, 564 Text, 589–591

JavaFX CSS, 573 JavaFX UI controls

ScrollBar, 665-689 BounceBallSlider.java, 692–693

Z11_LIAN1878_11_GE_IDX.indd 1219 1/2/18 10:54 PM

1220 Index

Linear search algorithm, 897 comparing growth functions, 870 recurrence relations and, 869

Linear searches, arrays, 289–290 Linked data structures

binary search trees, 976–977 hash maps. see LinkedHashMap class hash sets. see LinkedHashSet class lists. see LinkedList class

Linked hash map, 853, 854 Linked hash set, 842, 847 LinkedHashMap class

concrete implementation of Map class, 850–852 implementation of Map class, 1032 overview of, 853 TestMap.java example, 853–855 types of maps, 850–851

LinkedHashSet class implementation of Set class, 1050 ordering elements in hash sets, 840 overview of, 842 SetListPerformanceTest.java example, 847–848 types of sets, 838

LinkedList class animation of linked lists, 940, 941 compared with ArrayList class, 806–808 defined under List interface, 804 Dequeue interface, 822–824 implementing buckets, 1039 implementing linked lists, 951–965 implementing MyLinkedList class, 955–963 implementing queues using linked lists. see Queues MyArrayList compared with MyLinkedList, 963 MyLinkedList, 940–941, 953, 963 representing edges in graphs using linked lists, 1070 SetListPerformanceTest.java example, 847–848 TestArrayAndLinkedList.java, 807–808 TestMyLinkedList.java example, 954–955 variations on linked lists, 964–965

Linux OS, 31 List interface

common features of lists defined in, 940 methods of, 804–806 overview of, 804 Vector class implementing, 820, 821

ListIterator interface, 805 Lists

adjacency lists for representing edges, 1068–1070 array lists. see ArrayList class as collection type, 798 comparing performance with sets, 846–848 implementing, 940–944 linked lists. see LinkedList class List interface, 804–806 ListViewDemo.java, 686–687 methods of List interface, 804–806 MyList.java example, 942–944 singleton and unmodifiable, 857 static methods for, 813–816

ListView, 684–686 Literal values, not using as identifiers, 1179 Literals

Boolean literals, 99 character literals, 147 constructing strings from string literal, 410 defined, 70

Keyboards, 27–28 KeyEvents

ControlCircleWithMouseAndKey.java, 637–638 KeyEventDemo.java, 636–637 overview of, 635–636

Keys hashing functions, 1032 maps and, 1057

keySet method, Map interface, 852 Key/value pairs, in maps, 850–851 Keywords (reserved words)

break and continue, 208–211 case study: counting, 849–850 extends, 546 final, 65 list of Java keywords, 1179 super, 440 throws, 484, 485 transient, 731 in Welcome.java, 35

Kilobytes (KBs), 26 Knight’s Tour, 769–770 Koch snowflake fractal, 769 Kruskal’s algorithm, 1134

L Label, 666–668 Labeled, 666–668 Labeling vertices, 1066 Labels, LabelWithGraphic.java, 666–668 Lambda expression

LambdaHandlerDemo.java, 629–631 overview of, 627–629

Landis, E. M., 1012 LANs (local area networks), 28 lastIndexOf method

implementing MyLinkedList, 962 List interface, 804 MyArrayList.java example, 947, 949 MyList.java example, 942 strings, 158–159

lastModified method, File class, 501 Latin square, 342–343 Layout panes

BorderPane, 585–586 FlowPane, 581 GridPane, 583–585 HBox and VBox, 586–587

Lazy operator, 118 Leaf, 976

deleting, 990 Left subtree, of binary trees, 976 Left-heavy, balancing AVL nodes, 1012, 1019 Length, 976 length method, for checking file instances, 500–501 Length, strings, 152–153, 419 Letters, counting, 285–288 Level, 976 Libraries, APIs as, 33 Line

overview, 591 ShowLine.java, 591–592

Line comments, in Welcome.java, 35 Line numbers, in Welcome.java, 34 Linear probing, collision handling, 1035–1036

Z11_LIAN1878_11_GE_IDX.indd 1220 1/2/18 10:54 PM

Index 1221

AVLTree class, 1019 balancing nodes on a path, 1016 options for balancing AVL nodes, 1013, 1014

M Mac OS, 31 Machine language

bytecode compared with, 38 overview of, 29 translating source program into, 30, 31

Machine stacks. see Call stacks Main class

defined, 347 in TestSimpleCircle.java example, 348

main method in Circle.java (AlternativeCircle.java) example, 350, 351 in ComputeExpression.java, 36–37 invoking, 232 main class vs., 347 receiving string arguments from command line, 296–297 in TestSimpleCircle.java example, 348 in TestTV.java example, 352–353 in Welcome.java, 35 in WelcomeWithThreeMessages.java, 36

Maintenance, in software development process, 82 MalformedURLException class, 509 Map interface

methods, 851–852, 854 overview of, 851

map method, 1150–1151 Maps

case study: counting occurrence of words using tree map, 855–856 containers supported by Java Collections Framework, 798 hash maps. see HashMap class key terms, 857 linked hash maps. see LinkedHashMap class overview of, 838, 850–855 quiz and exercises, 858–860 singleton and unmodifiable, 857 summary, 858 TestMap.java example, 853–855 tree maps. see TreeMap class

Maps, implementing with hashing MyHashMap.java example, 1043–1048 MyMap.java example, 1042–1043 overview of, 1041 TestMyHashMap.java example, 1048–1049

mapToInt method, 1153, 1154 Marker interfaces, 540 Match braces, in Welcome.java, 35 matches method, strings, 412 Math class

BigInteger and BigDecimal classes, 408–409 case study: computing angles of a triangle, 145–146 complex numbers, 560–561 exponent methods, 143 invoking object methods, 355 methods generally, 142 random method, 109–110, 120–121, 144 rounding methods, 143–144 service methods, 142 trigonometric methods, 142–143

Matrices adjacency matrices for representing edges, 1068–1070 case study: designing class for matrix using generic types, 788–789

floating-point literals, 71 integer literals, 71

LL imbalance, AVL nodes, 1012 LL rotation

AVLTree class, 1019 balancing nodes on a path, 1016 implementing, 1017 options for balancing AVL nodes, 1012

Load factor hash sets and, 838 rehashing and, 1039–1041

LoanCalculator.java, 632–633 Loans

Loan calculator case study, in event-driven programming, 631–633 Loan.java object, 392–394

Local area networks (LANs), 28 Local variables, 246 Locker puzzle, 304 Logarithmic algorithm, 868–870 Logic errors (bugs), 43, 128 Logical operators (Boolean operators)

overview of, 115 TestBooleanOperators.java example, 116–118 truth tables, 115–116

Long, numeric types converting characters and numeric values to strings, 413–414 hash codes for primitive types, 1033 integer literals and, 71 java.util.Random, 359 overview of numeric types, 67

LongStream, 1152–1155 Loop body, 182 Loop-continuation-condition

do-while loop, 192–193 loop design and, 188 in multiple subtraction quiz, 188 overview of, 182–183

Loops break and continue keywords as controls in, 208–211 case study: displaying prime numbers, 213–215 case study: finding greatest common denominator, 204–205 case study: guessing numbers, 185–188 case study: multiple subtraction quiz, 188–190 case study: predicting future tuition, 205–206 creating arrays, 281 deciding which to use, 198–200 design strategies, 188–190 do-while loop, 192–195 examples of determining Big O, 864–867 graph edges, 1064 input and output redirections, 192 iteration compared with recursion, 762 key terms, 215 for loops, 195–198 minimizing numeric errors related to, 202–203 nesting, 200–202 overview of, 182 quiz and exercises, 216–226 sentinel-controlled, 190–192 summary, 215–216 while loop, 182–185

Lottery game, 831 Lower-bound wildcards, 784 Low-level languages, 30 LR imbalance, AVL nodes, 1013, 1014 LR rotation

Z11_LIAN1878_11_GE_IDX.indd 1221 1/2/18 10:54 PM

1222 Index

overview of, 228 passing arrays to, 281–284 passing objects to, 371–375 passing parameters by values, 236–239 passing to two-dimensional arrays, 317–318 quiz and exercises, 258–268 recursive methods, 742 returning arrays from, 284–285 rounding, 143–144 static. see Static methods stepwise refinement, 249–256 summary, 257 top-down and/or bottom-up implementation, 251–253 top-down design, 250–252 tracing or stepping over as debugging technique, 128 trigonometric, 142–143 variable scope and, 246–247 void method example, 233–235

MHz (Megahertz), clock speed, 25 Microsoft Windows, 31 Million bits per second (Mbps), 28 min method

finding minimum element in lists, 815 Math class, 144

Minimum spanning trees (MSTs) MST algorithm, 1117–1118 overview of, 1108 Prim’s minimum spanning tree algorithm, 1115–1117 TestMinimumSpanningTree.java, 1118–1120 weighted graphs and, 1104 WeightedGraph class, 1109–1111

Mnemonics, in assembly language, 29 Modeling, graphs and, 1070–1078 Modems (modulator/demodulator), 28 Modifier keys, on keyboards, 27 Modifiers

list of, 1184–1185 method modifiers, 229

Modularizing code GreatestCommonDivisorMethod.java, 239–240 overview of, 239 PrimeNumberMethod.java, 240–241

Monitors (displays), 28 Motherboard, 25 Mouse, as I/O device, 28

ControlCircleWithMouseAndKey.java, 637–638 event-driven programming, 633–634 MouseEvent, 633–634

MouseEvent, 633–634 MST algorithm, 1117–1118 MST class, 1118–1119 MSTs. see Minimum spanning trees (MSTs) Multi-dimensional arrays. see Arrays, multi-dimensional Multimedia. see JavaFX UI controls Multiple-choice test, 318–320 Multiplication (*=) assignment operator, 76 Multiplication operator (*), 37, 68, 72 Multiplication table, 200, 201 Multiplicities, in object composition, 397 Multiprocessing, 32 Multiprogramming, 32 Multithreading, 32 Multi-way if-else statements

in computing taxes, 112–115 overview of, 103–105

Mutator methods. see Setter (mutator) methods

Matrices (Continued) GenericMatrix.java example, 789–791 IntegerMatrix.java example, 791 RationalMatrix.java example, 791–792 TestIntegerMatrix.java example, 792 TestRationalMatrix.java example, 792–793 two-dimensional arrays for storing, 312–313

max and min method, 1148, 1150 max method

defining and invoking, 230–232 finding minimum element in lists, 815 GeometricObjectComparator.java example, 810 MaxUsingGenericType.java example, 782–783 overloading, 243 overview of, 144

maxRow variable, for finding largest sum, 316 Mbps (million bits per second), 28 MBs (megabytes), of storage, 26 Media, 698–701 MediaPlayer, 698–701 MediaView, 698–701 Megabytes (MBs), of storage, 26 Megahertz (MHz), clock speed, 25 Memory, computers, 26 Merge sorts

algorithms, 909 heap sort compared with, 922 merge sort algorithms, 909 MergeSort.java example, 909–912 overview of, 909–912 quick sorts compared with, 916 recurrence relations and, 869 time complexity of, 912

mergeSort method, 911 Mersenne prime, 264 MessagePanel class

ClockPane.java, 604–606 DisplayClock.java, 603–604

Meters, converting to/from feet, 260 Method header, 229 Method modifiers, 229, 1184–1185 Method reference, 811 Method signature, 229 Methods

abstraction and, 249–256 accessing object methods, 354–355 calling, 230–232 case study: converting decimals to hexadecimals, 206–208 case study: converting a hexadecimal digit to a decimal value, 164–166 case study: generating random numbers, 247–249 case study: generic method for sorting array, 780–781 class, 361–366 Collection interface, 800 commenting, 41 Comparator interface, 809, 810 defining, 228–230 generic, 778–780 identifiers, 62 implementation methods, 253–256 invoking, 230, 231, 355, 779 key terms, 256 modularizing code, 239–241 naming conventions, 66 object actions defined by, 346–348 overloading, 243–246 overriding, 1016–1017

Z11_LIAN1878_11_GE_IDX.indd 1222 1/2/18 10:54 PM

Index 1223

abstract number class, 527–529 binary. see Binary numbers case study: converting hexadecimals to decimals, 164–166, 241–243 case study: displaying prime numbers, 213–215 case study: generating random numbers, 247–249 case study: guessing numbers, 185–188 casting to/from char types, 149 conversion between strings and, 159–160 converting to/from strings, 413–414 decimal. see Decimal numbers double. see double (double precision), numeric types floating-point. see Floating-point numbers (float data type) generating random numbers, 109–110 GreatestCommonDivisorMethod.java, 239–240 hexadecimal. see Hexadecimal numbers integers. see Integers (int data type) LargestNumbers.java, 528–529 overview of, 67–70 PrimeNumberMethod.java, 240–241 processing large numbers, 408–409 types of number systems, 1187–1190

Numerators, in rational numbers, 548 Numeric keypads, on keyboards, 28 Numeric literals, 70–72 Numeric operators

applied to characters, 149 overview of, 68–69

O Object class, 446, 455–456 Object I/O. see ObjectInputStream/ObjectOutputStream classes Object member access operator (.), 354, 453 Object reference variables, 354 ObjectInputStream/ObjectOutputStream classes

overview of, 728–729 serializable interface, 730–731 serializing arrays, 731–733 TestObjectInputStream.java, 730 TestObjectOutputStream.java, 729–730

Object-oriented programming (OOP), 346, 354, 394–397 Objects

accessing data and methods of, 354–355 accessing via reference variables, 354 array of, 283 ArrayList class, 456–457 arrays of, 375–377 automatic conversion between primitive types and wrapper class types,

407–408 BigInteger and BigDecimal classes, 408–409 cannot be created from abstract classes, 526 case study: designing class for stacks, 402–404 case study: designing Course class, 400–402 casting, 451–455 Circle.java (for CircleWithStaticMembers) example, 362–363 in Circle.java (for CircleWithPrivateDataFields) example,

369–370 class abstraction and encapsulation, 390–394 class design guidelines, 553–556 classes from Java Library, 358–361 comparing primitive variables with reference variables, 356–358 composing, 398–399 constructors, 353 creating, 348–353 data field encapsulation for maintaining classes, 368–371 Date class, 358–359 defining classes for, 346–348

N Named constants. see Constants Naming conventions

class design guidelines, 554 interfaces, 546 programming and, 66 wrapper classes, 404

Naming rules, identifiers, 62 NavigableMap interface, 853 N-by-n matrix, 262 Negative angles, drawing arcs, 599 Neighbors

depth-first searches (DFS), 1084 vertices, 1064

Nested classes. see Inner (nested) classes Nested if statements

in computing body mass index, 111–112 in computing taxes, 112–115 overview of, 103

Nested loops, 200–202, 315, 865 NetBeans

built in debugging, 128 creating/editing Java source code, 37

Network interface cards (NICs), 28 new operator

creating arrays, 271 creating objects, 353

next method, whitespace characters and, 155 nextLine() method, whitespace characters and, 155 Next-line style, block styles, 41 NICs (network interface cards), 28 Nine tails problem

graphic approach to, 1093–1099 reducing to shortest path problem, 1130–1133

No-arg constructors class design guidelines, 554 Loan class, 392 wrapper classes not having, 405

Node, 564–567 Nodes, AVL trees

balancing on a path, 1016–1017 creating, 1019 creating and storing in AVLTreeNode class, 1015–1016 deleting elements, 1018 rotation, 1019–1020

Nodes, binary trees deleting leaf node, 990–991 overview of, 976 representing binary search trees, 977–978

Nodes, JavaFX, 573–574 Nodes, linked lists

creating, 956, 957 deleting, 957–959 overview of, 951–953 storing elements in, 955, 956

noneMatch method, 1150 Nonleaves, finding, 1006 Not (!) logical operator, 115–119 Not equal to (!=) comparison operator, 98 NotSerializableException, 731 null values, objects, 355–356 NullPointerException, as runtime error, 356 Number class

case study: abstract number class, 527 as root class for numeric wrapper classes, 527

Numbers/numeric types

Z11_LIAN1878_11_GE_IDX.indd 1223 1/2/18 10:54 PM

1224 Index

case study: copying files, 727 DataOutputStream, 720–723 DetectEndOfFile.java, 723 FileOutputStream, 717–718 FilterOutputStream, 720 ObjectOutputStream, 728–729 overview of, 716–717 serialization and, 731 TestDataStream.java, 721–722 TestFileStream.java, 718–720 TestObjectOutputStream.java, 729–730

Overflows Rational class, 552 variables, 87

Overloading methods, 243–246 Overriding methods, 443–444, 1016–1017

P π (pi), estimating, 261 Package-private (package-access) visibility modifiers, 366 Packages

organizing classes in, 367 organizing programs in, 40

Page Down key, on keyboards, 28 Page Up key, on keyboards, 28 Pair of points, algorithm for finding closest, 883–886 Palindromes

case study: checking if string is a palindrome, 211–213 case study: ignoring nonalphanumeric characters when checking palin-

dromes, 420–422 palindrome integers, 258 palindromic primes, 264 RecursivePalindrome.java, 750–751 RecursivePalindromeUsingSubstring.java, 749–750

Panels ButtonInPane.java, 568 MessagePanel class. see MessagePanel class

Parallel edges, 1064 Parallel execution, order of, 1157 Parallel streams

overview of, 1155 ParallelStreamDemo.java example, 1155–1157 vs. sequential streams, 1156–1157

Parameters actual parameters, 229 defining methods and, 228–229 generic classes, 778 generic methods, 780 generic parameters not allowed in static context, 787–788 as local variable, 246 order association, 236 passing by values, 236–239 variable-length argument lists, 288–289

Parent, 567 Parentheses (( ))

defining and invoking methods and, 249 in Welcome.java, 36

Parsing methods, 406 Pascal, high-level languages, 30 Pass-by-sharing

arrays to methods, 282 objects to methods, 372

Pass-by-value arrays to methods, 282 Increment.java example, 236

Objects (Continued) edges defined as, 1068 equals method of Object class, 455–456 event listener object, 619 event objects, 618 immutable, 377–378 inheritance. see Inheritance key terms, 383, 422 Loan.java, 392–394 null values, 355–356 Object class, 446 object-oriented thinking, 394–397 ObservablePropertyDemo.java, 638–639 overview of, 346, 390 passing to methods, 371–375 processing primitive data type values as, 404–407 quiz and exercises, 384–388, 423–432 Random class, 359–360 reference data fields and, 355–356 representing edges, 1067 ResizableCircleRectangle.java, 639–640 static variables, constants, and methods and, 361–366 summary, 383–384, 422–423 TestCircle.java example, 348–349 in TestCircleWithPrivateDataFields.java example, 370–371 in TestCircleWithStaticMembers.java example, 363–366 in TestTV.java example, 352–353 this reference and, 380–383 TotalArea.java example, 376–377 in TV.java example, 351–352 variable scope and, 379–380 vertices as object of any type, 1065 visibility modifiers, 366–368

Off-by-one errors arrays and, 275 in loops, 184

OOP (Object-oriented programming), 346, 354, 394–397 Open addressing, hashing

collision handling using, 1035–1039 double hashing, 1037–1039 linear probing, 1035–1036 quadratic probing, 1036–1037

Operands defined, 117 incompatible, 117

Operators assignment operator (=), 64–65 augmented assignment operators, 76–77 bit operators, 1191 comparison operators, 98 increment and decrement operators, 77–78 numeric operators, 68–69 precedence and associativity, 126–127 precedence and associativity chart, 1182–1183 processing, 825 unary and binary, 70

Option buttons. see Radio buttons Or (||) logical operator, 115–119 OSs (operating systems)

overview of, 31 tasks of, 32

Output redirection, 192 streams, 714

OutputStream classes BufferedOutputStream, 723–726

Z11_LIAN1878_11_GE_IDX.indd 1224 1/2/18 10:54 PM

Index 1225

writing data to file using, 502–503 for writing text data, 714

Priority queues implementing, 969–970 MyPriorityQueue.java example, 969–970 overview of, 821 PriorityQueue class, 823 for storing weighted edges, 1105 TestPriorityQueue.java example, 970

PriorityQueue class, 823 private

encapsulation of data fields and, 368–369 visibility modifier, 367–368, 464–466

Problems breaking into subproblems, 214 creating programs to address, 56 solving with recursion, 748–750

Procedural paradigm, compared with object-oriented paradigm, 396–397 Procedures, 229. see also Methods Processing arrays, 273–275 Processor, 1163 Programming errors. see also Exception handling

ClassCastException, 452 debugging, 128 logic errors, 43 minimizing numeric errors related to loops, 202–203 runtime errors, 43 selections, 105–109 syntax errors, 35, 36, 42 using generic classes for detecting, 774–776

Programming languages assembly language, 29 high-level languages, 30–31 Java. see Java programming machine language, 29 overview of, 24

Programming style block styles, 41 comments and, 40–41 indentation and spacing, 41 overview of, 40

Programs/programming assignment statements and expressions, 64–65 case study: counting monetary units, 85–87 case study: displaying current time, 74–75 character data type, 147–151 coding incrementally, 186 evaluating expressions and operator precedence rules, 72–73 exponent operations, 70 identifiers, 62 increment and decrement operators, 77–78 introduction to, 56 with Java language. see Java programming key terms, 89 modularizing code, 239–241 named constants, 65–66 naming conventions, 66 numeric literals, 70–72 numeric operators, 68–70 numeric type conversions, 79–81 numeric types, 67 overview of, 24 questions and exercises, 91–96 reading input from console, 59–61 recursive methods in, 742 software development process, 81–85

objects to methods, 371 overview of, 236 TestPassByValue.java example, 237–239

Passwords, checking if string is valid password, 262 PaswordField, 677 PathTransition, 641–644 Pentagonal numbers, 258 Perfect hash function, 1032 Perfectly balanced trees, 1012 Pivot element, 912 Pixels (picture elements), measuring resolution in, 28 Points

algorithm for finding closest pair of, 883–886 finding convex hull for a set of points, 889–891

Polygon and Polyline overview, 599 ShowPolygon.java, 600–602

Polymorphism CastingDemo.java example, 452–453 overview of, 447 PolymorphismDemo.java example, 447

Polynomial hash codes, 1034 Postfix decrement operator, 77–78 Postfix increment operator, 77–78 Postfix notation, 833–834 Postorder traversal

time complexity of, 993 tree traversal, 979

Posttest loops, 198 pow method, Math class, 70 Precedence, operator, 126–127, 1182–1183 Prefix decrement operator, 77–78 Prefix increment operator, 77–78 Preorder traversal

time complexity of, 993 tree traversal, 979

Pretest loops, 198 Prime numbers

algorithm for finding, 877–883 case study: displaying prime numbers, 213–215 comparing prime number algorithms, 883 EfficientPrimeNumbers.java example, 879–882 PrimeNumberMethod.java, 240–241 PrimeNumbers.java example, 878–879 SieveOfEratosthenes.java example, 882–883 types of, 264

Primitive types (fundamental types) automatic conversion between primitive types and wrapper class types,

407–408, 775 casting, 453 comparing parameters of primitive type with parameters of reference

types, 373 comparing primitive variables with reference variables, 356–358 converting wrapper object to/from (boxing/unboxing), 407 creating arrays of, 375 hash codes for, 1033

Prim’s minimum spanning tree algorithm Dijkstra’s algorithm compared to, 1121 overview of, 1115–1117

print method, PrintWriter class, 60, 502–503 printf method, PrintWriter class, 502–503 Printing arrays, 315 println method, PrintWriter class, 60, 502–503 printStackTrace method, 487–488 PrintWriter class

case study: replacing text, 506–508

Z11_LIAN1878_11_GE_IDX.indd 1225 1/2/18 10:54 PM

1226 Index

case study: designing class for matrix using generic types, 788–789 overview of, 548–549 Rational.java example, 550–553 RationalMatrix.java example, 791–792 TestRationalClass.java example, 549–550 TestRationalMatrix.java example, 792–793

Rational numbers, representing and processing, 548–550 Raw types, backward compatiblity and, 782–783 readASolution() method, applying to Sudoku grid, 324 Read-only streams, 733. see also InputStream classes Read-only views, Collections class, 857 Rebalancing AVL trees, 1012–1014 Rectangle

overview, 592–593 ShowRectangle.java, 593–594

Recurrence relations, in analysis of algorithm complexity, 869 Recursion

binary searches, 752 case study: computing factorials, 742–745 case study: computing Fibonacci numbers, 745–748 case study: determining directory size, 753–754 case study: fractals, 758–761 case study: Towers of Hanoi, 755–758 ComputeFactorial.java, 743–745 ComputeFactorialTailRecursion.java, 763 ComputeFibonacci.java, 746–748 depth-first searches (DFS), 1084 DirectorySize.java, 753–754 displaying/visualizing binary trees, 995 helper method, 750 iteration compared with, 762 key terms, 763 overview of, 742 problem solving by thinking recursively, 748–750 questions and exercises, 764–772 RecursivePalindrome.java, 750–751 RecursivePalindromeUsingSubstring.java, 749–750 RecursiveSelectionSort.java, 751 selection sorts, 751 SierpinskiTriangle.java, 758–761 summary, 764 tail recursion, 762–763 TowersOfHanoi.java, 756–758

Recursive methods, 742 Red–black trees, 1032 reduce method, stream reduction using, 1157–1160 Reduction, characteristics of recursion, 748 Reference data fields, 381 Reference types

classes as, 354 comparing parameters of primitive type with parameters of reference

types, 373 comparing primitive variables with, 356–358 generic types as, 774 reference data fields, 355–356 string data type as, 152

Reference variables accessing objects with, 354 array of objects as array of, 375 comparing primitive variables with, 356–358

Register listeners ControlCircle.java, 622–623 ControlCircleWithMouseAndKey.java, 620–621, 637–638 KeyEventDemo.java, 636–637 LoanCalculator.java, 632–633 overview of, 619–620

Programs/programming (Continued) string data type, 152–161 summary, 89–90 variables, 62–64 writing a simple program, 56–59

protected data and methods, 464–466 visibility modifier, 367, 464–466

Protected data fields, 1015 Pseudocode, 56 Public classes, 349 public method, 370 public visibility modifier, 366–368, 464–466 Python, high-level languages, 30

Q Quadratic algorithm, 865, 870 Quadratic probing, collision handling, 1036–1037 Query methods, Map interface, 851 Query operations, Collection interface, 799 Queue interface, 822 Queues

breadth-first search algorithm, 1091 bucket sorts and, 924–925 as collection type, 798 Dequeue interface, 822–824 GenericQueue.java example, 967 implementing, 965–969 overview of, 821 priority queues. see Priority queues Queue interface, 822 TestStackQueue.java example, 967–969 WeightedGraph class, 1109–1110

Quick sorts algorithm, 912–913 merge sorts compared with, 916 overview of, 912 QuickSort.java example, 913–916

Quincunx, 308 Quotients

Quotient.java example, 476 QuotientWithException.java example, 478–480 QuotientWithIf.java example, 477 QuotientWithMethod.java example, 477–478

R Radio buttons, 673–675 RadioButtonDemo.java, 674–675 Radix sorts, 923–925 Ragged arrays, 314–315, 1068 RAM (random-access memory), 26 Random class, java.util, 359 random method

case study: generating random numbers, 247–249 case study: lottery, 120–121 Math class, 109–110, 144

Random numbers case study: generating random numbers, 247–249 case study: lottery, 120–121 generating, 109–110

Random-access files overview of, 733–734 TestRandomAccessFile.java, 735–736

Random-access memory (RAM), 26 Rational class

Z11_LIAN1878_11_GE_IDX.indd 1226 1/2/18 10:54 PM

Index 1227

reading data from file using, 504–505 for reading text data, 714

Scanners case study: replacing text, 506–508 creating, 480

Scene, 564–567 Scheduling operations, 32 Scientific notation, of floating-point literals, 72 Scope, of variables, 63, 246–247 Screen resolution, 28 Scroll bars

overview of, 687–688 ScrollBarDemo.java, 688–689

Scroll panes DescriptionPanel.java, 680 overview of, 678 scrolling lists, 686

search method, AVLTree class, 1027 Searches

arrays, 289 binary search trees. see Binary search trees (BST) binary searches, 290–293, 752 linear searches, 289–290 recursive approach to searching for words, 742 search keys, 1032, 1057

SearchTree class as inner class of UnweightedGraph class, 1077 MST class extending, 1117–1118 ShortestPathTree class extending, 1126 traversing graphs and, 1083

Secondary clustering, quadratic probing issue, 1037 Segments, merging, 928–929 Selection sort algorithm

analyzing, 868 recurrence relations and, 869

Selection sorts arrays, 293–294 RecursiveSelectionSort.java, 751 using recursion, 751

Selection statements, 98, 100 determining Big O for, 864–867

Selections Addition.Quiz.java example, 99–100 boolean data type, 98–100 case study: computing Body Mass Index, 111–112 case study: computing taxes, 112–115 case study: determining leap year, 119–120 case study: guessing birthdays, 161–164 case study: lottery, 120–121 common errors, 105–109 conditional operators, 125–126 debugging, 128 formatting output consoles, 167–171 generating random numbers, 109–110 if statements, 100–101 if-else statements, 102–103 key terms, 129 logical operators, 115–119 nested if statements and multi-way if-else statements, 103–105 operator precedence and associativity, 126–127 overview of, 98 questions and exercises, 130–140 summary, 129 switch statements, 122–125

Semicolons (;), common errors, 106 Sentinel-controlled loops, 190–192

Regular expressions matching strings with, 412, 1192 replacing and splitting strings, 1195–1196 syntax, 1192–1195

Rehashing load factor and, 1039–1041 time complexity of hashing methods and, 1048

Relative file names, 499–500 Remainder (%=) assignment operator, 76–77 Remainder (%) or modulo operator, 68, 72 remove method, linked lists, 951, 961 Repetition

determining Big O for repetition statements, 864–867 loops. see Loops

replace method, strings, 412 replaceAll method, strings, 412, 1195 replaceFirst method, strings, 412, 1195 Requirements specification, in software development process, 81 Reserved words. see Keywords (reserved words) Resources, role of OSs in allocating, 32 Responsibilities, separation as class design principle, 554 return statements, 231 Return value type

constructors not having, 353 in defining methods, 229

Reusable code benefits of stepwise refinement, 256 code modularization and, 239 method enabling, 232 methods for, 228

reverse method applying to lists, 814 returning arrays from methods, 284

Right subtree, of binary trees, 976 Right-heavy, balancing AVL nodes, 1012, 1020 RL imbalance, AVL nodes, 1013, 1014 RL rotation

AVLTree class, 1020, 1021 balancing nodes on a path, 1016 options for balancing AVL nodes, 1013, 1014

Root, of binary trees, 976, 977 Rotation

AVLTree class, 1019–1020 balancing nodes on a path, 1016–1017 implementing, 1017 methods for performing, 1023 options for balancing AVL nodes, 1012–1014

Rounding methods, Math class, 143–144 RR imbalance, AVL nodes, 1012, 1014 RR rotation

AVLTree class, 1020, 1021 balancing nodes on a path, 1016 options for balancing AVL nodes, 1012, 1014

Runtime errors debugging, 128 declaring, 483 exception handling and, 61, 476 NullPointerException as, 356 programming errors, 43

Runtime stacks. see Call stacks

S Scanner class

obtaining input with, 89 for reading console input, 59–61

Z11_LIAN1878_11_GE_IDX.indd 1227 1/2/18 10:54 PM

1228 Index

Sibling, 976 Sierpinski triangle

case study, 758–761 computing recursively, 766, 769, 771 SierpinskiTriangle.java, 758–761

Sieve of Eratosthenes, 881–883 Simple graphs, 1064 sin method, trigonometry, 142–143 Single abstract method (SAM) interface, 629 Single precision numbers. see Floating-point numbers (float data type) Single-dimensional arrays. see Arrays, single-dimensional Single-source shortest-path algorithm, Dijkstra’s, 1122–1127 Singly linked lists. see LinkedList class Sinking sorts, 303, 906–908 Sliders

overview of, 690 SliderDemo.java, 691–692

Software development process, 81–85 programs as, 24

sort method Arrays class, 295, 296 ComparableRectangle.java example, 537–538 lists and, 814 SortRectangles.java example, 538–539 using recursion, 751

sorted method, 1149–1150 SortedMap interface, 852, 853 Sorting

adding nodes to heaps, 917 arrays using heaps, 921–922 bubble sort, 906–908 bucket sorts and radix sorts, 923–925 complexity of external sorts, 932 complexity of heap sorts, 922–923 CreateLargeFile.java example of external sorts, 925–926 external sorts, 925–932 Heap class and, 920–921 heap sort, 916–923 Heap.java example, 920–921 HeapSort.java example, 922 implementation phases of external sorts, 926–931 insertion sorts, 904–906 key terms, 932 merge sorts, 909–912 overview of, 904 quick sort, 912–916 quiz and exercises, 933–937 removing root from heap, 918–919 storing heaps, 917 summary, 932–933

Sorting arrays bubble sorts, 303 case study: generic method for, 780–781 insertion sorts, 904–906 overview of, 293 selection sorts, 293–294

Source objects, event sources and, 618–619 Source program or source code, 30 Space complexity, 863 Spacing, programming style and, 41 Spanning trees

graphs, 1064 minimum spanning trees, 1115–1117 MST algorithm, 1117–1118 Prim’s minimum spanning tree algorithm, 1115–1117

Separate chaining handling collision in hashing, 1039 implementing map using hashing, 1041

Sequence statements, determining Big O for, 864–867 Sequential files, input/output streams, 733 Sequential streams, 1155

parallel streams vs., 1156–1157 Serialization

of arrays, 731–733 of objects, 731

set method, List interface, 805 Set operations, Collection interface, 799 setLength method, StringBuilder class, 419 setRadius method

Circle example, 349 CircleWithPrivateDataFields.java example, 370

Sets case study: counting keywords, 849–850 as collection type, 798 comparing list performance with, 846–848 HashSet class, 838–842 key terms, 857 LinkedHashSet class, 842 overview of, 838 quiz and exercises, 858–860 singleton and unmodifiable, 857 summary, 858 TestHashSet.java example, 839–840 TestLinkedHashSet.java example, 842 TestMethodsInCollection.java example, 840–841 TestTreeSet.java example, 843 TestTreeSetWithComparator.java example, 843–845 TreeSet class, 842–846

Sets, implementing with hashing MyHashSet.java example, 1050–1056 overview of, 1050 TestMyHashSet.java example, 1056–1057

Setter (mutator) methods ArrayList class and, 460 encapsulation of data fields and, 369–370 implementing linked lists, 951

Seven Bridges of Königsberg problem, 1063 Shallow copies, clone method and, 542–543 Shapes, 567–570

Arc, 597–599 Circle and Ellipse, 594–596 Line, 591–592 Polygon and Polyline, 599–602 Rectangle, 592–594 Text, 589–591

Sharing code, 232 short, numeric types

hash codes for primitive types, 1033 overview of, 67

Short-circuit operator, 118 Shortest path tree, 1125 Shortest paths

case study: weighted nine tails problem, 1130–1133 Dijkstra’s algorithm, 1122–1127 finding with graph, 1062, 1065 nine tails problem, 1093–1099 overview of, 1121–1122 TestShortestPath.java, 1127–1128 WeightedGraph class and, 1111

ShortestPathTree class, 1126 Shuffling arrays, 274, 316–317

Z11_LIAN1878_11_GE_IDX.indd 1228 1/2/18 10:54 PM

Index 1229

case study: occurrences of words, 1172–1174 case study: processing all elements in two-dimensional array,

1169–1170 CollectDemo.java example, 1161–1163 CollectGroupDemo.java example, 1164–1166 CountKeywordStream.java example, 1171–1172 CountLettersUsingStream.java example, 1167–1168 CountOccurrenceOfLettersInAString.java example, 1168–1169 CountOccurrenceOfWordsStream.java example, 1173–1174 DirectorySizeStream.java example, 1170–1171 DoubleStream, 1152–1155 grouping elements using groupingby collector, 1163–1166 IntStream, 1152–1155 IntStreamDemo.java example, 1152–1155 LongStream, 1152–1155 overview of, 1146 parallel streams, 1155–1157 ParallelStreamDemo.java example, 1155–1157 quiz and exercises, 1174–1175 Stream class, 1147 stream pipelines, 1146–1152 stream reduction using collect method, 1160–1163 stream reduction using reduce method, 1157–1160 StreamDemo.java example, 1148–1149 StreamReductionDemo.java example, 1158–1160 summary, 1174 TwoDimensionalArrayStream.java example, 1169–1170

String class, 410 String concatenation operator (+), 58 String literals, 410 String variables, 410 StringBuffer class, 416–417 StringBuilder class

case study: ignoring nonalphanumeric characters when checking palindromes, 420–422

modifying strings in, 417–418 overview of, 416, 417 toString, capacity, length, setLength, and charAt

methods, 419 Strings

in binary I/O, 720–721 case study: checking if string is a palindrome, 211–213 case study: converting hexadecimals to decimals, 241–243 case study: counting the occurrences of each letter in a string,

1168–1169 case study: ignoring nonalphanumeric characters when checking palin-

dromes, 420–422 case study: revising the lottery program, 166–167 Character class, 211–213 command-line arguments, 296–299 concatenating, 58, 152 constructing, 410 conversion between numbers and, 159–160 converting to/from arrays, 413 finding characters or substrings in, 412–413 formatting, 414–416 generic method for sorting array of Comparable objects, 780–781 hash codes for, 1033–1034 immutable and interned, 410–411 matching, replacing, and splitting by patterns, 412–413, 1195–1196 overview of, 410 replacing, and splitting, 411–412 string data type, 152 StringBuilder and StringBuffer classes, 416–422 substrings, 59, 157–158 in Welcome.java, 35

TestMinimumSpanningTree.java, 1118–1120 traversing graphs and, 1083

Special characters, 36 Specific import, 60 split method, strings, 412, 413, 1195, 1196 Stack class, 821 StackOfIntegers class, 402–403 StackOverflowError, recursion causing, 762 Stacks

case study: designing class for stacks, 402–404 case study: evaluating expressions, 825–826 EvaluateExpression.java example, 826–829 GenericStack class, 777–778 implementing, 965–969 Stack class, 821 TestStackQueue.java example, 967–969

Stage, 564, 567 State, of objects, 346 Statements

break statements, 123 continue statements, 209–210 executing one at a time, 128 executing repeatedly (loops), 182 in high-level languages, 30 if. see if statements if-else. see if-else statements return statements, 231 switch statements, 122–125 terminators, 35

Static methods in Circle.java (for CircleWithStaticMembers), 362–363 class design guidelines, 555 declaring, 362 defined, 362 for lists and collections, 813–816 Stream interface, 1146 when to use instance methods vs. static, 363–366 wrapper classes and, 406

Static variables in Circle.java (for CircleWithStaticMembers), 362–363 class, 361–366 class design guidelines, 555 declaring, 362 instance variables compared with, 361–363 in TestCircleWithStaticMembers.java example, 363 when to use instance variables vs. static, 363–366

Stepwise refinement benefits, 256 implementation methods, 253–256 method abstraction, 249–256 top-down and/or bottom-up implementation, 251–253 top-down design, 250–252

Storage devices CDs and DVDs, 27 disks, 27 overview of, 26–27 USB flash drives, 27

Storage units, for measuring memory, 25 Stream.of method, 1149 Streams, 1146

Z11_LIAN1878_11_GE_IDX.indd 1229 1/2/18 10:54 PM

1230 Index

TextArea, 677–680 TextAreaDemo.java, 680 TextField, 676–677 TextFieldDemo.java, 676–677 .txt files (text), 716

Text I/O vs. binary I/O, 715–716 handling in Java, 714–715 overview of, 714

TextPad, for creating/editing Java source code, 34 thenComparing method, 812 this reference

invoking constructors with, 382 overview of, 380 referencing data fields with, 381

Three-dimensional arrays. see Arrays, multi-dimensional throw keyword

chained exceptions, 495–496 throw ex for rethrowing exceptions, 494 for throwing exceptions, 489

Throwable class generic classes not extending, 788 getting information about exceptions, 487 java.lang, 481–483

Throwing exceptions, 484–485, 489–492 CircleWithException.java example, 489 QuotientWithException.java example, 478–480 rethrowing, 494–495 TestCircleWithCustomException.java example, 498 throw keyword for, 485

throws keyword chained exceptions, 496 for declaring exceptions, 484, 489 IOException, 717, 718 for throwing exceptions, 485

Tic-tac-toe game, 332 Time complexity, 863

AVL trees, 1027 BST class, 993 bubble sort, 908 heap sorts, 922–923 insertion sorts, 906 merge sorts, 912 rehashing, 1048

toArray method, 1148, 1151–1152 toCharArray method, converting strings into arrays, 413 ToggleButton, 673 ToggleGroup, 674 Token reading methods, Scanner class, 505–506 Top-down design, 250–252 Top-down implementation, 251–253 toString method

ArrayList class, 459 Arrays class, 296 Date class, 359 implementing MyLinkedList, 961 MyArrayList.java example, 947, 949 Object class, 455 StringBuilder class, 419

total variable, for storing sums, 316 Towers of Hanoi problem

analyzing algorithm for, 868–869 computing recursively, 766 nonrecursive computation, 835 recurrence relations and, 869

Tracing a program, 58

Subclasses abstract methods and, 523 abstracting, 526 constructors, 440 of Exception class, 482–483 inheritance and, 434–440 of RuntimeException class, 483

Subdirectories, 753 Subgraphs, 1064 Subinterfaces, 546 substring method, 157, 158, 750 Substrings, 157–158 Subtraction (-=) assignment operator, 76–77 Subtraction (-) operator, 68, 72 Subtrees

of binary trees, 976 searching for elements in BST, 978

Sudoku puzzle, 322–325, 899–900 sum method, 228, 1153, 1154 super keyword, 440 Superclass methods, 442–443 Superclasses

of abstract class can be concrete, 526 classes extending, 545 inheritance and, 434–440 subclasses related to, 523

Supplementary characters, Unicode, 147 swap method

swapping elements in an array, 283–284 in TestPassByValue.java example, 237, 238

switch statements ChineseZodiac.java example, 124–125 with enumerated types, 1199 overview of, 122–123

Syntax errors (compile errors) common errors, 35, 36 debugging, 128 programming errors, 42

Syntax rules, in Welcome.java, 36 System activities, role of OSs, 32 System analysis, in software development process, 81–83 System design, in software development process, 82, 83 System errors, 482 System.in, 59 System.out, 59, 167–171

T Tables, storing, 312 Tail recursion

ComputeFactorialTailRecusion.java, 763 overview of, 762–763

tan method, trigonometry, 142–143 TBs (terabytes), of storage, 26–27 Teamwork, facilitated by stepwise refinement, 256 Terabytes (TBs), of storage, 26–27 Terminal method, Stream interface, 1146 Testing

benefits of stepwise refinement, 256 in software development process, 82, 84–85

TestShortestPath.java, 1127–1128 Text

case study: replacing text, 506–508 files, 714 overview, 589 ShowText.java, 590–591

Z11_LIAN1878_11_GE_IDX.indd 1230 1/2/18 10:54 PM

Index 1231

Unary operators, 70 Unbounded wildcards, 784 Unboxing, 407 Unchecked exceptions, 483 Unconditional AND operator, 118 Underflow, floating point numbers, 88 Undirected graphs, 1063 Unicode

character data type (char) and, 147–150 data input and output streams, 720–721 generating random numbers and, 247 text encoding, 714 text I/O vs. binary I/O, 715–716

Unified Modeling Language. see UML (Unified Modeling Language) Uniform Resource Locators. see URLs (Uniform Resource Locators) Unique addresses, for each byte of memory, 26 Universal serial bus (USB) flash drives, 27 UNIX epoch, 74 Unweighted graphs

defined, 1064 modeling graphs and, 1064, 1071 UnweightedGraph.java example, 1074–1080

Upcasting objects, 451 Update methods, Map interface, 851 URL class, java.net, 508 URLs (Uniform Resource Locators)

ReadFileFromURL.java example, 509–510 reading data from Web, 508–510

USB (universal serial bus) flash drives, 27 UTF, 721. see also Unicode

V valueOf methods

converting strings into arrays, 413 wrapper classes and, 406

Value-returning methods return statements required by, 231 TestReturnGradeMethod.java, 233–235 void method and, 229

Values hashing functions, 1032 maps and, 1057

values method, Map interface, 852 Variable-length argument lists, 288–289 Variables

Boolean variables. see Boolean variables comparing primitive variables with reference variables, 356–358 control variables, in for loops, 195–196 declaring, 57–58, 63 declaring array variables, 270 declaring for two-dimensional arrays, 312–313 displaying/modifying, 128 hidden, 379 identifiers, 62 naming conventions, 66 overflow, 87 overview of, 62–64 reference variables, 354 scope of, 63, 246–247, 379–380 static variables, 361–366

Vector class methods, 820–821 overview of, 820 Stack class extending, 821

Vertex-weighted graphs, 1105

transient keyword, serialization and, 731 Transistors, CPUs, 25 Traveling salesperson problem (TSP), 1135 Traversing binary search trees, 979–980 Traversing graphs

breadth-first searches (BFS), 1091–1093 case study: connected circles problem, 1088–1092 depth-first searches (DFS), 1084–1088 overview of, 1084 TestWeightedGraph.java, 1113

Tree interface, BST class, 981 Tree traversal, 979–980 TreeMap class

case study: counting occurrence of words, 855–856 concrete implementation of Map class, 850–852 implementation of Map class, 1032 overview of, 853 TestMap.java example, 853–855 types of maps, 850–851

Trees AVL trees. see AVL trees binary search. see Binary search trees (BST) connected graphs, 1064 creating BFS trees, 1091 Huffman coding. see Huffman coding trees overview of, 976 Red–black trees, 1032 spanning trees. see Spanning trees traversing, 979–980

TreeSet class implementation of Set class, 1050 overview of, 842–843 TestTreeSet.java class, 843 TestTreeSetWithComparator.java class, 843–845 TestTreeSetWithComparator.java example, 530–532 types of sets, 838

Trigonometric methods, Math class, 142–143 trimToSize method, 948 True/false (Boolean) values, 98 Truth tables, 115–116 try-catch blocks

catching exceptions, 483, 485–487 chained exceptions, 495–496 exception classes cannot be generic, 788 InputMismatchExceptionDemo.java example, 480–481 QuotientWithMethod.java example, 477–478 rethrowing exceptions, 494–495 TestCircleWithException.java example, 490–492 when to use exceptions, 493–494

Twin primes, 264 Two-dimensional arrays. see Arrays, two-dimensional Type casting

between char and numeric types, 149 generic types and, 776 loss of precision, 87 for numeric type conversion, 79–80

Type erasure, erasing generic types, 786–787

U UML (Unified Modeling Language)

aggregation shown in, 398 class diagrams with, 347 diagram for Loan class, 391 diagram of StackOfIntegers, 402 diagram of static variables and methods, 361–362

Z11_LIAN1878_11_GE_IDX.indd 1231 1/2/18 10:54 PM

1232 Index

summary, 1134 TestMinimumSpanningTree.java, 1118–1120 TestShortestPath.java, 1131–1134 TestWeightedGraph.java, 1112–1114 weighted adjacency matrices, 1106 weighted edges using edge array, 1105–1106 WeightedGraph class, 1107–1108 WeightedGraph.java, 1108–1112

WeightedEdge class, 1106 WeightedGraph class

getMinimumSpanningTree method, 1119–1120 overview of, 1139 ShortestPathTree class as inner class of, 1126 TestWeightedGraph.java, 1112–1114 WeightedGraph.java, 1108–1112

Well-balanced trees AVL trees, 1012 binary search trees, 1032

while loops case study: guessing numbers, 185–188 case study: multiple subtraction quiz, 188–190 case study: predicting future tuition, 206 deciding when to use, 198–200 design strategies, 188 do-while loops. see do-while loops input and output redirections, 192 overview of, 182–183 RepeatAdditionQuiz.java example, 184–185 sentinel-controlled, 190–192 syntax of, 182

Whitespace characters, 154, 155 as delimiter in token reading methods, 505

Wildcard import, 60 Wildcards, for specifying range of generic types, 783–786 Windows. see Frames (windows) Windows OSs, 31 Wireless networking, 28 Worst-case input

heap sorts and, 922 measuring algorithm efficiency, 862, 876 quick sort and, 916

Wrapper classes automatic conversion between primitive types and wrapper

class types, 775 File class as, 499 numeric, 548 primitive types and, 404–407

Wrapping lines of text or words, 678, 680 Write-only streams, 733. see also OutputStream classes

X Xlint:unchecked error, compile time errors, 782

Vertical scroll bars, 688 Vertical sliders, 690, 691 Vertices

adjacent and incident, 1064 depth-first searches (DFS), 1084 Graph.java example, 1072 on graphs, 1064 Prim’s algorithm and, 1115 representing on graphs, 1064–1066 shortest paths. see Shortest paths TestBFS.java, 1092 TestGraph.java example, 1072 TestMinimumSpanningTree.java, 1118 TestWeightedGraph.java, 1112 vertex-weighted graphs, 1105 weighted adjacency matrices, 1106 WeightedGraph class, 1107–1108

Video, MediaDemo.java, 699–701 Virtual machines (VMs), 38. see also JVM (Java Virtual Machine) Visibility modifiers, 1185 Visibility (accessibility) modifiers

classes and, 366–368 protected, public, and private, 464–466

Visual Basic, high-level languages, 30 Visualizing (displaying) graphs

Displayable.java example, 1080 DisplayUSMap.java example, 1081–1083 GraphView.java example, 1080–1081 overview of, 1080

VLSI (very large-scale integration), 742 VMs (virtual machines), 38. see also JVM (Java Virtual Machine) void method

defined, 229 defining and invoking, 233 TestVoidMethod.java, 233

W Web, reading file data from, 508–510 Weighted graphs

case study: weighted nine tails problem, 1130–1133 defined, 1064 Dijkstra’s single-source shortest-path algorithm, 1122–1127 getMinimumSpanningTree method, 1118 key terms, 1133 minimum spanning trees, 1115 modeling graphs and, 1064 MST algorithm, 1117–1118 overview of, 1061–1062 Prim’s minimum spanning tree algorithm, 1115–1117 priority adjacency lists, 1106–1107 questions and exercises, 1134–1140 representing, 1105 shortest paths, 1121

Z11_LIAN1878_11_GE_IDX.indd 1232 1/2/18 10:54 PM

Front Cover
Title Page
Copyright Page
Preface
Contents
Chapter 1 Introduction to Computers, Programs, and Java?

1.1 Introduction
1.2 What Is a Computer?
1.3 Programming Languages
1.4 Operating Systems
1.5 Java, the World Wide Web, and Beyond
1.6 The Java Language Specification, API, JDK, JRE, and IDE
1.7 A Simple Java Program
1.8 Creating, Compiling, and Executing a Java Program
1.9 Programming Style and Documentation
1.10 Programming Errors
1.11 Developing Java Programs Using NetBeans
1.12 Developing Java Programs Using Eclipse

Chapter 2 Elementary Programming

2.1 Introduction
2.2 Writing a Simple Program
2.3 Reading Input from the Console
2.4 Identifiers
2.5 Variables
2.6 Assignment Statements and Assignment Expressions
2.7 Named Constants
2.8 Naming Conventions
2.9 Numeric Data Types and Operations
2.10 Numeric Literals
2.11 Evaluating Expressions and Operator Precedence
2.12 Case Study: Displaying the Current Time
2.13 Augmented Assignment Operators
2.14 Increment and Decrement Operators
2.15 Numeric Type Conversions
2.16 Software Development Process
2.17 Case Study: Counting Monetary Units
2.18 Common Errors and Pitfalls

Chapter 3 Selections

3.1 Introduction
3.2 boolean Data Type
3.3 if Statements
3.4 Two-Way if-else Statements
3.5 Nested if and Multi-Way if-else Statements
3.6 Common Errors and Pitfalls
3.7 Generating Random Numbers
3.8 Case Study: Computing Body Mass Index
3.9 Case Study: Computing Taxes
3.10 Logical Operators
3.11 Case Study: Determining Leap Year
3.12 Case Study: Lottery
3.13 switch Statements
3.14 Conditional Operators
3.15 Operator Precedence and Associativity
3.16 Debugging

Chapter 4 Mathematical Functions, Characters, and Strings

4.1 Introduction
4.2 Common Mathematical Functions
4.3 Character Data Type and Operations
4.4 The String Type
4.5 Case Studies
4.6 Formatting Console Output

Chapter 5 Loops

5.1 Introduction
5.2 The while Loop
5.3 Case Study: Guessing Numbers
5.4 Loop Design Strategies
5.5 Controlling a Loop with User Confirmation or a Sentinel Value
5.6 The do-while Loop
5.7 The for Loop
5.8 Which Loop to Use?
5.9 Nested Loops
5.10 Minimizing Numeric Errors
5.11 Case Studies
5.12 Keywords break and continue
5.13 Case Study: Checking Palindromes
5.14 Case Study: Displaying Prime Numbers

Chapter 6 Methods

6.1 Introduction
6.2 Defining a Method
6.3 Calling a Method
6.4 void vs. Value-Returning Methods
6.5 Passing Parameters by Values
6.6 Modularizing Code
6.7 Case Study: Converting Hexadecimals to Decimals
6.8 Overloading Methods
6.9 The Scope of Variables
6.10 Case Study: Generating Random Characters
6.11 Method Abstraction and Stepwise Refinement

Chapter 7 Single-Dimensional Arrays

7.1 Introduction
7.2 Array Basics
7.3 Case Study: Analyzing Numbers
7.4 Case Study: Deck of Cards
7.5 Copying Arrays
7.6 Passing Arrays to Methods
7.7 Returning an Array from a Method
7.8 Case Study: Counting the Occurrences of Each Letter
7.9 Variable-Length Argument Lists
7.10 Searching Arrays
7.11 Sorting Arrays
7.12 The Arrays Class
7.13 Command-Line Arguments

Chapter 8 Multidimensional Arrays

8.1 Introduction
8.2 Two-Dimensional Array Basics
8.3 Processing Two-Dimensional Arrays
8.4 Passing Two-Dimensional Arrays to Methods
8.5 Case Study: Grading a Multiple-Choice Test
8.6 Case Study: Finding the Closest Pair
8.7 Case Study: Sudoku
8.8 Multidimensional Arrays

Chapter 9 Objects and Classes

9.1 Introduction
9.2 Defining Classes for Objects
9.3 Example: Defining Classes and Creating Objects
9.4 Constructing Objects Using Constructors
9.5 Accessing Objects via Reference Variables
9.6 Using Classes from the Java Library
9.7 Static Variables, Constants, and Methods
9.8 Visibility Modifiers
9.9 Data Field Encapsulation
9.10 Passing Objects to Methods
9.11 Array of Objects
9.12 Immutable Objects and Classes
9.13 The Scope of Variables
9.14 The this Reference

Chapter 10 Object-Oriented Thinking

10.1 Introduction
10.2 Class Abstraction and Encapsulation
10.3 Thinking in Objects
10.4 Class Relationships
10.5 Case Study: Designing the Course Class
10.6 Case Study: Designing a Class for Stacks
10.7 Processing Primitive Data Type Values as Objects
10.8 Automatic Conversion between Primitive Types and Wrapper Class Types
10.9 The BigInteger and BigDecimal Classes
10.10 The String Class
10.11 The StringBuilder and StringBuffer Classes

Chapter 11 Inheritance and Polymorphism

11.1 Introduction
11.2 Superclasses and Subclasses
11.3 Using the super Keyword
11.4 Overriding Methods
11.5 Overriding vs. Overloading
11.6 The Object Class and Its toString() Method
11.7 Polymorphism
11.8 Dynamic Binding
11.9 Casting Objects and the instanceof Operator
11.10 The Object?s equals Method
11.11 The ArrayList Class
11.12 Useful Methods for Lists
11.13 Case Study: A Custom Stack Class
11.14 The protected Data and Methods
11.15 Preventing Extending and Overriding

Chapter 12 Exception Handling and Text I/O

12.1 Introduction
12.2 Exception-Handling Overview
12.3 Exception Types
12.4 More on Exception Handling
12.5 The finally Clause
12.6 When to Use Exceptions
12.7 Rethrowing Exceptions
12.8 Chained Exceptions
12.9 Defining Custom Exception Classes
12.10 The File Class
12.11 File Input and Output
12.12 Reading Data from the Web
12.13 Case Study: Web Crawler

Chapter 13 Abstract Classes and Interfaces

13.1 Introduction
13.2 Abstract Classes
13.3 Case Study: the Abstract Number Class
13.4 Case Study: Calendar and GregorianCalendar
13.5 Interfaces
13.6 The Comparable Interface
13.7 The Cloneable Interface
13.8 Interfaces vs. Abstract Classes
13.9 Case Study: The Rational Class
13.10 Class-Design Guidelines

Chapter 14 JavaFX Basics

14.1 Introduction
14.2 JavaFX vs Swing and AWT
14.3 The Basic Structure of a JavaFX Program
14.4 Panes, Groups, UI Controls, and Shapes
14.5 Property Binding
14.6 Common Properties and Methods for Nodes
14.7 The Color Class
14.8 The Font Class
14.9 The Image and ImageView Classes
14.10 Layout Panes and Groups
14.11 Shapes
14.12 Case Study: The ClockPane Class

Chapter 15 Event-Driven Programming and Animations

15.1 Introduction
15.2 Events and Event Sources
15.3 Registering Handlers and Handling Events
15.4 Inner Classes
15.5 Anonymous Inner Class Handlers
15.6 Simplifying Event Handling Using Lambda Expressions
15.7 Case Study: Loan Calculator
15.8 Mouse Events
15.9 Key Events
15.10 Listeners for Observable Objects
15.11 Animation
15.12 Case Study: Bouncing Ball
15.13 Case Study: US Map

Chapter 16 JavaFX UI Controls and Multimedia

16.1 Introduction
16.2 Labeled and Label
16.3 Button
16.4 CheckBox
16.5 RadioButton
16.6 TextField
16.7 TextArea
16.8 ComboBox
16.9 ListView
16.10 ScrollBar
16.11 Slider
16.12 Case Study: Developing a Tic-Tac-Toe Game
16.13 Video and Audio
16.14 Case Study: National Flags and Anthems

Chapter 17 Binary I/O

17.1 Introduction
17.2 How Is Text I/O Handled in Java?
17.3 Text I/O vs. Binary I/O
17.4 Binary I/O Classes
17.5 Case Study: Copying Files
17.6 Object I/O
17.7 Random-Access Files

Chapter 18 Recursion

18.1 Introduction
18.2 Case Study: Computing Factorials
18.3 Case Study: Computing Fibonacci Numbers
18.4 Problem Solving Using Recursion
18.5 Recursive Helper Methods
18.6 Case Study: Finding the Directory Size
18.7 Case Study: Tower of Hanoi
18.8 Case Study: Fractals
18.9 Recursion vs. Iteration
18.10 Tail Recursion

Chapter 19 Generics

19.1 Introduction
19.2 Motivations and Benefits
19.3 Defining Generic Classes and Interfaces
19.4 Generic Methods
19.5 Case Study: Sorting an Array of Objects
19.6 Raw Types and Backward Compatibility
19.7 Wildcard Generic Types
19.8 Erasure and Restrictions on Generics
19.9 Case Study: Generic Matrix Class

Chapter 20 Lists, Stacks, Queues, and Priority Queues

20.1 Introduction
20.2 Collections
20.3 Iterators
20.4 Using the forEach Method
20.5 Lists
20.6 The Comparator Interface
20.7 Static Methods for Lists and Collections
20.8 Case Study: Bouncing Balls
20.9 Vector and Stack Classes
20.10 Queues and Priority Queues
20.11 Case Study: Evaluating Expressions

Chapter 21 Sets and Maps

21.1 Introduction
21.2 Sets
21.3 Comparing the Performance of Sets and Lists
21.4 Case Study: Counting Keywords
21.5 Maps
21.6 Case Study: Occurrences of Words
21.7 Singleton and Unmodifiable Collections and Maps

Chapter 22 Developing Efficient Algorithms

22.1 Introduction
22.2 Measuring Algorithm Efficiency Using Big O Notation
22.3 Examples: Determining Big O
22.4 Analyzing Algorithm Time Complexity
22.5 Finding Fibonacci Numbers Using Dynamic Programming
22.6 Finding Greatest Common Divisors Using Euclid?s Algorithm
22.7 Efficient Algorithms for Finding Prime Numbers
22.8 Finding the Closest Pair of Points Using Divide-and-Conquer
22.9 Solving the Eight Queens Problem Using Backtracking
22.10 Computational Geometry: Finding a Convex Hull

Chapter 23 Sorting

23.1 Introduction
23.2 Insertion Sort
23.3 Bubble Sort
23.4 Merge Sort
23.5 Quick Sort
23.6 Heap Sort
23.7 Bucket and Radix Sorts
23.8 External Sort

Chapter 24 Implementing Lists, Stacks, Queues, and Priority Queues

24.1 Introduction
24.2 Common Operations for Lists
24.3 Array Lists
24.4 Linked Lists
24.5 Stacks and Queues
24.6 Priority Queues

Chapter 25 Binary Search Trees

25.1 Introduction
25.2 Binary Search Trees
25.3 Deleting Elements from a BST
25.4 Tree Visualization and MVC
25.5 Iterators
25.6 Case Study: Data Compression

Chapter 26 AVL Trees

26.1 Introduction
26.2 Rebalancing Trees
26.3 Designing Classes for AVL Trees
26.4 Overriding the insert Method
26.5 Implementing Rotations
26.6 Implementing the delete Method
26.7 The AVLTree Class
26.8 Testing the AVLTree Class
26.9 AVL Tree Time Complexity Analysis

Chapter 27 Hashing

27.1 Introduction
27.2 What Is Hashing?
27.3 Hash Functions and Hash Codes
27.4 Handling Collisions Using Open Addressing
27.5 Handling Collisions Using Separate Chaining
27.6 Load Factor and Rehashing
27.7 Implementing a Map Using Hashing
27.8 Implementing Set Using Hashing

Chapter 28 Graphs and Applications

28.1 Introduction
28.2 Basic Graph Terminologies
28.3 Representing Graphs
28.4 Modeling Graphs
28.5 Graph Visualization
28.6 Graph Traversals
28.7 Depth-First Search (DFS)
28.8 Case Study: The Connected Circles Problem
28.9 Breadth-First Search (BFS)
28.10 Case Study: The Nine Tails Problem

Chapter 29 Weighted Graphs and Applications

29.1 Introduction
29.2 Representing Weighted Graphs
29.3 The WeightedGraph Class
29.4 Minimum Spanning Trees
29.5 Finding Shortest Paths
29.6 Case Study: The Weighted Nine Tails Problem

Chapter 30 Aggregate Operations for Collection Streams

30.1 Introduction
30.2 Stream Pipelines
30.3 IntStream, LongStream, and DoubleStream
30.4 Parallel Streams
30.5 Stream Reduction Using the reduce Method
30.6 Stream Reduction Using the collect Method
30.7 Grouping Elements Using the groupingby Collector
30.8 Case Studies

Online Chapters

Chapter 31 Advanced JavaFX and FXML
Chapter 32 Multithreading and Parallel Programming
Chapter 33 Networking
Chapter 34 Java Database Programming
Chapter 35 Advanced Java Database Programming
Chapter 36 Internationalization
Chapter 37 Servlets
Chapter 38 Javaserver Pages
Chapter 39 Javaserver Faces
Chapter 40 Remote Method Invocation
Chapter 41 Web Services
Chapter 42 2-4 Trees and B-Trees
Chapter 43 Red-Black Trees
Chapter 44 Testing Using Junit

Appendixes

Appendix A Java Keywords
Appendix B The ASCII Character Set
Appendix C Operator Precedence Chart
Appendix D Java Modifiers
Appendix E Special Floating-Point Values
Appendix F Number System
Appendix G Bitwise Operations
Appendix H Regular Expressions
Appendix I Enumerated Types

Quick Reference
Index
Back Cover