Walter, kalev, tobler, snaith, kossoroukov, roberts c++ how to

Đây là quyển sách tiếng anh về lĩnh vực công nghệ thông tin cho sinh viên và những ai có đam mê. Quyển sách này trình về lý thuyết ,phương pháp lập trình cho ngôn ngữ C và C++.

<iterator> Def ines a number of classes, class t emplat es, and f unct ion t emplat es used f or t he declarat ion and

manipulat ion of it erat ors—it erat ors and it erat or support

LANGUAGE SUPPORT

<cfloat> Includes t he St andard C header <float.h> w it hin t he std namespace, providing C-st yle f loat ing-point limit

macros.

<ciso646> Includes t he St andard C header <iso646.h> w it hin t he std namespace.

<climits> Includes t he St andard C header <limits.h> w it hin t he std namespace, providing C-st yle numeric scalar-limit

macros.

<csetjmp> Includes t he St andard C header <setjmp.h> w it hin t he std namespace, providing C-st yle st ack unw inding.

<csignal> Includes t he St andard C header <signal.h> w it hin t he std namespace, providing C-st yle signal handling.

<cstdarg> Includes t he St andard C header <stdarg.h> w it hin t he std namespace, providing variable-lengt h f unct ion

argument list s.

<cstddef> Includes t he St andard C header <stddef.h> w it hin t he std namespace, providing C library language support

<exception> Def ines several t ypes and f unct ions relat ed t o except ion handling.

<limits> Def ines t he class t emplat e numeric_limits , w hose specializat ions describe arit hmet ic propert ies of scalar

dat a t ypes.

LOCALIZATION

<clocale> Includes t he St andard C header <locale.h> w it hin t he std namespace, represent ing cult ural diff erences

C-st yle.

<locale> Def ines many class t emplat es and f unct ion t emplat es t hat encapsulat e and manipulat e locales, w hich repre

sent cult ural diff erences.

NUM ERICS

<cmath> Includes t he St andard C header <math.h> w it hin t he std namespace, providing st andard mat h f unct ions.

<complex> Provides a class t emplat e describing an object t hat st ores bot h t he real part and t he imaginary part of a

complex number.

<cstdlib> Includes t he St andard C header <stdlib.h> w it hin t he std namespace, providing pseudo-random numbers.

<numeric> Def ines several f unct ion t emplat es usef ul f or comput ing numeric values—generalized numeric operat ions.

<valarray> Def ines t he class t emplat e valarray along w it h many support ing t emplat e classes and f unct ions, providing

numeric vect ors and operat ions.

STRINGS

<cctype> Includes t he St andard C header <ctype.h> w it hin t he std namespace—charact er classsif icat ion.

<cstdlib> Includes t he St andard C header <stdlib.h> w it hin t he std namespace, providing C-st yle st ring and charact er

<cwctype> Includes t he St andard C header <wctype.h> w it hin t he std namespace, providing C-st yle w ide charact er

st ring f unct ions.

Def ines t he cont ainer t emplat e class and various support ing t emplat es—a st ring of T.

How -To

C++

Jan Walter Danny Kalev

M ichael J Tobler Paul Snaith Andrei Kossoroukov

Scott Roberts

A Division of Macmillan Computer Publishing

201 West 103rd St., Indianapolis, Indiana, 46290 USA

Copyright 1999 by Sam s Publishing

All rights reserved No part of this book shall be reproduced,

stored in a retrieval system, or transmitted by any means,

elec-tronic, mechanical, photocopying, recording, or otherwise,

without written permission from the publisher No patent liability

is assumed with respect to the use of the information contained

herein Although every precaution has been taken in the

prepara-tion of this book, the publisher and author assume no

responsi-bility for errors or omissions Neither is any liaresponsi-bility assumed for

damages resulting from the use of the information contained

herein.

International Standard Book Number: 1-57169-159-6

Library of Congress Catalog Card Number: 98-86976

Printed in the United States of America

First Printing: January, 1999

02 01 00 99 4 3 2 1

Tradem arks

All terms mentioned in this book that are known to be

trade-marks or service trade-marks have been appropriately capitalized Sams

Publishing cannot attest to the accuracy of this information Use

of a term in this book should not be regarded as affecting the

validity of any trademark or service mark.

Warning and Disclaim er

Every effort has been made to make this book as complete and as

accurate as possible, but no warranty or fitness is implied The

information provided is on an “as is” basis The authors and the

publisher shall have neither liability or responsibility to any

person or entity with respect to any loss or damages arising from

the information contained in this book or from the use of the CD

or programs accompanying it.

S OFTWARE D EVELOPM ENT

C O N T E N T S A T A G L A N C E

PRACTICE 51

CHAPTER 4: STRUCTURES VERSUS CLASSES 105

CHAPTER 6: TEM PLATE CLASSES 217

CHAPTER 10: C-STYLE ERROR HANDLING 371

CHAPTER 11: EXCEPTION HANDLING IN C++ 407

INTRODUCTION 1

PA RT I

La n g u a g e To p i cs

C H A P T E R 1

A QUICK INTRODUCTION TO THE LANGUAGE 9

1.1 Write the simplest C++ program without knowing the language 12

1.2 Create a program to perform calculations 17

1.3 Use the various loop statements that are available in C++ 21

1.4 Create a program that uses one or more functions 25

1.5 Use the derived data types such as arrays and pointers 29

1.6 Create data when the program is running 37

1.7 Create a program to perform error handling, specifically exception handling 41

C H A P T E R 2 OBJECT ORIENTATION—THEORY AND PRACTICE 53

2.1 Understand the object-oriented paradigm 55

2.2 Learn the concept of inheritance so that I can apply it programmatically 57

2.3 Learn the concept of encapsulation 62

2.4 Learn the concept of polymorphism 64

C H A P T E R 3 OBJECT ORIENTATION—C++ SPECIFICS 71

3.1 Create a simple class in C++ 72

3.2 Implement the use of inheritance in C++ 80

3.3 Apply the use of encapsulation in a C++ program 87

3.4 Implement polymorphism in C++ 91

3.5 Implement static members of a class 97

PA RT I I D a t a St r u ct u re s C H A P T E R 4 STRUCTURES VERSUS CLASSES 107

4.1 Create my own data type 110

4.2 Hide my data from external programs 118

4.3 Use encapsulation? What steps are required to encapsulate data? 129

4.4 Create my own operators 134

4.5 Overload relational and equality operators 139

4.6 Provide access to encapsulated data to certain classes 144

4.7 Maintain global data in my program 148

4.8 Know when I should use structures and when I should use classes 151

C H A P T E R 5 COM M ON M ISTAKES M ADE W ITH CLASSES 159

5.1 Know when to take an object-oriented approach or a procedural approach 162

5.2 Use and access a class’s data members 169

5.3 Use the scope resolution operator 175

5.4 Use dot notation to access the member functions of an object 181

5.5 Know which constructor to use when there are several to choose from 186

5.6 Implement function overloading 191

5.7 Correctly make use of inheritance 194

5.8 Pass parameters back through the C++ inheritance mechanism to parent classes 200

5.9 Distinguish between virtual classes and nonvirtual classes? How do I know when to use virtual classes and what does the word virtual mean in C++? 206

C H A P T E R 6 TEM PLATE CLASSES 219

6.1 Create a template class to represent any simple data type and understand how to use the template in a working C++ program 221

6.2 Create a template class to represent any simple data type and extend it to read in data to a variable of any data type 225

6.3 Create a template class to represent a compound data type and understand how to use the template in a working C++ program 228

6.4 Write a template class that has two undefined data types that can be resolved at a later time 233

6.5 Use a template class to handle a structure 238

C H A P T E R 7 THE STANDARD TEM PLATE LIBRARY’S CONTAINER CLASSES 247

7.1 Create a container object that automatically grows or shrinks as needed 250

7.2 Read a single element of a container 254

7.3 Modify a single element of a container 257

7.4 Use a generic LIFO data model 261

7.5 Prevent automatic reallocation of a container 268

7.6 Traverse through a container’s elements 273

7.7 Implement a queue data model 278

PA RT I I I

A l g o r i t h m s

C H A P T E R 8

THE STANDARD C LIBRARY’S INCLUDED ALGORITHM S 287

8.1 Sort an array 289

8.2 Find an element in an array 298

8.3 Locate an element in a nonsorted array 302

8.4 Choose between _lfindand _lsearch 305

8.5 Generate a sequence of random numbers 309

C H A P T E R 9 THE STANDARD TEM PLATE LIBRARY’S INCLUDED ALGORITHM S 315

9.1 Create classes for sequential containers 320

9.2 Use predicates with sequence operations 328

9.3 Repeat an action with all elements in a container range 333

9.4 Compare two sequences 337

9.5 Search for a sequence of values in a container 341

9.6 Accumulate all container elements and create a sequence of accumulated sums 345

9.7 Sort elements in a container using different sorting indexes 351

9.8 Change the order of the container elements 362

PA RT I V Er ro r Ha n d l i n g C H A P T E R 1 0 C-STYLE ERROR HANDLING 373

10.1 Handle runtime errors in my programs 376

10.2 Use the standard C library functions perrorand strerrorand the predefined C macros to report runtime errors in my programs 381

10.3 Use assertto catch errors in my code when running in debug mode 385

10.4 Use raiseand signalto indicate errors in my programs 387

10.5 Use abortto terminate my application if a serious error occurs 391

10.6 Use exitand atexittogether to perform some action when my program terminates normally 393

10.7 Detect errors that occur when reading from or writing to a file using the file functions provided with the Standard C Library 395

10.8 Use setjmpand longjmpto maintain state when handling errors 397

10.9 Use a C++ class to handle runtime errors in a more maintainable fashion 400

CO N T EN T S

ix

C H A P T E R 1 1

EXCEPTION HANDLING IN C++ 409

11.1 Utilize the exception handling mechanism to process potential error conditions 411

11.2 Use the various catchconstructs such as multiple catchclauses, catchordering, and the rethrowing of exceptions 419

11.3 Implement and use an exception class 425

11.4 Specify exceptions that a function will throw 430

11.5 Handle exceptions that are not caught or not expected 434

PA RT V M e m o r y M a n a g e m e n t C H A P T E R 1 2 NEW AND DELETE VERSUS MALLOC() AND FREE() 443

12.1 Use newand deletewith the C malloc()and free()routines 445

12.2 Use other C mem routines on objects allocated with new 446

12.3 Find out how much memory my structures and classes really take 447

12.4 Prevent memory leaks caused by not using delete[]to delete arrays 450

12.5 Override the newor deleteoperators for my classes 451

12.6 Overload the newand deletefunctions for arrays 455

C H A P T E R 1 3 M EM ORY M ANAGEM ENT TECHNIQUES USING CLASSES 473

13.1 Make a simple class to clean up dynamically allocated memory automatically 475

13.2 Make a class that automatically cleans up objects allocated with new 479

13.3 Make an object that deallocates itself when there is no more code referencing it 483

PA RT V I I /O C H A P T E R 1 4 UNDERSTANDING THE I/ O STREAM S LIBRARY 499

14.1 Use the C Standard I/O Library with the C++ I/O streams library 500

14.2 Make my own classes compatible with cinand cout 503

14.3 Perform complex formatting with coutor another ostreamobject 504

14.4 Make my own stream manipulators 515

C H A P T E R 1 5

FILE I/ O 525

15.1 Open a file stream 526

15.2 Continually read data until the end of file 529

15.3 Handle stream errors 532

15.4 Read and write binary files 536

15.5 Read from or write to different positions in a file 542

PA RT V I I A p p e n d i x e s A P P E N D I X A NAM ESPACES 549

The Rationale Behind Namespaces 550

A Brief Historical Background 550

Large-Scale Projects Are Susceptible to Name Clashes 550

Properties of Namespaces 551

A Fully Qualified Name 551

A using-Declaration and a using-Directive 551

Namespaces Are Open 552

Namespace Aliases 553

Koenig Lookup 554

Namespaces Do Not Incur Additional Overhead 554

The Interaction of Namespaces with Other Language Features 554

::Operator Should Not Be Used to Designate a Global Function 555

Turning an External Function into a File-Local Function 555

Standard Headers Names 556

Restrictions on Namespaces 557

Namespace stdMay Not Be Modified 557

User-Defined newand deleteCannot Be Declared in a Namespace 557

Comments 558

A P P E N D I X B RUNTIM E TYPE INFORM ATION 559

Static Binding 559

Dynamic Binding 560

Historical Background 563

RTTI Constituents 563

Comments 568

INDEX 569

A B O U T T H E A U T H O R S

Jan Walter is a freelance computer consultant living around Vancouver, B.C.,

Canada He spends most of his time doing network design and implementation,

but “would rather be coding.” Jan has fulfilled programming contracts with

several large companies in the area, mostly using C++ Code optimization and

performance tuning are a distinct area of interest, as is compiler technology The

latter came about after spending a lot of time hunting down bugs introduced by

parser problems in Borland C++ in one of his projects

This is his second project with Macmillan Computer Publishing The first was

as contributing author to Que’s Using Linux, by Bill Ball, 1998.

Jan Walter can be reached at jwalter@rogers.wave.ca

Danny Kalev is a certified system analyst and software engineer with 10 years

of experience, specializing in C++ and object-oriented analysis and design He

is now finishing his M.A degree in applied linguistics at the University of Tel

Aviv, Israel His research focuses on speech recognition algorithms He is a hi-fi

enthusiast and likes all kinds of music He is also interested in natural

languages and philology

His technical interests involve generic programming, networking, compiler

technology, artificial intelligence, and embedded systems He has contributed

several articles to C++ magazines and Web publishers He is also a member of

the ANSI C++ standardization committee He can be reached at

Michael J Tobler is a senior technical specialist with BSI Consulting in

Houston, Texas He has more than 16 years experience working on software

development projects, specializing in architecting, designing, and developing

multitier systems using C++ and Java He is currently the president of the

Houston Java Users Group Michael is an advocate and practitioner of the

Unified Modeling process from Rational and a proponent of patterns and

pattern languages He is a contributing author for C++ Unleashed He has

discovered that skydiving is a very addicting sport Michael can be reached at

mtobler@ibm.net

Paul Snaith works in the I.T industry in the UK He is currently a project

leader in software development for a well-known British airline dealing mainly

with Web development and electronic commerce Previously, he spent several

years lecturing in modern computer languages such as C++, Java, VB, and

Delphi In the far distant past (only a few years ago but it seems much longer),

Paul was involved in the rock music industry as a musician and played with the

wonders of electronic computer generation of his work Paul has just finished

another book called The Complete Idiot’s Guide to C++ and is planning other such

in 1985 Working with computers for the last 25 years, he has developed cial intelligence systems, financial software, and distributed messaging systems.

artifi-He is an author of more than 30 scientific and technical articles, papers, andbooks Currently he works for I.T Systems, Inc.(Vancouver, B.C., Canada) as asenior software consultant and specializes in the development of Internet-basedsystems with distributed data

Scott Roberts works as a developer support engineer on the Internet Client

Development team at Microsoft He assists developers who are using Visual C++and COM to create Internet applications Scott has been a developer for 9 yearsand has worked at a number of companies on a wide range of products, fromback-end communications systems to end-user software applications Scott hasbeen a frequent contributor to many technical publications, including MicrosoftInteractive Developer and Microsoft Systems Journal In addition, he hasspoken at many technical conferences in the United States and Europe Scottcan be reached at scottrobe@hotmail.com

xii

D E D I C A T I O N

I dedicate this book to my grandparents In their time, people went through so much, and many gave their

lives to let us have what we have now It is easy for us today to look back 60 years and say, “Times were tough

then,” but we have little comprehension of what actually went on

“Lest we forget” is an apt phrase to take to heart.

—Jan Walter

A C K N O W L E D G M E N T S

I do also need to thank the staff at Macmillan Computer Publishing for their

seemingly infinite patience Many thanks for encouragement go to Michelle

Newcomb, and thanks, Bryan, for your insightful feedback

—Jan Walter

As the reader of this book, you are our most important critic and commentator.

We value your opinion and want to know what we’re doing right, what wecould do better, what areas you’d like to see us publish in, and any other words

of wisdom you’re willing to pass our way

As the executive editor for the Advanced Programming and Distributed

Architectures team at Macmillan Computer Publishing, I welcome your

comments You can fax, email, or write me directly to let me know what youdid or didn’t like about this book—as well as what we can do to make ourbooks stronger

Please note that I cannot help you with technical problems related to the topic of this book, and that due to the high volume of mail I receive, I might not be able to reply to every message.

When you write, please be sure to include this book’s title and author as well asyour name and phone or fax number I will carefully review your commentsand share them with the author and editors who worked on the book

Email: programming@mcp.com

Mail: Tracy Dunkelberger

Executive EditorAdvanced Programming and Distributed Architectures TeamMacmillan Computer Publishing

201 West 103rd StreetIndianapolis, IN 46290 USA

I N T R O D U C T I O N

Despite the development of several new languages over the past five years, C++has held its own in the development world Originally developed at AT&T BellLaboratories by Bjarne Stoustrup, the language has evolved to encompass evermore object-oriented concepts

With the latest ANSI revision of C++, version 3, the language gets some tuning in memory management, solid exception support, and new typecastingoperators The biggest improvement, however, was the introduction of the

fine-Standard Template Library (STL), which provides a standard set of containerclasses and the means to work with them Before this, most compiler vendorsincluded their own container classes and other objects in their libraries, posing

an impediment to porting code from one platform to another where the samevendor’s compiler was not supported

What makes C++ so useful? Some people say that its C language nings are great for getting as close to the hardware as possible without usingassembler code Developers find that the strong typing by the C++ languagehelps reduce the number of bugs in programs Some people use C++ as “just abetter C,” but the point is missed then: The object-oriented programming(OOP) style makes it easier to translate from the problem language and problemworld to the solution language

underpin-The great misconception of many people is that object-oriented ming produces slower code This is not necessarily so If a problem can be put

program-so that a human mind can understand it better, it is likely that this mind canproduce a better solution Algorithms, and their efficiency, have a much greaterimpact on program performance than the language does C++ is only marginallyslower than C code overall, and this makes it one of the fastest executinglanguages for object-oriented programming

This book is intended to be an eye-opener for C++ programmers I admitfreely that I learn the most from reading other peoples’ code, as I learn theirapproaches to problems It’s easy to get stuck in one way of doing things, andthis is dangerous with any creative exercise Computer programming is noexception What’s written in this book is not gospel, but yet another approachthat can be added to your arsenal

The Waite Group’s C++ How-Tois divided into 15 chapters, each of which

covers a specific feature of the C++ programming language:

• Chapter 1, “A Quick Introduction to the Language”—Before diving into the

“deep end” of C++ programming, this chapter provides a gentle introductioninto the basic concepts of the language Loops, calculations, and error handlingare all covered here for the beginning C++ programmer

• Chapter 2, “Object Orientation—Theory and Practice”—To fully make use ofthe C++ language, object-oriented programming must be understood Many C,Pascal, and COBOL programmers are assigned to C++ projects and never takethe time—or get the opportunity—to fully learn the benefits of OOP Withoutfocusing on the specific syntax used in C++, this chapter covers the basicobject-oriented concepts required to become a great C++ programmer

• Chapter 3, “Object Orientation—C++ Specifics”—After learning or reviewingthe basic concepts of object-oriented programming in Chapter 2, you will learnhow these concepts are specifically applied to C++ in this chapter The basictenets of OOP—inheritance, polymorphism, and encapsulation—are all coveredhere as well as the fundamental C++ unit: the class

• Chapter 4, “Structures Versus Classes”—C++ provides two primary devices thatcan be used to contain data: structures and classes Although the two datastructures can be identical, classes are much more powerful because they fullysupport the primary OOP constructs This chapter compares and contraststhese two data types and provide numerous examples of their use

• Chapter 5, “Common Mistakes Made with Classes”—Because of the tremendousdifference between traditional structured programming and object-orientedprogramming, many beginning-to-intermediate–level programmers makeseveral common mistakes This chapter introduces these mistakes, explains whythey are mistakes, and offers solutions to them

• Chapter 6, “Template Classes”—The template class is a mechanism that enablesyou to write a single solution to a problem that can satisfy all data types Theactual data type required can be specified later, and the template class can then

be used for a wide range of data types, all using the same C++ template Thischapter introduces the template class, or templates as they are commonlyknown

• Chapter 7, “The Standard Template Library’s Container Classes”—In thischapter, you will explore the following containers of the Standard TemplateLibrary: vector, string, stack, list, and queue I also discuss iterators and theirrole in the STL framework Finally, you will survey some other containers aswell as “almost container” classes of the Standard Library

2

• Chapter 8, “The Standard C Library’s Included Algorithms”—In this chapter, I

survey the algorithms of the Standard C Library These algorithms enable you tosort an array and find an element in it In addition, I discuss the random

number generation functions of the Standard C Library

• Chapter 9, “The Standard Template Library’s Included Algorithms”—STL

algo-rithms are represented by template functions and provide copying, searching,

sorting, and merging functions, as well as other operations on data Algorithmsare not member functions; they are separate from the container classes You willexamine a number of these algorithms in this chapter I give examples, where

possible, showing their use in typical situations

• Chapter 10, “C-Style Error Handling”—This chapter covers different

error-handling methods that work in C and C++ programs and some methods that

don’t work well in C++ programs Although this is a C++ book, I discuss C-typeerror-handling techniques so that you will know what you are dealing with if

you run into them in older C and C++ programs

• Chapter 11, “Exception Handling in C++”—The exception-handling mechanismfurnished by the standard provides a common and standard interface for

handling program anomalies Without exception handling, error handling is

performed using return values from functions and global status variables Eachdeveloper has his or her own style of handling errors, leading to inconsistency

among applications and library packages In this chapter, you will see various

ways to apply the exception-handling mechanisms

• Chapter 12, “newand deleteVersus malloc()and free()”—This chapter

compares and contrasts the C++ newand deleteoperators and the C functions

malloc()and free(), and it covers their place in C++ memory management

You must understand memory management using C and C++ techniques

because of the existence of legacy code and because many existing libraries usethe older C language techniques

• Chapter 13, “Memory Management Techniques Using Classes”—Whereas

C-style memory management basically requires the programmer to free

variables that have been created, C++-style memory management is much morepowerful, although a bit more challenging This chapter focuses on many

advanced memory management techniques that you can use to improve the

reliability and performance of your applications

I N T RO D U CT I O N

3

• Chapter 14, “Understanding the I/O Streams Library”—So many programmersseem to have difficulty with the C++ I/O Streams library that it is quite common

to see C++ code still using the old C stdiofunctions to handle program I/O.Although some programmers might consider this approach practical, or evensuperior, the C stdiolibrary does not do much to help good programmingpractice or catch oversights made by the programmer This chapter introducesthe basic concepts required to use the Streams library and shows why it is apreferred solution to the standard C method

• Chapter 15, “File I/O”—This chapter covers IOStreamswith a focus on filestreams Examples are provided that demonstrate how to seek through files,read/write binary files, and manage file I/O errors

This book also includes two appendixes, one on namespaces and another onruntime type information (RTTI)

WHO IS THIS BOOK INTENDED FOR?

The Waite Group’s C++ How-Tois designed for C++ programmers of all skill

levels, from beginning to advanced However, I anticipate that readers who pick

up this book already have a solid understanding of C++ programming and arelooking for “just the answers” to specific programming questions This is also agreat supplemental resource for new C++ programmers searching for informa-

tion and instruction that goes beyond a typical tutorial

4

P A R T I

LANGUAGE TOPICS

C H A P T E R 1

A QUICK INTRODUCTION TO THE LANGUAGE

1.2 Creat e a program t o perf orm calculat ions?

1.3 Use t he various loop st at em ent s t hat are available

in C++?

1.4 Creat e a program t hat uses one or m ore f unct ions? 1.5 Use t he derived dat a t ypes such as arrays and

point ers?

1.6 Creat e dat a w hen t he program is running?

1.7 Creat e a program t o perf orm error handling,

specif ically except ion handling?

The C++ programming language is one of the most significant languages toemerge in the past twenty years The C++ language is also the most popularobject-oriented language in use today I will examine the object-orientedfeatures of C++ beginning with Chapter 3, “Object-Orientation—C++ Specifics.”Everything from word processors to spreadsheets, graphics applications, andoperating systems is written in the C++ language, and a majority of all custom-written solutions are implemented in C++ as well Despite this, manyprogrammers have yet to make the switch to C++.

This chapter introduces you to the C++ programming language Theintended audience for this chapter is programmers currently using otherlanguages, such as C and Pascal, among others If you have experience in the Cprogramming language, much of what you see in the chapter will be familiar,with a few exceptions, of course This chapter will introduce the basics of theC++ programming language It is not designed to be an exhaustive tutorial orreference

Chapter 2, “Object Orientation—Theory and Practice,” introduces the threemajor concepts of object-oriented programming It also introduces some of thecommon terms used Chapter 3 addresses object-oriented programming usingC++ You will not find any How-Tos in this chapter addressing object-orientedprogramming

1.1 Writ e t he Sim plest C++ Program Wit hout Know ing

t he Language

The quickest way to learn to program in an unfamiliar language is to write asimple program and then analyze it In this How-To, you will write a verysimple program This program is really a template for a C++ program Theprogram in this How-To demonstrates the minimum a C++ program mustconsist of You are not required to know C or C++ to create this program Youwill, although, need a C++ compiler and linker and a text editor

1.2 Creat e a Program t o Perf orm Calculat ions

Most useful programs perform a repetitive task And almost every programperforms calculations of one type or another In this How-To, a program ispresented that will perform some simple calculations The program will introduce the use of program variables to hold data It also demonstrates some

of the popular mathematical operators Output operations (displaying text) ispresented, as well as acquiring input from the user

1.3 Use t he Various Loop St at em ent s That Are

Available in C++

In this How-To, you will write a program that uses the three C++ looping

constructs The three loops discussed are for, do-while, and while Input andoutput functionality is also demonstrated Looping statements are very

important for repetitive operations, such as accessing elements of an array, or

for visiting some finite or unknown number of objects, or for obtaining

continual input from a user

1.4 Creat e a Program That Uses One or M ore Funct ions

Most programming languages support the use of functions and/or procedures

In this How-To, you will be introduced to functions, including the declarationand definition of functions Argument passing is also shown; additionally,

returning values from functions is presented You will create a program that

uses the facilities of a function to perform a calculation

1.5 Use t he Derived Dat a Types Such As Arrays and

Point ers

Arrays and pointers are considered by many developers as difficult types to

work with This How-To should dispel that myth A program is presented in

this How-To to demonstrate the more common uses of these derived types Youwill create a function to perform a calculation and one to display a message andgather input from the user The C++ structis also introduced

1.6 Creat e Dat a When t he Program Is Running

The creation of data at runtime is introduced in this How-To You will learn

how to dynamically allocate memory for various data types Dynamically

creating memory at runtime is an important feature of C++; it allows you to

create only the objects that are required by your application The C++ operators

newand deleteare introduced

1.7 Creat e a Program t o Perf orm Error Handling,

Specif ically Except ion Handling

Error handling is important to the success of a program Invalid input, accesses

to hardware that is not available, and other program anomalies must be testedfor and handled gracefully This How-To presents the C++ exception handlingmechanism The C++ Standard defines a standardized method for handling

exceptions at runtime

CH A PT ER 1

A Q U I C K I N T R O D U C T I O N T O T H E L A N G U A G E 11

C O M PL EX I T Y

B E G I N N I N G1.1 How do I…

Writ e t he sim plest C++ program

w it hout know ing t he language?

Problem

I would like to be able to write a C++ program without initially knowinganything about the language Is there a program I can write without having toknow what is required?

Technique

Most programmers investigate a new language by writing a simple program andthen go back to understand how it all works You can do this in C++, too At aminimum, you will need

• A C++ compiler and linker

• A text editor

You will use the editor to write the source text, referred to as source code.

Then you will use the compiler and linker to produce the program

Steps

1. You should create a base source code directory named SOURCEand thenchange to this directory Next, create a work directory named FIRSTandchange to it

2. Start your text editor In a pinch, you could use Notepad if you arerunning under Windows 95/NT If you are using UNIX, you could use ed

or vi Refer to your system documentation for details

3. Type in the following source code, exactly as shown Pay special attention

to symbols, such as brackets and semicolons

// filename: first.cpp - my first C++ program

#include <iostream>

using namespace std ; int main( ) /* main function */

{ cout << “My first C++ program!” << endl ; return( 0 ) ;

}

4. Save the file, naming it FIRST.CPP Then exit the editor and return to thecommand line.

5. At the command line, type the command required to run the compiler

and linker For example, if you are using Microsoft Visual C++, youwould type cl first.cpp The compiler will run and (if you typed alltext correctly) return to the command line without any error messages Ifthe compiler does report an error, refer to the following “Comments”

section

6. At the command prompt, type the name of the executable (called first)

If you are on a UNIX system and you do not see a file named first, lookfor a file named a.out; if you find this filename, execute it You shouldsee the following message on your screen:

My first C++ program!

In the next section, “How It Works,” I will discuss how this programoperates

How It Works

Now it is time to examine the process in more detail First, I will review the

processes that are required to create a program Then I will examine the sourcefile itself, line by line

The procedure to create a program is always the same First, you create

source code and save it in a file so that the compiler can parse it You can use

any text editor to type your source code If you are using Windows, you can

use Notepad or WordPad If you must use a word processor, be sure to save thefile as a pure ASCII text file If you are using UNIX, you can use vi, emacs, or

ed I recommend that you use a dedicated programmer’s editor for serious

source editing

The next step is to type the source code to satisfy the functionality for the

program Visual style for the source layout is important for both you and otherswho will be maintaining your source code A number of styles are accepted Seethe following “Comments” section for examples

After you have typed the source code and have saved the file, the compiler

must be run against the source file The compiler reads the source file, performsinterpretation, and produces an object file The object file, in its current form,cannot be executed directly

Next, the linker must be run to produce the executable The linker combinesthe object file(s), plus any required library functions and classes to produce theprogram The output of the linker, if everything is successful, is an executableprogram This program is now ready to run

1 1

W R I T E A S I M P L E C + + P R O G R A M W I T H O U T K N O W I N G T H E L A N G U A G E 13

Now shift your attention to the source code itself and discover what eachline does.

The first line in the file is identified as a comment A comment in C++ can

be denoted in one of two ways One way, as is demonstrated in the example,consists of the two-character sequence //(two forward slashes) This style is

referred to as the line comment The compiler disregards any text beyond this

sequence The comment in this source text// filename: first.cpp - my first C++ program

simply tells any readers that the name of this file is first.cppand that this isyour first C++ program The second form of comment in C++ consists of twoindividual character sequences /*(slash-star) and */(star-slash) and is

commonly referred to as a block comment The sequence /*begins the commentand the sequence */ends the comment The difference between the two styles

is that block comments can span many lines, whereas the line comment cannot

In addition, program statements can exist after the block comment’s endsequence A block comment is shown in the following example:

int main( ) /* main function */

The second line of code in the file

The third line in the source fileusing namespace std ;

is a usingdirective and is used in conjunction with the namespace feature.Namespaces are used to partition the global namespace This eliminates, or atleast reduces, name conflicts Refer to the following “Comments” section for adiscussion of the usingdirective if your compiler complains

The function found on the fourth line of codeint main( ) /* main function */

is the starting point for a C++ program; the mainfunction is a requirement ofevery C++ program The intis a C++ data type and designates that the returnvalue from mainis an integer value In short, the operating system loads a

program into memory and then calls the mainfunction to start the ball rolling.The body of the mainfunction is delineated by the opening and closing braces{} The “Comments” section that follows discusses an optional declaration of

main Functions will be discussed in How-To 1.4

The statement on the fifth line

cout << “My first C++ program!” << endl ;

displays a message to standard output (the screen) The insertion operator <<isused to put data to the coutobject Notice that the insertion operator can be

chained In the example, the first data item sent to the coutobject is a string oftext, followed by the endlmanipulator (Note that the last character in endlis

L, not the numeric value 1.) Finally, a semicolon is used to end the statement.The last line of code

return( 0 ) ;

is the returnstatement The returnstatement directs a function to return to itscaller; in this context, mainreturns to the operating system A return

expression can optionally return a value to its caller In this instance, the

returnstatement returns a zero to the operating system

That’s all that is required for a C++ program In this exercise, you have built

a very simple, yet straightforward C++ program All C++ programs must followthis “template.” The remainder of a program’s functionality is up to you to

develop This book consists of many examples you need to expand on this basicexample

Comments

Source code style is a hotly debated topic Every programmer develops his or

her own style of source code formatting The sample that follows is the

traditional K&R style:

The following sample is a variation of the block style and is known as

indented block style:

The most widely accepted style is the second form, block style Block style ismore visually pleasing and the braces are easier to visually align.

The convention used for the filename identified in the #includecomplieswith the Standard C++ convention No extension is specifically identified forthe filename If your compiler complained that it could not open iostream, tryreplacing the statement:

#include <iostream>

with the following:

#include <iostream.h>

If your compiler does not conform to the new convention, it will try to find

a file named iostream Changing the name to iostream.hshould help If yourcompiler still complains with No such file or directory, you have one of

two problems Either you have a very old compiler or your compiler’s settings

are incorrect If you are using a very old compiler, I strongly suggest that youupgrade to one of the newer packages Why? The examples throughout thisbook assume you are using a current compiler package—older compilers mightnot understand some of the constructs used I also mentioned that the

compiler’s settings might be incorrect Specifically, check the settings that thecompiler uses for file locations, such as path settings or environment variables

If your compiler complains about the following source code line:

using namespace std ;

simply remove that line from the source file and recompile If the compiler nolonger complains, it does not support the namespacekeyword Alternatively,you can precede the text with the line comment //characters, as in thefollowing example:

//using namespace std ;

Remember that the compiler will ignore any source code beyond the linecomment sequence It is the decision of the compiler vendor whether to imple-ment a language feature or not; the only way to know is to try the feature orcheck the vendor’s documentation

The optional declaration for mainis the following:

int main( int argc, char *argv[] ) ;

The first argument, int argc, represents the number of command-line ments provided to the program The second argument, char *argv[], is anarray of pointers to char(specifically, to each of the arguments) Each argument

argu-to a C++ program is treated as a string of text You can pass information argu-to aprogram using arguments on the command line For example, the followingshows a program that accepts input to multiply two numbers:

Prompt> mult 114 23

This example will read the two arguments, multiply the two values, and

display the result to the standard output

C O M PL EX I T Y

B E G I N N I N G1.2 How do I…

Creat e a program t o perf orm

calculat ions?

Problem

I would like to write a program that can perform some calculations I know that

I will have to use data variables, but I am unfamiliar with the various data typesand operators in C++

Technique

Programming languages provide data type variables that can be used to hold

values A value can be stored, accessed, and replaced in a variable In order touse a variable, you must decide the data type(s) that you need The C++

language defines a number of variable data types to choose from The programpresented in this How-To will calculate the number of inches for a specified

number of feet

Steps

1. Change to your base source directory and create a new directory named

TOINCH Then, start up your text editor

2. Type in the following source code, exactly as shown:

// toinch.cpp - converts feet to inches.

int numberOfInches = numberOfFeet * inchesInFoot ; cout << numberOfInches << “ inches are in “ ; cout << numberOfFeet << “ feet” << endl ;

1 2

C R E A T E A P R O G R A M T O P E R F O R M C A L C U L A T I O N S 17

return( 0 ) ; }

3. Save the file, naming it TOINCH.CPP Then, exit out of the editor andreturn to the command line

4. Run the compiler and linker, naming TOINCH.CPPas the input file

5. Run the TOINCH program The following message will be displayed:Please enter the number of feet, I will tell you the number of inches

6. The cursor will be on the next line, awaiting input Enter the value 2 Theprogram will perform the calculation and display the following message:

24 inches are in 2 feet

7. To perform another calculation, simply rerun the program

This source line also introduces the assignment operator = The assignmentoperator is used to place the value on the right-hand side of the operator to thevariable on the left-hand side In this example, the value 12is placed in thevariable named inchesInFoot The constant (variable) inchesInFootwill holdthis value throughout the life of the program (its value cannot be changed) Thenext line of code is the following:

int numberOfFeet = 0 ;

This int, named numberOfFeetis declared and initialized to a value of 0.

The purpose of this variable is to hold the number of feet that the user specifies

at runtime Note that the constmodifier is not applied to this variable

definition because numberOfFeet’s contents will need to be changed The nexttwo lines

cout << “Please enter the number of feet, “ ;

cout << “I will tell you the number of inches” << endl ;

are used to output a message (an instruction) to the user Notice that althoughthe two statements are independent and exist on two separate lines, the

message is actually displayed to the user as a single line The endlmanipulatorperforms two operations First, it flushes the output buffer and outputs a

newline character Because output is to the screen, the cursor will be placed onthe next line, flush with the left margin Because the first coutstatement doesnot end with an endl, the second message is appended to the first The next

line of code

cin >> numberOfFeet ;

awaits input from the user After the user enters a value and presses the Enter

key, the cinobject will take the value and place it in the variable named

numberOfFeet The cinstatement actually performs a conversion behind the

scenes You press the keyboard character 2; cintakes this character and

converts it to a C++ integer type and places the value in the variable This

brings you to the next line of code:

int numberOfInches = numberOfFeet * inchesInFoot ;

This statement introduces the multiplication operator * This operator is one

of many defined by the C++ language This source line is referred to as a

compound statement First, the intvariable numberOfInchesis declared for use.Next, the value stored in numberOfFeetis multiplied by the value stored in

inchesInFoot The result of this expression is then assigned to the variable

numberOfInches The next two source lines

cout << numberOfInches << “ inches are in “ ;

cout << numberOfFeet << “ feet” << endl ;

display the result of the conversion to the standard output Notice that you canstream different data types to the coutobject First, the value stored in

numberOfInchesis sent to cout, followed by a string In the next statement,

numberOfFeetis sent to cout, followed by another string, and finally the endl

manipulator

The last program statement, the returnstatement, returns the value 0

returns to the operating system

1 2

C R E A T E A P R O G R A M T O P E R F O R M C A L C U L A T I O N S 19

The C++ language defines a number of data types: bool, char, int, long,

float, and double A boolis used to represent a Boolean value, either trueor

false The chartype is used to contain characters, such as 3, #, or m The types

intand longare integral types and can hold whole numbers only The types

floatand doubleare used for storing floating-point numbers

In C++, strings are implemented as an array of char In fact, any data typecan be declared as an array You will look at arrays in a How-To later in thischapter

The constkeyword is used to qualify a declaration You must remember toinitialize the constvariable at the point of declaration; if you don’t, you will not

be able to assign to it For example, the compiler will not allow the following:const int inchesInFoot ;

inchesInFoot = 12 ; // error right here!

The constvariable is used as a symbolic constant (or alias) for a value.Using a symbolic name provides a descriptive name for some value An additional benefit of using constis that the value stored in the variable isprotected from modification It also helps with code maintenance Why? Let’slook at a brief example The following example defines a charthat is constant.Assume that the charis used as a choice within a menu system

const char REPLACE_OPTION = ‘S’ ; // replace

words, you would have to search for all occurrences of the letter Sand change

them to the new value Instead, by using a constant, you only have to changethe initialization value of the constant declaration The following line of codeshows a decision to change the value of REPLACE_OPTIONto R:

const char REPLACE_OPTION = ‘R’ ;

The change is only required at the declaration The next step would be torecompile all the code associated with this change You do not have to worryabout searching for every occurrence of Rbecause the compiler will do all thework required

The cinobject is used to accept data from the standard input; the default isthe keyboard The extraction operator >>directs the input from the cinobject

to the named variable(s) In the previous example, numberOfFeetis the ient of data from cin

recip-C O M PL EX I T Y

B E G I N N I N G1.3 How do I…

Use t he various loop st at em ent s

t hat are available in C++?

Problem

I have started to sink my teeth into C++ now by writing very simple programs

I need to extend my knowledge of the language and have considered using a

loop statement as I’ve seen in other languages I need to know the various loopsavailable in C++ and how to use them

Technique

Looping techniques in C++ are essentially the same as they are in other

languages The loop names might be different, but for the most part, they servethe same purposes The C++ language serves up three different types of loopingstatements: for, do-while, and while The following example will show all

three in action

Steps

1. Change to your base source directory and create a new directory named

LOOP Next, fire up your text editor

2. Type in the following source code, exactly as shown:

// loop.cpp - program to demonstrate C++ loops

#include <iostream>

using namespace std ; int main()

{ int value = 0 ; cout << “Beginning of while loop” << endl ; while( value < 5 )

{ cout << “The value of value is: “ << value << endl ; ++value ;

} cout << “End of while loop” << endl ; cout << “\nBeginning of do-while loop” << endl ;

do { cout << “The value of value is: “ << value << endl ; }while( value < 5 ) ;

cout << “End of do-while loop” << endl ;

1 3

U S E T H E V A R I O U S L O O P S TA T E M E N T S T H A T A R E A V A I L A B L E I N C + + 21

cout << “\nBeginning of for loop” << endl ; for( value = 0; value < 5; value++ ) {

cout << “The value of value is: “ << value << endl ; }

cout << “End of for loop” << endl ; return 0 ;

}

3. Save the file as LOOP.CPPand exit the editor to the command line

4. Compile and link the LOOP.CPPsource file

5. Run the program; the output should be as follows:

Beginning of while loop The value of value is: 0 The value of value is: 1 The value of value is: 2 The value of value is: 3 The value of value is: 4 End of while loop Beginning of do-while loop The value of value is: 5 End of do-while loop Beginning of for loop The value of value is: 0 The value of value is: 1 The value of value is: 2 The value of value is: 3 The value of value is: 4 End of for loop

How It Works

Starting with the sixth line of code, an inthas been declared and initializedwith a value of 0 This integer variable will be used by all three loops In addition to demonstrating loops, this How-To introduces you to the C++ increment (++) and less-than (<) operators

The next statement sends a message to the screen:

cout << “Beginning of while loop” << endl ;

It is always a good idea to let the user know what is going on within aprogram You can consider these messages as “progress reports.”

You begin your investigation of C++ loops with the next line of code:while( value < 5 )

The whileloop is a precondition (entry condition) loop First, the test

expression within the parentheses is evaluated If the result of the expression isnon-zero (true), the statement(s) within the whileloop’s body are executed Ifthe result of the expression is 0(false), then the loop exits and control returns

to the first statement after the whileloop’s body This loop can be verbally

expressed as follows: “While the value stored in valueis less than 5, execute

the statements within the loop’s body.” It is possible for a loop to never execute

if the expression is initially false For example, if the variable valueis initiallyset to 20, the whileloop will never execute

An opening brace begins the body of a whileloop and a closing brace endsthe body All statements within the body are executed as long as the loop’s

expression is true The whileloop contains two statements, a message that

displays the contents of the variable valueand an increment (++) expression

You have seen the coutstatement before; its job is to display data to the

standard output The next statement uses the increment operator ++ This

operator simply increments the value of the operand by one It is shorthand forthe following statement:

value = value + 1 ;

Two versions of the increment operator exist: prefix increment and postfix

increment The version used in the whileloop is the prefix increment operator.The prefix increment version increments a variable first and then evaluates thebalance of the expression The postfix increment version does the opposite: Theexpression is evaluated first, and then the variable is incremented The

following example shows the difference:

Line 1: if( ++check < 6 )

Line 2: if( check++ < 6 )

Assume the value of check is 5before the ifstatement In line 1, the value

of checkis incremented first, and then the expression is evaluated Because

checkis now 6, the expression evaluates to false(check < 6) and the

statements within if’s body are not executed In line 2, the expression within

the ifstatement (check < 6) evaluates to true, checkis incremented by 1

Because the expression evaluates to true, the statements within the if’s body

are executed

After valueis incremented, program control is returned to the top of the

whileloop The whileloop’s expression is again evaluated and if it is non-zero,the body is again executed; otherwise, the program exits the loop and control

jumps to the first statement after the whileloop’s body In this example, it is amessage to the user that the whileloop has ended

The next statement displays a message announcing the beginning of the

do-whileloop A do-whileloop is a postcondition (exit condition) loop A

do-whileloop is always executed once, even if the test expression evaluates to

1 3

U S E T H E V A R I O U S L O O P S TA T E M E N T S T H A T A R E A V A I L A B L E I N C + + 23

false This behavior is in contrast to the whileloop Note that the valuecontained in valueis still 5, yet the coutstatement is still executed The testexpression within the whileportion is evaluated as falseand the loop isterminated Control resumes at the coutstatement announcing the end of the

Next, you come to the forloop The forloop consists of three expressions.The first expression is initialization, followed by a test expression, and finally,the update (or change) expression; each expression is separated by a semicolon.The following is the program’s forloop:

for( value = 0; value < 5; value++ )

First, valueis initialized to 0 The middle expression (value < 5) is the testexpression The test expression is evaluated; if the result is true, the loop’sbody is executed, otherwise, the program exits the loop After all statements inthe body have executed, the third expression is evaluated After the thirdexpression is evaluated, control returns to the test expression

Comments

Each of the C++ loops can have either a single statement or multiple statementsenclosed within braces All of the loop examples in this How-To use braces todelineate the loop’s body For example, the forloop can be written this way:for( value = 0; value < 5; value++ )

cout << “The value of value is: “ << value << endl ;

You can do this because this forloop only has one statement associated with

it It is more appropriate to use braces, even if you only have one statementassociated with a loop The reason is that if you come back to add anotherstatement to the single-statement loop, you might forget to add the braces Todemonstrate, assume you want to add a calculation that is performed for everyiteration of the loop

for( value = 0; value < 5; value++ )

cout << “The value of value is: “ << value << endl ;

paycheckAmount = hours * value ;

Everything appears fine; the indentation shows that the new statement ispart of the loop’s body Don’t be fooled by appearances By visual inspection, it

appearsthat the program will execute both statements following the forsion In reality, only the first statement will execute for every iteration of theloop The second statement will only execute after the loop is finished Tocorrect the problem, the source code needs to look like the following:

expres-for( value = 0; value < 5; value++ )

{

cout << “The value of value is: “ << value << endl ;

paycheckAmount = hours * value ;

}

Now, everything will work as expected.

The third (update) expression in a forloop is not restricted to using the

increment operator ++ You might want to use the decrement operator Youcan also increment the value by two, three, or more The expression can also bethe result of a multiplication The following will increment valueby 20in thethird expression:

for( value = 1; value < 100; value = value + 20 )

Creat e a program t hat uses one or

m ore f unct ions?

Problem

I am ready to move on to the more advanced features of the C++ language,

specifically functions I know that functions are used to accomplish specific

tasks I need to know how to declare and define functions

Technique

Functions in C++ are the basic building blocks to modularize a program The

technique for creating functions in C++ is the same as it is in any other

language You must decide the specific functionality required for your

application If possible, you should make your functions as general as possible.This allows you to use a function in other programs

Steps

1. Change to the base source directory and create a directory named FUNC

2. Start up your text editor and type in the following source code:

// func.cpp - program to demonstrate a function in C++

#include <iostream>

using namespace std ; long multiply( long left, long right ) ; int main()

{ long value1 = 0, value2 = 0 ;

1 4

C R E A T E A P R O G R A M T H A T U S E S O N E O R M O R E F U N C T I O N S 25

cout << “\nProgram to multiply two integers “ ; cout << “and display the result “ << endl ; for( long result = 1; result != 0; /*empty expression*/ ) {

value1 = value2 = result = 0 ; cout << “\nTwo zeros will end the program.” << endl ; cout << “Enter the first integer (and enter key): “ ; cin >> value1 ;

cout << “Enter the second integer (and enter key): “ ; cin >> value2 ;

if( value1 == 0 && value2 == 0 ) break ;

result = multiply( value1, value2 ) ; cout << value1 << “ multiplied by “ ; cout << value2 << “ = “ << result << endl ; }

return 0 ; }

// Multiplies the two arguments // together and returns the results long multiply( long left, long right ) {

long result = left * right ; return( result ) ;

}

3. Save the file as LOOP.CPPand exit the editor to the command line

4. Compile and link the LOOP.CPPsource file

5. Run the program; the output should be as follows:

Program to multiply two integers and display the result.

Two zeros will end the program.

Enter the first integer: 4 Enter the second integer: 4

4 multiplied by 4 = 16

How It Works

If you have been following the previous How-Tos, you will notice a new line ofcode after the using namespace stdline:

long multiply( long left, long right ) ;

This is referred to as a function declaration It is the function’s signature The

declaration specifies the number or arguments, if any, and their data type Thedeclaration also shows the return type (if the function returns a value) In this