"Herb Schildt tells his

programmers what they

want and need to know-
simply, clearly, concisely.
and authoritatively."
- ACM Computing Re\,'i,,,,,.,.s*_

Third
 •

Teach
Yourself
c
Third Edition
"'\ct. io..~ ~ ~
II'(';)C\oJ}

\.u..~. "'+ c..S L ( f\--crc; (' ~ ')

~l\.AC IJ~~,-~

OsbOrne McGraw-HIM
8ef1itele) New York St louis Sen Ft~ncis!:o AuclQInG ~ Hamburg london Medrid Me.voo Cit)'
Milan MontnseI j.,;aw Delhi ?tI11IO/NI Oty Pant SIo P.uIo S:ogapore S')IdI)8y Tokro TOI'Onto
•
 Teach
Yourself
c
Third Edition

. OsbOrne McGraw-HIH
8e~ ~ YOt1l St lOUis Sen Fr.nciso:) AucklIII'IO ~ HIomOuI'Q london M.and Me~~ Cit)<
SIo....,
M:1an Montreal l.oew Delhi ' nT'I5IM City Pw.s S:fIG8pOfe: Sydney Tok)'O Toronto

,".,. ' ,
Contents

Preface, Xl
For Ftmher study, XV"

,,,,'.)' C Fundamentals . . . 1
1.1 UNDERSTAND THE COMPONENTS Of A C
PROGRAM, 2
1.2 CREATE AND COMPILE A PROGRAM. 7
V3 DECLARE VARIABLES AND ASSIGN VALLIES,
• 10
1.4 INPUT NUMBERS fROM THE KEYBOARD. 15
1.5 PERFORM CALCULATIONS USING
ARITHMETIC EXPRESSIONS, 17
1.6 ADD COMMENTS TO A PROGRAM. 20
1.7 WRITE YOUR OWN FUNCTIONS, 23
1.8 USE FUNCTIONS TO RETURN VALUES, . 27
~ USE fUNCTION ARGUMENTS, 32
1.10 REMEMBER THE C KEYWORDS, 35
2 Introducing C's Program Control Statements . . . 39
2.1 BECOME fAMILIAR WITH THE if, 41
2.2 ADD THE else, 44
2.3 CREA'rE BLOCKS OF CODE, 46
2.4 USE THE for LOOP, 49
2.5 SUBSTITUTE C'S INCREMENT AND
DECREMENT OPERATORS, 54
2.6 EXPAND printf( )'S CAPABILITIES, 58
2.7 PROGRAM WITH C'S RELATIONAL AND
LOGICAL OPERATORS, 61

3 More C Program Control Statements • . • 69

3.1 INPUT CHARACTERS, 70
3.2 NEST if STATEMENTS, 75
3.3 EXAMINE for LOOP VARIATIONS, 79
3.4 UNDERSTAND C'S while LOOP, 82
3.5 LISE THE do LOOP, 84
3.6 CREATE NESTED, LOOPS, 87
 3.7 USE break TO EXIT A LOOP, 89
3.8 KNOW WHEN TO USE THE continue
STATEMENT, 92
3.9 SELECT AMQNG ALTERNATIVES WITH THE
switch STATEMENT, 94
3.10 UNDERSTAND THE goto STATEMENT, 100

-v<A' A Closer Look at Data Types, Variables,

and Expressions . . . 105
4.1 USE C'S DATA-TYPE MODIFIERS, 107
4.2 LEARN WHERE VARIABLES ARE DECLARED,
)]2
4.3 TAKE A CLOSER LOOK AT CONSTANTS, 119
'V!'A INITIALIZE VARIABLES, 123
4.5 UNDERSTAND TYPE CONVERSIONS IN
EXPRESSIONS, 126
4.6 UNDERSTAND TYPE CONVERSIONS IN
ASSIGNMENTS, 129
47 PROG RAM WITH TYPE CASTS, 132

L~ Exploring Arrays and Strings . . • 137

5.1 DECLARE ONE-DIMENSIONAL ARRAYS, 139
5.2 USE STRINGS, 145
5.3 CREATE MULTIDIMENSIONAL ARRAYS, lSI
5.4 INITIALIZE ARRAYS, 154
5.5 BUILD ARRA YS OF STRINGS, 159

4 uSing Pointers . . . 165

6. I UNDERSTAND POINTER BASICS, 167
6.2 LEARN RESTRICTIONS TO POINTER
EXPRESSIONS, 172
6.3 USE POINTERS WITH ARRAYS, 176
6.4 USE POINTERS TO STRING CONSTANTS, 183
6.5 CREATE ARRAYS OF POINTERS, 186
6.6 BECOME ACQUAINTED WITH MULTIPLE
INDIRECTION, 188
6.7 USE POINTERS AS PARAMETERS, 191

~ A Closer Look at Functions ••• 195

7.1 UNDERSTAND FUNCTION PROTOTYPES, 196
7.2 UNDERSTAND RECURSION, 207
7.3 TAKE A CLOSER LOOK AT PARAMETERS, 211
7.4 PASS ARGUMENTS TO main(), 215
 7.5 COMPARE OLD-STYLE TO MODERN FUNCTION
PARAMETER DECLARATIONS, 220
8 Console 110 _ _ _ 227
8.1 LEARN ANOTHER PREPROCESSOR
DIRECTIVE, 229
8.2 EXAMINE CHARACTER AND STRING INPUT
AND OUTPUT, 233
8.3 EXAMINE SOME-NON-STANDARD CONSOLE
fUNCTIONS, 235
8.4 TAKE A CLOSER LOOK AT gets() AND
puts(), 238

g 8.6 8.5 MASTER printf(), 241

MASTER seanf(), 246
9 File I/O _ _ _ 257
9.1 UNDERSTAND STREAMS, 259
9.2 MASTER fiLE-SYSTEM BASICS, 260
9.3 UNDERSTAND fcof() AND ferror(), 269
94 LEARN SOME HIGHER-LEVEL TEXT
FUNCTIONS, 274
9.5 LEARN TO READ AND WRITE lllNARY DATA,
278
9.6 UNDERSTAND RAl'lDOM ACCESS, 285·
9.7 LEARN ABOUT VARIOUS fILE-SYSTEM
FUNCTIONS, 290
9.8 LEAR!" ABOUT THE STANDARD STREAMS, 293
4 Structures
vro)
~nd
Unions _ _ _ 299
MASTER STRUCTURE BASICS, 300
'vf6.2 DECLARE POINTERS TO STRUCTURES, 314
10.3 WORK WITH NESTED STRUCTURES, 318
vrQ4 UNDERSTAND BIT-fiELDS, 324
/ 10.5 CREATE UNIONS, 329

~{,Advanced Oala Types and Operators _ • . 337

::..rf.1 USE THE STORAGE CLASS SPECIfIERS, 339
11.2 USE THE ACCESS MODIFIERS, 349
11.3 DEFINE ENUMERATIONS, 352
vl1.4 UNDERSTAND typcdef, 356
11.5 USE C'S BITWISE OPERATORS, 358
11.6 MASTER THE SHIFT OPERATORS, 363
~.7 UNDERSTANDTHE?OPERATOR, 365
-vrf.8 DO MORE WITH THE ASSIGNMENT
OPERATOR, 367
 11.9 UNDERSTAND THE COMMA OPERATOR, 370
11.10 KNOW THE PRECEDENCE SUMMARY, 372
// .
'01 12 The C Preprocessor and Some Advanced Topics . 375
12.1 LEARN MORE ABOUT #define AND #inc1ude,
377
12.2 UNDERSTAND CONDITIONAL COMPILATION,
381
12.3 LEARN ABOUT #error, #undef, #line, AND
#pragma, 388
12 .4 EXAMINE C'S BUILT·IN MACROS, 391
12.5 USE THE # AND ## OPERATORS, 393
02.6 UNDERSTAND FUNCTION POINTERS, 395
12.7 MASTER DYN.'\MIC ALLOCATION, 402

A Some Common C Library Functions . . . 411

A.l STRlNG AND CHARACTER FUNCTIONS. 412
A.2 THE MATHEMATICS fUNCTIONS, 424
A .3 TIM": AND DATE FUNCTIONS, 434
A.4 DYNAMIC ALLOCATION, 440
A.S MISCELLANEOUS FUNCTIONS, 444

B C Keyword Summary . . . 457

C Building a Windows Skeleton . . . 469
WHICH VEkSION OF WINDOWS? 470
WINDOWS PROGRAMMING PERSPECTIVE, 470
HOW WINDOWS AND YOUR PROGRAM
INTERACT, 473
WINDOWS IS MULTITASKiNG, 474
THE WIN32 API, 474
THE COMPONENTS OF A WINDOW, 475
SOME WINDOWS APPLiCATION BASICS, 476
THE WINDOW FUNCTION, 489
A SHORT WORD ABOUT DEFINITION FILES,
490
NAMING CONVENTIONS, 490
TO LEARN MORE, 490

o Answers .. 493

Index • .. 633
Preface

This book teaches you how to program in what is usually regarded as

the world's most important professional programming language: C.
One J'eason for C's success and staying power is that programmers
like it. C combines subtlety and elegance with raw power and
flexibility . It is a structured language that does not confine. It is
a high-performance language that does not constrain . C is also a
language that puts you, the programmer, firmly in charge . C was
created by a programmer for programmers. It is not the contrived
product of a committee, but rather the outcome of programmers
seeking a better programming language.
C is important for another reason. It is the gateway to the world's
two other professional programming languages: C++ and Java. C++ is
built upon C, and Java is built upon C++. Thus, C is at the foundation
of all modern programming, and knowledge of C is fundamental to the
successful creation of high-performance, high-quality software. Simply
put, to be a professional programmer today means that you are
competent in C.

A Short HIst"" of C
C was invented and first implemented by Dennis Ritchie on a DEC
PDP-II using the UNIX operating system. C is the result of a
development process that started with an older language called BCPL,
developed by Martin Richards. BCPL influenced a language called B
that was invented by Ken Thompson and that led to the development
of C in the 1970s.
For many years, the de facto standard for C was the one described
in The C Programming Language by Brian Kernighan aJ;ld Dennis
Ritchie (Prentice-Hall, 1978). However, as C grew in popularity, a
committee was organized in 1983 to create an ANSI (American
National Standards Institute) standard for C. The standardization
process took six years (much longer than anyone reasonably
expected). The ANSI C standard was finally adopted late in 1989 and Ix
."
•... TEACH YOURSELF

the first copies became generally available in 1990. The standard was
amended slightly in 1996. Today, virtually all C compilers comply with
ANSI standard C and that is the version of C you will learn in this
book. (That is, this book teaches ANSI standard C.)
C is often referred to as a mllidle-Iel'ellanguage. Before C th~re "'ere
basically two types of languages used to program computers. One is
called assembly language, which is the symbolic representation of the
actual machine instructions executed by the computer. Assembly
language is a low-Ievellangllage because the programmer is working
with (in symbolic form) the actual instructions that the computer will
execute. Assembly language can be used to create very efficient
programs, but it provides no built-in control structures or 110
functions. All such items must be manually constructed by the
programmer. By contrast, a hlgh-levellal1guage b~ffers the programmer
from the computer. A high-level language typically supplies various
control structures, input and output commands, and the like , which
make programming easier and faster. However. the elements of a
high-level language may not relate directly to the way that the
computer ultimately carries out the program. This separation often
causes programs written using a high-level language to be less efficient
than those written in assembly language. Because many people find
assembly language programming to be a tedious, difficult task, there
was a need for a language that balanced ease-of-use with efficiency.
Many programmers feel that C provides this balance. It successfully
combines the structure of a high-level language with the power and
efficiency of assembly language. Since it spans the gap between
• assembly language and high-level languages, it is called'a middle-Ie"el
language.
Initially, C was used primarily for creating systems sojtu'are. Systems
software consists of those programs that help run the computer These
include programs such as operating systems, compilers , and editors
However, as C gained in popularity, it began to be used for general
purpose programming. Today, C is used by programmers for virtually
any programming task. It is a language that has survived the test of
time and proven itself to be as versatile as it is powerful.
C IrS. c++

Newcomers are sometimes confused about the differences between

C and c++ and how they relate to each other. In short, c++ is an
extended version of C that is designed to support object-oriented
programming (OOP). c++ contains and supports the entire C language
in addition to a set of object-oriented extensions. (That is, C++ is a
superset of C.) Because C++ is built upon the foundation of C, you
cannot learn C++ without learning the basics ofC. Therefore, if you
think that you will someday move on to C++, your knowledge of C
will not only be useful, it will be necessary.

About TIlls Book

This book is unique because it teaches you the C language by applying

mastery learning. It does so by presenting one idea at a time, followed
by numerous examples and exercises to help you thoroughly
understand each topic. This approach ensures that you master each
topic before moving on.
The material is presented sequentially. Therefore, you should work
carefully through each chapter because each chapter assumes that you
know the material presented in all preceding chapters.
This book teaches ANSI standard C. This ensures that your
knowledge will be applicable to the widest range of C environments.
This book also uses contemporary syntax and structure, which means
that you will be learning the right way to write C programs from the
very beginning.

How This Book is Organized

This book is composed of 12 chapters and 4 appendices. Each chapter

(except Chapter I) begins with a Review Skills Check, which consists
of q'u estions and exercises covering the previous chapter's material.
The chapters are divided into sections. Each section covers one topic.
At the end of each section are examples followed by exercises that test
your understanding of the topic. At the end of each chapter, you will
find a Mastery Skills Check, which checks your knowledge of the
material in the chapter. Finally, a Cumulative Skills Check is
For Further Study

Teach YOlU-self C, T/.ird Edition is your gateway into the "Herb

Schildt' series of programming books. Here is a partial list of Schildt's
other programming books published by Osborne/McGraw-Hili.
If you want to learn more about C, you will find these books
especially helpful.

C: The Complete Reference

The Annotated ANSI C Standard

If you will be moving on to C++ (C's object-oriented extension),

then you will find that Schildt's C++ books provide excellent coverage
of this important language. We recommend

Teach Yourself C++

C++: The Complete Reference
C++ from the Ground Up

If you will be developing programs for the Web, you will want to
read

Java: The Comple·" Reference

,co-authored by Herbert Schildt and Patrick Naughton.

finally, if you want to program for Windows, we recommend

Schildt's Windows 95 Programming in C and C++

Schildt's Advanced Windows 95 Programming in C and C++
Windows NT 4 from the Ground Up
MFC Programming from the Ground Up

When you need solid answers, fast, turn to

Herbert Schildt, the recognized authority on
programming. . .
TEACH YOURSELF

.IH E individual elements of a computer language such as C do

not stand alone, but rather 111 cortjunction with one another.
Therefore, it is necessary to ul1detstand several key aspects
of C before examining each element of the language in detail.
To this end, this chapter presents a qUick overview of the C
language. Its goal is to give you sufficient working knowledge of C so
that you can understand the examples in later chapters.
As you work through this chapter, dOl1't worry if a few points are
not entirely clear. The main thing you need to understand is how and
why the example programs execute as they do. Keep in mind that
most of the topics introduced in this chapter will be discussed in
greater detail later in this book. 111 this chapter, you will learn about
the basic structure of a C program; what a C statement is; and what
variables, constants, and functions are. You will learn how to display
text on the screen and input illfonnation from the keyboard.
To lise this book to the fullest, you must have a computer, a C
compiler, and a lext editor. (YOli may also use a c++ compiler. c++
compilers (an dlso compile C programs.) Your compiler may include
its 0\'\'11 text editor, in which case you ",.'on't need a separate one. For
the best results, you should work along with the examples and try the
exercises .

· 1.1 UNDERSTAND THE COMPONENTS OF A

C PROGRAM

.All C programs share certain essential components and iraits. All C

programs consist of one or more functIons, each of which contains one
or more statements. In C, a function is a named subroutine that can be
cailed by other parts of the program . Functions are the building blocks
of C. A statement specifies an action to be performed by the proglam
III other words, statements are the parts of your program that actually
perform operations.
All C statements end with a semicolon. C does not recognize the
end of the line as a terminator. This means there ate no constraints on
the position of statements withm a line . Also, you may place two or
lnore statenlents on one linc .
 C FUNDAMENTALS 3
1. 1 UNDERSTAND THE COMPONENTS OF A C PROGRAM
'"
Tha general form of a C function is shown here:

ret-type function-name(param-list)
{
statement sequence
}

Here , ret-type specifies the type of data returned by the function.

As you will see it is possible for a function to return a value. The
{!lI1ctiol1-nIlI11C is the name of the function. Information can be passed
to a tunction through its parameters, which are specified in the
function's parameter list, paral11-list. The statement sequence l11ay be
one or more statements. (Technically, a function can contain no
statenlents, but since this means the function performs no action, it is
a degenerative case.) If return types and parameters are new concepts,
don't worry, they will be explained later in this chapter.
(\"ith few exceptions, you can call a function by any nanle you like.
It must be composed of only the upper- and lowercase letters of the
alphabet, the digits 0-9, and the underscore. A digit cannot start a
funct·ion nanle, however. C is cllsc-sel1sitwe, \\'hich means that C
recognizes the difference between upper- and lo"-ercase letters:)
Thus, as far as C is concerned, Myfunc and myfunc are entirely
dIfferent nanleo,.
(Although a C program may contam several functions, the only
function that it must have is main ( ). The main( ) function is where
execution of yctur program begins. That is, when your program begins
running, it starts executing the statements inside m ain e), beginning
\"ith the first statement after the opening curly brace. Your program
ends whe'n main( )'S closing curly brace is reached. Of course, the
curly brace does not actually exist in the compiled version of your
program, but it is helpful to think of it in this way.
Throughout this book, when a function is referred to in text, it will
be printed in bold and followed by parentheses. This way, you can see
immediately that the name refers to a function , not some other part
of the program.
Another important component of all C programs is library functions.
The ,ANSI C standard specifies a set oflibrary functions to be supplied
by all C compilers, which your program may use. This colle£tion of

•
.,
4 TEACH YOURSElF
C

lunctions 'is usually referred to as the C standard library. The standard

library contains functions to perform disk 110 (input/ output), string
manipulations, mathematical computations, and much more. When
your program is compiled, the code for each library function used by
your program is automatically included. This differs from the way
some other computer languages work. For example, in BASIC or
Pascal, operations such as writing to a file or computing a cosine are
performed using keywords that are built into the language. The
advantage C gains by having them as library functions is increased
flexihility. Library functions can be enhanced and expanded as needed
to accommodate changing circumstances The C language itself does
not need to change. As you will see. virtually all C programs you
create will usc functions from the C standard library.
One of the most common library functions is called printf'( ). This
is C's general-purpose output function. The printf( ) function is quite
"ersatile, allowing many variations. Its simplest form is shown here:

printf( "string-to-output ");

TI". printf( ) function outputs the characters that are contained
between the beginning and ending double quotes to the screen. (The
douhle quotes are nO! displayed on the screen.) In C, one or more
characters enclosed between double quotes is called a stJing. The
quoted string between printf( )'s parentheses is said to be an argument
to printf( ). II] gene!"l, infomlation passed to a function is called an
--"rgument. In C, calling a library function is a statement; therefore, it
must end with"a semicolon.
To call a function. you specify its name followed by iparenthesized
list of arguments that you will be passing to it. If the function does
not require any arguments. no arguments will be specified-and the
parenthesized list will be empty. If there is more than one argument.
the arguments must be separated by commas.
Another component common to most C programs is the header file.
In C, information about the standard library functions is found in
various files supplied with your compiler. These files all end with a
.H extension. The C compiler uses the information in these files to
handle the library functions properly. You add these files to your •
program using the #inc1ude preprocessor directive. All C compilers use
as their first phase of compilation a preprocessor, which performs
various manipulations on your source file before it is compiled.
 C FUNDAMENTALS 5
1.1 UNDERSTAND THE COMPONENTS OF A C PROGRAM
"
Preprocessor directives are not actually part of the C language, but
rather instructions from you to the compiler. The #include directive
tells the preprocessor to read in another file and include it with your
program. You will learn more about the preprocessor later in this book.
The most commonly' required header file is called STOIO.H. Here is
the directive that includes this file:
#include <stdio.h>

You can specify the file name in either upper- or lowercase, but
lowercase is the traditional method. The STDlO.H header file
contains, among other things, information related to the printf( )
library function. Notice that the #include directive does not end with
a semicolon. The reason for this is that #include is not a C keyword
that can define a statement. Instead, it is an instruction to the C
compiler itself.
One last point: With few exceptions, C ignores spaces. That is, it
doesn't care where on a line a statement, curly brace, or function
name occurs. If you like, you can even put two or more of these itenlS
on the same line. The examples you wiII see in this book reflect the
way C code is normally written; it is a form you should follow. The
actual positioning of statements, functions, and braces is a stylistic,
not a programming, decision.

EXAMPLES

I. Since all C programs share certain common trailS,

understanding one program will help you understand many
others. One of the simplest C programs is shown here:
#include <stdio.h>

int rnain(void)
{
printf(~This is a short C program.");

return 0;
)
..
8 TEACH YOURSD.f
C

When compiled and executed, this Progrilm displays the message

This is a short C program. on the screen of your computer.
Even though this program is only six lines long, it illustrates
those aspects common to all C programs. Let's examine it line
by line.
The first line of the program is
#include <stdio . h>

It causes the file STDlO.H to be read by the C compiler and to

be included with the program . This file contains information
related to printf( ).
The second line,
int rnain(void)

begins the maine ) function. As stated earlier, all C programs

must ha"e a maine ) function. This is where program execution
begins. The int specifies that maine ) returns an integer value .
The ,"aid tells the compiler that maine ) does not have any
pararnercrs.
After maine ) is an opening curly brace. This marks the
beginning of statements that make up the function .
The next line in the program is
printf("This is a short C program.");

. function,
This is a,C statement. It calls the standard library
printf( ), which causes the string to be displayed.
The following line causes maine ) to return the value zero. In
this case, the value is returned to the calling process, which is
usually the operating system.
return 0;

By convention, a return value of zero from maine ) indicates

normal program termination. Any other value represents an
error. The operating system can test this value to determine
whether the program ran successfully or experienced an error.
return is one of C's keywords and is examined more closely
later in this chapter.
Finally, the program is formally concluded when maine )'S
closing curly brace is encountered.
 C FuNDAMENTALS 7
.,.
12 CREATE AND COMPILE A PROGRAM

2. Here is another simple C program:

#include <stdio.h>

int main(void)
(
printf("This is ") i
printf(-another C R) i
printf("program. R ) ;

return OJ

This program displays This is another C program. on the

screen. The key point to Ehis program is that statements are
executed sequentially, beginning with the opening curly brace
and ending with the closing curly brace.

1 .2 < • CREATE AND COMPILE A PR O GRAM

How you vdll create and compile a program is determined to a very'
large extent by the compiler you are using and the operating system
under which it is running. If you are using a PC or compatible, you
have your choice of a number of excellent compilers, such as those by
Borland and Microsoft, that contain integrated program-developmem
environnlents. ,If you are using such an environment, you can edit,
compile, and run your programs directly inside this environment.
This is an excellent option for beginners-just follow the instructions
supplied with your compiler.
If you are using a traditional command-line compiler, then you
need to follow these sieps to create and compile a program:

1. Create your program using an editor.

2. Compile the program.
3. Execute your prograln .

The exact method to accomplish these steps \l'i11 be explained in the

user's manual for your compiler.
10
.., TEACH YOURSELF
C

EXERCISE
-
I. Enter into your Gomputer the example programs from Section
1.1. Compile them and run them.
. .

ECLARE VARIABLE AND ASSION

VALUES

A vGllable is a named memory location that can hold various values.

Only the most trivial C programs do not include variables. In C, unlike
some computer languages, all variables must be declared before they
can be used. A variable's declaration serves one important purpose: It
tells the compiler "'/>at type of valla/'!e is bemg used C supports five
different basic data types, as shown in Table 1-1 along with the C
keywords that represent them. Don't be confused by void. This is a
special-purpose data type that we will later examine closely.
A variable of type char is 8 bits long and is most commonly used to
hold a single character. Because C is very flexible, a variable of type
char can also be used as a 'little integer' if desired.
Integer variables (int) may hold signed whole numbers (numbers
with no fr!ctional part). For 16-bit environments, such as DOS or
Windows 3.1, integers are usuallv 16 bits long and may hold values in
the range -32,168 to 32,767. In 32-bit environments, such as Windows
NT or Windows 95, integers are typically 32 bits in length. In this case,
they may store "alues in the range -2,147,483,648 to 2.147,483,647.

~ Keyword
character data char
signed whole numbers int
floating-point numbers float
double-precision floating-point numbers double
valueless void
TABLE 1-1 C's Five Basic Data Types T
 1.3
C FUNO_ENTALS

DECLARE VARIABLES AND ASSIGN VALUES

.
11

Variables of types float and double hold signed floating-point

values, which may have fractional components. One difference
between flo ..t aJld double i~ that double provides about twice the
precision (number of significant digits) as does float. Also, for most
uses of C, a variable of type double is capable of storing values with
absolute magnitudes larger than those stored by variables of type
float. Of course, in all cases, variables of types float and double can
hold very large values.
To declare a variable, use this general form.

type var-name-

where type is a C data type and var-name is the name of the variable.
For example , this declares counter to be of type int:
int counter ;

In C. a variable declaration is a statement and it must end in a

semicolon.
(There are two places where variables are declared: inside a functIOn
or outside all functions,. Variables declared outside all functions are
called global variables and they may be accessed by any function in
your program. Global variables exist the entire time your program
is executing.) .
(Variables declared inside a function are called local variables. A local
variable is known to-and may be accessed by-only the function in
•
which it is declared. It is common practice to declare all Jocal variables
ased by a function at the start of the function, after the opening curly
brace. There are two important pomts you need to know about local
variables at this time. First, the local variables in one function have
no relationship to the local vanables in another function. That is, if
a variable called count is declared in one function, another variable
called count may also be declared in a seGOnd function - the two
variables are completely separate from and unrelated to each other.
The second thing you need to know is that local variables are created
when a function is called, and they are destroyed when the function is
exited. Therefore, local variables do not maintain their values between
function calls. The examples in this and the next few chapters will
use
only local variables. Chapter 4 discusses more thoroughly the issues
and Imp];catlons of global and local variables.
12 ltACHYOURSELf

'" c

You can declare more than one variable of the same type by using a
comma-separated list. for example, thIs declares three floating-point
variables, x, y, and z:
float x, y, z;

(Like function names, variable names in C can consist of the letters

of the alphabet, the digits 0 through 9, and the underscore. (But a digit
may not start a vanable's name. ) Remember, C is case-sensitive;
count and COUNT are two completely different variable names)
To assign a value to a variable, put its name to the left of an equal
sign. Put the value you want to give the variable to the right of the
equal sign. In C, an assignment operation is J statement; therefore, it
must be terminated by a semicolon. The general form of an
assignment statement is:

variable-name = value;

For e.ample, to assign an Integer variable named num the value 100,
you can use this statement:
num = 100;

In the preceding assignment, 100 is a constant. Just as there are

different types of variables, there are different types of constants. A
constant is " fixed value used in your program. Constants are often
used to inlllalize variables at the beginning of a progra~'s execution.
A character constant is specified by placing the character between
single quotes. For example, to specify the letter" A," you would use
'.\' lntegen are specified as whole numbets Floatmg-point values
must include a decimal point For example, to specity 100.1, you
"auld use 1 00.1. If the floating-pOint value you wish to specity does
nct have any digi(s to the nght of the decimal point, then you must
use O. ~-')r e,,~.nple, to HI the compIler that 100 is a floating-point
number, u<e 100.v.
You can u:; pri •. tf() to dIsplay values of characters, integers, and
floating-point values. To do so. however, requires that you know more
l"A " the printf( ) functlOn. Lets look first at an examp.le. This

%d" . 991;
pr~ntt ( 'T1_~ prl.nts the number '
 C FIJNOAMEJfI'ALS 13
1.3 DEClARE VARIABUS AND ASSIGN VALUES
'"

displays This prints the number 99 on the screen. As you can see,
this call'to printf( ) contains not one, but two arguments. The first
is the quoted string and' the other is the constant 99. Notice that the
arguments are separated from each other by a comma. In general,
when there is more than one argument to a function, the arguments
are separated from each other by commas. The operation of the
printf( ) function is as follows. The first argument is a quoted string
that may contain either normal characters or forma! specifiers that
begin with the percent sign. Normal characters are simply displayed
as-is on the screen in the order in which they are encountered in the
string (reading left to right). A format speCifier, also called a format
code, informs printf( ) that a different type item is to be displayed. In
this case, the %d means that an integer is to be output in decimal
format. The value to he displayed is found in the second argument.
This value is then output to the screen at the point where the format
specifier is found in the string. To understand the relationship
between the normal characters and the format codcs, examine this
statcment:
printf (-This displays %d. too", 99);

Now the call to printf( ) displays This displays 99, too. The key
point is that the value associated with a format code is displayed at the
point where that format code is encountered in the string.
If you want to sped/)' a character value, the format speciffer is %c .
•
To specify a floating-point value, use %f. The %f works fpr both float
and double. As you will see, printf( ) has many more capabilities.
Keep in mind that the values matched with the format specifier
need not be constants; they may be variables, too.

1. TIle program shown here illustrates the three new concepts

introduced in this section. First, it declares a variable named
num. Second, it assigns this variable the value 100. Finally, it
uses printf( ) to display The value is 100 on the screen.
Examine this program closely:
'4
•
#include <stdio.h>

int main(void)
(
int num;

num = 100;
printi ("The value is %d ", num) ;

return 0;
)

The statement
i nt num ;

declares num to be an integer va n','''"

To dIsplay the value 01 Ilu,n, the program uses thIs statement.
pr1nt .c \ "The value is %d ', nu.q);

2. ThIs program creates variables of types char, float, ancl double;

assigns each ~val ue; and outputs these values to the screen.
fi nclude <stdio.h>

1n t ma1n(void)
(
c har ch
float f;
double d;

ch = 'X';
f = 100 . 123;
d = 123. 00 9;
printf("ch is %c ... , chI;
printf("f is %f, ., f);
printf("d is tEo, dl;

return 0:
)
 C RlNDAM£NTALS 15
~
,.. ,NPUT NUMBERS FROM TIlE KEYBOARD

1. Enter, compile, and run the example programs in this section.

2. Write a program that declares one integer variable called num.
Give this variable the value 1000 and then, using one printf( )
statement, display the value on the screen like this:
1000 is the value of num

NPUT NUMBERS FROM THE KEYBOARD

Although there are actually several ways to input numeric values from
I"
the keyboard, one of the easiest is to use another of C's standard
library functions called seanf( ). Although it possesses considerable
versatility, we will use it in this chapter to read only integers and
floating-point numbers entered from the keyboard.
To u~e seanft ) to read an integer value from the keyboard. call it
using the general form

scanfC'%d". &int-var-name);
where im-uar-l1ame is the name of the integer variable you wish to
receive he value. The first argument to scanf( ) is a string that
determines how the second argument will be treated. In this case, the
%d specifies that the second argument will be receivir g an integer
value entered in decimal format. This fragment. for example, reads an ·
integer entered from the keyboard.
int num;
scanf ("'d·. &num) i

The & preceding the variable name is essential to the operation

of scanf( ). Although a detal' ed explanation will have to waIt until
later. loosely . the & allows a function to place a value into one of
its arguments.

J
16 TEACH YOURSELF

'" C

It Is important to understand one key point: When you enter a

number at the keyboard, you are simply typing a string of digits. The
seanf( ) function waits until you have pressed ENTER before it
converts the string into the internal binary format used by .the
computer.
To read a floating-point number from the keyboard, call seanf( )
using the general form

scanfC'%f', &floal-var-name);

where jloat-var->1ame is the name of a variable that is declared as being

of type float. If you want to input to a double variable, use the %If
specifier.
"'otice that the format specifiers for seanf( ) are similar to those
used for printf( ) for the corresponding data types except that %If is
used to read a double. This is no coincidence-printf( ) and seanf( )
are complementary functions.

- EXAMPLE :

1. This program asks you to input an integer and a floating-point

number. It then displays the values you enter.
#include <stdio.h>

int main(void) •
{
int n\lln:
float f;

printf ("Enter an integer: ") i

scanf("%d", &numJ i

printf("Enter a floating point number: ");

scanf ("%f". &f);

printf ("%d •• num);

printf("%f", f) i/

return 0:
}
 C FUNDAMENTAlS 17
1.5 PERFORM CALCULAnONS USING ARfTHME77C EXPRESSIONS
"

EXERCISES

1. Enter, compile, and run the example program.

2. Write a program that inputs twO floating-point number5 (use type float)
and then displays theit sum.

RFORM CALCULA TlONS USING

ARITHMETIC EXPRESSIONS

In C, the expression plays a much more important role than it does in

most "ther programming languages. Part of the reason for this is that C
defines many more operators than do most other languages. An
'expression is a combination of operators a'n d operands. C expressions
follow the rules of algebra, so, for the most part, they \dB be familiar.
In this section we will look only at arithmetic expressions.
C defines these five arithmetic operators:

Operator Meaning

+ addition
subtraction
• multiplication
t division
% modulus

The +, -, I, and' operators may be used with any of the basic data
types. However, the % may be used with integer types only. The
modulus operator produces the remainder of an integer division.
This has no meaning when applied to floating-point types.
The - has two meanings. First, it is the subtraction operator.
Second, it can be used as a unary minus to reverse the sign of a
number. A unary operator uses only one operand.
An el'pression may appear on the right side of an assignment
statement. For example, this program fragment assigns the integer
variable answer the value of) 00' 31.
1.
.~
_YOURSElf
C

int answer;
answer = 100 • 31;

The 0, I, and % are higher in precedence than the + and the -.

However, you can use parentheses to alter the order of evaluation. For
example, this expression produces the value zero, .

10 - 2 0 5
but this one produces the value 40.

(10-2)05

A C expression may contain variables, constants, or both . For

example, assuming that answer and count are variables, this
expression is perfectly valid:
answer = count - 100;

Finally, you may use spaces liberally within an expression .

EXAMPLES

1. As stated earlier, the modulus operator returns the remainder of

an integer division. The remainder of 10 % 3 equals 1, for
example. This program shows the outcome of some integer
divisions and their remainders:
tinclude.<stdio.h>

int main(void)
{
print! ("'d", 5 / 2) ;
printf(" %dOl , 5'2) ;
printf(" %d· , 4/2) ;
printf(" %dOl , 4%2) ;

return 0; .
)

'This program displays 2 1 2 0 on the screen.

 1.5
CRlNDIlMENTALS
PfRFOIIMCALCUIATIONS USING AAITHMETIC EXPRESSiONS
.
1.

2. In long expressions, ,the use of par.entheses and spaces can add

clarity, even if they are not necessary. For example, examine
this expression:
count *nurn+88/val-19%count

Thisexpression produces the same result, but is much easier

to read:
(count * nurn) + (88 I val) - (19 " count)

3. This program computes the area of a rectangle; given its

dimensions. 'It fi~t prompts the user for the length and width of
the rectangle and. then displays the area.
*include <stdio.h>

int main(void)
(
int len, width;

printf(-Enter length: M);

scanf("'d", &lenl;
printf ("Enter width: .. );
scanf ("%d", &width);

printf ( .. Area is %d"", len • width);

return 0:

4: As stated earlier, the - can be used as a unary operator to

reverse the sign of its operand. To see how this works, try
. thffi program:
.include <stdio.h>

int main (void)

(
int i;

i = 10;
i = -i:
printf(-This is i: %d", i)i
22
'r'

printf("Enter number of Earth days: "):

scanf("%f", &e_daysJ i

/* now, compute Jovian years */

Lyears. = e_days I (365.0 • 12.0);

/* display the answer *f

printf{WEquivalent Jovian years: %f-, j-years);

return 0 i
)

Notice that comments can appear on the same line as other C

program statements.
Comments are often used to. help describe what the program
is doing. Although this program is.easy to understand even without
the comments, many programs are very difficult to understand
even with the liberal use of comments. For more complex
programs, the general approach is the same 'as used here:
simply describe the actions of the program. Also, . notice the
comment at the start of the program. In general, it is a good idea
to identify the purpose of a program auhe top of its source file.
2. You cannot place a comment inside the name of a function or
variable name. For example, this is an incorrect statement:
pri/* wrong * / ntf("this won't work");

EXERCISES ·

1. Go back and add comments to the programs developed in

previous sections.
2. Is this comment correcU
/ ** /

3. Is this comment correct?

/* printf(~this is a test"); */
 C FUNDAM£NTAlS 23
1.7 WRITE YOUR OWN RJNCnONS •

ITE YOUR OWN FUNCTIONS

Functions are the building blocks of C. So far, the programs you have
seen included only one function: maine ). Most real-world programs,
however, will contain mlmy functions. In this section you will begin to
learn how to write programs that contain multiple functions.
The general form of a (;-program that has multiple functions is
shown here:

r include header files here '/

r function prototypes here '/

int main(void)
{
r ... '/
}

ret-type f1(param-lisf)
{
r' ... */
}

ret-type f2(param-lisf)
(
r ... '/
)

ret-type fN(param-lisf)
{
r ...*/
}

Of course, you can call your functions by different names. Here,

ret-type specifies the type of data returned by the function. If a .
function does not return a value, then its return type should be void.
If a function does not use parameters, then its param-list should
contain the keyword void.
24 "TEACH YOURS[[F
... C

Notice the comment about prototypes. A function prototype nec1ares

a function before it is used and prior to its definition. A prototype
consists of a function's name, its return type, and its parameter Jist.
It is terminated by a semicolon. The compiler needs to know this ·
information in order for it to properly execute a call to the function .
. for example, given this simple function:
void myfunc(voidl
(
printf("This is a t.est.")":
}

Its prototype is
void myfunc(void);

The only function that does not need a prototype is maine) since
it is predefined by the C language .
Prototypes are an important part of C programming, but you will
. need to learn more about C before you can fully understand their
purpose and value .. for the next few chapters we will be using prototypes
without any further explanation. They will be included as needed in
all of the example programs shown in this book. You should also
include them in programs that you write. A full explanation of
prototypes is found in Chapter 7.
When a function is called, execution transfers to that function. ;Vhen
the end of that ,function is reached, execution returns to a point
immediately after the place at which the function was called. Put
differently, when a function ends, execution resumes at the point in
your program immediately following the call to the function. Any
function inside a program may call any other function within the same
program. Traditionally, maine) is not called by any other function ,
but there is no technical restriction to this effect.
In the examples that follow, you wi11learn to create the simplest
type of C functions: those that that do not return values and do not use
parameters. The skeletal form of such a function is shown here:
void FuncName(void) {
/* body of function here * /
}
 C FUNDAMENTAlS 25
1.7 KW1C YOUR OWN FUNCnONS
'"

Of course, the name of the function will vary. Because the function
doe. not return a value, its return rype is void. Because the function
does not have parameters, its parameter list is void.

EXAMPLES

I. The following program comains rwo functions: main() and func1(). Try to
determine what it displays on the screen before reading the description that
follows it.

/ * A program with t ....·o functions :It /

#include <stdio.h>

void funcl(void): /* prototype for funcl() *j

int main(void)

printf.("I H);
ftmcl() ;
printf("C."li

return 0;
)

void fuftcl(void)

printf("like ");
)

This program displays I like C. on the screen. Here is how it

works. In maine ), the first call to printf( ) executes, printing
the I. Next, Cuncl ( ) is called. This causes the printf( ) inside
CuneI ( ) to execute, displaying like. Since this is the only
statement inside funcI( ), the function returns. This causes
execution to resume inside maine ) and the C. is printed.
Notice that the statement that calls funcl ( ) ends with a
semicolon. (Remember a function call is a statement.)
26 TEACH YOURSELF

" C

A key point to understand about writing your own functions is

that when the closing curly brace is reached the function will
return, and execution resumes one line after the point at which
the fu nction was caJled.
Notice the prototype for funci ( ). As you can see, it consists
of its name, return type, and parameters list, but no body. It is
terminated by a semicolon.
2. This program prints 1 2 3 on the screen:
/* This program has three functions. */

#include <stdio.h> •

void funcl(void); /* prototypes */

void func2(void);

int main(void)
{
func2 () ;
printf{"3~) ;

return 0;
}

void func2(void)
{
funcl (J ;
printf("2 ~);

void funcl(void)
{
printf("! ");
}

I n this program, maine ) first caJls func2 ( ), which then calls

funcI() . Next, funcl() displays I and then returns to func2( ) ,
which prints 2 and then returns to maine ) , which prints 3.
 C FUNDAMENTALS 27
T
. 1.8 USE FUNcnONS TO RETURN VALUES

EXERCIIU
1. Enter, compile, and run the two example programs in this
section.
2. Write a program that contains at least two functions and prints
the message The summer soldier, the sunshine patriot.
3. Remove the prototype from the first example program and then
compile it. What happens?

E FUNCTIONS TO RETURN VALUES

In C, a function may return a value to the calling routine. For

example , another of C's standard library functions is sqrt( ), which
returns the square root of its argument. For your program to obtain
the return value, you must put the function on the right side of an
assignment .statement. For example, this program prints the square
root of 10:
#include <stdio.h>
#include <math.h> / * needed by sqrt() *1

int rnain (void )

(
double answer;

answer = sqrt(lO.O);
printf ( ~%f". answer);

return 0;
1

This program calls sqrt( ) and assigns its return value to answer.
Notice that sqrt() uses the MATH.H header file.
28 TEACH YOURSELF
... C

Actually, the assignment statement in the preceding program is not

technically necessary because sqrt( ) could simply be used as an
argument to printf( ), as shown here:
#include <stdio.h>
. #include <math.h> /* needed by sqrt() */

int main(void)
(
printf("%f",sqrt(lO.O));

return 0;
}

The reason this works is that C will automatically call sqrt( ) and
obtain its return value before calling printf( ). The return value then
becomes the second argument to printf( ). If this seems strange,
don't worry: you will understand this sort of situation better as you
learn more about C.
The sqrt( ) function requires a floating-point value for its
argument, and the value it returns is of type double. You must match
the type of value a function returns with the variable that the value
will be assigned to. As you learn more about C, you ",ill see why this is
important. It is also important that you match the types of a function's
arguments to the types it requires.
When writing your own functions, you can return a value to the
calling routine using the return statement. The return statement
takes the general form

return value;

where value is the yalue to be returned. For example, this program

prints 10 on the screen:
#include <stdio.h>

int func(voidl; / * prototype *1

int rnain{void)
(
int num;

nwn = func();
 C FUNDAMENTALS
1.8 USE FUNcnONS TO RETU~ .vALUES

printf(~%d·, num);

return 0;
r
int func(void)
{
return 10;
)

In this example, func( ) returns an integer value and its return type
is specified as into Although you can create functions that return any
type of data, functions that return values of type int are quite common.
Later in this book, you will see many examples of functions that return
other types. Functions that are declare' 3S void may not return values.
If a function does not explicitly specify a return type, it is assumed
to return an integer by default. For example, func( ) could have been
coded like this:
func(void)
{
return 10;
)

In this case, the int is implied. The use of the "default to int" rule is
- .
very common in older C code. However, recentlv there. has been a
move away from using the integer default. Whether this trend will
continue is unknown. In any event, to avoid misunderstandings, this
book will always explicitly specify into
, One important point: When the return statement is encountered,
the function returns immediately. No statements after it will be
executed. Thus, a return statement causes a function to return before
its closing curly brace is reached.
The value associated with the return statement need not be a
constant. It can be any valid C expression.
A return statement can also be used bv . itself, without a return
·value. This form of return looks like this:
return ;

It is used mostly by void function·s (i.e., functions that have a void

reUlrn type) to cause the function to return immediately, before the
function's closing curly brace is reached. While not recommended ,
30 TEACH YOURSElf

'" C

you can also use this form of return in functions that are supposed to
return values. However, doing so makes the returned value undefined.
There can be more than one return in a function. You will see
examples of this later in this book.
Even though a function returns a value, you don't necessarily have
to assign that value to anything. If the return value of a function is not
used, it·is lost, but no harm is done.

EXAMPLES

I. This program displays the square of a number entered from the

keyboard. The square is computed using the gecsqr( )
function. Its operation should be clear.
#include <stdio.h>

int get_sqr(void) i

int main(void)
(
int sqr;

sqr = get_sqr();
printf(nSquare: %d K
, sqr);

return 0;
}

int get_sqr(void)

int nurn~

printf("Enter a number: n);

scanf (" %d", &num);
return num*nurn; /* square the number */
}

2. As mentioned earlier, you can use return without specitying a

value. This allows a function to return before its closing curly
. brace is reached. For example, in tlte following program, the
line This is never printed. will not be displayed. .
 C FUNO_AlI 31
1.8 USE FUNcnONS TO RETURN VALUES
"
#include <stdio.h>

void funcl(void);

int main{void)
{
funcl {I;

return 0;
)

void funcl(void)
{
printf("This is printed. "I;
return; / * return with no value */
printf ("This is never printed.");

EXERCISES

1. Enter, compile, and run the example programs in this section.

2. Write a program that uses a function called convert( ), which
prompts she user for an amount in dollars and returns this value
converted into pounds. (Use an exchange rate of $2.00 per
pound.) Display the conversion.
3. What is wrong with this program?
#include <stdio.h>

int £1 (void) ;

int main(void)
{
double answer;

answer = fl();
printf("%f-, answer);

return 0;
32 TEACH YOURSELF
~
C

int fl(void)

return 100;
)

4. \'<Ihar is ,... rong with this function?

void func(void)
{
int i;

printf("Enter a number: ~);

scanf ("%od". &i);

return i;
)

E FUNCTION ARGUMENTS
As stated earlier, a function's argument is a value that is passed to the
fimction when the functiQn is ~alled. A function in C can have from
,ero to several arguments. (The upper limit is determined by the
compiler you are using, but the ANSI C standard specifies that a
function must be able to take at least 31 arguments.) For a function
to be able to take arguments, special variables to receive argument
values must be declared. These are called the formal parameters of the
function. The parameters are declared between the parentheses that
follow the function's name. For example, the function listed below
prints the sum of the two integer arguments used to call it.
void surn(int x. int y)
{
printf("%d· x + y);
)
 C FUNDAMENTAlS 33
1.9 uSE FUNCnON ARGUMENTS •
Each time sum( ) is called, it will sum the value passed to x with the
\'alue passed to y. Remember, howe"er, that 11: and y are simply the
function's operational variables, which receive the "alues you use
when calling the function. Consider the following short program,
which illustrates how to call sum( ) .
/* A simple progr~~ that demonstrates sum(). -,

~include <sLdio.h>

'loic sl;.m(ir.t x, i:-:t y);

int main(voidl
,.
sum(l. 2')):
surr.(9, f: ;
s'.lrniE!. 9':

prlr.:.f \ "\0..;::' >: - Yj ;

This program "'ilI print 21. 15. and 90 on the screen. When sum( )
is called, the \'alue of each argument is copied into its matching
parameter. That is, in the tirs, rail to sum( ) , 1 is copied illlO x and 20
. is copied illlO y, In the second call, 9 is copied into x and 6 into y , In
the third call. 81 is copied illlO 11: and 9 into y .
If you ha"e ne"er worked with a language that allows parameterized
functions. the preceding process may seem strange. Don't worry-as
you see more examples of C programs. the concept of arguments.
par~meters, and functions wi11 become clear.
Qt is Important to keep two terms straight. First, mgument refers to
Ihe ",lillc that is passed to d function. The variable that rcc.ei'·cs the
,'alul' ul the argument inside the function tS the tomw/ parame,e,' of
the functiun Functions that take arguments are called parame,enzed
34 TEACH YOURSELF
~
C

("nctiolls. Remember, if a variable is used as an argument to a function, it

has nothing to do with the formal parameter that receives its value.
In C functions, arguments are always separated by commas. In this
book. the term ""gwl1e>1t l,st will refer to comma-separated arguments.
All function parameters are declared in a fashion similar to that
used by sum( ). You must specify the type and name of each
parameter and. if there is more than one parameter, you must use a
comma to separate them. functions that do not have parameters
should use the keyword void in their parameter Ibt.

EXAMPLES

1. An argument to a function ran consist of an expression For

eXdmple. it is perfectly "alid to call sum( ) as shown here:

'} This progran1 uses {he outchar( ) fUllction tv output characters

[(J the SLreen. The program prints ABC.

#include <stdio.h>

veid out.char (char ch);

•
int !"'1ain(void)

outcl)ar( 'h');
outchar( 'B');
outchar( 'e'):

return 0;

\:oid outchar (char cr.)

printf("%c", ch);
 35
1.10
C FUNDAMENTALS

REMEMBER THE C KDWORDS

..,

EXERCISES

1. Write a program that uses a function called outnum( ) that

takes one integer argument and displays it on the screen.
2. \Vhat is wrong with t11is program?

#include <stoio.h>

void sqr_it{int num);

int main(void}
(

rett.!rn 0;
)

(
p::-if'.cf ("%d", num * nu!':1);
,\

1.10 EMEMBER
, THE C KEYWORDS
Before concluding this chapter, you should familiarize yourself with
the keywords that make up the C langu,l ge A0!SI C standard has 32
keyu'ords that may not be used as variable or function names. These
words, combined with the formal C syntax, form the C programming
language. They are listed in Table 1-2.
Many C compilers have added several additional keywords that are
used to take better advantage of the environment in which the
compiler is used, and that give support for interlanguage program-
ming, interrupts, and memory organization. Some commonly used
extended keywords are shown in Table 1-3.
The 10\\'ercase lettering of the keywords is significant. C requires
that all kevwords be in lowercase form. For example, RETURN will
not be recognized as the kevword return . Also, no kevword mav be
• y -

used as a \'ariable or function name.

 TEACH YOURSElF
36
" C

---
-- double Int struct
auto switch
else long
break typedef
enum register
case union
extern return
char unsigned
float short
canst void
for signed
continue volatile
goto sizeof
default
static while
do if
- - - -- - ~-~----

- -- --
The 32 Key.vords as Deflfled by the ANSI C Standard ...
TIUIlE 1-2

ds es
asm cs
far huge
_55 cdec!
near pascal
Interrupt

SOfTIe Common C Extended Keywords •

Check
1. The moon's gra\'1t\' Pi about 17 percent of Earth's Writp a
progranl {hat allm\'s you to enter your weight and cOlnpuIes
your effecti\'e \\'C1ghr on the moon .
2. \\'hat is \\Tong \\' ith thi!-. program fragnlent'?
? L~~S inputs a nUITDer
scar.f ( ~%d", &m:m ) ;

3. There are 8 ounces in a cup. \"rite a progranl that converts

ounces to ClipS . L:se a function called o_to_c( ) to perform the
converSIOn . Calln \\'ith the number of ounces and haYe it feInT
the number of ClipS.
4. \,'llar arc the fh'C' bask dJtd types 111 C?
 "".
'f'
1.10 REMEMBER THE C KEYWORDS

5. What is wrong with each of these variable names?

a) short-fall
b) Sbalance
c) last + name
d) 9times
 •

2
Introducing CIS
Program Control
Statements

chapter objectlv..

2.1 Become familiar with the if

2.2 Add the else

2.3 Create blocks of code

2.4 Use the lor loop

2.5 Substitute C's increment and decrement
operators

2.6 Expand prinU( ) 's capabilities

2.7 Program with C's relatIOnal and logical

operators

39
"
40
., TEACH YOURSELF

'" thiS ehdpter you will learn about t\\'o of C's most important
program control statcments: if and fOI". In gencral, program
c011trol statements determine your progr<l1l1's flow of execution .
.As such, they.form the backbone of your prugrmns. In addition to
these, yuu ,,·ill also learn about blocks of cocle. the relational and
logical operators, anumore about the printf( ) function.

\ Review
Skills Check
Before proceeding, you should he able to correctly ans"·cr these
qUCStlOllS and do these exercises:
,
1. ;\011 C programs are c;olnposcd of one or more functions. \\'hat is
the name of the function that all programs must llave? rurtl1cr,
\"har special purpose doe~ it perform:
') The printf( ) fUllctlon IS lIsed to outpUt information to the
screen WritC' J program rhdt tlisplJ~' s This is the nurnbcr 100.
(Output the 100 as a number, not as a ,,!·ing.)
3. Header tilt,!'> (untain infonn,l1ion used by tll!" srallcbrd library
fUIlUlOIlS. Hm,· do you tell the cnrnpilcr [0 incllldl~ onc ill your
' program-:' Gi\"e all example.
4. C supports five basic types of data. Name the Ill .
J. Which (~I these vdriable nameS are ill\"alid in C~
a. count
h. 123Lount
•
c. $test
d. This_is_aJong_nanle
e. new-',-onJ
6. What is seanfe ) used for"
7. Write a program that inputs an integer from the kcvboard anJ
displays its square.
8 . How arC' comments entered into a C progranl'? Give an eX31nple
9. 1{o,," docs a function return a value to the routine that called it?
 INTRODUCING C'S PROGRAM CONTROL STATEMENTS 41
21 BECOME FAMILIAR WITH THE if
...

10. A function called Myfunc( ) has these three parameters: an int

called count, a float called balance, and a char called ch. The
function does not return a value. Show how this function is
,prototyped,

ECOME FAMILIAR , WITH THE if

The if statement is one of C's selection srmemenrs (sometimes c.alled

condltiorml'Siarcmenrs). Its operation is governed by the outcome of a
conditional test that evaluates to either true or false. Sinlply put,
selection Statelnents mak,e dcci~ion~ based upon the outc0111e uf some
condition.
In its simplest form, the if statement allows vour program to
conditiOJially execute a statement, This form of the if is shown here:

if(expression) statement:

Tlw .expression lnay he any valid C expres~ion. If the expression

('va!uates as true, the statement "'ill be, executed. If it does not, the
statement is bypassed, and the line of code fulluwing the if is executed.
In C, an" explession is_tru~ if it evaluates to Clny nonzero value. If it
evaluates to zero, it is false. The 'statetnent that follows an if is usually
reterred to as the larger of the if statement.
Commonly, the expression inside the if compares one value with
another using a','c/atiOJwl operator. Althougll~'Ou. will learn about all
the relational operators later in thIS chapter, th;'ee are introduced here
so that we can create some example programs, A relational operator
tests how one value relates to another. For example, to see if one value
is greater than another, C uses the> relational operator. The outcome
of this comparison is either true or false, For example, 10 > 9 is true,
but 9 > ]0 is false, Therefore, the following if will cause the message
true to be displayed,
i::(lO> 9) printf("true");

However, because the expression in the following statement is false,

[he if does not execute its target statement.

if(5 > 9) printf("this will not print");

42 TEACH YOURSELf

'" C

C uses < as its less than opemtor. For example, 10 < 11 is true . To
test for equality, C provides the = = operator. (There can be no space
between the two equal signs.) Therefore, 10 == 10 is true, but 10 ==
11 is not.
Of course, the expression inside the if may involve variables. for
example. the toll owing program tells whether an integer entered from
the keyboard is negative or non-negative .
#include <stdio.h>

int main(void)
(
int num;

printf("E:l.ter an integer: ");

scanf("%d", &r.\..:m);

iftnum< C) printfl"Number is negative." ) ;

if(num> -11 pr::'ntf("rJumber is .. on-negative.");

return 0;
i

Remember, in C. true 'is al'Y nen'zero value-and·false is zero.

Therefore, it is perfectly valid to have an if statement such as the one
shown here: . ..... .

if(count+l) pri:l.tf(-Not Zero");

EXAMPLES

1. This program forms the basis for an addition drill. It displays

two numbers and asks the user what the answer is . The
program then tells the user if the ans,,'er is right or wrong.
#include <stdio.h>

int main{void)
(
int answer;
 · ;,
INTRODUCING C'S PROGRAM CONTROL STATEMENTS 43
2.' BECOME FAMIUAR WITH THE If
...

printf("What is 10 + 14? ");

scanf("%d", &answer);
if(answer == 10 ... 14) printf("Right!");

return 0;

2. This program converts either feet to meters or meters to feet,

depending upon what the user requests.
;include <stdio.h>

inc main(voidl

float nurn;
int choice:

princf("Enter value: "1;

scanf("~f", &:1um):

print f (" 1: Feet to ~eter5, 2: !1eters to Feet. ");

printf ("EntEr cr.oic€:: ");
scanf ("%d", &choice);

if/choice -- 1) printf{"%f", num 13.28):

if (choice -- 2) printf("%f", num'" 3.281:

return 0;
)

EXERCISES

1. Which of these expressions are true?

a. 0
b. J
c. 10 • 9 < 90
d. 1== 1
e. -1
44 T<ACH YOURS£Lf

~ C'

2. Write a program that asks the user for an integer and then tells
the user if that number is even or odd. (Hint, use Cs modulus
operator' %.) .

DO THE.else

You can add an else statement to the if. When this is done, the if
statement looks like this:

ifCexpression) statement1:
else statement2;
If the expressi~n is true, then the target of the if will execute, and the
else portion will be skipped. However, if the expression is false, then
.'
the target of the if is bypassed, and the target of the else will execute.
Untl,'r no circumstances will both statements execute. Thus, the
additlon of the else-provides a two-way decision path.

EXAMPLES

I. YOll can ,lise the else to create more efficient code in some
cases. For example, here the else is used in place of a second if
in the program from the preceding section, which determines
whether a number is negative or non·negative.
~include <stdio.h>

int main(void)
{
int num;

printf(ftEnter an integer: ");

scanf("%d", &nurn);

if(num < 0) printf("Nurr~er is negative,");

else printf{"Number is non-negative,-);
 INTAODtlCING C'S _ CONTROl. SlIlTlMlMTS . 45
2.2 ADD THE else

return 0;
)

Recall that the original version of this program explic.itly t,,,ted

,for non·ncgative numhers by cotllparing nurn to -1 using J
second i,f statement. But since there are only tWO possihilitics-
Dum is either negative Of non·ncgativc-there is no reason for
tllis second test. Because of the way a C compiler generates
code, the else requires far fewer l1lachil1t.~ instructions th<ll1 an
additional jf and is, therefore, more "ftki('nt.
2. This program prompts the user for t"·o numhers. divides the
first hy the second. and dispbys the result. How"ver. division h\
·zero is undctlncd, so the program uses an if and an else
statement to pf(~vent di\'ision by zero from occurring.
#include <stdio.h>

pr:':1tf("Enter first number: "'.

scanf ("lid", &numl):

printf("Enter second number: "J;

scanf("'d", &num2);

::(nurn...2 == 0) printf{"Cannot divide by ::1210,");

else printf(HAnswer is: %d.", numl / num2);

return 0;
)

EXERCISES

1. Write a program that requests two numhers and thell displays

either their sum or product, depending on what the user selects.
TEACH YOURSELF

2. Rewrite Exercise 2 from Section 2.1 so that it uses an else

statement.

EA TE BL OCKS OF CODE

In C, you can link two or more statements together. This is called a

block of COli<' or a code block To create a block of code, you surround
the statements in the block with opening and closing curly braces.
Once this is d()n{~,
the statements form one logical unit, H'hich nlay
be used anywhere [hat a single statement may.
For example, the general form of the if using blocks of code is

if[express;on) (
statement/:
statement2:

statement N:
)
else (
statement/;
statement2,~

statement N;
)

If the expression evaluates to true, then all the statements in the block
of code associated with the if will be executed. If the expression is false,
then all the statements in the else block will be executed. (Remember,
the else is optional and need not be present.) For example, this
fragment prints the message This is an example of a code block,
if the tlscr cntr:rs any positi\'f~ l111l11her. -'
 INTRODUCING C'S PROGRAM CONTROL STATEMENTS 47
..,
23 CREATEBLOCKSOFCODE

scanf ( "%d" I &num);

if{num> 0) {
printf ("This is ");
printf ("an example of ");
printf("a code block.~);
)

Keep in mind that a block of code represents one indivisible logical

unit. This means that under no circumstances could one of the
printf( ) statements in this fragment execute without the others
also executing.
In the example shown, the statt!I11Cnts that appear within the hlock
of code are indented. Although C does not care wl",re a statement
appears on a line, it is common practice to indent one level at the
start of a block. Indenting makes the structure of a program easier
to understand. Also, the placement of the curl\' hraces is arhitran·.
However, the v,ray they are shown ill the example is a COn1m0l1
. method and will be uscu bv the examples in this book
In C, as you \\'i11 see, YPu-c~ln_l:'I"se...a blork of cock any\\·hcrc you can
use p .single statement.

EXAMPLES
•
I. This program is an improved version of the feet-la-meters,
meters-to-feet conversion program. Notice hO\v the use of code
blocks allows the program to prompt specifically for each unit.
#include <stdio.h>
•
int main (void)
{
float num;
int choice;

printf (" 1: Feet to Meters, 2: Meters to Feet . ");

printf ("Enter choice: "):
scanf ("%d", &choice) i

if(choice == 1) {
printf ("EntEtr number of feet: ");
48 W\CH YOURSELF

'" C

scanf(~%f·, &num);
printf(nMeters: %f", num / 3.28);
}
else {
printf(~Enter number of meters: .);
scanf(-%f", &num);
printf("Feet: %f~, num· 3.29);

return 0;

2. Using code blocks, we can improve the addition drill program

so that it also prints the correct answer when the user makes
a mistake.
~include <stdio.h>

int main{void)

.
p=ir,tf ("vJhat is . 10. + :4? .. ) ;
scan: ("%d", &answer) i

if (answer == 10+14) printf ("Right!") ;

else (
printf("Sorry. you're wrong . • );

} .
printf{"The ans\o..'er is 24 .... );

return ·0 ;
•
)

This 'example illuStrates an important point: it is not necessary

for targets of both the' if and the else statements to be blocks of
code . In this case, the target of if is a single statement, while the
target of else is a block. Remember, you are free to use either a
single statement or a code block at either place.
 INTROOUCINO C'S PROGRAM CONTROL STATtMENTS 49
y
2.4 USE ntHor LOOP

). Write a program that either adds or subtracts two integers. First,

prompt the user to choose an operation; then prompt for the
two numbers and display the result.
2. Is this fragment correct?
if(count < 100)
printf(-Number is less than 100.~);
printf(-Its square is %d.-, count" co"..mt);
)

SET H,E for L 00 P

The for loop is one of C's three loop statements, It allows one or more
statements to be repeated. lfyoll have programmed in any other
computer language, sllch as BASIC or Pascal, you will be pleased to
learn that the for behaves mllch like its equivalent in other languages.
The for loop is considered by many C programmers to be it<; most
flexible loop. Although the for loop allows a large number of
variations, we will examine only its most common form in this section.
The for loop is used to repeat a statement or block of statements a
specified number of times. Its general form for repeating a single
statement is shown here,

for(initialization: conditional-test: increment) statement:

The initialization section is used to give an initial value to the variable

that controls the loop. This variable is usually referred to as the
loop-control variable. The initialization section is executed only once,
before the loop begins. The conditional-test portion of the loop tests the
!oop-control variable against a target value. )fthe conditional test
50 TEACH YOURSW

• C

evaluates true, the loop repeats. If it is false, the loop stops, and
program execution picks up with the next line of code that follows the
loop. The conditional test is performed at the start or top of the loop
each time the loop is repeated. The increment portion of the for is
executed at the bottom of the loop. That is, the increment portion is
executed after the statement or block that forms its body has been
executed. The purpose of the increment portion is to increase (or
decrease) the loop-control value by a certain amount.
As a simple first example, this program uses a for loop to print the
numbers 1 through 10 on the screen.
#include <stdio.h>

int main(void)
(
int nurn;

for(num=li num<11; num=num+l) printf(-%d· nurn);

prin~f ( "terminating,") :
•
return 0;
}

This program produces the following output:

1 2 3 4 5 6 7 8 9 10 lenninating
The program "'arks like this: First, the loop control variable Dum is
initialized to I. Next, the expression num < 11 i$ evaluated. Since it
is true, the for loop begins running. After the number is printed, Jnlm
is incremented by one and the conditional test is evaluated again. This
process continues u!,til num equal' 11. When this happens, the for
loop stops, and tenninating is displayed. Keep in mind that (he
initialization portion of the for loop is only executed once, when
the loop is first entered.
As stated earlier, the conditional test is performed at the stan of
each iteration. This means that if the test is false to begin with, the
loop will not execute even once. For example, this program only
displays terminating because num is initialized to 11, ca~ing the
. conditional test to fail.
 1N11I00UCING C'S PROGRAM CONTROL STATtMENTS. &1
2.4 USE THEfor LOOP
...
.

#include <stdio. h >

int main{void)
{
int num;

* this loop will not execute * /

for{num=11; num<ll; num=nu!Tl+l) p rin tf("%d" n um);

printf ("terminating");

return 0;
J

To re peat several statements, use a block of code as the target of the

for loop, For example, thi~ program computes the product and sum of
the numbers from-} to 5:
.inc l ud e <stdi o. h>

int ma in( vo i d)

i nt num, sum , pr o d ;

sum = 0;
prod = 1;
for (num=l; num<6; num=num+l) (
sum = sum + num;
,
prod = prod * num;
)
printf (·product and sum.: %d %d- .. prod. sum);

return 0;
J

A for loop can run negatively. For example. this fragment

decrements the loop-control variable.
for (num=20; num>O; num=num-l) ...
\-
,
_.
Further. the loolKOntrol vaj"iable may be incremented or decremented
by more than one, For eXarl\Ple. this program counts to 100 by fives:
52 TIACH YOURSELf

• C

#include <stdio.h>

int main (void)

(
int i;

for(i=O; i<101; i=i+5 printf ( -td i);

return 0;
)

EXAMPLES

I. The addition-drill program created earlier can be enhanced

using a fOT loop. The version shown here asks for the sums of
the numbers between 1 and 10 That IS. It ash tor I + 1 [hen 2
+ 2. and so all. This program would be useful to a first grade,
who is learning to add.
~include <stdio.h>

int main {void}

(
int answer, count,

for(cgunt=l; count<11; count=count~l)' {

printf{-What is %d + %d? -, count. count};
scant (-%d-, &answer) ~

if(answer == count+count) printf{-Right! .);

else {
printf(-Sorry. you're wrong . • );
printfC-The answer is 'd . • count+count);
)
)

return 0:
)

Notice that this program has an if sta tCmenL& part of the for
block. Notice further that the target of elae is a block of code.
This is perfectly valid. In C. a code block may contain
 IJI11IOI)UCING C'S _ CON1ROI. STATtMEHTS 53
2.4 USE THE for LOOP
"

statements that create other code blocks. Notice how the

indentation adds clarity to the structure of the program.
2. We can use a for loop to create a program that determines if a
number is prime. The fol1owing program asks the user to enter
a number and then checks to see if it has any factors.
I * Prime number tester .... ,'
#include <stdio.h>

int main(void)
(
int num, L is-prime;

printf("Enter che number to test: "J;

scanf ( "%d". &num);

... now test for factors •

is-prime = 1;
for ( i=2; i<=:1um : 2; i=i+1)
if I (nl.!~%iJ==Ol is-prirr.e = 0;

iffis-prime==l) printf("The number is prime."):

else printf("The number is not prime.~):

return 0;
)

I. Create a program that prints the numbers from 1 to 100.

2. Write a program that pIints the numbers between 17 and 100 •
that can be evenly divided by 17..
3. Write a program similar to the prime-number tester, except that
it displays al1 the factors of a number entered by the user. For
example, if the user entered 8, it would respond with Z and 4_

I
TfACH YOURSELF
C

UBSTITUTE C'S INCREMENT AND

DECREMENT OPERATORS
When you learned about the for in the preceding section, the increment
portion of the loop looked more or less like the one shown here:
for{num=O; num<some_valuei num=num+l) ...

Although not incorrect, you will almost never see a statement like
num = num + 1 in professionally written C programs because C
provides a special operator that increments a variable by one. The
iJ1C>"cmcl1t opacao,- is ++ (two pluses with no intervening space). Using
the increment operator, you can change this line of code:
i=i+l;

into this:
i++;

111erefore, the for shown earlier will normally be written like this:
for (num=O; n~~<some_value; num++J ...

In a similar fashion, to decrease a variable by one, you can use C's

decrement operator. - -. (There must be no space between the two
minus signs.) Therefore,
count ; count - 1:

can be rewritten as
count--,.

Aside from saving you a little typing effort, the reason you will want
to use the increment and decrement operators is that, for I1106t C
compilers, they will be faster than the equivalent assignment
statements. The reason for this difference is that the C compiler can
often avoid separate load-and-store machine-language instructions and
substitute a single increment or decrement Instruction in the
executable version of a program.
The increment and decrement operators do not need to follow the
variable; they can precede it. Although the effect on the variable is the
 2..
INIROOUCING C'S _ CONTROL STA1IMUITS

SJJBSTTTUTE C'S INCREMENT AND DECREMENT OPfIIATORS

.,
5&

same, the position of the operator does affect when the operation is
performed. To see how, examine this program:
#include <stdio . h>

int main(void)
(
int i. j;

i = 10,
j = i+.;

/ * this will print 11 10 . /

pr int f ( • i and j: %d %d", i. j);
•
return 0;
)

Don't let the j = i++ statement trouble you. The increment

operator may bt! used as part of any valid C expression. This statement -
works like this. First, the current value of i is assigned to j. Then i is
incremented , This is why j has the value 10, not 11. When the
increment or decrement operator follows the variable, the operation
is performed after its value has been obtained for use in the
expression. Therefore, assuming that max has the value 1, an
expression such as this:
count = 10 • max++;
•
assigns the value 10 to count and increa&es mu by one.
'If the variable is preceded by the increment or decrement operator,
the operation is performed first, and then the value of the variable is
obtained for use in the expressiol\. For example. rewriting the
previous program as ililows causes j to be 11,
tinclude <stdio.h>

in!: main (void)

{
int' i. j;

i = 10;
j = ... i;
.
&8 1IACII YOUR8(JS
C

j- this will print 11 11 */

printfl"i and j: %d %d", i, j);

return 0;
)

If you are simply using the increment or decrement operators to

replace equivalent assignment statements, it doesn't matter if the
operator precedes or follows the variable. This is a matter of your own
personal style.

EXAMPLES

I. Here is the addition drill program developed in Section 2. It has

been rewritten using the increment operator .
• include <stdio.h>

int mainlvoidl
{
int enswer. count;

for (count=!; count<11; count++) (

printf (-What is ltd . , %d? ., count, count);
scanf c.. %d·. &answer);

if(answer == count+count) printf(-Right! .);

else (
printf{-Sorry. you're wrong . • );
printf(-The answer is %d . • , count+count);
I
I

return 0;
I

2, This program illustrates the use of the increment and

decrement operators:
'include <stdio.h>

int main (void )

 INTROOUCING C'S PROOIIM' CONTROL STATEMENTS 57
25 SUBSTTTlffE COS INCREMENT AND DECREMENT OPERATORS
...

.-
{
~ ..... ,-
" int ii

i = 0;

i++;
printf (" %d " i) ; j' prints 1 ' j
i--·,
printf ("'d " i) ; j' prints 0 ' j

return 0,
J

-.- .., -

I. Rewrite the answer to the for loop exercises in the previous

section so that they use the increment or decrement operators.
2. Change all appropriate assignment state ments in this program
to increment or decrement statements.
#include <stdio.h>

int main(vcid)
(
int a,~ bi

a = 1;

a = a + 1,

b = a,

b = b - 1,

printf(-'d %d", a, b);

return 0;
)
60 TEACH YOURSELF

'" C

return 0;
)

2. You can enter any special character by specifYing it as an octal

or hexadecimal value following the backslash. The octal number
system is based on 8 and uses the digits 0 through 7. In octal,
the number lOis the same as 8 in decimal. The hexadecimal
number system is based on 16 and uses the digits 0 through 9
plus the letters 'A' through 'F', which stand for 10,11,12,13,
14, and 15. For example, the hexadecimal number 10 is 16 in
decimal. When specifYing a character in hexadecimal, you must
follow the backslash with an 'x', followed by the number.
The ASCI! character set is defined from 0 to 127. However,
many computers, including most PCs, use the values 128 to 255
for special and graphics characters. If your computer supports
these extra characters, the following program will display a few
of them on the screen.
#include <stdio.h>

int main(void)
{
printfl"\xAO \xAl \xA2 \xA3");

return 0;

3. The \ n newline character does not have to go at the end of the

string that is being output by printf{ ); it can go anywhere in
the string. Further, there can be as many newline characters i~
a string as you desire. The point is that there is no connection
between a newline and the end of a string. For example, this
program:
#include <stdio.h>

int main(void)
•
{
printf(Mone\ntwo\nthree\nfour-) ;

return 0;
)
 INTROOUCING C'S PROGRAM CONTROL STATEMENTS 61
27 PROGRAM WTTH C'S RELAnONAL AND LOGICAL OPERA TORS
"

displays

one
two
three
four
on the screen.

EXERCISES

1. Write a program that outputs a table of numbers. Each line in

the table contains three entries: the number, its square, and its
cube. Begin with 1 and end with 10. Also, use a for loop to
generate the numbers.
2. Write a program that prompts the user for an integer value.
Next, using a for loop, make it count down from this value to 0,
displaying each number .on its own line. When it reaches 0, have
it sound the bell.
3. Experircent on your own with the backslash codes.

OGRAM WITH C'S RELA TlONAL AND

LOGICAL OPERATORS

The C language contains a rich set of operators. In this section you will
learn about C's relational and logical operators. As you saw earlier, the
relational operators compare two values and return a true or false
result based upon that comparison. The logical operators connect
together truelfalse results. These operators are shown in Table 2-2
and Table 2-3.
62 TEACH YOURSELf
... C

Operator Action

> Greater than

>= Greater than or equal
< Less than
<= Less than or equal
Equal
!= Not equal

TABLE 2-2 RelaUonal Operators ...

The logical operators are used to support the basic logical operations
of AND, OR, and NOT according to this truth table. The table uses 1
for true and 0 for false.

p q p&&q pliq !p

o o o o 1
o 1 o 1 1

1 1 1 1 o
1 o o 1 o
•
The relational and logical operators are both lower in precedence
than the arithmetic operators. This means that an expression like
10 + count > a + 12

Operator Action

&& AND
II OR
NOT

TABLE 2-3 Logical Operators ...

 INTRODUCING C'S ~OGRAM CONTROL STATtMEHTS 63
T
27 PROGRAM Vt'ITH COS RELAnoNAL AND LOGICAL OPERATORS

is evaluated as if it were written

(10 + count) > (a + 12)

You may hnk any number ofralational operations together using

logical operators. For example, this expression joins three relational
operations.
var > max 11 ! (max==lOO) && a <= item

The table below shows the relative precedence of the relational and
logical operators.

Highest
> >= < <=
== !=
&&
Lowest II
There is one Important fact to remember about the values produced .
by the relational and logical operators: the result is either 0 or I. Even
though C defines true as any nonzero value, the relational and logical
operators always produce the value 1 for true. Your programs may
make use of this fact.
You can use the relational and logical operators in both the if and
for statementS'. For example, the following statement reports when
both a and b are positive:
if(a>O && b>O) printf(~Both are positive.·);

, EXAMPLES

I. In professionally written C code, it is uncommon to find a

statement like this:
ii(count 1= 0) ...

The reason is that in C, true is any nonzero value and false is zero.
Therefore, the preceding statement is generally written as this:
if(count} .. .
TEACH YOURSELF
C

Further, statements like this:

if{count == 0) ...

are generally written as:

if (!count) ...

The expression Icount is true only if count is zero.

2. It is important to remember that the outcome of a relational or
logical operation is 0 when false and 1 when true. For example,
the following program requests two integers, then displays the
outcome of each relational and logical operation when applied
to them. In all cases, the result will be a 0 or a 1.
#include <stdio.h>

int main(void)

int i, j;

printf("Enter first number: ~);

scanf("%d", &i);
printf("Enter second number: M) i
scanf ( " %d" , &j I:

/ * relational operations */
printf(ni < j td\n-, i < j);
printf(-i <= j %d\n", i <= j);
printf"("i == J %d\n", i == j);
printf("i > j %d\n", i > j);
printf("i >= j %d\n", i >= j);

/ * logical opera~ions */
printf("i && j %d\n", i && j);
printtl"i II j %d\n", i II jl;
printf ( ~!i !J %d %d\n", !i, !j);

return 0;
)

3. C does not define an exclusive-OR (XOR) logical operator.

However, it is easy to create a function that performs the
operation. The XOR operation uses this truth table:
 INTRODUCING C'S PROGRAM CONTROL SY"TEMENTS 65
V PROGRAM WITH C'S RELATIONAL AND LOGICAL OPERATORS ""

p q XOR
o o o
o 1 1

1 o 1
1 1 o
That is, the XOR operation produces a true result when one and
only one operand is true. The following function uses the &&
and II operators to construct an XOR operation. It compares the
values of its two arguments and returns the outcome of an XOR
operation.
int xor(int a, int b)
(
return (a II b) && ! (a && b);
)

The following program uses this function. It displays the result

of an AND, OR, and XOR on the values you enter.
, . This program demonstrates the xor() function . • /
#include <stdio.h>

int xor(int a, int b);

int main(void)
(
int p. q;

printf(-enter P (0 or 1): .):

scanf (-'d-, &p);
printf (-enter Q (0 or 1·): .);
scanf ("%d", &q);
printf ("P AND Q, %d\n" , p && q);
printf ("P OR Q, %d\n", p II q);
printf("P XOR Q, %d\n", xor(p, q»;

return 0;
)

int xor(int a, int b)

6& TEACH YOURSElf

" C

(
.return la II b) && ! (a && b);
)

I. What does this loop do?

for(x=O; x<100; x++) printf(~%d· x);

2.. Is this expression true?

! 110==9)

3. Do these two expressions evaluate to the same outcome?

a.O&&1111
b. 0 && (I III)
4. On your own, experiment with the relational and logical
operators.

1. Write a program that plays a computerized form of the 'guess

the magic number' game. It works like this: The player has ten
nies to guess the magic l1umber. If the number entered is the
value you have selected for your magic number, have the
program print the message 'RIGHT!' and then terminate.
Othem"ise, have the program report whether the guess was
high or low and then let the player enter another number. This
process goes on until the player guesses the number or the· ten
tries have been used up. For fun, you might want to report the
number of tries it takes to guess the number.
• INTRODUCING COS PROGRAM CONTROL STATtMENTS 67
21 PROGRAM WITH C'S RELIt nONAL AND LOGICAL OPERA TORS

2. Write a program that computes the square footage of a house

given the dimensions of each room. Have the program ask the
user how many rooms are in the house and then request the
dimensions of each room. Display the resulting total square
footage.
3. What are the. increment and decrement operators and what
do they do?
4. Create an improved addition-drill program that keeps track of
the number of right and wrong answers and displays them when
the program ends.
5. Write a program that prints the numbers 1 to 100 using 5
columns. Have each number separated from the next by a tab.
3
A10re C Progran7
Control Staten7ents

chapter objectives

3.1 Input characters

3.2 Nest if statements

3.3 Examine for loop variations

3.4 Understand C's while loop

3.5 Use the do loop

3.6 Create nested loops

3.7 Use break to exit a loop

3.8 Know when to use the continue statement
3.9 Select among alternatives with the switch
statement

3.10 Understand the goto statement

69
•
70 TEACH YOURSELF

" C

--HIS chapter continues the discussion of C's program control

statements. Before doing so, however, the chapter begins
by explaining how to read characters from the keyboard.
Although you know how to input numbers, it is now time
for you to know how to input individual characters because
several examples in this chapter will make use of them. Next. the
chapter finishes the discussion of the if and for statements. Then it
presents C's two other loop statements, the while and do. Next you
"'ill learn about nested loops and two more of C's control statements,
the break and continue. This chapter also covers C's other selection
statement. the switch. It ends with a short discussion of C's
unconditional jump statement, goto.

Before proceeding. you should be able to answer these questions

and perform these exercises:
I. What are C's relational and logical operators?
2. What is a block of code? Ho\\' do you make one?
3. How do you output a newline using printf( )?
4. Write a program that prints the numbers -100 to 100.
5. Write a program that prints 5 different proverbs. The program
prompts the user for the number of the proverb to print and
then displays it. (Use any proverbs you like.)
6. How can this statement be rewritten?
count = count ~ 1;

7. What values are true in C? What values are false?

NPUT CHARACTERS

Although numbers are important, your programs will also need to read
characters from the keyboard. In C you can do this in a variety of
ways. Unfortunately. this conceptually simple task is complicated by
 MORE C PROGRAM CONTROL STATEMENTS 71
3.1 INPUT CHARACTERS
"

some baggage left over from the origins of C. However, let's begin with
the traditional way characters are read from the keyboard. Later you
will learn an alternative.
C defines a function called getchar( ), which returns a single
character typed on the keyboard. When called, the function waits for a
key to be pressed. Then getchar( ) echoes the keystroke to the screen
and returns the value of the key to the caller. The getchar( ) function
is defined by the ANSI C standard and requires the header file STDlO.H.
This program illustrates its use by reading a character and then telling
you what it received. (Remember, to display a character, use the %c
printf( ) format specifier.)
#include <stdio.h>

int main(void)
(
char chI

• ch = getchar(); /* read a char * /

printf(~ you typed: %e", ch);

return 0;
)

If you try this program, it may behave differently than you

expected. The trouble is this: in many C compilers, getchar( ) is
implemented in such a way that it line buffers input. That is, it does not
immediately return as soon as you have pressed a key, but waits until
you have entered an entire line, which may include several other
characters. This means that even though it will read and return only
one character, getchar( ) waits until you enter a carriage return (i.e.,
press ENTER) before doing so. When getchar( ) returns, it' will return
the first character you typed. However, any other characters that you
entered, including the carriage return, will still be in the input buffer.
These characters will be consumed by subsequent input requests, such
as through calls to scanf( ). In some circumstances, this can lead to
trouble. This situation is examined more closely in Chapter 8. For
now, just be aware that getchar( ) may behave differently than your
intuition would suggest. Of course, the programs shown in this book
behave properly.
72 TEACH YOURSELF

• C

The reason that getchar( ) works the way it does is that the version
of UNIX for which C was developed line-buffered input. When C
compilers were created for other interactive environments, developers
had to decide ho'" to make getchar( ) behave. Many C compiler
developers have decided, for the sake of compatibility, to keep
getchar( ) line-buffered, even though there is no technical reason for
it. (In fact, the ANSI C standard states that getchar( ) need not be
line-buffered.) When getchar( ) is implemented in a line-buffered
fashion in a modern interactive environtnent, its use is severely
limited.
Because many compilers have implemented line-buffered versions
of getchar( ), most C compilers supply another function to perform
interactive console input. Although it is not defined by the ANSI C
standard, most compilers call this function getche( ). You use it just
like getchar( ), except that it will return its value immediately after a
key is pressed; it does not line-buffer input. for most compilers, this
function requires a header file called CONIO.H, but it might be called
something different in your compiler. Thus, if you want to achieve
interactive character input, you will usually need to use the getche( )
function rather than getchar( ).
Since all readers will have access to the getchar( ) function, it will
be used by most of the examples in this book that require character
input. However, some examples will use the getche( ) function. If
your compiler does not include this function, substitute getchar( ).
You should feel free to experiment with getche( ) on your own.

At the time of this writing, when using Microsoft's Visual C++ compiler,
getche() is not compatible with C's standard input funcyons, such as scanf().
Instead, you must u~e special console versions of these of these functions,
such as cscanf(). This and other non-standard liD functions are described in
Chapter 8. The examples in this book that use getche() work correctly with
Visual C++ because they avoid the use of the standard input functions.

Virtually all computers use the ASCII character codes when

representing characters. Therefore, characters returned by either
getchar( ) or getche( ) will be represented by their ASCII codes. This
is useful because the ASCII character codes are an ordered sequence;
each letter's code is one greater than the previous letter, each digit'S
code is one greater than the previous digit. 1'l1's means that 'a' is less
 MORE C PROGRAM CONTROL STATEMENTS 73
3.1 INPUT CHARACTERS
...
"

than 'b', '2' is less than '3', and so on. You may compare characters just
like you compare numbers. For example.
ch = getchar() i
if(ch < 'f' printf("character is less than flo);

is a perfectly valid fragment that will display its message if the user
enters any character that comes before f.

EXAMPLES

l. This program reads a character and displays its ASCII code. This
illustrates an inlponant feature of C: You can use a character as
if it were a "little integer." The program also demonstrates the
use of the getche( ) function.
~include<conio.h>
#include <stdio.h>

int main(void)

char chi

printf ("Enter a character: ");

ch = getche();
printf("\nlts ASCII code is %d", ch);

return 0;
}

Because this program uses getche( ). it responds as soon as you

press a key. Before continuing. try substituting getchar( ) for
getche( ) in this program and observe th~ results. As you will
see, getchar( ) does not return a character to your program
until you press ENTER.
2. One of the most common uses of character input is to obtain a
menu selection. For example. this program allows the user to
add. subtract, multiply, or divide t\\'o numbers.
#include <stdio.h>
74 TEACH YOURSW
y
C

int main(void)
{
int a, bi
char chi

printf(-Do you want to:\n~};

printf("Add, Subcract, Multiply, or Divide?\n");
printf("Enter'first letter: It);
eh = getchar ( ) ;
printf("\n", ;

printf ( Enter first number: ");

scanf ("'d", &a); ~
printf("Enter second number: ");
scanf("'d", &h);

if(ch=='A'1 pnntfl"td", a+b) ;

if(ch=='$' ) printf ("'d", a-bl;
if(ch=='M') printf("%d", a-b);
if(ch=='O' && bl=O) printf{"%d", a/b);

return 0;
}

One point to keep in mind is that C makes a distinction between

upper- and lowercase letters, So, if the user enters an s, the
program will not recognize it as a request to subtract. (Later,
you will learn how to convert the case of a character.)
3. Another common reason that your program will need to read a
character from the keyboard is to obtain a yes' no response from
the user. For example, this fragment determines if the user
wants to proceed.
printf("Do you wish to continue? (YJN : .);
ch = getche I I ;
i f Ich=='Y' I (
/* continue with something *1

)
 MORE C PROGRAM CO~ROl STATEMENTS 75
3.2 NESTif STATEMENTS
'"

EXERCISES

I. Write a program that reads ten letters. After the letters have
been read, display the one that comes earliest in the alphabet.
(Hint: The one with the smallest value comes first.)
2. Write a program that displays the ASCII codes for the characters
A through Z and a through z. How do the codes differ between
the upper- and 10wer~ase characters?

EST if STATEMENTS

When an if statement is the target of another if or else, it is said to be

nested within the outer if. Here is a simple example of a nested if:
if(count > max) / * outer if * /
if (error princf {"Error, try again. -) - * nested if *

Here, the printf( ) statement "'ill only execute it count is greater

than max and if error is nonzero. Notice how the nested jf is
indented. This is common practice. It enables anyone reading "our
progralTI to know quickly that thf' if IS nested and what actions are
nested. A nested if may also appear inSIde a blo~k of statements that
are the target of the outer if.
An ANSI-standard compiler "'ill allow you to nest ifs at least 15
levels deep . (However, it would be rare to find such a deep nesting.)
One confus(ng aspect of nested ifs is illustrated by the following
fragment.
ifp
if ' q ) printf("a and b are true");
else printf ( "To which statement does this else apply?");

The question suggested by the second printf( ) is: which if is

associated with the else' Fortunately , the answer is quite easy: An
else always associates with the nearest if in the same block that does
not already have an else associated with it. In this example, the else is
associated with the second if.
76 TEACH YOURSELF

• C

EXAMPLES

1. 1t is possible to string together several ifs and elses into what is

sometimes called an if-else-if ladder or if-else-if stair'case because
of its visual appearance. In this situation a nested if has as its
target another if. The general form of the if-else-if ladder is
shown here;

if(expression) statement,
else
if(expression) statement,
else
if(expression) statement,

else statement,

The "xpressions arc c\'aluated from the top dow nward . As soon
as a true condition is found. the statement assori<lted with it is
executed , and the rest of the ladder is bypassed . If none of the
" xpressions arc true, the final else will be executed. That is, if
all other conditional tests fail, the last else statement is
performed . If the final else is not present, no action will take
place if all expressions are false .
Although the indentation of the general form of the if-else-if
ladder JUSt shown is technically correct, it can lead to overly
deep indentation. Because of this, the if-else-if ladder is
ge nerally written like this;

if(expression) statement,
else if(expression) statement,
else if(expression) statement;

else statement;

We can improve the arithmetic program developed in

Section 3.1 by using an if-else-if ladder, as shown here;
 MORE C PROGRAM CONTROL STATEMENTS 77
3.2 NEST if STATEMENTS •

#include <stdio.h>

int main(void)
(
int a, b;
char chi

printf("Do you want to:\n") i

printf{-Add, Subtract, Multiply, or Divide?\n");
printf("Enter first letter~ ");
ch = getchar ( ) i
printf (."\n");

printf("Enter first number: fl);

scanf ("%d", &a);
printfC-Enter second number: ");
scanf ("%d", &bJ;

if(ch=='A') printf("%d", a+b);

else if(ch=='S') printf("%d", a-b);
else iftch=='M') printi("%d", a~b);
else if(ch=='D' && b!=O) printf("%d", a b)i

return 0;
J

This is an inlprovement over the original version because oncr

a match is found, any remaining if statements are skipped. This
means th.;lt the program isn't wasting time on needless
operations. While this is not too important in this example, you
will encounter situations where it will be.
2. Nested if statements are very common in programming. For
example, here is a further improvement to the addition drill
program developed in the preceding chapter. It lets the user
have a second try at getting the right answer.
.include <stdio.h>

int main(voidl
{
int a n swe r, count ;
78 TEACH YOURSELF
~
C

i:1t again;

for(count=l; cQunt<11; count++) {

printf("y,Jhat is %d ... %d? ", count, count);
scanf{"%9", &answer);

if(answer == counL+count) printf(MPight!\n");

else {
printf("Sorry, you're wrong\n" 1 ;
printf("Try again.\n oil;

printf ("\nvlhat is %d + %d? count. count I ;

scanf ( "%d", &answer);

/* nested if .. /
if (answer count+count) printf("Right!\n");
else
printf("Wrong, the answer is %d\n" ,
count+count) ;
)
)
return 0;
)

Here, the second if is nested within the outer ifs else block.

EXERCISES

I. To which if does the else relate to in this example?

if(ch=='S') { /* first if */
printf("Enter a number: oil;
scanf("%d", &y);

/.. second if '* /

if(y) printf{ "I ts square is %d.", y*y);
)
else printf("Make next selection.");

2. Write a program that computes the area of either a circle,

rectangle, or triangle . Use an if-else-ifladder.
 MORE C PROGRAM CONTROL STATEMENTS 79
3.3 EXAMINEfor LOOP VARIA nONS
"
.
XAMI N E f or LOOP VARIATIONS
The for loop in C is significantly more powerful and flexible than in
most other computer languages. When you were introduced to the for
loop in Chapter 2, you were only shown the form similar to that used by
other languages. However, you will see that for is much more flexible.
The reason that f;'r is so flexible is that the expressions we called
the initialIzatIOn, conditional-test, and increment portions of the loop are
not linllted to these narrow roles. The for loop places no li mits on the
types of expressions that occur inside it. For example. you do not have
to use the initialization sectio'n to initialize a loop-control variable.
Further, there does not need to be any loop-control variable because
the conditional expression may use some other means of stopping the
loop. Finally, the increment portion is technically just an expression that
is evaluated each time the loop iterates. It does not have to increment
or decrement a variable.
Another important reason that the for is so flexible is that one or
more of the expressions inside it may be empty. For example, if the
loop-control variable has already beerl initialized outside the for, there
is no need for an initialization expression,

EXAMPLES

I. This program continues to loop until a q is entered at the

keyboard. Instead of testing a loop-control variable, the
conditional test in this for checks the value of a character
entered by the user.
~include <stdio:h>
.include <conio.h>

int main{void)
{
int i;
char chi

ch = 'a'; /* give ch an initial value .,

for(i=O; .ch != 'q'; i++) {

printf("pass: %cl\n" , i);
.,
80 TEACH YOURS£LF
C

ch = getche () ;
)

return 0;
)

Here, the condition that controls the loop has nothing to qo with
the loop-control variable. The reason ch is given an initial value
is to prevent it from accidentally containing a q when the
program begins.
2. As stated earlier, it is possible to leave an expression in a loop
empty. For example, this program asks the uSer for a value and
then counts down to zero from this number. Here, the
loop-control variable is initialized by the user outside the loop,
so the initialization portion of the loop is empty.
#include <stdio.h>

int main!voidl
{
int i;

printf (~Enter an integer: ");

scanf (~%dft, &i);

forf: i; i--) printf("%d i);

return 0;
)

3. Another variation to for is that its target may be empty. For

example, this program simply keeps inputting characters until
the user types q.
#include <stdio.h>
#include <conio.h>

int main{void)
{
char chi

. for(ch=getche(); ch!='q'; ch=getche(»;

printf("Found the q.~);
 MORE C PROGRAM CONTROL STATEMENTS 81
~
3.3 EXAM/NEfor LOOP VARJAnONS

return 0 i .
}

Notice that the statements assigning ch a value have been

moved into the loop. This means that when the loop stans,
getche( ) is called. Then, the value of ch is tested against q .
Next, conceptually, the nonexistent target of the for is executed,
and the call to getche( ) in the increment portion of the loop is
executed. This process repeats until the user enters a q.
The reason the target of the for can be empty is because C
allows null statements.
4. Using the for, it is possible to create a loop that never stops.
This type of loop is usually called an infinite loop. Although
accidentally creating an infinite loop is a bug, you will sometimes
Want to create one on purpose. (Later in this chapter, you will
see that there are ways to exit even an infinite loop') To create
an infinite loop, use a for construct like this:
for ( ; ; ) (

ASyou can see, there are no expressions in the for . When

there is no expression in the conditional portion, the compiler
assumes that it is true. Therefore , the loop continues to run.
5. In C, unlike most other computer languages, it is perfectly valid
fo r the loop-control variable to be altered outside the increment
section. For example, the following program manually
increments i at the bottom of the loop.
#include <stdio.h>

int rnain{void)

int i;
'-
for(i=O; i<10; ) {
printf ("%d ". i);
i++:
82 TEACH YOURSELF
~
C

return 0;
)

.
1. Write a program that computes driving time when given the
distance and the average speed. Let the user specifY the
number of drive time computations he or she wants to perform.
2. To create time-delay loops, for loops with empty targets are
often used. Create a program that asks the user for a numher
and then iterates until zero is reached . Once the countdown is
done , sound the hell, but don't display anything on the screen.
3. Even if a for loop uses a loop·control variable , it need not he
incremented or decremented by a fixed amount . Instead, the
amount added or subtracted may vary . Write a program that
begins at 1 and runs to 1000. Have the program add the
loop·controI variable to itself inside the increment expression .
This is an easy way to produce the arithmetic progression 1 2 4
816, and so on .

NDERSTAND C'S while LOOP

Another of C's loops is while. It has this general form:

while(expression) statement:

Of course, the target of while may also be a block of code. The while
loop works by repeating its target as long as the expression is true .
When it becomes false, the loop stops. The value of the expression is
checked at the top of the loop. This means that if the expression is
false to begin with, the loop will not execute even once.
 MORE C _ CONTROl. STATEMENTS 83
14 UNDERSTAND CSwhi&e LOOP
"
EXAMPLES

I. Even though the for is flexible enough to allow itself to be

controlled by factors not related to its traditional use, you should
generally select the loop that best fits the needs of the situation.
For example, a better way to wait for the letter q to be typed is
shown here using while. If you compare it to Example 3 in
Section 3.3, you will see how much clearer this version is.
(However, you will soon see that a better loop for this job exists.)
#include <stdio.h>
~include <conio.h>

int main(voidJ
{
char ch;

ch = getche();

while(ch! :'q') ch = getche();

printf(~Found the q.");

return 0;
)

2. The following program is a simple code machine. It translates

the characte rs you type into a coded form by adding I to each
letter. That is, 'A' becomes'S: and so forth. The program stops
when you press ENTER. (The getche( ) function returns \r
when ENTER is pressed.)
F.include <stdio.h>
#include <conio.h>

int'" main (void)

{
char chi

printf(-Enter your message.\n"J;

ch = getche ( ) ;
while(ch != '\r') {
84 TEACH YOURSELF
... C

printf("%c", ch+l ) i
cr getche ( ) i

return 0;

EXERCISES

1. In Exercise I of Section 3.3, vall created a program that

COlllputed driving time, given distance and average speed. You
used a for loop to Jet the user C0111pUte.: several drive times.
Rework that progri.l1l1 so that it uses a whi1c loop.
2. Write a program that \\-ill uecode messag{~ s that have been
encoded using tlw 'c ode machine prugram in the second
example.: in this section.
"

SE THE do LOOP

Cs final loop is do, which has this general form:

do (
statements
) w hile(express;on);

If only one statement is being repeated, the curly braces are not
necessary. Most progrzllllmers include thenl. however, so that they can
easily recognize that the while that ends the do is part of a d o loop,
not the beginning of a while loop.
The do loop repeats the statement or statements while the
expression is true. It stops when the expression hecomes false. The do
loop is unique because-it ",ill always execute the code within the loop
at least once, since the (~:\pre~sion controlling the loop is tested at the
bottom of the loop.
 MORE e PROGRAM CONTROL STATEMENTS 85
3.5 USE THE Ito LOOP
"

EXAMPLES

1. The fact that do will alwavs execute the body of its loop at least
once makes it perfect for ~hecking menu input. For example,
this version of the arithmetic program reprompts the user until
a valid response is entered.
~include <stdio.h>

int main(void)
(
int a, b;
char chi

printf("Do you want to:\n");

printf("Add, Subtract, Multiply, or Divide?\n");

r force user to enter a valid response

do {
print f ( .. Enter first let ter: ~);

ch = getchar ( ) ;
} while{ch!='A' && ch!='S' && Ch!='H' && ch!='D');
printf("\n") ;

printft"Enter first number: ") ;

u
scanf ("'td , &a);
printf("Enter second number : ~);

scan f ~" 'td ", &b);

if(ch=='A') printf("%d". a+b);

else if(ch=='S') printf("%d", a-b);
else if(ch=='M') printf{h%d", a*b);
else if(ch=='D' && b!=Ol printf(-%d", a 'b);

return 0;

2. The do loop is especially useful when your program is waiting

for some event to occur. For example, this program waits for the
user to type a q . Notice that it contains one less call to getche( )
than the equivalent ·program described in the section on the
while loop.
#include <stdio.h>
#include <conio.h>
aa
...
TtACH YOURSELf
C

int main(void)
{
char Chi

do {
ch = getche ( ) ;
} while(ch!='q')i

printf("Found the q.M);

return 0;
)

Since the loop condition is tested at the bottom, it is not

necessary to initializp. ch prior to entering the loop.

EXERCISES
•
1. Write a program that coilVerts gallons to liters. Using a do loop,
allow the user to repeat the conversion. (One gallon is
approximately 3.7854 liters.)
2. Write a program that displays the menu below and uses a do
loop to check for valid responses. (Your program does not need
to implement the actual functions shown in the menu.)

Mailing list menu:

1. Enter addresses

2. Delete address

3. Search the list

4. Print the list

5. Quit

Enter the number of your choice (1-5).

 MORE C PROGRAM CONTROL STIITEMENTS 87
~
3.6 CREATE NESTED LOOPS

•
REA tE NESTED LOOPS

When the body of one loop contains another, the second is said to
be nested inside the first. Any of C's loops may be nested within any
other loop. The ANSI C standard specifies that loops may be nested
at least 15 levels deep. However, most compilers allow nesting 10
virtually any level. As a simple example of nested fors , this fragment
prints the numbers 1 to lOon the screen ten times.
for(i=O; i<10; i++) {
for(j=l; j<11; j++) printf("%d" j); j* nested loop */
printf ("'n ft
) ;

EXAMPLES

1. You can use a nested for to make another improvement to the

arithmetic drill. In the version shown below, the program will
give the user three chances to get the right answer. Notice the
use of the variable right to stop the loop early if the correct
answer 15 given.
'include <stdio.h>

int main{voidl
(
int answer , count, chances, right;

for (count=l; count<11; count++) {

printf{"What is %d + %d?", count. count);
scanf ( "%d", &answer);

if (answer == count+count) printf ( "Right! \n" ) ;

else {
printf("Sorry. you're wrong.\n");
printf ("Try again. \n") ;

right = 0;
88 .TEACH Y~ELF
". C

/* nested for */
for (chances=O; chances<3 && !right; chances++) (
printf("What is %d .... %d? ", count, count);
scanf{"%d", &answer);

if (answer == cQunt+cQuntl {
printf("Right!\n") ;
right = 1;
}
}

f* if answer still wrong. tell user */

if(!right)
printf("The answer is %d.\n". count+count);
}
}

return 0;

2. This program uses three for loops to print the alphabet three
times, each time printing each letter twice:
#include <stdio.h>

int main (void)

{
inc i. j, k;
for(i;O; i<3; i++)
for(j=O; j<26; j++)
for{k=O; k<2; k++) printf("%c·, 'A'+j);

return 0;
}

The statement
printf("%c". 'A'+j);

works because ASCII codes for the letters of the alphabet are
strictly ascending-each one is greater than the letter that
precedes it.
 _C_~cotrINK.ITA-
.

11 USE_ roEXTTALOOP
..'"
,.

I. Write a program that finds aJl the prime numbers between Z

and 1000.
2. Write a program that reads ten characters from the keyboard.
Each time a character is read, use its ASCII code value to output
a string of periods equal in number to this code. For example, .,
given the letter 'A', whose code is 65, your program would
output 65 periods.

E break TO EXIT A LOOP

The break statement aJlows you to exit a loop from any point within
its budy. bypassing its normal termination expression. When the break
statement is encountered inside a loop, the loop is immediately
stopped, and program control resumes at the next statement follOWing
the loop. For example, this loop prints only the numbers 1 to 10:
.include <stdio.h>

int main (void)

(
int i;

for (i=l; 1<100; i++' {

printf('\d " i);
if(i= =l O) break.; j * exit the loop,·!
)

return 0:

The break statement can be used with all three of C's loops.
You can have as many break statements within a loop as you
desire . However, since too many exit points from a loop tend to
destructure your code, it is generally best to use the break for special
purposes, not as your normal loop exit.
,
80 1IACII YOUMUf
y
c·

EXAMPLES

1. The break statement is commonly used in loops in which a

special condition can cause immediate tennination. Here is an
example of such a situation. In this case, a keypress can stop the
execution of the program.
'include <stdio.h>
'include <conio.h>

int main(void)
(
int i;
char ch;
.
'* display all numbers
for(i=l: i<10000; i++)
that are .nultiples of 6 '*/
(
if(!(i\6» (
print! (-'d, more? CY IN) ", i 1;
ch = getche ( J ;
if(ch=='N') break ; /. stop the loop + j
printf(-\n") :

return 0 i
}

7.. A break" will cause an exit from only the innermost loop. For
example, this program prints the numbers 0 to 5 five times:
Finclude <stdio.h>

int iTIain(voidl
(
int .i. j j

for(i=O; i<5; i++) {

for(j=O; j<lOO; j++) (
printf ("\d", j);
i f (j==5) break;
)
printf (. \n') ;
 MORE C PROORAM CONTROL STAnMENTS 91
3.7 USE break TO EXIT A LOOP
'"
)

return 0;
)

3. The reason C includes the break statement is to allow your

programs to be more efficient. For example, examine this
fragment:
do (
printf(-Load. Save. Edit, Quit?\n·);
do (
printf(-Enter your selection: .);
ch = getchar();
} while(ch!='L' && ch!='S' && ch!='E' && ch!='Q');

iflch != 'Q') (
/ * do something */

.if~ch != 'Q') (
.' * do something else·/
)
etc. */
/ *
) whilelch != 'Q')

In this situation, several additional tests are performed on ch to

see if it is equal to 'ct to avoid executing certain sections of code
,,-hen the. Quit option is selected. Most C programmers would
write the preceding loop as shown here:
fore j ; ) { /* infinite for loop *;
printf(-Load. Save. Edit. Quit?\n");
do { ,
print f (. Enter your selection: .. );
ch = get char II ;
} while(ch!='L' && ch!='S' && ch!:'E' && ch!='Q');

iflch == 'Q') break;

•

'* dodo something

/* something else
*/
*/

)
'* etc. ftj
82
...

In this version, ch need only be tested once to see if it contains

a 'Q'. As you can see, this implementation is more efficient
because only one if statement is required.

EX'IICIIEB
I . On your own, write several short programs that use break to
exit a loop. Be sure to try all three loop statements.
.
2. Write a program that prints'a table showing the proper amount
of tip to leave. Start the table at $1 and stop at SIOO, using
increments of SI. Compute three tip percentages: \0%, 15% ,
and 20% . After each line, ask the user ifhe or she wants to
continue. If not, use break to stop the loop and end the program.

OW WHEN TO USE THE continue

STATEMENT
•
The' continue statement is somewhat the opposite of the break
statement. It forces the next iteration of the loop to take place,
skipping any code in between itself and the test condition of the loop.
For example, this program never displays any output:
'include <stdio.h>

int main (void)

(
int x;

for(x=O; x<lOO; x++) {

continue;
printf(-'d· Xl; / * this is never executed * /
)
 MORE C PROGRAM CONTROCSTAl'E .~.- . ._ ----
3.8 KNOW WHEN TO USE THE continue STATEMENT

return 0;
)

Each time the continue statement is reached, it causes the loop to

repeat, skipping t\1e printf( ) statemenl
In while and do-while loops, a continue statement will cause
control to go directly to the test condition and then continue the
looping process. In the case offor, the increment part of the loop is
performed, the conditional test is executed, and the loop continues.
Frankly, continue is seldom used, not because it is poor practice to
use it, but simply because good applications for it are not common.

EXAMPlE

1. One good use for continue is to restart a statement sequence

when an error occurs. For example, this program computes a
running total of numbers entered by the user. Before adding
a value to the running total, it verifies that the number was
correctly entered by having the user enter it a second time.
If the t\\'o numbers don't match, the program uses continue
to restart the loop .
• include <stdio.h>

int main(void)
{
int total, i, j;

total = 0;
do {
printf ("Enter next number (0 to stop): .. ) i
scanf{"'d", &i);
printf(-Enter number again: ");
scanf {"'d", &j);
if(i != j) {
printf("Mismatch\n") ;
continue;
)
total = total + i;
,.
94 TfACH YOURSElf
C

) while(i);

printf(-Total is %d\n-, total);

return 0;
)

I. Write a program that prints only the odd numbers between 1

and 100. Use a for loop that looks like this:
for(i=l; i<101; i++) . . .

Use a continue statement to avoid printing even numbers.

ELECT AMONG AL TERNATIVES WITH

THE switch STATEMENT

While if is good for choosing between two alternatives. it qUickly

becomes cumbersome when several alternatives are needed. C's
solution to this problem is the switch statement. The switch statement
is C's multiple selection statement. It is used to select one of several
alternative paths in program execution and works as follows. A value
is successively tested against a list of integer or character constants.
When a match is found. the statement sequence associated with that
match is executed. The general form of the switch statement is this:
switch(value) (
case constantl:
statement sequence
break;
 MOIliC_CONTROlSTATEMUmI
.9 SEUCT AMONGAL7FIW41MS WITH THE_ STATEMENT

case .constant2:
statement sequence
break;
case constant3;
statement sequence
break:

default
statement sequence
break:
I
The default statement sequence is performed if no matches are
found. The default is optional. If all matches fail and default is
absent, no action takes place. When a match is found, the statements
associated with that case are executed until break is encountered or,
in the case of default or the last case, the end of the switch is reached.
As a very simple example, this program recognizes the numbers 1,
2. 3, and 4 and prints the name of the one you enter. That is, if you
enter 2, the program displays two.
#include <stdio.h>

int mainlvoidl
(
int i;

printf(MEnter a number between 1 and 4: ~) ;

scanf(-%d-, &i);

switch(i) (
case 1:
printf(ft one ") ;
break;
case 2:
printf (-two·);
break;
case 3:
printf{"three K );

break;
 TEAQt YOU!lSW
86
... C

case 4: •
printf ("four");
break:
default: .
printf(-Unrecognized Number-);
}

return 0;
}

The switch statement differs from if in that 8witch can only test
for equality, whereas the if conditional expression can be of any type.
Also, switch will work with only int or char types. You cannot, for
example, use floating-point numbers.
The statement sequences associated with each case are not blocks;
they are not enclosed by curly braces.
The ANSI C standard states that at least 257 case statements will be
allowed. In practice, you should usually limit the amount of case
statements to a much smaller number for effiCiency reasons. Also. no
two case constants in the same switch can have identical values.
It is ~ossible to have a switch as part of the statement sequence of
an outer switch. This is called a nested suitcll. If the case constants of
the inner and outer switch contain common values, no conflicts will
arise. For example, the following code fragment is petfectly acceptable:
switch(a) (
case 1:
switch(b)" {
case 0: printf(-b is false~);
break;
case 1: printfC-b is true-);
}
break ;
case 2:

An ANSI-stanQard compiler will allow at least 15 levels of nesting

for switch statements.
 MORE C PROGRAM CONTROL STATlMENTS 97
3.9 SELECT AMONG ALTERNAT1VfS WITH THE switch STATEMENT •

EXAMPLES

1. The switch statement is often used to process menu

commands. For example, the arithmetic program can be
recoded as shown here. This version reflects the way
professional C code is written.
#include <stdio.h>

int main(void)
(
int a, bi
char chi

printf("Do you want to: ",n") j

printf("Add, Subtract, Multiply. or Divide?\n");

1- force user to enter a valid response */
do (
printf(~Enter first letter: ~):
ch = getchar ( ) ;
} while(ch!='A' && ch!='S' && ch!='M' && ch!='O');
printf(" \ n");

printf(nEnter first number: ") i

. scanf("'d", &al;
printft"Enter second number: .);
scanf("M", &b);

5witch(ch) (
case 'A': .printf("'d", a+b);
break;
case '5': printf("'d", a-b);
break.;
case 'M': printf ("%d" a-b) ;
I

break;
case 'D': if(b!=O) pr'n1;f("'''", alb);
)

return 0;
)
9a TEACH YOURSElf'
y
C

2. Technically, the break statement is optional. The break

statement, when encountered within a switch, causes the
program flow to exit from the entire Iwitch statement and
continue on to the next statement outside the Iwitch. This is
much the way it works when breaking out of a loop. However,
if a break statement is omitted, execution continues into the
following case or default statement (if either exists). Thahs,
when a break statement is missing, execution "falls through'
into the next case and stops only when a break statement or
the end of the switch is encountered. For example, study this
program carefully:
'include <stdio.h>
'include <conio.h>

int main(void)
(
char chi

do (
printf("\nEnter a character. q t.o quit: ");
ch = getche () ;
printf ("'n"';

switch (ch) {
case 'a':
printf ("Now is .. );
case 'b':
printfl"the time .);
case 'c':
printf("for all good men");
. break;
case 'd':
printf("The summer .);
case 'e':
printf("soldier .);
)
} whilelch != 'q');

return 0;
)

If the user types a, the entire phrase Now is the time fc:.r all
good men is displayed. Typing b displays the rime for all
 _ C PROGRAM CONTROl STATlM£NTS 99
~
3.9 SELECT AMONG ALTERNA nVES WITH THE switch STATEMENT

good men. As you can see, once execution begins inside a case,
it continues until a break statement or the end of the switch is
encountered.
3. The statement sequence associated with a case may be empty.
This allows two or more cases to share a common statenlent
sequence without duplication of code. For example, here is a
program that categorizes letters into vowels and consonants:
#include <stdio.h>
#include <conio.h>

int rnain(void)
{
char Chi

printf (~Enter the letter: ") i

ch = getche () ;

5wicch(ch) (
case 'a' :
case . e':
case . i' :
case '0' :
case 'u' :
case 'y' :
printf(" is a vowel\n");
break;
default:
printf(" is a consonant");
)

return 0;
)

1. What is wrong with this fragment?

float f;

scanfl"U", &f);
 .,
100 TEACH YOURSElf

switch(f) 1.
case 10.05:

2. Write a program that counts the numbers of letters, digits,

and common punctuation symhols entered hy the user. Stop
inputting when the user presses ENTER. Use a switch statement
to categorize the characters into punctuation, digits, and letters.
When the program ends, report the number of characters in
each category. (If you Jike ,si mply assume that, if a character is
not a digit or punctuation, it is a letter. Also, just use the most
common punctuation symbols.)

3.10 . UNDERSTAND THE goto STA TEMENT

C suppons a non-conditional jump statement, called the goto. Because

C is a replacement for assembly code, the inclusion of goto is necessa,,'
because it can be used to create »ery fast routines. However, most
programmers do not use goto because it destructures a program and,
if frequently u~ed, can render the program virtually impossible to
understimd later. Also, there is no routine that requires a goto. For.
these reasons, it is not used in this book outside of this section.
The goto statement can perform a jump within a function. It cannot
jump between functions. It works with a label. In C, a label is a valid
identifier name followed by a colon. For example, the folldwing goto
jumps around the printf( ) statement:
goto myl ..bel;
printf(MThis will not print,-);
mylabel: printf ("This will prine. ~) ;

About the only good use for goto is to jump out of a deeply nested
rOll tine when a catastrophic error occurs.
 MORE C PROGRAM CONTROL STATlMEII'IS 101
3.10 UNDERSTAND THEgoto STATEMENT
"

EXAMPLE

1. This program uses goto to create the equivalent of a for loop

running from 1 to 10. (This is just an example ofl!lS~. In actual
practice, you should use a real for loop when one is needed.)
#include <stdio.h>

int main(void)
(
int i;

i = 1;
again:
printf("%d" i);
i++;
if(i<10) gato again:

return 0;

1. Write a program that uses goto to emulate a while loop that

counts fWIl1 1 to 10.

At this point, you should be able to answer these questions and

perform these exercises:

1. As illustrated by Exercise 2 in Section 3.1, the ASCII codes for

the lowercase letters are separated from the uppercase letters
by a difference of 32. Therefore, to convert a lowercase letter to
102 TEACH YOURSElf

• C

an uppercase one, simply subtract 32 from it. Write a program

that reads characters from the keyboard and displays lowercase
letters as uppercase ones. Stop when ENTER is pressed.
2. Using a nested if statement, write a program that prompts the
user for a number and then reports if the number is positive,
zero, or negative.
3. Is this a valid for loop?
char Chi

ch = ' x' ;
for ( ; ch ! = ' , ; ) ch = getche () ;

4. Show the traditional way to create an infinite loop in C.

5. Using the three loop statements, show three different ways to
count from 1 to 10.
6. What does the break statement do when used in a loop?
7. Is this switch statement correct?
switch(i) (
case 1: printf ("nickel") ;
break;
case 2: printf(ndime b
);

break;
case 3: printf ("quarter M
) ;

} •

8. 1s this goto fragment correct'

goto alldone;

alldone

This section checks how well you have integrated the material in
this chapter with that from earlier chapters.
 _C_COIIIROLsrll_ 103
•
l. Using a switch statement, write a program that reads characters
from the keyboard and watches for tabs, newlines, and
backspaces. When one ia received, diaplay what it is in words.
For example, wilen the uaer preues the TAB key, print tab.
Have the _ enter a q to stop the program.
2. While tbia program ia not incorrect, show how it would look if
written by an experienced C programmer.
'include <stdio.h>

int main (void)

(
int i. j. k;

forlk=O; k<10; k~k+l) (

printfl"Enter first number: ");
ecanf(-'d- ... it;

printf(-Enter second number: .):

scanf(-'d-, &j):

iflj !. 0) printf I "'d\n" , i/~l,

iflj == 01 printfl"cannot divida by zero\n"l;
1

return 0;
1
 ,

4
A Closer Look at
Data Types, Variables,
and Expressions
•

chapter obIec .....

4.1 Use C's dala-type modifiers

4.2 Learn where variables are declared
4.3 Take a closer look al constants
4.4 Initialize variables
4.5 Understand type conversions in expressions
4.6 Undersland type conversions in assignments
,
4.7 Program wilh tyre casts

105
~
 •

4
A Closer Look at
Data Types, Variables,
and Expressions
•

chapIer objeLtl_

4.1 Use C's data-type modifiers

4.2 learn where variables are declared

4.3 Take a closer look at constants

4.4 Initialize variables

4.5 Understand type conversions in expressions

c-
4.6 Understand type conversions in assignments

4.7 Program with tyre casts

105
'"
 -IH IS chapter more fully examines several concepti presented
in Chapter 1. It covers COs data-type modifiers, global and
local variables, and constants. It also dlscuaaes how C handles
various type conversions.

( " ~.
Before proceeding, you should be able to answer these questions
and perfonn these exercises:
1. Using C's three loop statements, show three ways to write a loop
that counts from 1 to 10.
2. Convert this series of ifs into an equivalent switch.
if(ch=='L') load() ,
else if(ch=='S') save(),
else if(ch=='E') enter(),
else if(ch=='O') display()
else if(ch=='Q') quit(),

3. Write a program that inputs characters until the user strikes the
ENTER key.

4. What does break do?

5. What does continue do?
6. Write a program that displays this menu, performs the selected
operation, and then repeats until the user selects Quit.
Convert .
1. feet to meters
2. meters to feet
3. Otinces to pounds
•• pounds to ounces
5. Quit
Enter the number of your choICe:
 E C'S DA TA - TYPE MODIFIERS
In Chapter 1 you learned that C has five basic data types: void, char,
int, float, and double. These basic types, except type void, can be
modified using C's type modifiers to more precisely fit your specific
need. The type modifiers are

long
short
signed
unsigned

The type modifier precedes the type name. For example, this declares
a long integer:
long int i;

The effect of each modifier itexamined next.

The long and short m<jlifiers may be applied to into As a general
rule, short ints are ofti~ ·smaller than ints and long ints are often
larger than ints. For example, in most 16-bit environments, an int
is 16 bits long and a long int is 32 bits in length. However, the precise
meaning oflong and short is impleme,nratioh
• I
dependent. When the
ANSI C standard was created, it specified p1inimum ranges for integers,
short integers, and long integers. It. did not set fixed sizes for these
items.1See Table 4-1.) For example, using the minimum ranges set
forth in the ANSI C standard, the smallest acceptable size for an int is
16 bits and the smallest acceptable size for a short int is also 16 bits.
Thus, it is permissjble for integers and short integers to be the same '
size! In fact, in most I &bit environments, there is no difference between
an int and a short into Further, in many 32-bit environments, you will
find that integers and long integers are the same size. Since the exact
effect of long and short on integers is determined by the environment
in which you are working and by the compiler you are using, you will
I need to check your compiler's documentation for their precise effects.
The long modifier may-also be applied to double. Doing so roughly
doubles the precision of a floating point variable. .
108 TtACM YOUGW
G

The signed modifier is used to specify a signed integer value. (A

signed number means that it can be positive or negative .) However
the use of signed on integers is redundant because the default integer
declaration automatically creates a signed variable . The main use
of the signed modifier is with char. Whether char is signed or
unsigned by itself is implementation dependent. In some
implementations char is unsigned by default; in others, it is signed .
To ensure a signed character variable in all environments, you must
declare it as signed char. Since most compilers implement char as
signed, this book simply assumes that characters are signed and will
not use the signed modifier.
The unsigned modifier can be applied to char and into It may also
be used in combination with long or short. It is used to create an
unsigned integer. The difference between signed and unSigned
integers is in the way the high-order bit of the integer is interpreted. If
a signed integer is specified.then the C compiler will generate code
that assumes the high-order bit is used as a sign flag . If the sign flag is
0, the number is positive; if it is 1, the number is negative . Negative
numbers are generally represented using the two's complement
approach. In this method, all bits in the number (except the sign flag)
are reversed, and 1 is added to this number. Finally, the sign flag is set
to 1. (The reason for this method of representation is that it makes it
easier for the CPU to perform arithmetic operations on negative values.)
Signed integers are important for a great many algorithms , but they
only have half the absolute magnitude of their unsigned relatives . • or
example, here is 32,767 shown in binary:

0111111111111111

If this is a signed value and the high-order bit is set to 1, the number
would then be interpreted as -1 (assuming two's complement format).
However, if this is an unsigned value, then when the high-order bit
is set to I, the number becomes 65,535.
Table 4-1 shows all allowed combinations of the basic types and
the type modifiers. The table also shows the most common size and
minimum range for each type as specified by the ANSI C standard .
It is important to understand that the ranges shown In Table 4-1 are
just the minimums that all compilers must provide. The compiler is
free to exceed them, and most compilers do for at least some data
types. As mentioned, an int in a 32-bit environment will usually have
a range larger than the minimum . Also, in environme!1ts that use
 A CLOSER LOOK AT DATA TYPES. VARIABLES. AND EXPRESSIONS 109
.:i. I USEC'SDATA-TYPEMODIFIERS •

Typical Size in Bits Minimal Range

8 -127 to 127
siolnerl char 8 o to 255
char 8 -127 to 127
16 or 32 -32.767 to 32.767
[.inner! int 16 or 32 o to 65.535
16 or 32 same as int •
16 same as int
i si!~ne,d short int 16 o to 65.535
16 same as short int
32 -2.147,483.647 to 2,147,483.647
hn ••rl long int 32 same as long int
ned long int 32 o to 4,294.967.295
32 Six digits of precision
64 Ten digits of precision
80 Ten digits of precision

All Dala Types Defined by lhe ANSI C Standard ""

two's complement arithmetic (which is the case for the vast majority
of computers), the lower bound for signed characters and integers is
one greater than the minimums shown. For instance, in most
environments, a signed char has a range of -128 to 127 and a short
int is typically -32,768 to 32,767. You will need to check your
compiler's documentation for the specific ranges of the data types
as they apply to your compiler.
C allows a shorthand notation for declaring unsigned. short, or
long integers. You may simply use the word unsigned, short, or l~ng
without the into The int is implied. For example,
unsigned count;
unsigned int num;

both declare unsigned int variables.

It is important to remember that variables of type char may be used
to hold values other than just the ASCII character set. C makes little
distinction between a character and an integer, except for the
110
'II'

magnitudes cif the values eaclt may hold. Therefore, as m'e ntioned
earlier, a signed char variable can also be used as a 'small' integer
when the situation does not require larger numbers.
When outputting integers modified by short, long, or unai~
using printf( ), you cannot simply use the %d specifier. The re<\SOn is
that printf( ) needs to know precisely what type of data it is
receiving. To_use printf( ) to output a short, use %hd. To output a
long, use %ld. When outputting an unsigned value, use %u. To ·
output an unsigned long int, use %lu. Also, to output a long double
use %Lf.
The seanf( ) function operates in a fashion similar to printf( ).
When reading a short int using scanf( ), use 'l6hd. When reading a
long int, use %ld. To read an unsigned long int, use 'l6lu. To read a
double, use %If. To read a long double, use %Lf.

EXAMPLES

I. This program shows how to input and output short, IOllg, and
unsigned values.
'include <stdio.h>

int main (void)

(
unsigned u;
long ~;
short s;

printf(-Enter an unsigned: .);

seanf(-'u-, &u);
printf(-Enter a long: .);
scanf ('Ud', &1);
printf(-Enter a short: .);
scanf('%hd', &s);

printf('%u \ld \hd\n', u, 1, s);

return 0;
)

2. T;) understand the difference bttween the way that signed and
unsigned integers are interpreted by C, run the following short
program. (This program assumes that short integers are J 6 bits wide.)
 Aca.IlIiUlOlAfIlAfAftIIII,W...... AllliA.IIIUU. " ,
, ' ., IotIfCSD4TA'nI'f__ •

'include <stdto,h_

iftt main(voidl
t
short int i, /" • • igned abort integer "'
unsigned ahort int ", '" an unalgned abort integer *'
u = 33000;
i = u;
printf("hd 'hu', i, ul/

return 0,
}

When thil Jll'OIII'am It run, the output iI -3253& 33000. The

reason for this Is that the 'bit pattern that 33000 represents
as an anaiped ~ int is interpreted as -32536 as a aipted
abort into
3. In C, you may Ule a char variable any place you would use an
int variable (lIIuml"s the differences in their (anges is not a
factor). For example, the following program uses a char variable
to control the loop that il summing the numbers between 1 and
100. In lOme buellt takel the computer less time to access a
aina\e byte (one character) than it doe,
to access two bytes.
Therefore, many prof_lonal Programmers use a character
variable rather than an Integer one when the range permitS'.
'iDcludea catdio.h>

int aain(voidl
(
int i;
char j;

i = 0;
for(j~l: j<lOl; j++) i • j + i;

printf(OTotal is: (dO, il;

return 0;
l
112 TEACH YOURSW

" c

:) t :1 . ' ••

I. Show how to declare an"unsigned "short int called loc_counter"

2. Write a program that prompts the user for a distance and
computes how long it takes light to travel that distance" Use an
unsigned long int to hold the distance. (Light travels at
approximately 186,000 miles per second.)
3. Write this statement another way:
short .int i;

EARN WHERE VARIABLES ARE

DECLARED

As you learned in Chapter 1, there are two basic places where a

variable will be declared: inside a function and outside all functions.
These variables are called local variables and global variables,
respectively. It is now time to take .a closer look at these two types
of variables and the scope rules that govern them.
Local variables (declared Inside a function) may be referenced only
by statements that are inside that function. They are not known
outside their own function. One of the most important things to
understand about local variables is that they exist only while the function
in which they are declared is executing. That is, a local variable is created
upon entry ·into its function and destroyed upon exit.
Since local variables are not known outside their own function, it is
perfectly acceptable for local variables in different functions to have
the same name. Consider the follOWing program:
.include <stdio.h>

void f1 (void), f2 (void) ;

int main(void)
(
f1 () ;

return 0;
)
•
 A CI 0IlR LOOK AT DATA TYPES, VARIABLES, AND EXPRESSIONII 1'" :-:
4.2 LEARN WHERE VARJABL£S ARE DECLARED
v

void fl (void)
(
int count;

for{count=O; count<10; .count++) f2();

)

void f2 (void)
(
int count;

for{count=O; count<10; count++) printf(~%d· count);

)

This program prints the numbers 0 through 9 on the screen ten times.
The fact that both functions use a variable called count has no effect
upon the operation of the code, Therefore, what happens to count
inside f2() has no effect on count in fI().
The C language contains the keyword auto, which can be used to
declare local variables. However, since all local variables are, by
det:,ult, assumed to be auto, it is virtually never used. Hence, you will
not see it in anv of the examples in this book
Withm a function, local variables can be deciared at the start of any
block They do not need to be declared only at the start of the block
that defines the function. For example, the following program is
perfectly valid:
#incl~je <stdio.h>

int main (void)

{
int i:

for(i=O; i<10; i++) {

i f (i==5) {
int j: /* declare j within the if block *

j ::: i * 10;
printf ("%d", j);

return 0;
114 TEACH VOURSElf
c

A variable declared within a block is known only to other code within

that block. Thus, j may not be used outside of its block. Frankly, most
C programmers declare all variables used by a function at the start of
the function's block because it is simply more convenient to do so.
This is the approach that will be used in this book.
Remember one important point: You must declare aU local variables
at the start of the block in which they are defined, prior to any
program statements. For example, the following is incorrect:
.include <stdio.h>

int main(void)
(
printf{~This program wo~'t compile.- ) ;
int ii / * this should come f i rst *'
i = 10:

return 0;
)

When a function is called, its local variables are created, and upon
its return, they are destroyed . This means that local variable s cannot
retain their values between calls.
The formal parameters to a function are also local variables. Even
though these variables perform the special task of receIving the v~lue
of the arguments passed to the function , they can be used like any
other local variable' within that function .
Unlike local variables, global variables are known throughout the
entire program and may be used by ony piece of code in the program .
Also, they will hold their value during the entire execution of the
program. Global variables are created by declaring them outside any
function. For example , consIder this program:
.include <stdio.h>

voj.d f1 (void) ;

int max; /* this is a global variable * j

int main (void)

{
max = 10;
 A CLOSER LOOK AT DATA TYPES, VARIABLES, AND EXPR£SSK)NS 115
'.2 LEARN WHERE VARlNlLES ARE DECU'oRED

fl ( I ;

return 0;
)

void f1 (void)
(
int i;

for{i=O; i<max; i++) printf("%d i);

)

Here, both main() and fI() use the global variable m ax. The maine )
function sets the value of max to 10, and fI ( ) uses this value to
control its for loop.

EXAMPLES

1. In C, a local variable and a global variable may have the same

name. For example, this is a valid program:
, ....
.1nclude <stdio.h>

void f1 (void);

int count; /* global count */

int main(void)
•
count = 10;
fl () ;
printf(."count in main(): %d\n", count);

return 0;
)

void fl(voidl
(
int count; /* local count */

count = 100;
printf(-count in fl() %d\n". count);
116 TfACH YOURSELF
•
c

The program displays this output:

count in f1 () : 100
count in main() : 10

In maine ), the reference to count is to the global variable.

Inside fl( ), a local variable called count is also defined. When
the assignment statement inside fl ( ) is encountered, the
compiler first looks to see if there is a local vanable called
count. Since there is, the local variable is used. not the global
one with the same name. That is, when local and global
variables share the same name, the compiler will always use the
local variable.
2. Global variables are very helpful when the same data IS used by
many functions in your program. However, you should always
use local variables where you can because the excessive use of
global variables has some negative consequences. First, global
variables use nlen10ry the entire,time your program 1S
executing. not just when they are needed. In situations where
memory is in short supply, this could be a problem. Second,
using a global where a local \'ariable will do makes a function
less general, because it relies on something that must be defined
outside itself. For example, here is a case where global variables
are being used for no reason:
#include <stdio.h>

int power (void) ;

int m t e;

int main(void)
(
m :;: 2;
e = 3;

printf(M~d raised to the %d power is %d", m. e, power(»;

return 0;
)

/ * Non-general version of power. "'/

int power(void)
 A CLOSIR LOOK AT DATA TYPH, VARIABLU, AND EXPRiIISIDNII 117
'.2 LEARN WHERE VARIABLES ARE DECIARBJ

(
int temp, temp2;

temp :; 1;
ternp2 = e;
fore ; temp2> 0; temp2--) temp:; temp· m;

return temp;
)

Here, the function power( ) is created to compute the value of

'm raised to the e'h power. Since m and e are global, the function
cannot be used to compute the power of other values. It can
only operate on those contained within m and e. However, if
the program is rewritten as follows, power( ) can be used with
any two values.
#include <stdio.h>

int power(int m, int e) i

int mai:'1(voidl

int m, e;
m = 2·,
e = 3;

printf("%d to the %d is %d\n". m, e, power(m, e»;

printf..( "4 to the 5th is %d\n", power (4, 5»;
print!{"3 to the 3rd is %d\n", power '(3, 3»;

return 0;
)

, - Parameterized version of power. */

int power(int m, int e)
(
int temp;

temp = 1:
fort e> 0; e--) temp = temp * m;

return temp;
)
11 • TfACH YOURSElf
. " C

By parameterizing power( ), you can use it to return the

result of any value raised to some power, as the program
now shows.
The important point is that in the non-generalized version,
any program that uses power( ) must always declare m and e
as global variables and then load them with the desired values
each time power( ) is used. In the parameterized form, the
function is complete within itself-no extra baggage need be
carried about when it is used.
Finally, using a large number of global variables can lead to
program errors because of unknown and unwanted side effects.
A major problem in developing large programs is the accidental
modification of a variable's value because it was used elsewhere
in the program. This can happen in C if you use too many glohal
variables in your programs.
3. Rememher. local variahles do not maintain their values between
functions calls. For example. the following program will not
work correctly:
#include <stdio.h>

int series (void) ;

int main(void)

int i;
•
for(i=O; i<10; i'++) printf("%d series(») ;

return 0;
)

/ * This is incorrect. *I
int series(void)

int total:

total = (total + 1423) % 1422;

return total;
)
 A CI.OSEII LOOK AT DATA TYP£S, VAR!AIIUS, A M I _ 119
'.3 TAK£ACLOSERLOOKATCONST/INTS
...

This program attempts to use series( ) to l:enerate a number

series in which each number is based upon the value of the .
preceding one. However, the value total will not be maintained
between function calls, and the function fails to carry out its
intended task.

EXERCISES

J. What are key differences between local and global variables?

2. Write a program that contains a function called soundspeed( ),
which computes the number of seconds it will take sound to
travel a speCified distance . Write the program two ways: first,
with soundspeed( ) as a non-general function and second, with
soundspeed( ) parameterized. (For the speed of sound , use
1129 feet per second).

TAKE A CLOSER LOOK AT CO~STANTS

Constants refer to fixed values that may not be altered by the program .
For example , the number 100 is a constant. We have been using
constants in the preceding sample programs without much fanfare
because , in moS( cases, their use is intuitive. However, the time has
come to cover them formally.
Integer constants are specified as numbers without fractional
components. For example, 10 and -100 are integer constants.
Floating-point constants require the use of the decimal point followed
by the number's fractional component. For example, J J. J 23 is a
floating-point constant. C also allows you to use scientific notation for
floating-point numbers. Constants using scientific notation must follow
this general form:

number E sign exponent

The sign is optional. Although the general form is shown with spaces
between the component parts for clarity, there may be no spaces
between the parts in an actual number. For example, the following
defines the value 1234.56 using >'Cientific notation:
120 TEACH YOUAIIlU'
c

123.456El

Character constants are enclosed between single quotes. For

example 'a' and '%' are both charKtEl"r constants. As some of the
examples have shown, this means that if you wish to assign a
character to a variable of type ""ar, you will use a statement similar to
ch = 'z·';

However, there is nothing in C that prevents you from assigning a

character v/l.riable a value using a numeric constant. For example, the
ASCII code for 'A' is 65. Therefore, these two assignment statements
are equivalent.
char chi

ch = 'A ' ;
ch = 65;
When y{)U enter numeric constants into your program. the compiler ·
must decide what type of constant they are. for example, is J 000 an
int, an unsigned, or a short? The rcason ,. .'e haven't worried about
this earlier is that C automatically converts the type of the right side of
an assignment statement to that of the variable on the left. (We will
examine this process more fully later in this chapter.) So, for many
situations it doesn't matter what the compiler thinks 1000 is. However,
this can be important when you use a constant as an argument to a
function, such'ils in a call to printf( ).
By default, the C compiler fits a numeric constant into the smallest
compatible data type that will hold it. Assuming 16-bit integers, 10 is
an int by default and 10(1003 is a long. Even though the value 10 could
be fit into a char, the compiler will not do this because it means
crossing type boundaries. The only exceptions to the small<:st-type
rule are floating-point constants, which are assumed to be doubles.
For virtually all programs you will write as a beginner, the compiler
defaults are perfectly adequate. However, as you will see later in this
book, there will come a point when you will need to specify precisely
the type of constant you want.
In cases where the assumption that C makes about a numeric
constant is not what you want, C allows you to specify the exact type
 II CLD5(R LOOK AT !lATA TYPES, VARIABLES, AND EXPRESSIONS 121
~
4.3 TAKE A CLOSER LODKAT CONSTANTS

by using a suffix. For floating-point types, if you follow the number

with an 'F', the number is treated as a float. If you follow it with an 'L',
the lIumber becomes a long double. For integer types, the 'U' suffix
stands for unsigned and the 'L' stands for long.
As you may know, in programming it is sometimes easier to use a
number system based on 8 or 16 instead oflO. As you learned in
Chapter 2, the number system based on 8 is called octal and it uses the
digits 0 through 7. The base-16 number system is called hexadecimal
and uses the digits 0 through 9 plus the letters 'A' through 'F', which
stand for IO through 15. C allows you to specify integer constants as
hexadecimal or octal instead of decimal if you prefer. A hexadecimal
constant must begin with 'Ox' (a zero followed by an x) then the
constant in hexadecimal form. An octal constant begins with a zero.
For example, OxAB is a hexadecimal constant, and 024 is an octal
constant. You may use either upper- or lowercase letters when
entering hexadecimal constants.
C supports one other type of constant in addition to those of the
predefined data types: the string. A string is a set of characters
enclosed by double quotes. You J,ave been working with strings since
Chapter I because both the printf( ) and scanf( ) functions use
them. Keep in mind one important fact: although C allo\\"s you to
de tine string constants, it does not formally have a·string data type.
Instead, as you will see a little later in this book, strings are supported
in Cas character arrays. (Arrays are discussed in Chapter 5.)
To display a string using printf( ) you can either make it part of
the control string or pass it as a separate argument and display it using
the %s format code. For example, this program prints Once upon a
time on the screen:
#include <stdio.h>

int m.a,in(void)
(

printf{"%s %5 %s", "Once", "upon", "a time");

return 0 i
)

Here, each string is passed to printf( ) as an argument and

displayed using the %8 specifier.
122 TEACH YOURSELF
~
C

,
EXAMPLES

1. To see why it is important to use the correct type ~pecifie1:

with printf( ), try this program. (It assumes that short
integers are 16 bits.) Instead of printing the number 42340,
it displays -23196, because it thinks that it is receiving a
signed short integer. The problem is that 42,340 is outside the
'range of a short into To make it work properly, you must use
the %hu specifier .
• include <stdio.h>

int main (void)

{
printf(~'hd·. 42340); /- this won't work right -/
c
return 0;
)

2. To see why you may need to explicitly tell the compiler what
type of constant you are using, try this program. For most
compilers, it will not produce the desired output. (Ifit does
work, it is only by chance. )
#include <stdio.h>

int main(void)
( .
printf("%f", 23091;

return 0;
)

This program is telling printf( ) to expect a floating point

value, but the compiler assumes that 2309 is simply an into
Hence, it does not output the correct value . To fix it, you must
specify 2309 as 2309.0. Adding the decimal point forces the
/
compiler to treat the value as a double .

•
 A CLOSER LOOK AT DATA TYPES, VARJABLES, AND EXPRESSIONS 123
4.4 INfTlALIZE VARIABLES

EXERCISES
1. How do you tell the C compiler that a floating-point constant
should be represented as a float instead of a double?
2. Write a program that reads anq writes a long int value.
3. Write a program that outputs I like C using three
separate strings.
"

NITlALlZE VARIA LES

A variable may be given an initial value when it is declared. This is

called mriable initializatiol1. The general form of variable initialization
is shown here:

type var-name = constant;

For example, this statentent declares count as an int and gives it an

initial valne of 100.
int count = 100;

The main advantage of using an initialization rather than a separate

assignment statement is that the compiler may be able to produce
faster code. Also, this saves some typing effort on your part.
Global variables may be initialized using only constants. Local
variables can be initialized using constants, variables, or function calls
as long as each is valid at the time of the initialization. However, most
often both global and local variables are initialized using constants.
Global variables are initialized only once, at the start of program
execution. Local variables are initialized each time a function .
is entered.
Global variables that are not explicitly initialized are automatically
set to zero. Local variables that are not initialized should be assumed
to contain unknown values. Although some C compilers automatically
initialize un-initialized local variables to 0, you should not count
on this.
124 TtACH YOURSW
c

EXAMPLES

I. This program gives i the initial value of -I and then displays

its value.
'include <stdio.h>

int main (void)

(
int i = -1;

printf(-i is initialized to %d~. i);

return 0;
)

2. When you declare a list of variables, you may initialize one or

more of them. For example, this fragment initializes min to 0
and max to 100. It does not initialize count.
int min=O, count. max=100;

3. As stated earlier, local variables are initialized each time the

function is entered. For this reason, this program prints 10
three times .
#include <stdio.h>

void f (void) ;

int main(void)
(
f{);
•
t{);
t () ;

return 0;
)

void f (void)
{
int i = 10;
 A CIIISEII LOOK AT DATA TYPES, VAlUABLES, AND EXPRESSIONS 125
...4 INmALiZE VARIABLES
...

4. A local variable can be initialized by any expression valid

at the time the ~ariable is declared. For example, consider
this program:
.include <stdio.h>

int x = 10; /* initialize global variable */

int myfunc (int i);

int main (void)

(
/* initialize a local variable using
a global variable */
int y = X;

/ * initialize a local variable using another

local variable and a function call */
int z = myfunc(y);
printfl"%d %d", y, z);

return 0;
}

int myfunc(int i)
(
return' i/2;
)

The local variable y is initialized using the value of the glQbal

variable x. Since x is initialized before main( ) is called, it is
valid to use its value to. initialize a local variable. The value of z
is initialized by calling myfllnc( ) using y as an argument.
Since y has alteady been initialized, it is entirely proper to use it
as an argument to myfllnc( ) at this point.

I. Write a program that gives an integer variable called i an initial

value of 100 and then uses i to control a for loop that displays
the numbers 100 down to 1.
12& TEACH YOURSELf
~ C

2. Assume that this line of code declares global variables. Is

it correct?
int a=l, b=2. c=a;

3. If the preceding declaration was for local variables, would

it be correct?

NOERSTANO TYPE CONVERSIONS IN

EXPRESSIONS

Unlike many other computer languages, C lets you mix different types
of data together in one expression. For example, this is perfectly valid
C code:
char chi
int i;
float fi
double outcome;

ch = ' 0';
i = 10;
f : 10.2;

outcome = ch * i / f;

C allows the mixing of types within an expression because it has a

strict set of conversion rules that dictate how type differences are
resolved. Let's look closely at them in this section.
One portion of C's conversion rules is called integral promotion. In C,
whenever a char or a short int is used in an expression, its value is
automatically elevated to int during the evaluation of that expression .
This is why you can use char variables as 'little integers' anywhere ,
you can use an int variable. Keep in mind that the integral promotion
is only in effect during the evaluation of an expression. The variable
does not become physically larger. (In essence, the compiler just uses
a temporary copy of its value.)
After the automatic integral promotions have been applied, the C
compiler will convert all operands 'up' to the type of the largest
operand. This is called type promotion and is done on an operation-
 A CLOSER LOOK AT DATA TYPES, VARIABLES, AND EXPRESSIONS 127
..,
4.5 UNDERSTAND TYPE CONVERSIONS IN EXPRESSIONS

by-operation basis, as described in the following type-conversion

algorithm.

IF an operand is a long double

THEN the second is converted to long double
ELSE IF an operand is a double
THEN the second is converted to double
ELSE IF an operand is a Hoat
THEN the second is converted to ftoat
ELSE II' an operand is an unsigned long
THEN the second is converted to unsigned long
ELSE IF an operand is long
THEN the second is converted to long
ELSE IF an operand is unsigned
THEN the second is converted to unsigned

There is one additional special case: If one operand is long and the
other is unsigned int, and if the value of the unsigned int cannot be
represented by a long, both operands are converted to unsigned long.
Once these conversion rules have been applied, each pair of
operands will be of the same type and the result of each operation will
be the same as the type of bot" operands.

, EXAMPLES

1. In this program, i is elevated to a float during the evaluation of

the expn;ssion i'f. Thus, the program prints 232.5.
#include <stdio.h>

int main(void)
(
int i;
float f:

i = 10;
f = 23.25;
printf(~%f~, i*f);

return 0;
128 TfACHYOURSW
c

2. This program illustrates how short ints are automatically

promoted to ints. The p~tf( ) statement works correctly even
though the %d modifier is used because i is automatically
elevated to int when printf( ) is called.
#include <stdio.h>

int main(void)
(
~ short int i;

i = -10:
printfC"'d", i);

return 0;
)

3. Even though the final outcome of an expression will be of the

largest type, the type conversion rules are applied on an
operation-by-operation basis. For example., in iM expression
100.0I(10/3)

the division of 10 by 3 produces an integer result, since both are

integers. Then this value is elevated to 3.0 to divide 100.0.

DEPC~.

I. Given these variables,

char chi
short i;
unsigned long ul;
float f;

what is the overall type of this expression:

f/ch - (i*ul)

2. What is the type of the subexpression i'ul, above?

 A CLOSER LOOK AT DATA TYPES. VARIABLES. AND EXPRESSIONS 129
4.6 UNDERSTAND TYPE CONVERSIONS IN ASSIGNMENTS
"
I
I 4.6 NDERSTAND TYPE CONVERSIONS IN
ASSIGNMENTS

In an assignment statement in which the type of the right side differs

from that of the left, the type of the right side is converted into that of
the left. When the type of the left side is larger than the type of the
right side, this process causes no problems. However, when the type
of the left side is smaller than the type of the right, data loss may
occur. For example, this program displays -24:
#include <stdio.h>

int main(void)

char chi
int i;

i =
1000;
ch = i;

printf ("%d", ch);

return 0;

The reason for this is that only the low·order eight bits of i are
copied into ch .. Since this sort of assignment type conversion is not an
error in C, you will receive no error message. Remember, one reason
C was created was to replace assembly language, so it must allow all
sorts of type conversions. For example, in some instances you may
only want the low-order eight bits of i, and this sort of assignment is an
easy way to obtain them.
When there is an integer-to-character or a longer-integer to
shorter-integer type conversion across an assignment, the ba'sic rule is
that the appropriate number of high-order bits will be removed. For
example, in many environments, this means 8 bits will be lost when
going from an int to a char, and 16 bits will be lost when going from a
long to an into
When convel1ing from a long double to a double or from a double
to a float, precision is lost. When converting from a floating-point
130 TEACH YOURSELF
... C

value to an integer value, the fractional part is lost, and if the number
is too large to fit in the target type, a garbage value will result.
Remember two important points: First, the conversion of an int to a
float or a float to double, and so on, will not add any precision or
accuracy. These kinds of conversions will only drange the form in
which the value is represented. Second, some C compilers will always
treat a char variable as an unsigned value. Others will treat it as a
signed value. Thus, what will happen when a character variable holds
a value greater than 127 is implementation-dependent. If this is
important in a program that you write, it is best to declare the variable
explicitly as either signed or unsigned.

EXAMPLES

I. As stated, when converting from a floating-point value to an

integer value, the fractional portion of the number is lost. The
follOWing program illustrates this fact. It prints 1234,0098 ]234.
.
#include <stdio.h>

int main(void)
(
int i;
float f;

f = 1234.0098;
i = f;· / * convert to int ,.. I
printf("%f %d", f. i);

return 0;
)

2. When converting from a larger integer type to a smaller one, it

is possible to generate a garbage value, as this program
illustrates. (This program assumes that short integers are 16 bits
long and that long integers are 32 bits long.)
#include <stdio.h>

int main(void)
(
short int si;
long int Ii:
 A CLOSER LOOK AT DATA TVPfS, VARIABLES, AND EXPRESSIONS
4.6 UNDERSTAND TYPE CO/NfRSIONS IN ASSIGNMENTS
.,
131

U = 100000;
si = Ii; j* convert to short int */

printf("%hd", 5i);

return 0;
)

Since the largest value that a short integer can hold is 32,767, it
cannot hold 100,000. What the compiler does, however, is copy
the lower-order J 6 bits of Ii into si. This produces the
meaningless value of -31072 on the screen.

EXERCISES

J. What will this program display?

#include <stdio.h>

int main(voidl
(
int i;
long double Id;

1d = 10.0;
i = Id;

printf'("%d", il;
)

2. What does t~is program display?

~include <stdio.h>

int main(void)
(
float f;

f = 10 I 3,
printf("%f", f);

return 0;
)
132 TEACH YOURSElF

'" C

OGRAM WITH TYPE CASTS

Sometimes you may want to transform the type of a variable

temporarily. For example, you may want to use a floating-point value
for one computation, but wish to apply the modulus operator to it
elsewhere. Since the modulus operator can only be used on integer
values, you have a problem. One solution is to create an integer
variable for use in the modulus operation and assign the value of the
floating-point variable to it when the time comes. This is a somewhat
inelegant solution, however. The other way around this problem is to
use a type cast, which causes a temporary type change.
A type cast takes this general form:

(type) value

where type is the name of a valid C data type. For example,

float f;

f = 100.2;
;--- print f as an integer */
printf(~%d". (int) f);

Here, the type cast causes the value of f to be converted to an into

EXAMPLES

I. As you learned in Chapter I, sqrt( ) , one of C's library

functions, returns the square root of its argument. It uses the
MATH.H header file. Its single argument must be of type
double. It also returns a double value. The following program
prints the square roots of the numbers between 1 and 100 using
a for loop. It also prints the whole number portion and the
fractional part of each result separately. To do so, it uses a type
cast to convert sqrt( )'s return value into an into
 A CLOSER LOOK AT DATA TYPES, VARIABLES, AND EXPRESSIONS
4.7 PROGRAM WITH TYPE CASTS
.,
133

#include <stdio.h>
#include <math.h>

int main(void)
{
double i ;

for (i=l. 0; i<101. 0; i++ I {

printf("The square root of %1f is %If\n'', i. sqrt(il I;
printf ("vI,lhole number part: %d ", (int) sqrt (i I) ;
printf("Fractional part: %If\n'', sqrt(il-(int)sqrt(i));
printf (" \n") ;
}

return 0;
}

2. You cannot cast a variable that is un the lett side of an

assigmTIcnt statement. For example. this is an in\'.Jlid
statement in C:
int nurn;

( float) nurn = 123.23; / - this is incorrect */

UERCISES
I. Write a program that uses fo r to print the numbers 1 to 10 b"
tenths. Use a floating-point variable to control the loop.
However, use a type cast so that the conditional expression is
evaluated as an integer expression jn the interest of speed.
2. Since a floating point value cannot be used with the % operator,
how can you fix this statement?
x = 123 . 23 % 3; / * fix this statement * /
 .,
134 TEACH YOURSELF

At this point you should be able to answer these questions and

perform these exercises:

1. What are C's data-type modifiers and what function do

they perform?
2. How do you explicitly define an unsigned constant, a long
constant , and a long double constant?
3. Show how to give a float variable called balance an initial value
of 0 .0.
4 . What are C's automatic integral promotions?
j . What is the difference hetween a signed and an unsigned
integer?
6 . Give one reason why you might want to use a global variable in
your progra m.
7. Write a program that contains a function called senese ). Have
this function generate a series of numbers, based upon this
formula:

next-num~er = (previous-number' 1468) % 467

Give the number an initial value of 21. Use a global variable to
hold the last value between function calls. In maine )
demonstrate that the function works by calling it ten times and
displaying the result.
8. What is a type cast? Give an example.
 A CLOSER LOOK AT DATA TYPES. VARIABLES. AND EXPRESSiONS 135
4.7 PROGRAM WITH TYPE CASTS
"

This section checks how well you have ilHegrated the material in
this chapter with that from earlier chapters.

1. As you know from Chapter 3, no two cases with the same

switch may :Jse the same value. Therefore. is this switch valid
or invalid? Why? (Hint: the ASCII code for 'A' is 65.)
switch(x) {
case 'A' : printf("is an A");
break;
case 65 : printf (" is the number 65") i
break;
}

2. Technically, far traditional reasons the getchar( ) and getche( )

functions are declared as returning integers, not character
values. However, the character read from the kevboard is
contained in the low-order byte. Can you explain why this value
can be assigned to char vilriables?
3. In this fragment, will the loop ever terminate? Why? (Assume
integers are 16 bits long.)
int .i
for(i=O; i<33000; i++);
5
Exploring Arrays
and Strings

dI.........&.II_

5.1 Declare one-dimensional arrays \

5.2 Use strings 'I
5.3 Create multidimensional arrays "2-
5.11 Initialize arrays :l
5.5 Build arrays of strings 5

137
...
138 TEACH YOURSELF •
c

N this chapter you will learn about arrays. An array is essentially

a list of related variables and can be very useful in a variety of
situations. Since in C strings are simply arrays of characters, you
will also learn about strings and several of C's string functions.

B~fore proceeding, you should be able to answer these questions

and perform these exercises:

I. What is the difference between a local and a global variable?

2. What data type will a C compiler assign to these numbers?
(Assume 16-bit intege~s.)
a. 10
•
b. 10000
c. 123.45
d. 123564
e. -45099
3. Write a program that inputs a long, a short, and a double and
then writes these values to the screen.
4. What dOBS a type cast do?
5. To which if is the else in this fragment associated? What is the
general rule?
if (i)
if(j) printf(~i and j are true");
else printf{"i is false-);

6. Using the following fragment, what is the value of a when i is I?

What is a's value when i is 4?
switch(i) {
case L a = 1·,
case 2, a = 2;
break;
case 3, a = 3;
break;
 Exploring _ ond Strings 139
5.' DECLAAE ONE·DIMENSIONAL ARRAYS

case 4:
case 5: a = 5:
)

ECLARE ONE-DIMENSIONAL ARRA YS

( In C, a one-dimensional array is a list of variables that are all of the

same type and are accessed through a common name. An individual
variable in the array is called an arTay elem<J1t) Arrays form a
convenient way to handle groups of related data.
( To declare a one-dimensional array, use the general form

type va,_name[s;ze);
·<there type is a valid C data type, var_name is the name of the array,
~nd size specifies the number of elements in the arraY) For example, to
declare an integer array with 20 elements called myarray, use this
staren1ent.

int myarray[201;

(An array element is accessed by indexing the array using the

number of the element. In C, all arrays begin at zero) This means that
if you want to access the first element in an array, use zero for the
index .(To index an array, specify the index of the element you want
inside square brackets! For example, the following refers to the second
element of myarray:
myarray (1]

Remember, arrays start at zero, so an index of I references the second

element.
( TO assign an array element a value, put the array on the left side of
an assignment statemenl)For example, this gives the first element in
myarray the value 100:
myarray[Oj = 100;

(Cstores one-dimensional arrays in one contiguous memory location

with the first element at the lowest addressJFor example, after this
fragment executes,
142 TEACH YOURSElf
~
C

printf(ftAverage temperature: %d\n-, avg/days);

I * find min and max wI

min 200;
= /* initialize min and max */
max = 0;
for(i=O; i<days; i++) (
if(min>temp[i]) min = temp[i];
ifCmax<temp[i]) max = temp[i};

printf(~Minimum temperature: %d\n-, min);

printf ("Maximum temperature: %d\n· I max);

return 0 i
)

2. As stated earlier, to copy the contents of one array to another,

you must expliCitly copy each element separately. For example,
this program loads al with the numbers I through 10 and then
copies them into a2.
#include <stdio.h>

int main(voidl
{
int a1[10], a2[10];
int i;

for(i:;::1; i < 11 ; i++ ) a1 [i-11 =i ;

for(i=O: i<10; i.+) a2[il = a1 [i] ;

for(i=O; i<10; i++) printf (-%d • , a2[il];

return 0;

3. The following program is an improved version of the code-

machine program developed in Chapter 3. In this version, the
user first enters the message, which is stored in a character
array. When the user presses ENTER, the entire message is then
encoded by adding I to each letter.
 bploring _ end String. 1 43
5.1 DECLARE ONE-DIMENSIONAL ARRAYS
"

#include <stdio.h>
#include <conio.h>

int main(void)
(
char mess [80J;
int i;

printf(MEnter message (less than 80 characters)\n") i

for(i=O; i<80; i++) (
mess[i] = getche();
if(mess(i]=='\r') break;
)
printf("\n~) ;

for(i=O; mess{i] !='\r'; i-++) printfC"%c", mess[i]+l);

return 0;

4. Arrays are especially useful when you want to sort information.

For example, this program lets"the user enter up to 100
numbers and then sorts them. The sorting algorithm is the
bubble sort. The bubble sort algorithm is not very efficient, but
it is simple to understand and easy to code. The general concept
behind the bubble sort, indeed how it got its name, is the
repeated comparisons and, if necessary, exchanges of adjacent
elements. This is a little like bubbles in a tank of water with
each bubble, in turn, seeking its own level.
'include <stdio.h>
#include <stdlib.h>

int main (void)

(
int item[lOO];
int a, b, t;
lnt count;

/ * read in numbers */
printfC"How many numbers? ");
scanf("%d", &count);
for(a=O; a<count; a++) scanf("%d", &item[a});
144 TEACH YOURSELF

'" C

j* now, sort them using a bubble sort */

for(a=l; a<count: ++a)
for(b=count-!; b>=a; --b) (
/* compare adjacent elements */
if(item[b-l) > item[b)) (
/* exchange elements */
t = item[b-l);
item[b-l] = item[bJ;
item[b] = ti
)
)

/* display sorted list */

for(t=O; t<cQunt; t+.) printf{-%d· item[t);

return 0;
)

EXERCISES
). What is wrong with this program fragment?
#include <stdio.h>

int main(void)

int i: count[lO};

for(i=O: i<100; i++) (

printf("Enter a number: .);
scanf("%d", &count[i]);
)

2. Write a program that reads ten numbers entered by the user and
reports if any of them match.
3. Change the sorting program shown in the examples so that it
sorts data of type float.
 Exploring Armys and Strings 145
5.2 USE STRINGS
T

E STRINGS

The most common use of the one-dimensional array in C is the string.

Unlike most other computer languages, C has no built-in string data
type. Instead,Ca string is defined as a m<ll-temlloated character array. In
C, a null is zero. The fact that string must be terminated by a null
means that you must denne the array that is going to hold a string to
be one byte larger than the largest string it will be required to hold, to
make room for the null. A string constant is null-terminated by the
compiler automatically.)
(There are several ways to read a string from the keyboard)The
method we will use in this chapter employs another of C's standard
library functions: gets( ). Like the other standard 110 functions,
gets( ) also uses the STDlO.H header t1le.(To use gets( ), call it using
the name of a character array without any index. The gets( ) function
reads characters until you press ENTER. The ENTER key (Le., carriage
return) is not stored, but is replaced by a null, which terminates the
string. For example, this program reads a string entered at the
keyboard. It then displa\'s the comems of that string one character
at a time.
#include <stdio.h>

int main(void)
{
char str[80] i
int i;

printf(~Enter a string (less than 80 chars): ");

gets (str) i
for(i=O; str[ili iTT) printf(~'tc", str[i);

return 0;

Notice how the program uses the fact that a null is false to control the
loop that outputs the string.)
There is a potential problem with gets( ) that you need to be aware
of. The gets( ) function performs no bounds checking, so it is possible
for the user to enter more characters than the array receiving them
can hold. For example, if you call gcts( ) with an array that is 20
characters long, there is no mechanism to stop you from entering
146 TEACH YOURSELF

'" C

more than 20 characters. If you do enter more than 20 characters, the

array will be overrun. This can obviously lead to trouble, including a
program crash. Later in this book you will learn some alternative ways
to read strings, although none are as convenient as using gets( ). For
now, just be sure to call get8( ) with an array that is more than large
enough to hold the expected input.
In the previous program, the string that was entered by the user
was output to the screen a character at a time. There is, of course, a
much easier way to display a string using printf( ), as shown in this
version of the program:
#include <stdio.h>

int main{void)
{
char str [80] ;

printf(REnter a string (less than 80 chars): -);

gets (str);
printf(str); / ~ output the string -/

return 0;
)

Recall that the first argument to printf( ) is a string. Since str

contains a string it can be used as the first argument to printf( ).
The contents of str will then be displayed.
If you wanted to output other items in addition to str, you could
display str using the %8 format code. For example, to output a newline
after str, you could use this call to printf( ).
printf(-'s\n-, str);

This method uses the %s format specifier followed by the .newline

character and uses str as a second argument to be matched by the %s
speCifier.
The C standard library supplies many string-related functions.
The four most important are strcpy( ), strcat(), strcmp( ), and
strIen( ). These functions require the header file STRlNG.H. Let's
look at each n,o\\'.
The strcpy( ) function has this general form:

strcpy(to, from);
 Exploring _ and Strings 147
5.2 USE STRlNGS

•
It copies the contents of from to to. The contents of from are
unchanged. For example, this fragment copies the string "hello" into
str and displays it on the screen:
char str[80);

strcpy (str, "hello" 1 ;

printf("%s", str);

The strcpy( ) function performs no bounds checking, so you must

make sure that the array on the receiving end is large enough to hold
what is being copied, including the null terminator.
The strcat( ) function adds the contents of one string to another.
This is called concatel1anol1. Its general form is

streatCto, from);
It adds the contents of from to the contents of to. It performs no
bounds checking-. so you must make sure that to is large enough to
hold its current contents plus what it will be receiving. This fragment
displays hello there.
char str[BO];

strcpy{str, "hello"};
strcat(str. " there");
printf(str);

The strcmp{ ) function compares two strings. It takes this general

form:

strempes 1, s2);
It relurns zero if the strings are the same. It returns less than zero if 51
is less than 52 and greater than zero if 51 is greater than s2. The strings
are compared lexicographically; that is, in dictionary order. Therefore,
a string is less than another when it would appear before the other in a
dictionary. A string is greater than another when it would appear after
the other. The comparison is not based upon the length of the string.
Also, the comparison is case-sensitive, lowercase characters being
greater than uppercase. This fragment prints 0, because the strings are
the same:
printf("%d", strcmp("one", "one"»;
 .,
148 TEACH YOURSELF

The strlen( ) function returns the length , in characters, of a string.

Its general form is

strlen(str);

The strlen( ) function does not count the null terminator. This means
that ifstrlen() is called using the string 'test', it will return 4.

EXAMPLES
1. This program requests input of two strings, then demonstrates
the four string functions with them.
#include <str1ng.h~

#include <stdio.h~

inc main(void)

char strl[BO], str2(BO];

int i;

printf("Enter the first string: .) i

gets (strl);
printf("Enter the second string: ");
gets(str2) ;

/* see how long the strings are */

printf("%s is %d chars 10ng\n", strl, strlen(strl»;
printf("%s 1S %d chars 10ng\n-, str2, strlen(str2»;

/*compare the strings */

i = strcmp(str!, str2);
if(!il printf(-The strings are equa1.\n-);
eh:e if(i<O) printf(-%s is less than %s\n", strl, str2);
else printf("%s is greater than %s\n-, strl, str2);

1*concatenate str2 to end of strl if

there is enough room * /
if(strlen(strl) + strlen(str2) < 80) {
st:-cat (str!, str2);
printf ("%s\n", strl) i
 bploring Anap and String. 149
5.2' USE STRINGS
'"
)

/* copy str2 to strl */

strcpy(strl, str2);
pr'intf("%s %s\n", strl, str2);

return 0:
)

2. One common use of strings is to support a command-based

inte'ince. Unlike a menu. which allows the user to make a
selection, a command-based interface displays a prompting
message , waits for the user to enter a command, and then does
what the command requests. Many operating systems, such as
Windows or DOS, support command-line interfaces, for
example. The following program is similar to a program
developed in Section 3.1. It allows the user to add. subtract,
multiply, or divide, but does not use a menu. Instead, it uses a
command-based interface.
#include <stdlib.h>
iinclude <stdio.h>
#include <string.h>

int main(void)

char command(80], temp(80];

int i. j;

for ( ; ; ) (
printf(MOperation? ") i
gets (command);

/ * see if user wants to stop * /

if (!strcmp(command. "quit")) break;

printf(MEnter first number: "J;

gets (temp) ;
i = atoi (temp) i

printf("Enter second number: H);

gets (temp) ;
j = atoi (temp);
150 TEACH youRSELF

'" C

j'" now, perform the operation *j

if (!strcmp(command, ~ac:id"»
printf("%d\n", i+j);
else if ( ! strcmp (comrnar.d. .. subtract" ) )
printf("%d\n", i-jl;
else if(!strcmp(comrnand. "divide"» {
if(jl printf("%d\n", iijl;
I
else if{!strcmp(command, "multiply"»
printf("%d\n", i*j);
else printf ("Unknown command. \n");
I

return 0:
I

Notice that this example also introduces another of Cs standard

library functions: atoi( ). The atoi( ) function returns the
integer equivalent of the number represented by its string
argument. For example, atoi("100") returns the value 100. The
reason that scanf( ) is not used to read the numbers is because,
in this context, it is incompatible with gets( ). (You will need to
know more about C before you can understand the cause of this
incompatibility.) The atoi( ) function uses the header file
STDLlB.H.

3. You can create a zero-length string using a strcpy( ) statement

like this:
strcpy(str, "");

Such a string is called a nul/ string. It contains only one element:

the null terminator.

EXERCISES

1. Write a program that inputs a string, then displays it backward

on the screen.
2. What is wrong with this program?
 Exploring Amtys and Strings 151
~
5.3 CREATE MULnOIMENSIONAL ARRAYS

#include <string.h>
#include <stdio.h>

int main(void)
{
char str[5];

strcpy(str. "this is a test");

printf (str):

return 0:
}

3. Write a program that repeatedly inputs strings. Each time a

string is input, col1catenate it with a second string called bigstr.
Add newlines to the end of each string. If the user types quit,
stop inputting and display bigstr (which will contain a record of
all strings input). Also stop ifbigstr will be overrun by the next
concatenation.

EATE MUL TlO/MENS/DNAL ARRA YS

CIn addition to one-dimensional arrays, you can create arrays of two or

more dimensions. for example, to create a I Oxl2 two-dimensional
integer array called count, you would use thi·s statement:
int count[10] [12];)

As you can see, to add a dimension, you simply specify its size inside
square brackets.
A two-dimensional array is essentially an array of one-dimensional
arrays and is most easily thought of in a row, column format. for
exampl~(given a 4x5 integer array called two_d, you can think of it
looking like that shown in figure 5-1. Assuming this conceptual view,
a two-dimensional array is accessed a row at a time, from left to right.
This means that the rightmost index will change most quickly when
the array is accessed sequentially from the lowest to highest memory
address)
152 TEACH YOURSElF
y
C

fiGURE 5-2

A conceptual view 0 I 2 3 4
ofa 4x5
two-dimensional 0
array
I
y
2

Two-dimensional arrays are used like one-dimensional ones. for

example, this program loads a 4x5 array with the products of the
indices, then displays the array in row, column formal.
#include <stdio.h>

int main ("Joid)

{
int twod{4) (S);
int i. j;

for(i=Oi i<4; 1+ .... )

for(j=:Oi j<5; j++)
twod [i J [j J = i * j ;

for{i=O; i<4; i++) {

for(j=O; j<5; j++)
printf("%d", twod{i){j));
printfC-'n-) ;
)

return 0;
)

The program output looks like this:

0 0 0 0 0
0 1 2 3 4
0 2 4 6 8
0 3 6 9 12
 Exploring _ - SCrIngo 153
5.3 CREATE MULnOIMENSJDNAL ARRA\'S

LTO create arrays of three dimensions or greater, simply add the size
of the additional dimension)For example, the following statement
creates a 10x12x8 three-dimensional array.
float values[10) [12) [B);

A three-dimensional array is essentially an array of two-dimensional

arrays.
( You may create arrays of more than three dimensions, but this is
seldom done because the amount of memory they consume increases
exponentially with each additional dimension. For example, a
1DO-character one-dimensional array requires 100 bytes of memory. A
1OOx1 00 character array requires 10,000 bytes, and a 1OOx1 OOxlOO
array requires 1,000,000 bytes. A 100xlOOxl00xlOO four-dimensional
array would require 100,000,000 bytes of storage-large even by
today's standards)

EXAMPLE

I. A good use of a two-dimensional array is to manage lists of

numbers. For example, you could use this two-dimensional
array to hold the noontime temperature for each day of the
year, grouped by month.
float yeartemp[121 [311;

In the same vein, the following program can be used to keep

track of the number of points scored per quarter by each
member of a basketball team .
• include <stdio.h>

int main(void)
[
int bball(4115);
int i. j:

for(i=O; i<4; i++)

for(j=O; j<5: j++) {
printf(·Quarter %d, player %d, ., i+1. j+l);
printf(-Enter number of points: -);
154 TEACH YOURSELF
". C

scanf ("%d", &bball [il [i I l;

)

j *display results . j
for ( i=O; i<4: i++J
for(j=O; j<S; j++) (
printf("Quarter %d. player %d, i+l. j+l ) i

printf("%d \ n", bball[il [i l l;

)

return 0;
)

EXERCISES

1. Write a program that defines a 3x3x3 three-dimensional array,

and load it with the numbers I to 27.
2. Have the program from the first exercise display the sum of its
elements.

INITIALIZE ARRAYS
Like other types of variables, you can give the elements of arrays
initial values. This is accomplished by specifying a list of values the
array elements will have.(The general form of array initialization for
one-dimensional arrays is shmvn here:

type array-name[sizej = (value-list I)

The value-list is a comma-separated list of constants that are type
compatible with the base type of the array. Moving from left to right,
the first constant will be placed in the first position of the array, the
second constant in the second position, and so on. Note that a
semicolon follows the }. In the following example, a five-element
integer array is initialized with the squares of the numbers I through 5.
 Exploring Amoys and Strings 155
504 INmALIZE ARRAYS

int i[5J = (1, 4, 9, 16, 25J;

This means that i[O) will have the value 1 and i[4) will have the
value 25 .
./ (You can initialize character arrays two ways. First, if the array is not
holding a null-terminated string, you simply specify each character
using a comma-separated list. For example, this initializes a with the
letters 'A', 'B', and 'C'.

char a[3J = ('A', 'B', 'C'J;

If the character array is going to hold a string, you can initialize the
array using a quoted string, as shown here:
char narne[Sl = HHerb~;

Notice that no curly braces surround the string. They are not used in
this form of initialization. Because strings in C must end with a null,
you must make sure that the array you declare is long enough to
include the null. This is why name is 5 characters long, even though
'Herb' is only 4. When a string constant is used, the compiler
automatically supplies the null terminator.)
..,... (Multidimensional arrays are initialized in the same way as
one-dimensional arrays. For example, here the array sqr is initialized
with the values 1 through 9, using row order:
int sqr[3J [3J = (
1, 2, 3,
4, 5, 6,
7, 8, 9
J;

This initialization causes sqr[O) [0) to have the value 1, sqr[O) [1) to
contain 2, sqr[O) [2) to hold 3, and so forth.)
(If you are initializing a one-dimensional array, you need not specify
the size of the array-simply put nothing inside the square brackets. If
you don't specify the size, the compiler counts the number of
initializers and uses that value as the size of the array)For example,
int pwr[] = {1, 2, 4, 8, 16, 32, 64, 128J;

causes the compiler to create an initialized array eight elements long.

(Arrays that don't have their dimensions explicitly speCified are called·
ul1sized arrays)An unsized array is useful because the size of the array
156 TEACH YOURSElf

'" C

•
will be automatically adjusted when you change the number of its
initializers. It also helps avoid counting errors on long lists, which is
especially important when initializing strings. For example, here an
unsized array is used to hold a prompting message.
char prompt ~~ter y~)Ur name: •i

If, at a later date, you wanted to change the prompt to "Enter your last
name:", you would not have to count the characters and then change
the array size. The size of prompt would automatically be adjusted.
/(Unsized array initializations are not restricted to one-dimensional
arrays. However, for multidimensional arrays you must specifY all but
the leftmost dimension to allow C to index the array properly. In this
way you may build tables of varying lengths with the compiler
allocating enough storage for them automatically. For example, the
declaration of sqr as an unsized array is shown here:
int sqr[J [3J = (
1, 2, 3.
4. 5. 6,
7: B. 9
J;

The advantage to this declaration over the sized version is that tables
may be lengthened or shortened without changing the array
dimensions)

EXAMPLES

1. A common use of an initialized array is to create a lookup table.

For example, in this program a 5x2 two-dimensional array·is
initialized so that the first element in each row is the number of
a file server in a network and the second element contains the
number of users connected to that server. The program allows a
user to enter the number of a server. It then looks up the server
in the table and reports the number of users .
• include <stdio.h>

int main (void)

 Exploring Arrays end Strings
5.4 INITIALIZE ARRA YS
.,
157

2, 28,
3, 19,
4, 8,
5, 15
) ;

int server;
int i;

printf(MEnter the server number: -);

scanf("%d-, &server);

/ * look it up in the table * /

for(i=O; i<5; i++)
if (server == ServerUsers [i 1 (01) (
printf(MThere are %d users on server %d.\nM,
ServerUsers [i] [1]. server);
break;
)

/ * report error if not found * /

if(i==5) printf("Server not llsted.\n");

return . 0;
)

2. Even though an array has been given an in itial value, its

contents "may be changed. For example, this program prints
hello on the screen .
• include <stdio.h>
.include <string . h>

int main(void)
{
char str(BO) = -I like C·;

strcpy (str. "hello·);

printf (str);

return 0;
)
158 TEACH YOURSELF
.., C

As this program illustrates, in no way does an initialization fIx the

contents of an array.

EXERCISES

1. Is this fragment correct?

int balancer] = 10.0, 122.23, 100.0;

2. Is this fragment correct?

#include <stdio.h>
#include <5tring.h>

int ma~n(void)

char name[] = -Tom";

strcpy{name, ftTom Brazwell-) i

3. Write a program that initializes a IOx3 array so that the fIrst

element of each row contains a number, the second element
contains its square, and the third element contams its cube.
Stan with 1 and stop at 10. For example, the fIrst few rows will
look like this:

1, 1, 1,
2,4,8,
3,9,27,
4,16,64,

Next, prompt the user for a cube, look up this value in the table,
and report the cube's root and the root's square. Use an unsized
array so that the table size may be easily changed
 Exploring An'eys and Strings
5.5 BUILD ARRAYS OF STRINGS
,.
159

UILO ARRAYS OF STRINGS

(Arrays of strings, often called string tables, are very common in C

programming. A string table is created like any other two-dimensional
array. However, the way you think about it will be slightly different.)
For example, here is a small string table. What do you think it defines?

(char names(lO] (40];

This statement specifies a table that can contain 10 strings, each up to

40 characters long (including the null terminator). To access a string
within this table, specify only the left-most index. For example, to read
a string from the keyboard into the third string in names, use this
statenlent:
gets(narnes(2]);

By the same token, to output the first string, use this printf( )
statement:
printf (names! 0 J ) ;)

(The declaration that follows creates a three-dimensional table with

three lists of strings. Each list is five strings long, and each string can
hold 80 characters.
char animals[3J (5) [80J;

To access a specific string in this situation, you must specify the two
left-most indexes. For example, to access the second string in the third
list, specify animals[2)[l].)

EXAMPLES

1. This program lets you enter ten strings, then lets you display
them, one at a time, in any order you choose. To stop the
program, enter a negative number.
#include <stdio.h>

int main{void)
160 TEACH YOURSELF

'" C

(
char text [10J [80J;
int i;

for{i=O; 1<10; i++) {

printf("%d: ., i+1);
gets{text[iJ I;
}

do
printfC"Enter number of string (1-10) : ");
scanf("%d", &i);
1--; j* adjust value to match array index * 1
i£(i>=O && i<10) printf("%s \ n", textli);
} while (i>=O);

return 0;
}

2. You can initialize a string table as you would any other type of
array. For example. the following program uses an initialized
string table to translate between German and English. Notice
that curly braces are needed to surround the list. The only time
they are not needed is when a single string is being initialized .
/* English-to-German Translator. */

_include <stdio.h>
#include <string.h>

char words [J [2 J [40 J = {

"dog", "Hund".
"110·. "nein",
"year", "Jahr".
·child". "Kind".
"r", "Ich" ,
"drive". "fahren" I

"house", "Haus".
"to·, ·zu·,
.... . . .
};

int main(void)
{
char engllsh[80] ;
 Exploring Arrays and Strings 161
'f'
5.5 BUlW ARRA YS OF STRINGS

int i;

printf{MEnter English word: ~);

gets (english);

/* look up the word */

i ': 0;
/* search while null string not yet encountered */
while(strcmp{words[iJ to], ""» {
if(!strcmp(english, words[i] [0]»
printf(~German translation: %sn, words(i][l]);
break;
)
i++;
)
if (!strcmp(words [i] [0], ~"»
printf ("Not in dictionary\n") ;

return 0;
)

3. You can access the individual characters that comprise a string

within a string table by using the rightmost index. For example,
the following program prints the strings in the table one
character at a time.
#include <stdio.h>

int main (voi-d)

{
char text[J [80) = {
"when", -in", "the~,
"course", "OfM, "human",
~events", ,,~

) ;

int if j;

/* now, display them */

for(i=O; text[ij [0]; i++)
for{j=O; text[il [j]; j++)
printf("%c", text[ij(j]);
printf(" ");
)
,.
162 TEACH YOURSELF

return 0:
)

EXERCISE
1. Write a program that creates a string table containing the
English words for the numbers 0 through 9. Using this table,
allow the user to enter a digit (as a character) and then have
your program display the word equivalent. (Hint: to obtain an
index into the table, subtract '0' from the character entered.)
. "

At this point you should be able to perform these exercises and

answer these questions:
1. What is an array?
2. Given the array
int coul1t[lO) i

will this statement generate an error message?

for(i=O; i<20; i++J count[i] =~;

3. In statistics, the mode of a group of Lumbers is the one that

occurs the most often. For example, given the list 1, 2, 3, 6, 4, 7,
5,4,6, 9, 4, the mode is 4, because It occurs three times. Write a
program that allows the user to enter a list of 20 numbers and
then finds and displays the mode.
4. Show how to initialize an integer array called items with the
values 1 through 10.
5. Write a program that repeatedly reads strings from the keyboard
until the user enters quit.
6. Write a program that acts like an electronic dictionary. If the
user enters a word in the dictionary, the program displays its
 Exploring Arrays and Strings 163
5.5 BUILD ARM Y.S OF STRINGS
...

meaning. Use a three-dimensional character array to hold the

words and their meanings.

This section checks how well you have integrated the material in
this chapter with that from earlier chapters.
I. Write a program that inputs strings from the user. If the string is
less than 80 characters long, pad it with periods. Print out the
string to verify that you have correctly lengthened the string.
2. Write a program that inputs a string and then encodes it by
taking the characters from each end, starting with the left side
and alternating, stopping when the middle of the string has been
reached. For example, the string "Hi there" would be "Heir eth".
3. Write a program that counts the number of spaces, commas, and
periods in a string. Usey switch to categorize the characters."
4. What is wrong with this fragment?
char str[BO);
str = getchar();

5. Write a program that plays a computerized version of Hangman.

In the game of Hangman , you are shown the length of a magic
word (using hyphens) and you try to guess what the word is by
entering letters. Each time you enter a letter, the magic word is
checked to see if it contains that letter. If it does, that letter is
shown. Keep a count on the number ofletters entered to
complete the word. For the sake of simplicity, a player wins
when the magic word is entirely filled by characters using 15 or
fewer guesses. For this exercise make the magic word
·concatenation. "
\
 6
Using Pointers

--
, chapter objectives

6.1 Understand pointer basics

6.2 Learn restrictions to pointer expreSSions

6.3 Use pointers with arrays
6.4 Use pointers to string constants
6.5 Create arrays of pOinters
6.6 Become. acquainted with multiple indirection
6.7 Use pointers as parameters

c'
165
y
166 TEACH YOURSELF

• C

I S chapter covers one of C's most important and sometimes

most troublesome features: the pointer. A pointer is basically
the address of an object. One reason that pointers are so
important is that much of the power of the C language is
derived from the unique way in which they are implemented.
You will learn about the special pointer operators, pointer arithmetic,
and how arrays and pointers are related. Also, you will be introduced
to using pointers as parameters to functions.

Review
Skills Check

Before proceeding, you should be able to answer these questions

and perform these exercises:

I. Write a program that inputs 10 integers into an array. Then

have the program display the sum of the even numbers and the
sum of the odd numbers.
2. Write a program that simulates a log-on to a remote system . The
system can be accessed only if the user knows the password,
which in this case is 'Tristan: Give the user three tries to enter
the correct password. If the user succeeds, simply print Log-on
Successful and exit. If the user fails after three attempts to
enter the correct password, display Access Denied and exit.
3. What is wrong with this fragment>
char name[lO] = ftThomas Jefferson-;

4. What is a null string?

5. What does strcpy( ) do? What does strcmp( ) do?
6. Write a program that creates a string table consisting of names
and telephone numbers. Initialize the array with some nam~s of
people you know and their phone numbers. Next, have the
program request a name and print the associated telephone
number. In other words, create a computerized telephone book .
 6.7
USING POINTERS

UNDERSTAND POINTER BASICS

,.
167

DER S T AND POINTER BASICS

(A pointer is a variable that holds the memory address of another

object)For example, if a variable called p contains the address of
another variable called q, then p is said to point to q . Therefore if
q is at location 100 in memory, then p would have the value 100 .
. (To declare a pointer variable, use this general for m:

type 'var-name;
Here, type is the base rype of the pointer. The base type specifies the
type of the object that the pointer can point to)Notice that the variable
name is preceded by an asterisk. This tells the computer that a pointer
variable is being created. For example, the following statem ent creates
a pointer·to an inreger:
int *p;

( C contains t\"O special pointer operators: • and & . Th e & operator

rerurns the address of the variable it precedes. The ' operator returns
the value stored at the address that it prececles)(Th e • pointe r
operator has no relationship to the multiplication operator, which uses
the same symbol.) For example, examine this short program :
(#include <stdio.h>

int main(void)

int *p, q;

q = 199; /* assign q 199 */

p = &q; /* assign p the address of q */

printf("%d", *p); /* display q's value using pointer */

return 0;
)

This program prints 199 on the screen) Let's see why.

168 TEACH YOURSELf

" C

First, the line

int *p, q;

defines two variables: p, which is declared as an integer pointer, and

q , which is an integer. Next, q is assigned the value 199. In the next
line, p is assigned the address of q. You can verbalize the &' operator as
'address of.' Therefore, this line can be read as "assign p the address
of q ." Finally, the value is displayed using the' operator applied to p.
The • operator can be verbalized as "at address." Therefore, the
printf( ) statement can be read as "print the value at address q,"
which is 199.
L
"''hen a variable's value is referenced through a pointer, the process
is called indirection.
It is possible to use the' operator on the left side of an' assignment
statement in order to assign a variable a new value given a pointer to
it. For example, this program assIgns q a value indirectly using the
pointer p :
#include <stdio.h>

int rnain{void)
(
int *p, q;

p = &q: /* get q's address */

*p = 199; /* assign q a value using a pointer */

printf("q's value is %d", q);

return 0;
l)
In the two simple example programs just shown, there is no reason to
use a pointer. However, as you learn more about C, you will understand
why pointers are important. Pointers are used to support linked lists
and binary trees, for example.
(The base type of a pointer is very important. Although C allows any
type of pointer to point anywhere in memory, it is the base type that
determines how the object pointed to will be treated. To understand
the importance of this, consider the follOwing fragment:
 USING POINT£RS 169
."
6.1 UNDERSTAND POINTER BASICS

int q;
d o uble . . fp;

fp = &q;

/ . . what does this line do? . . /

*fp = 100.23;

Although not syntactically incorrect, this fragment is wrong)The

pointer fp is assigned the address of an integer. This address is then
used on the left side of an assignment statement to assign a floating-
point value . However, ints are usually shorter than doubles, and this
assignment statement causes memory adjacent to q to be overwritten.
for example, in an environment in which integers are 2 bytes and
doubles are 8 bytes, the assignment statement uses the 2 bytes
allocated to q as well as 6 adjacent bytes, thus causing an error.
( In general, the C compiler uses the base type to determine how
many bytes are in the object pointed to by the pointer) This is hal\' C
kno\\'s ho\·\, many bytes to copy when an indirect assignment is made ,
or ho\·\, many bytes to compare if an indirect comparison is made.
Therefore, it is very important that you always use the proper base
type for a pointer.(Except in special cases, never use a pointer of one
ty pe to point to an object of a different type)
( If you attempt to use a pointer before it has been assigned the
address of a variable, your program will probably crash . Remember,
declaring a pointer variable simply creates a variable capable of
holding a memory address. It does not give it any meaningful initial
value) This is why the following fragment is incorrect.
inc rnain(void)
(
ir: t . . p ;

.p = 10; /* incorrect - p is not pointing to

anything . . /

As the comment notes, the pointer p is not pointing to any l;nown

·object . Hence, trying to indirectly assign a value using p is
meaningless and dangerous.
-/(, AS pointers are defined in C, a pointer that contains a null value
(zero) is assumed to be unused and pointing at nothing. In C, a null is,
by convention, assumed to be an invalid memory address. However,
170 TEACH YOURSELf
c

the compiler will still let you use a null pointer, usually with
disastrous results)

Examples

I. To graphically illustrate how indirection works, assume these

declarations:
int "p, q:

Further assume that q is located at memory address 102 and

that p is right before it, at location 100. After this statement
p = &q;
the pointer p contains the value 102. Therefore , after this
assignment, memory looks like this:

Locatio" Contents

100 102
I-------j --"I p points to q
102 unknown .-J

After the statement

.p = 1000;
executes, memory looks like this:

Location Contents

100 102
1-----..., --"I p points to q
102 1000 .-J

Remember, the value of p has nothing to do with the value of q.

It simply holds q's address, to which the indirection operator
ma;- be applied.
 USING POINTERS 171
" UNDERSTAND POIWTER BASiCS
...

2. To illustrate why you must make sure that the base type of a
pointer is the same as the object it points to, try this incor~ect
but benign program. (Some compilers may generate a warning
message when you compile it, but none will issue an actual
error message and stop compilation.)
/ * This program is wrong. but harmless. *1

.include <stdio.h>

int main (void)

(
int ·Pi
double q. temp;

temp = 1234.34;
p = &temp; 1 * attempt to assign q a value using * 1
q = ·Pi /* indirection through an integer pointer'" /

printf { "'f", q); 1* this will not print 1234.34 * /

return OJ
)

Even though p points to temp, which does, indeed, hold the

value 1234.34, the assignment
q = ·Pi

fails to copy the number because only 2 bytes (assuming 2-byte

integers) will be transferred. Since p is an integer pointer, it
cannot be used to transfer an 8-byte quantity (assuming 8-byte
doubles).

EXERCISES

1. What is a pointer?
2. What are the pointer operators and what are their effects?
172 TEACH YOURSELF
y
C

3. Why is the base type of a pointer important?

4. Write a program with a for loop that counts from 0 to 9,
displaying the n].lmbers on the screen. Print the numbers
using a pointer.

LEARN RESTRICTIONS TO POINTER s:: L

EXPRESSIONS

( In general, pointers may be used like other variables. However, you

need to understand a few rules and restrictions.
A In addition to the' and lit operators, there are only four other
operators that may be applied to pointer variables: the arithmetic
operators +, ++, -, and - -. Further, you may add or subtract only
integer quantities. You cannot, for example. add a floating-point
number to a pointer.
Pointer arithmetic differs from "normal" arithmetic in one very
important way: it is performed relative to the base type of the pointer.
Each time a pointer is incremented, it will point to the next item , as
defined by its base type, beyond the one currently pointed to. For
example, assume that an integer pointer called p contains the address
200. After the statement
p++;

executes, p will have the value 202, assuming integers are two bytes
long. By the same token, ifp had been a float pointer (assuming
4-byte floats), then the resultant value contained in ·p would have
been 204 .
The only pointer arithmetic that appears as 'normal' occurs when
char pointers are used. Because characters are one byte long, an
increment increases the pointer's value by one, and a decre!TIent
decreases its value by one.
You may add or subtract any integer quantity to or from a pointer.
For example, the following is a valid fragment:
 6.2
USING POINTERS

LEARN RESTRICTIONS TO POINTER EXPRESSIONS

,.
173

int *p

p =p + 200;

This statement causes p to point to the 200th integer past the one to
which p was previously pointing.
Aside from addition and subtraction of an integer, you may not
perform any other type of arithmetic operations-you may not
multiply, divide, or take the modulus of a pointer. However, you may
subtract one pointer from another in order to find the number of
elements separating them)
(It is possible to apply the increment and decrement operators to
either the pointer itself or the object to which it points)However, you
must be careful when attempting to modify the object pointed to by a
pointer. For example, assume that p points to an integer that contains
the value 1. What do you think the following statement will do:

- *p++-;
Contrary to what you might think,~his statement first increments p
and then obtains the value at the new location. To increment what is
pointed to by a pointer, you must usc a form like this:
(*p) ++;)
The parentheses cause the value pointed to by p to be incremented .
v (You may cOlnpare two pointers using the relational operators.
However, pointer comparisons make sense only if the pointers relate
to each other-if they both point to the same object, for example.
(Soon you will see an example of pointer comparisons.) You may also
compare a pointer to zero to see if it is a null pointer.)
At this point you might be wondering what use there is for pointer
arithmetic. You will shortly see, however, that it is one of the most
valuable components of the C language.
174 TEACH YOURSELF

'" C

EXAMPLES

1. You can use printf( ) to display the memory address contained

in a pointer by using the %p format specifier. We can uSe this
printf( ) capability to illustrate several aspects of pointer
arithmetic. The following program, for example, shows how all
pointer arithmetic is relative to the base type of the pointer.
#include <stdio.h>
.,
int main(void)
{
char ·cp. chi
int *ip, i;
float *fp, f;
double *dp, d;

cp = &ch;
ip = &i;
fp = &f ;
dp = &d ;
/ * print the current values * /
printf {"%p %p %p %p \ n", cp, ip, fp. dp);

/ * now increment them by one * /

CP+T;
ip++;
fp++; .
dp++;

/ - print their new values */

printf("%p %p %p %p\n", cp, ip, fp, dp);

return 0;
)

Although the values contained in the pointer variables in this

program will vary widely between compilers and even between
versions of the same compiler, you will see that the address
pointed to by ch will be incremented by one byte. The others
will be incremented by the number of bytes in their base types.
For example, in a 16-bit environment this will typically be 2 for
ints, 4 for floats, and 8 for doubles.
 USING POINTtRS 175
."
6.2 LEARN RESTRlCnONS TO POINTER EXPRESSIONS

2. The following program illustrates the need for parentheses

when you want to increment the object pointed to by a pointer
instead of the pointer itself.
.include <stdio.h>

int main(voidl
{
int .p, q;

p = &q;
q =
1;
printf("%p pl;

*p++: /* this will not increment q */

printf ("%d %p", q, pi;

return 0;
)

After this program has executed, q still has the value 1, but p
has been incremented. However, if the program is written
like this:
.include <stdio.h>

int main(void)
{
int ·P. q;

p = &q;

q =
1;
printf("%p pl;

(*pl++; 1* now q is incremented and p is unchanged */

printf ("%d %p", q, pi;

return 0;
)

q is incremented to 2 and p is unchanged.

174 TfACH YOURSElF
y
C

EXAMPLES

J. You can use printf( ) to display the memory address con

in a pointer by using the %p format specifier. We can use tJ
printf( ) capability to illustrate several aspects of pointer
arithmetic. The following program, for example, shows how a
pointer arithmetic is relative to the base type of the pointer.
#include <stdio.h>
.,
iot main(void )

char ·ep. chi

i nt Rip , i;
float *fp, f;
double *dp . d;

cp = &ch;
ip = &L
fp = &f ;
dp = &d ;

/ * print the current values * j

pr i ntf (" %p %p %p %p \ n ", cpo ip, fp . dp );

/ * n ow increment them by o n e */
cp+"' ;
ip++;
fp++ ; ,
dp++ ;

/ * print their new values * /

printf ( "%p %p %p %p \ n", cp, ip, fp, dp);

return 0 ;
)

Although the values contained in the pointer variables in this

program will vary widely between compilers and even between
versions of the same compiler, you will see that the address
pointed to by ch will be incremented by one byte. The others
will be incremented by the number of bytes in their base types.
For example, in a 16-bit environment this will typically be 2 for
ints, 4 for floats, and 8 for doubles .
 USING POINTERS 175
6.2 LEARN RESTRlcnONS TO POINTER EXPRESSIONS
...

2. The following program illustrates the need for parentheses

when you want to increment the object pointed to by a pointer
instead of the pointer itself.
iinclude <stdio.h>

int main(void)
{
int *p, qi

p = &g;

g = 1;
printf("%p ~ pI ;

*p++; this will not increment q * 1

/*
printf (" %d %p", g, pi;

return 0;
I

After this program has executed, q still has the value I, but p
has been incremented. However, if the program is written
like this:
.include <stdio.h>

int main{void )
{
int *p. q;

p = &g;

q = 1;
printf("%p p);

(*p)++; /* now q is incremented and p is unchanged * 1

printf ("%d %p", q, pi;

return 0;
)

q is incremented to 2 and p is unchanged.

176 TEACH YOURSELF
y
C

EXERCISES

1. What is wrong with this fragment?

int .p, i;

p = &i;

p =P .. 8;

2. Can you add a floating-point number to a pointer?

3. Assume that p is a float pointer that currently points to locatio)!
100 and that floats are 4 bytes long. What is the value of p after
this fragment has executed?
p = p + 2;

USE POINTERS WITH ARRA YS

In C, pointers and arrays are closely related. In fact, they are often
interchangeable. It is this relationship between the two that makes
their implementation both unique and powerful.
When you use an array name without an index, you are generating
a pointer to the start of the array. This is why no indexes are used
when you read a string using gets( ), for example. What is being
passed to gets( ) is not an array, but a pointer. In fact, you cannot
pass an array to a function in C; you may only pass a pointer to the
array. This important point was not mentioned in the preceding
chapter on arrays because you had not yet learned about pointers,
However, this fact is crucial to understanding the C language. The
gets( ) function uses the pointer to load the array it points to with
the characters you enter at the keyboard. You will see how this
is done later.
(Since an array name without an index is a pointer to the start of
the array, it stands to reason that you can assign that value to another
pointer and acceRS the array using pointer arithmetic. And, in fact, this
is exactly what you can do. Consider this program:
 USING POINTtRS 177
..,
£3 USE POINTERS WITH ARRAYS

#include <stdio.h>

int main(void)
{
int a[lOJ = {lO, 20, 30, 40, 50, 60, 70, 80, 90, 100J;
int ""p;

p = a: ; * assign p the address of start of a */

/ * this prints a's first,second and third elements */

printf{"%d %d %d\n", 'p, '(p+1J, '(p+211;

j *this does the same thing using a */

printf ( "%d %d %d", a[OJ, a[lJ, a[2JI;

return 0;
J

Here , both printf( ) statements display the same thing. The

parentheses in expressions such as • (p + 2) are necessary because
the' has a higher precedence than the + operator.
Now you should be able to fully understand why pointer arithmetic
is done relative to the base type-it allows arrays and pointers to relate
to each other)
v' ( To use a pointer to access multidimensional arrays, you must
manually do what the compiler does automatically. For example, in
this array:
float balance[lOJ [5J;

each row is five elements long. Therefore, to access balance [3] [1]
using a pointer you must use a fragment like this:
float .p;

p = ( float *) balance:
*(p + (3*5) + 1)

To reach the desired element, you must multiply the row number by
the number of elements in the row and then add the number of the
element within the row. Generally, with multidimensional arrays it is
easier to use array indexing rather than pointer arithmetic.
178 TIACII YOURSELF
., C

In the preceding example, the cast of balance to float' was

necessary. Since the array is being indexed manually, the pointer
arithmetic must be relative to a float pointer. However, the type of
pointer generated by balance is to a two-dimensional array of floats .
Thus, there is need for the cast)
( Pointers and arrays are linked by more than the fact that by using
pointer arithmetic you can access array ele.ments. You might be
surprised to learn that you can index a pointer as if it were an array.
The follOWing program, for example, is perfectly valid:
.include <stdio.h>

int rnain(void)
(
char str[] = ·Pointers are fun-;
char "Pi
int i;

p = str:

/* loop until null is found * j

for(i=O; p[i); i++)
printfl"'c", prill;

return 0;
}

Keep one point firmly in mind: you should index a pointer only when
that pointer pOints to an array) While the following fragment is
syntactIcally correct, It IS wrong; If you tned to execute it, you would
probably crash your computer.
char ·P. Chi
int i;

p = &ch;
for(i=O; i<lO; i++) plil = 'A'+i; /* wrong */

Since ch is not an array, it cannot be meaningfully indexed.

V (Although you can index a pointer as if it were an array, you will
seldom want to do this because pointer arithmetic is usually more
col'lvenient. Also, in some cases a C compiler can generate faster
 63
USlNGPOIImRS
USE POINTERS WfTHAARAYS
.
179

executable code for an expression involving pointers than for a

comparable expression using arrays.) ,
( Because an array name Without an index is a pointer to the start of
the array, you can, if you choose, use pointer arithmetic rather than
array indexing to access elements of the array. For example, this
program is perfectly valid and prints c on the screen:
.include <stdio.h>

int main (void)

(
char strlaO];

-(str+3) = 'c';
printf{-%c·, *(str+3»;

return 0,
)

You cannot, however, modify the value of the pointer generated by

using an array name. For example, assuming the previous program,
this is an invalid statement:
str++;

The pointer that is generated bystr must be thought of as a constant

that always points to the start of the array. Therefore, it is invalid to
modifY it and the compiler will report an error)

EXAMPLES
•
1. Two ofC's library functions, toupper() andtolower(), are
cal1ed using a character argument. In the case of toupper( ),
if the character is a lowercase letter, the uppercase eqUivalent
is returned; otherwise the character is returned unchanged.
For tolower( ), if the character is an uppercase letter, the
lowercase equivalent is returned; otherwise the character is
returned unchanged. These functions use the header file
CTYPE.H. The fol1owing program requests a string from the
180 TEACH YOURSElF

'" C

user and then prints the string, first in uppercase letters and
then in lowercase. This version uses array indexing to access
the characters in the string so they can be converted into the
appropriate case.
#include <ctype.h>
#include <stdio.h>

int main{void)
(

char str[BO):
int i;

printf(~Enter a string: ");

gets (str) i

for(i=O; str[i]; i++)

strfi] = toupper(str[i]);

printf{"%s\n", str); /* uppercase string * /

for(i=O; str[i]; i++)

str[i] = to!ower(str[ij);

printf("%s\n", str); /* lowercase string *j

return 0:
)

The same program is shown below, only this time, a pointer is

used to access the string. This second approach is the way you
would see this program written by professional C programmers
because incrementing a pointer is often faster than indexing
an array.
#include <ctype.h>
#include <stdio.h>

int main(void)

char str [80), "Pi

printf("Enter a string: .);

 &3
USING PGINTtRS
USE PoiNTERS WfTH ARM YS
.,
181

gets(str);
p = str;

while(*p) (
*p = toupper (.p) ;
p++ ;
)

printf("%s\n", str); /* uppercase string */

p = str; /* reset p */

while(*p) (
*p = tolower (.p) i
p++;
)

printf("%s\n~, str); ;. lower. .case string */

return 0;
)

Before leaving this example, a small digression is in order.

The routine
while(*p) {
*p = toupper (.p) ;
p++;
)

will generally be written by experienced programmers like this:

while (.p)
*p++ = toupper (.p) ;

Because the ++ follows the p, the value pointed to by p is first

modified and then p is incremented to point to the next
element. Since this is the way C GOde is often written, this book
will use the more compact form from time to time when it
seems appropriate.
2. Remember that although most of the examples have been
incrementing pointers', you can decrement a pointer as well. For
example, the following program uses a pointer to copy the
contents of one string into ~ nother in reversed order.
182 TEACH YOURSELF
~ C

#include <stdio.h>
#include <string.h>

int main{void)
(
char strl[] = ~Pointers are fun to use~;
char str2(80], *pl. ·p2;

/* make p point to end of strl */

pl = strl + strlen(strll - 1;

p2 = str2;
while(pl >= strl)
·p2++ = ·pl--;

1* null terminate str2 */

*p2 = '\0' i

printf(~%s %5·, strl, str2);

return 0;
}

This program works by setting pI to point to the end of strl,

and p2 to the start of str2. It then copies the contents of otrl
into str2 in reverse order. Notice the pointer comparison in the
while loop. It is used to stop the copying process when the start
of strl is reached.
Also, notice the use of the compacted forms 'p2++ and
·pl--. The loop is the equivalent of this one:
while (pl >= strl) {
*p2 = ·pl;
pl--;
p2++;
}

Again, it is important for you to become familiar with the

compact form of these types of pointer operations.
 USING POINT1RS 183
54 USE POIWTERS TO STRING CONSTANTS

EXERCISES

1. Is this fragment correct?

int count [10];

count = count + 2;

2. What value does this fragment display?

int ternp[5] = (10. 19. 23. 8. 9);
int ·Pi

p = temp;

printf("%d", *(p ... 3 » ;

3. Write a program that inputs a string . Have the program look for
the first space. If it finds one. print the remainder of the string.

E POINTERS TO STRING CONSTANTS

As you know, C allows string constants enclosed between double

quotes to be used in a program. When the compiler encounters such a
string, it stores it in the program's string table and generates a pointer
to the string. For this reason, ~he folloWing program is correct and
prints one two three on the screen .
• include <stdio.h>

int main (void)

(
char "'P;

p = ·one two three-;

printf (p) ;

return 0;
) )
184 TtACH YOURSEl.f
~ C

Let's see how this program works. First, p is declared as a character

pointer. This means that it may point to an array of characters. When
the compiler compiles the line
p = ·one two three-;

it stores the string in the program's string table and assigns to p the
address of the string in the table. Therefore, when p is used in the
printf( ) statement, one two three is displayed on the screen.
(This program can be written more efficiently, as shown here:
#include <stdio.h>

int main {vo idl

(
char *P = ·one two three-;

printf (p) ;

return 0;
)

Here , p is initialized to point to the string';

1. This program continues to read strings until you enter stop:

linclude <stdio.h>
.include <string.h>

int main(voidl
(
char .p = ·stop·:
char str[BO);

do (
printf{-Enter a string: W);
gets(str) ;
} while (strcmp(p, str));
 USING POINTIRS 135
y
6.4 uSE POINTERS TO STRING CQNSTAfv7S

return 0;
)

2. Using pointers to string constants can be vcry hel pful when

those constants are quite long. For example, suppose that you
had a program that at various times would prompt the user to
insert a diskette into drive A. To save yourself some typing, you
might elect to initialize a pointer to the string and then simply
use the pointer when the message needed to be dispbyed; for
example:
char ;InsDisk = -Insert disk into drive A, then press C~~ER~;

printf(InsDisk) ;

printf(InsDiskl;

Another d(h'antage to thl~ <lppru<lch i.s tll,lt to ch'lIlgc the

prompt, you unly I1ced to ch~ll1ge it once, and eli! reft'rences
to 1t \\',11 refkct the change.

EXERCISE

1. Write a program that creates three character pointers and

initialize thcm so that one points to the string "one" , the second
to the string "two", and the third to the string "threc" Next, have
the program print all six permutations of these three strings,
(For exatnple, one pernllItation is "one two three', another is
"two one three".)
 186 TEACH YOURSELF

'" C

EATE ARRAYS OF POINTERS

( Pointers may be arrayed like any other data type. for example, the
following statement declares an integer pointer army that has 20
elements:
,
int "'pa[20];)

(The address of an integer variahle called myvar is assigned to the

ninth element of the array as follows:
palS] = &myvar:;

Because pa is an array of pointers, the only values that the array

elements may hold are the addresses of integer vari"hles(ro assigll
the integer pOInted to by the third element of pa the value IIlD, u,e
the statement:

EXAMPlES

1. Prohably the single 1110St Cotnnl0n use of arrays of pointers is to

create string tables in 111uch the same \\'ay that unsized arrays
were lIsed in the previous chapter. For cxanlp1c, this function
displays an error message based on the value of its parameter
CTT_num.
char 'P [I = (
Input exceeds field width"
H I

"Out of range".
"Printer not turned on",
"Paper out",
"Disk full",
"Disk write error"
) ;

void e~ror(int err_num)

(
 USING POINTERS 187
6.5 CREA TE ARRA YS OF POINTERS
...

2. The following program uses a two-dimensional array of pointers

to create a string table that links apple varieties with their
colors. To use the program, enter the name of the apple, and the
program will tell you its color.
#include <stdio.h>
.include <string.h>

char 'p11121 =(
~Red Delicious", "red",
MGolden Delicious·, ~yellow·, -,
M"Hnesap~, ~red",

"Gala", "reddish orange",

"Lodi" , "green",
"Mutsu", "yellow",
·Cortland", "red",
"Jonathan", "red",
"n, M" i * terminate the table wi th null strings * /
) ;

int rr.air. (void)

int i:
char apple (80 J;

printfl"Enter name of apple: "I;

gets (apple);

for(i=O; *p[i] [0]; i++) {

if( !strcmp(apple, p(i) [0)))
printf("%s is %5\n", apple, p[i] [1]);

return 0;

Look carefully at the condition controlling the for loop. The

expression 'p[i) [0) gets the value of the first byte of the ith
string. Since the list is terminated by null strings, this value will
he z~ro (false) when the end of the table is reached. In all o,ther
cases it will be nonzero, and the loop will repeat.
188 TEACH YOURSELf
"f' C

EXERCISE
I. In this exercise, you will create an 'executive decision aid.' This
is a program that answers yes, no, or maybe to a question
entered at the keyboard. To create this progral11 use an array
of character pointers and initialize them to point to these three
w·ings. 'Yes", "No', and 'Maybe. Rephrase the qlleStlOn'. Next,
input thc user's question and find the length of the string. Next,
use this formula to cmnpute an index into the pointer array:

index = length % 3

BECOME ACQUAINTED WITH

MULTIPLE INDIRECTION

(It IS possihle in ( \() ha\e a pointer pOillt to another pOIl1tcr This is

(ailed mulflple IIU;lIU,UOIv(sec figure 6-1). \"hen a pointc~r points to
another pOlllter. the tirst pointer contains the address of the second
pointer, which pOints to the location containing the ohjecr.
(To declare a pointer to J pointer, an additional asterisk is placed in
front of the pointe,', name. For example, this declaration tells the
complier that nlp 1~ a pOlIHer to a character pointer:

It IS important [() understand that mp is not a pointer to a character,

hut rather a pointer IO a character pointer)

Mulflple
ind;rectlon
T

Pointer to Pointer "ariable

pointC'r
 189
6.6
USING POINllRS

BECOME ACQUAINTED WITH MUL nPLE INDIRECTION

.,

(Accessing the target value indirectly pointed to by a pointer to a

pointer requires that the asterisk oper.ator be applied twice. For
e~ample.

char" -mp, .p, ch;

p = &ch; 1* get address of ch */

mp = &p: .. get address of p ./
.... mp = 'A'; .. assign ch the value A using multiple
indirection *i )

As the C01nments suggest, ch is assigned a value indirectly using two

pointers.
(~lultiple indirection is not limited to merely "a pointer to a pointer:
You can appl\' the' as often as needed. Howe\'er. multiple indirection
heyond a pointer to a pointer is very difficult to follow and is not
recommended')
You 1l1.l~' not see the need for multiple indirection at this time, but
as you karn more ilhout C, you will see sotne examplp.s in which it is
\'cr\' \'aillable.

EXAMPLES

1. The fonowing program assigns val a value using multiple

indirection. It displJ)"s the value first directly .. then through
the use of multiple indirection.
#include <stdio.h>

, int main (voirl)

I
float *fp, *'*mfp, val;

fp = &va1;
rnfp = &fp;
"""'mfp = 123.903;
printf("%f %f-, val. **mfp);

return 0;
)

•
 TEACH VOURS£lF
190
• C

2. This program shows how you can input a string using gets( ) by
using a pointer to a pointer to the string.
#include <stdio.h>

int main(voidl

char *p, .. *mp, str (80 J;

p = str;
mp = &Pi

printf ("Enter your name: ");

gets ("mp) ;
printf (~Hi %s", '"mp) i

return 0;

Notice that when mp is used as an argument to both gets( )

and printf( ), only on" • is used . This is because both of these
fUlluions require a pointer to a string fur their operatiun .
Rememher .• *mp is a pointer to p. J lo\\"c\"(~r, pis (] pointer to
[he string str. Therefore. *mp is a p{)int(~r to str. If you are a
little confused, don't worry. Over time, you will develop a
clearer concept of pointers to pointers.
"

EXERCISE

1. To help you understand multiple indirection better, write a

program that assigns an integer a value using a pointer to a
pointer. Before the program ends, display the addresses of the
integer variable, the pointer, and the pointer to the pointer.
(Remember, use %p to display a pointer \'alue .)
 6.7
USING POINTERS
USE POINTERS AS PARAMETERS
.,
191

E POINTERS AS PARAMETERS

Pointers may be passed to functions. For example, when you call a

function like strlen( ) with the name of a string, you are actually
passing a pointer to a function. When yoti pass a pointer to a function,
a
the function must be declared 3S receiving pointer of the same type.
In rhe case of strlen( ), this is a character pointer. A complete
discussion of using pointers as parameters is presented in the next
chapter. However, some· basic concepts are discussed here .
II'hen you pass a puinter to a function, the code inside that function
has access to the variable pointed to by the parameter. This means
that the function can change the variable used to call the function.
This is why functions like strepy( ). for example , can work. Because
it is passc9 il pointer, the function is able to modify the array that
receives the string.
No\\' you can llnOel'stalld why you need to precede il variable's
name with an & when tlsing seanf( ). In order for seanf( ) to modify
rhe vallie of one of its arguments. it must he passed a poi mer to that
argul11ellt

• EXAMPlES

1. Anhther of Cs standard library functions is called puts( ); it

writes its string argun,.,nt to the screen followed by a newline.
The program that follows creates its own version o(puts( )
called myptits( ) .
~~nc l ude <s t dio.h>

void myputs(char *p) ;

int main(void )

myputs (" this is a test"):

•
192 TEACH YOURSELF

• C

return 0;
)

void myputs{char .p)

(
while(*p) { /* loop as long as p does not point to the
null that terminates the string *'
prir.tf{~%c", .p):
p++; /* go to next character * /

printf(ft \ n~) i

This program i1lustrate.s a very illlportant point thilt Wi)S

tnentioned carlier in this chapter. When the compiler
encounters a string constant, it places it into the progriull 's
string table anel generates a pointer to it. Therefore. the
myputs( ) function is actually calleel with a character pointer.
and the p<JrJl11c(l~r p must be dccl;ucd JS J cl1.1r.u :(('[ pointer in
order to reLei\'(~ it.
2. The. follu\\'jng progr.lm sho\\'s one way to implement dlC
strcpy( ) function. called m~'strcpy( ).
#include <stdio.h>

void rr.ystrcpy(char -to, char ·from):

•

int ma:n i \'o:'dl

char str[BOj;

mystrcpy(str, -this is a test");

printf(str) ;

return 0;

void mystrcpy(char yto, char -fro m)

while(-from) Yto •• = ·from~+;

yto = ' '. 0'; / * null terminates the string - /
)
 USING POINTERS 193
6.7 USE PoiNTERS AS PARAMETERS
...

UERCISES
1. Write your own version of strcat( ) called mystrcat( ), and
write a short program that demonstrates it.
2. Write a program that passes a pointer to an integer variable to a
function. Inside that function, assign the variable the value -1.
After the function has returned, demonstrate that the variable
does, indeed, contain -1 by printing its value.

At this point you should be able to perform these exercises and

ansvo,'er these questions:
l. Show how to declare a pointer to a double.
2. Write a program that assigns a value to a variable indirectly by
using a pointer to that variable.
3. Is this fragment correct? If not, why not?
int main(void)
(
char .p;

printf("Enter a string: ");

gets(p) ;

return 0;
)

4. How do pointers and arrays relate to each other?

5. Given this fragment:
char .p, str[aO] = ~this lS a test";

p =: str;

show two ways to access the 'i' in "this."

194 TIACII YOURSW
c

6. Assume that p is declared as a pointer to a double and ttJfitains

the address 100. Further, assume that doubles ate 8 bytes long.
After p is incremented, what will its value be?

Skills

This section checks how well you have integrated the material in
this chapter with that from eatl!~r chapters.
1. What is the advantage of using pointers over array indexing?
2. Below is a program that counts the I1Umber of spaces in a string
entered by the user. Rewrite the program so that it uses pointer
arithmetic rather than array indeXing.
#include <stdio.h>

int main(void)
(
char str(801;
• int i. spaces;

printf("Enter a string: ");

gets(str) :

spaces = 0;
for(i=O: str(i]; i .... +)
if(strji]==' ') spaces++:

printf("Number of spaces: %d". spaces ) ;

return 0;
)

3. Rewrite the following array reference using pointer arithmetic.

int count(100) (10);

count(44) (8J = 99;

 7
A Closer Look
at Functions

chapter objectives

'?1 Understand function prototypes

7.2 Understand recursion

7.3 Take a closer look at pararyleters

7.4 Pass arguments to maine l
7.5 Comp.are old-style to modern function
parameter declarations

195
y
 TEACH YOURSELF
196
.., C

T the very foundation of C is the function. All action

statements must appear within one and an understanding
of its operation is crucial to successful C programming.
This chapter takes a close look at several important topics
,related to functions.

~~~P~~Be~~~o~r~e proceeding you should be able to answer these questions and

perform these exercises:

1 What does this fragment do?

int i .p;

p = &i;
.p = 19;
? is generated when you use an array nanle v:ithout an
\\rhc1t
index?
3. Is this fragment correct? If it is correct, explain why it works.
char *p = ~this is a string-;

~. Write a short program that assigns a floating-point value to a

variable indirectly using a pointer to the variable.
s. Write your own version of strlen( ), called mystrlen( ), and
demonstrate it in a program.
6. Is this fragment correct? If it is, ,\(hat does the program display?
char str[SJ;

strcpy(str, -ABCDEFG-);
•
printi(-%c·, *(str+2) j

UNDERSTAND FUNCTION PflOTOTYPES

In Chapter 1 you were briefly introduced to the function prototype.
No" it is time for you to understand precisely what a prototype does
and why it is important to C programming. Function prototypes were
 A CIooe<"-"- 197
~
7.1 UNDERSTAND FUNCTlQN PROTOTYPES

not supported by the original version ofC. They were added when C was
standardized in 1989. Many consider prototypes to be the single most
important addition made to the C language since its creation. Prototypes
are not technically necessary. However, for reasons that will become
self-evident, they should be used in all programs that you write.
The general form of a function prototype is shown here:

type function-name(type parameter-name/,

type parameter-name2,

type parameter-nameN);

A prototype declares three attributes associated with a function :

I . Its return type.

2. The number of its parameters.
3. The type of its parameters.

Prototypes provide several benefits. They inform the compileT

about the return type of a function . They allow the compiler to find
and report illegal type conversions between the type of arguments
used to caJl a function and the type definition of its parameters.
Prototypes also enable the compiler to report when the number of
arguments pas~ed to a function is not the same as the number of
parameters declared by the function. Let's look at each of these.
When you call a function, the compiler needs to know the type of
.data returned by that function so that it can generate the proper code
to handle that data. The reason for this is easy to understand: different
data types have different sizes. The code that handles an integer
return type will be different from that which handles a double, for
example. If you use a function that is not prototyped, then the
compiler will simply assume that it is returning an integer. However,
if it is actuaJly returning some other type, an error will occur. If the
function is in the same file as the rest of your program, then the
.
compiler will catch
. this error. But if the function is in another file or a
liorary, then the error will go uncaught-and this will lead to trouble
when your program is executed.
In the absence of a function prototype, it is not syntactically wrong
to call a function with incompatible arguments or with more or less
 .
"... ID\G~ YOURSELf

arg].lments than the function has parameters. Of course, doing either

of these is obviously incorrect eVen though the compiler ma)' accept
your program withoUI complilint. The use of a function prototype
prevents these Brrors by enabling the compiler to finp thefTI· It is
impprr,mt to understanq, however, that not all kinds of type conversipps
are illegal in a function call. In fact, C automatically converts most
types of arguments into the type of data specified by the parameter.
But a few type conversions are inherently wrong. For example, you
cannot convert an integer into a pointer. A function prototype allows
the compiler to catch and return this type of error.
As mentioned, as important as prototypes are, they <lre l1Pt currently
required. Because of the need to maintain compatibility with older
code, all C compilers still support non-prototyped programs. Of course,
at some point in the future, this situation may change.
In early versions of C, before prototypes were invented, it was still
necessa'ry to tell the compiler about the return type of a functiol1
.(unless it returned type int) for the reasons explained earlier. This
was done using a forerunner of the prototype, called a fonmr(i
dcc/al'{ltlOI1 or a fonmrd reference. A forward declaration is ess@nti<llly •
truncated form of a prototype that declares only the retljrp type of a
function-not the type and number of its parameters. AlthQugh
forward declarations are obsolete, they are still allowed fOf
compatibility with older code.
The following program demonstrates an old-style forward declaration,
It uses it to inform the compiler of voIume{ )'s ret\lrn type.

#include <stdio.h>

double volume(); /* old-style forward degl&ration for

volume() */

int main (void)

(
double vol;

vol = vo!urne(l2.2, 5.67, 9.03);

printf (~Volume: %f u·, voll;

return 0;
)

/* Compute the volume of a cube. */

 A Closer Look at Functions 188
I.'
.,.,--:--':'---==
UNDERSTAND FUNCtiON PROTOTYPES
~

double volurne(double 51. double 521 double sJ)

(
return sl * 52 • 53;
)

Since the old-style declaration does not inform the compiler about any
of volume( )'S parameters it is not a function prototype. Instead, it
r
simply states vl11btl1ll( g fefurn type. The trouble is that the lack of a
full prototype will allow volume( ) to be called using an incorrect
type and/or number of arguments. For example, given the preceding
program, the following will hot generate a compiler error message
even though it is wrong.
volume(120.2, 99.3); /* missing 1ast arg ttl

Since the compiler has not been given information about volume( )'s
parameters it won't catch the fact thtlt this call is wrong ..
Although the old-style fonvard declaration is no longer used in new
code, you "'ill still find it quite frequently in older programs. If you
will be updating older programs, you should consider adding
prototypes to be your first job.
When function prototypes were added to C, two minor compatibility
problems between the old version of C and the ANSI version of Chad
to be resolved. The first issue was how to handle the old-style forward
declaration, which does not use a parameter list. To do so, the ANSI C
standard specifies that when a function declaration occurs without a
parameter list, nothing whatsoever is being said ahout the parameters
to the function .. It might have parameters, it might not. This allows
old-style declarations to coexist with prototypes. But it also leads to a
question: how do you prototype a function that takes no arguments?
For example, this function simply outputs a line of periods:
void line ()
{
int i;

for(i=O; i<SO; i++) printf(".");

}

If you try to use the following as a prototype, it won't work because

the compiler will think that you are simply using the old-style
declaration method.
 ,.
200 TEACH YOURSW

void line () ;

The solution to this problem is through the use of the "oid

keyword. When a function has no parameters, its prototype uses "oid
inside the parentheses. For example, here is line( )'s proper
prototype:
void line(void);

This explicitly tells the compiler that the function has no parameters,
and anv call to that function that has parameters is an error. You must
make sure to also use "oid when the function is defined. For example,
line( ) must look like this:
void line(void)
(
int i;

Since we have been using void to specify empty parameter lists since
Chapter I, this mechanism is already familIar to you.
The second issue related to prototyping is the way it affects C's
automatic type promotions. Because of some features of the
environment in which C was developed, when a non-prototyped
function is called, all integral promotions take place (for example,
characters are converted to integers) and all floats are converted to
doubles. However, these type promotions seem to violate the purpose
of the prototype. The resolution to this problem is that when a
prototype eXIsts, the types specified in the prototype are maintained,
and no type promotions will occur.
There is one other special case that relates to prototypes: variable
length argument lists. We won't be creating any functions in this book
that use a variable number of arguments because they require the use
of some advanced techniques. But it is possible to do so, and it is
sometimes quite useful. For example, both printf( ) and seanf( )
accept a variable number of arguments. To sped!)' a variable number
of argumcnts, use ... in the prototype. For example,
int myfunc(int a, ... J;
 A ClaMr Look at Functions 201
7.1 UNDERSTAND FUNCTION PROTOTYPES
'"

specifies a function that has one integer parameter and a variable

number of other parameters.
In C programming there has been a long-standing confusion about
the usage of two terms: declaration and definition. A declaration
specifies the type of an object. A definition causes storage for an object
to be created. As these terms relate to functions, a prototype is a
declaration . The function, itself. which contains the body of the.
function is a definition.
In C, it is also legal to fully define a fun,pion prior to its first use,
thus eliminati"'g the need for a separate prototype. However, this
works only in very small programs. In real-world applications, this
option is not feasible. For all practical purposes, function prototypes
must exist for all functions that your program will use.
Remember that if a function does not return a value, then its return
type should be specified as void-both in its definition and in its
prototype .
Function prototypes enable you to write better, more reliable
programs because they help ensure that the functions in your
programs are being called with correct types and numbers of
arguments. Fully prototyped programs are the norm and represent the
current state of the art of C programming. Frankly, no professional C
programmer today would write programs without them. Also, future .
versions of the ANSI C standard may mandate function prototypes and •
c++ requires them now. Although prototypes are still technically
optional, their use is nearly universal. You should use them in all of the
programs you write.

1. To see how a function prototype can catch an error, try

compiling this version of the volume program, which includes
volume( )'S full prototype:
#include <stdio.h>

this is volume{) 's full prototype '*/

/ '*
double vo1ume(double s1. double s2. double 53);

int main(void)
{
 dQUble vol:

vol s volume(12.2. 5.67, 9.03, 10.2); /* error */

printf(-Volume: %f-, vol);

return 0;
}

/* Compute the volume of a cube. *'

double volume (double 81. double 52. double 53)
(
return 51 • 52 • s3;
)

As you will see, this program will not compile because the
compiler knows that volume( ) is declared as having only three
parameters, but the program is attempting to call it with four
parameters.
2. As explained, if a function is defined before it is called. it does
not require a separate prototype. for example. the following
program is perfectly valid:
#include <stdio.h>

/* define getnum{) prior to its first use -j

float getnum(void)
(
float x;

printf("Enter a number: .);

scanf (",t-, &xl;
return x;
)

int main (void)

(
float i;

i = getnum();
printf(-,r-, i);

return 0;
}
 1.1 UNDERSTAND RJNCTION PRDTOTYPfS

Since getnum( ) is defined before it is used, the compiler

knows what type of data it returns and that it has no
parameters. A separate prototype is not needed. The reason that
you wiJI seldom use this method is that large programs are
typically spread across several files. Since you can't define a
function more than once, prototypes are the only way to inform
all files about a function. (Multi-file programs are explained in
Chapter 11.)
3. As you know, the standard library function sqrt{ ) returns a .
double value . You might be wondering how the compiler knows
this. The answer is that sqrt( ) is prototyped in its header file
MATH.H. To see the importance of using the header file, try
this program:
jinclude <stdio.h>
/ - math.h is intentionally not included */

int main(void)
{
double answer;

answer = sqrt(9.0~;
printf ("%f". answe'rl;

return 0;
)

When you run this program, it displays something other than 3

because the compiler generates code that copies only two bytes
. (assuming two-byte integers) into answer and not the 8 bytes
that typically comprise a double. If you include MATH. H, the
program wiJI work correctly.
In general, each of C's standard library functions has its
prototype specified in a header file. For example, printf( ) and
scanf{ ) have their prototypes in STDIO.H. This is one of the
reasons that it is important to include the appropriate header
file for each library function you use.
4. There is one situation that you wiJI encounter quite frequently
that is, at first, unsettling. Some 'character-based'-functions
have a return type of int rather than char. For example, the
•
204 ltACH YOURSW
• C

getchar( ) function's return type is int, not char. The reason

for this is found in the fact that C very cleanly handles the
conversion of characters to integers and integers back to
characters. There is no loss of information. For example, the
following program is perfectly valid:
#include <stdio.h>

int main (void)

(
char chi

ch = get_a_char():
printf(-'c·, ch);

return 0:
)

int get_a_char(voidl
(
return 'a';
)

When geca_char( ) returns, it elevates the character 'a' to an

integer by adding a high-order byte (or bytes) containing zeros.
When this value is assigned to ch in main( ), the high-order
byte (or bytes) is removed. One reason to declare functions like
geca_char( ) as returning an integer instead of a character is
to allow various error values to be returned that are
intentionally outside the range of a char.
-LWhen a function returns a pointer, both the function and its
prototype must declare the same poihter return type. For
example, consider this short program:
#include <stdio.h>

int -init(int xl;

int count;

int main(void)
f
 A CIa.... L_ at Functions 205
7.1 UNDERSTAND FUNCTION PROTOTYPfS
"
int *p;

p = init(110); /* return pointer *1

printf("cQunt (through p) is %d", *p);

return 0;
}

int *init(int xl
{
count = x;

return &CQunt; / * return a pointer * 1

}

AS you can see, the function init( ) retufns a pointer to the

global variable count. Notice the way that the return type for
init( ) is specified. This same general form is used for any sort
of pointer return type. Although this example is trivial,
functions that return pointers are quite valuable in many
programming situations. One other thing: if a function returns a
pointer, then it must make sure that the object being pointed to
does not go out-of-scope when the function returns. This means
that you must not return pointers to local variables.
6. The maine ) function does not have (nor does it require) a
prototype. This allows you to define maine ) any way that is
supported by your compiler. This book uses
int main (void) ( ...

because it is one of the most common forms. Another

frequently used form of maine ) is shown here:
void main(void) { ...

This form is used when no value is returned by maine ). Later

in this chapter, you will see another form of maine ) that has
parameters.
The reason maine ) does not have a prototype is to allow C
to be used in the widest variety of environments. Since the
precise conditions present at program start-up and what actions
must occur at program termination may differ widely from one
 ,.
206 TEACH YOURSELF
C

operating system to the ~ext, C allows the acceptable forms of

maine ) to be determined by the compiler. However, nearly all
compilers will accept int main (void) and void main(void) .

I . Write a program that creates a function, called avg( ), that

reads ten floating-point numbers entered by the user and
returns their average. Use an old-style forward reference and
not a function prototype.
2. Rewrite the program from Exercise 1 so that it uses a function
prototype.
3. Is the following program correct? If not, why not? If it is, can it
be made better?
winclude <stdio.h>

double myfunc();

int main (void)

(
printf(-%f-, myfunc{lO.2»;

return 0;

double myfunc(double num)

{
return num / 2.0;
}

4. Show the prototype for a function called Purge( ) that has no

parameter~ and returns a pointer to a double.

5. On your own, experiment with the concepts presented in this

section. .
 7.2
ACIose<Lookot_
UNDERSTAND RECURSION
.,
207

DERSTAND RECURSION
Recursion is the process by which something is defined in terms of
itself. When applied to computer languages,C!'ecursion means that a
function can call itself)Not all computer languages support .recursive
functions, but C does. A very simple example of recursion is shown in
this program:
(#include <stdio.h>

void recurse(int i);

int main(void)
(
recurse (0) ;

return 0;
)

void recurse(int i)
(
ifCi<101 (
recurse(i+l); / * recursive call */
pnntf C"%d" il;
)
)

This program prints

9876543210
on the screen. Let's see why.
The recurse( ) function is first called with O. This is recurse( )'s
first activation. Since 0 is Ie.. than 10, recurse( ) ,then calls itself with
the value of i (in this case 0) plus I, This is the second activation of
recurse( ). and i equals I. This causes recurse( ) to be called again
using the value 2. This process repeats until recurse( ) is called with
the value 10. This causes recurse( ) to return. Since it returns to the
point of its call, it will execute the printf( ) statement in its previous
activation, print 9, and return. This, then, returns to the point of its
call in the previous activation, which causes 8 to be displayed. The
process continues until all the calls return, and the program
terminates)
208 TtACHYOURSa.f

" c

(It is important to understand that there are not multiple copies of a

recursive function. Instead, only one copy exists. When a function is
called, storage for its parameters and local data are allocated on the
stack. Thus, when a function is called recursively, the function begins
executing with a new set of parameters and local variables, but the
code that constitutes the function remains the same)
If you think about the preceding program, you will see that
recursion is essentially a new type of program control mechanism.
This is wh~very recursive function you write will have a conditional
statement that controls whether the function will call itself again or
return. Without such a statement, a recursive function will simply run
wild, using up all the memory allocated to the stack and then crashing
the program.)
Recursion is generally employed sparingly. However, it can be
quite useful in simplifying certain algorithms. For example, the
Quicksort sorting algorithm is diffitult to implement without the
use of recursion. If you are new to programming in general, you
might find yourself uncomfortable with recursion. Don't "'om" - , as
you become more experienced, the use of recursive functions will
become more natural.

EXAMPLES
1. The recursive program described above can be altered to print
the numbers 0 through 9 on the screen. To accomplish this,
only the position of the printf( ) statement needs to be
changed, as shown here:
.include <stdio.h>

void recurse(int i);

int main(void)
(
recurse (0) j

return 0;
)

void recurselint il
 "CIaM<~"~" 209
7.2 UNDERSTAND RECURSION
"
(
ifli<lO) (
printfl"%d" i);
recurse (i+1) j

)
}

Because the call to printf( ) now precedes the recursive call to

recurse( ), the numbers are printed in ascending order.
2. The following program demonstrates how recursion can be used
to copy one string to another.
#include <stdio.h>

void rcopy(char ·51. char *52);

int main(void)
(
char str[BO)j

rcopy(str, "this is a test");

printf (str);

return n:
}

; * Copy 52 to s1 using recursion. *;

void rcopy(char *51. char ·s2)
(
i£C*s2") ( /* if not at end of 52 */
*51++ = ·s2++:
rcopy{sl. 52);
}
else *s1 = '\0'; /* null terminate the string */
}

The program works by assigning the character currently-pointed

to by 82 to the one pointed to by 81, and then incrementing both
pointers. These pointers are then used in a reCU1'sive call to
rcopy( ). until 82 points to the null that terminates the string.
Although this program makes an interesting example of
recursion. no professional C programmer would actually code a
function like this for one simple reason: efficiency. It takes
more time to execute a function call than it does to execute a
 .
21 • lEACH YOURSII F
C

loop. Therefore, tasks like this will almost always be coded

using an iterative approach.
3. It is possible to have a program in which two or more functions
are mutually recursive. Mutual recursion occurs when one
function calls another, which in turn calls the first. For example,
study this short program:
linclude <stdio.h>

void f2 (int b);

void f1 (int a);

int main(void)
(
fl (30);

return 0;
)

"

void fl(int a)
(
H(a) f2(a-l);
., printf('%d " a);
)

void f2(int b)
(
printf(· ... );
H (b) fl (b-l) ;
)

This program displays

.. _......._0246810121416182022 24 26 28 30

on the screen. Its output is caused by the way the two functions
fI ( ) and f2( ) call each other. Each time fI ( ) is called, it
checks to see if a is zero. If not, it calls f2( ) with a-I. The f2( )
function first prints a period and then checks to see ifb is zero . .
Ifnot, it calls fI( ) with bol, and the process repeats.
Eventually, b is zero and the function calls start unraveling,
causing fI ( ) to display the numbers 0 to 30 counting by twos.
 211
~
7.3 TAKEACLOSERLOOKATP/WIMETERS

1. One of the best known examples of recursion is the recursive

version of.a function that computes the factorial of a number.
The factorial of a number is obtained by multiplying the original
number by all integers less than it and greater than 1.
Therefore, 4 factorial is 4x3x2, or 24. Write a function, called
fact( ), that uses recursion to compute the factorial of its
integer argument. Have it return the result. Also, demonstrate
its use in a program.
2. What is.wrong with this recursive function?
void f (void)
{
int i;

/ * call f()10 times . /

for(i=O; i<10; i++) f( );
)

3. Write a program that displays.a string on the screen, one

character at a time, using a recursive function.

AKE A CLOSER LOOK AT PARAMETERS

( For computer languages in general, a subroutine can be passed

arguments in one of two ways. The first is called call by value. This
method copies the value of an argument into the formal parameter of
the subroutine. Therefore, changes made to a parameter of the
subroutine have no effect on the argument used to call it. The second
way a subroutine can have arguments passed to it is through can by
reference. In this method, the address of an argument is copied into the
parameter. Inside the subroutine, the address is J.1sed to access the
actual argument. This means that changes made to the parameter will
affect the, argument')
By default, C uses call by value to pass arguments. This means that
you cannot alter the arguments used in a call to a function. What
212 TEACH yOURSllf
.., c

occurs to a parameter inside the function will have no effect on the

argument outside the function . However, as you saw in Chapter 6 , it
is possible to manually construct a call by reference by passing a
pointer to an argument. Since this causes the address of the argument
to be passed, it then is possible to change the value of the argument
outside the function.
(The classic example ofa call-by-reference function is swap() ,
shown here. It exchanges the value of its two integer arguments.
#include <stdio.h>

vo id swap(int "'i, int *j);

int main (void)

{
int numl . num2;

nurn! = 100:
num2 = 800;

printf("num!: %d num2: %d \ n " , numl. num2 1;

swap(&numl, &nurn2);
printf ( ~numl: %d num2 : %d \ n·, nurn!, num2 );

return 0:
)

/ * Exchange the values pointed to by two integer pointers . */

void swap{int wi. int "'j)
{
int temp;

temp wi;
==
*i = *j;
*j = temp;
) )
, Since pointers to the two integers are passed to the function , the actual
values pointed~ by the arguments are exchanged.
As you know,(when an array is used as an argument to a function ,
only the address of the array is passed, not a copy of the entire array,
which implies call-by-reference. This means that the parameter
declaration must be of a compatible pointer type. There are three
 ".-~.
1.3 TAKE A CLOSER LOOI! AT PNIMIf1EIIS

ways to declare a parameter that is to receive a pointer to an array.

First, the parameter may be declared as an array of the same type and
size as that used to call the function. Second, it may be specified as an
unsized array. Finally, and most commonly, it may be specified as a
pointer to the base type of the array. The following program
demonstrates all three methods:
'include <stdio.h>

void fl(int nurn[5]). f2(int nurn[]) , f3(int *nurn);

int main (void)

{
int count[5] = (l. 2, 3. 4. 5];

fl (count) ;
f2 (count);
f3 (count) ;

return 0;
)

/* parameter specified as array * /

void fl(int num[5])
(
int i;

for(i=O; i<5; i++) printfC-%d" num[i]);

]

/ *parameter specified as unsized array

void f2(int nurn[])
*'
(
int i:

for(i=O; i<5: i++' printfC"'d· num[i);

}

/* parameter specified as pointer

void f3(int ·nurn)
*'
{
int i:
214 TtACH YOURSELF
y
C

for(i=O; i<5; i++) printf (-%d· num[ill;

)

Even though the three methods of declaring a parameter that will

receive a pointer to an array look dIfferent, they all result in a pointer
parameter being created')

EXAMPLE

I. Some computer languages, such as BASIC, provide an input

function that allows you to specify a prompting message. C has
no counterpart for this type of function. However, you can
easily create one. The program shown here uses the function
prompt( ) to display a prompting message and then to read a
number entered by the user.
#include <stdio.h>

void prompt (char *msg. int *nurn);

int main (void)

(
int i:
prompt (-Enter anum: ., &i);
printf(-Your number is: %d", i);

return 0;
}

void prompt(char *msg, int ·num)

(
printf (msgl ;
scanf(-%d-, num);
}

Because the parameter num is already a pointer, you do not

need to precede it with an & in tlie call to scanf( ). (In fact, it
would be an error to do so.)
 A CIoMr look lit Function8 215
~
7.' PASS ARGUMENIS
, ro moinO

I. Is this program correct? If not, why not?

#include <stdio.h>

myfunc(int ~um, int min, int max);

int main(void)
(
int i;

pr1ntf ( "Enter a number between 1 and 10: H);

myfunc(&i, 1, 10) ;

return OJ
)

void myfunc(int num , int min, int max)

do (
scanf (" %d num );
M
,

} while(;num<min I I ;num>max);

2. Write a program that creates an input function similar to

prompt( ) described earlier in this section. Have it inpUt a
string rather than an integer.
3. Explain the difference between call by value and call by
reference .

SS ARGUMENTS TO main(}

Many programs allow command-line arguments to be specified when

they are run. A command-line argument is the information that follows
the program's name on the command line of the operating system.
Command-line arguments are used to pass information into a program.
F..or example, when you use a text editor, you probably specif'y the
name of the file you want to e~it after the name of the text editor.
 .,
216 TEACH YOURsELF
C

Assuming you use a text editor called EDTEXT, then this line causes
the file TEST to be edited.
EDTEXT TEST

Here, TEST is a command-line argument.

Your C programs may also utilize command-line arguments. These
are passed to a C program through two arguments to the main( )
function. The parameters are called argc and argv. As you probably
guessed, these parameters are optional and are not present ",hen no
command-line arguments are being used. Let's look at argc and argv
more closely.
The argc parameter holds the number of arguments on the
command-line and is an integer. It will always be at least J because the
name of the program qualifies as the first argument.
The argv parameter is an array of string pointers. The most common
method for declaring argv is shown here:
char ,.. argv [J ;

. The empty brackets indicate that it is an array of undetermined

length. AfI command-line arguments are passed to main( ) as strings .
. To access an individual string, index argv. For example, argv[O]
points to the program's name and argv(l] points to the first argument.
This program displays all the command-line arguments that are
present when it is executed.
#include <stdio.h>

.int main(int argc, char *argv[]l

{
int i;

for(i=l; i<argc; i++) printf("%s ~ argv[i);

return 0;
}

C does not specify what constitutes a command-line argument,

because operating systems vary conSiderably on this pOint. However,
the most common convention is as follows: Each command-line
argument must be separated by a space or a tab "character. Commas,
semicolons, and the like are not considered separators. For example,
 A_~ .. _217
7.4 pASS ARGUMENTS TOmainO V

This is a test

is made up of four strings, but

this,that,and,another

is one string.
If you need to pass a command-line argument that does, in fact,
contain spaces, you must place it between quotes, as shown in this
example:

The names of argv and argc are arbitrary-you can use any names
you like. However, argc and argv are traditional and have been used
since C's origin. It is a good idea to use these names so that anyone
reading your program can quickly identify them as the command-line
parameters. •
One last point: the ANSI C standard only defines the argc and argv
parameters. However, your compiler may allow additional parameters
to maine ). For example, some DOS or Windows compatible compilers
allow access to environmental infonnation through a command-line
argument. Check your compiler's user manual.

. EXAMPLES
-
1. When you pass numeric data to a program, that data will be
received in its string form. Your program will need to convert
it into the proper internal fonnat using one or another of C's
standard library functions. The most common conversion
functions are shown here, using their prototypes:
int atoiCchar ·str);

double atof(char ·str);

long atol(char ·str)i

These functions use the STDLlB.H header file. The atoi( )

function returns the int equivalent of its string argument. The
218 nACIf YOURSElf
T C

atof( ) returns the double equivalent of its string argument,

and the atol() returns the long equivalent of its string
argument. !fyou call one of these functions with a string that is
not a valid number, zero will be returned. The follOwing program
demonstrates these functions. To use it, enter an integer, a long
integer; and a floating-point number on the command line It
will then redisplay them on the screen.
'include <stdio.h>
'include <stdlib.h>

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

(
int i;
double d;
long 1;

i = atoi(argv[l);
1 = ato1(argv[2));
d = atof(argv[3));

printf("'d 'ld if". i, 1. d);

return 0:
)

2. This program coverts ounces to pounds. To use it, specity the

number of ounces on the command line .
• include <stdio.h>
'include <stdlib.h>

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

(
double pounds;

pounds = atof(argv[l)) I 16.0;

printfC-'f pounds·, pougds);

return 0;
)

3. Although the examples up to this point haven't done so,

you should verity in real programs, that the right numbE:r of
 219
ACI.-LookatFun_
..,

command-line arguments have been supplied by the user. The

way to do this is to test the value of argc. For example, the
ounces-to-pounds program can be improved as shown here:
linclude <stdio.h>
#include <stdlib.h>

int main(int argc, char *argv{])

{
double pounds:

if (arge! =2) (
printf("Usage: CONVERT <ounces>\n");
printf("Try Again");
)
else (
"-
pounds; atof(argv[l]) / 16.0;
printf!-'f pounds·, pounds);
)

return 0;
)

This way the program will perform a conversion only if a

command-line argument is present. (Of COlfrse, you may
'. choose.)
prompt the user for any missing information, if you
Generally, the preceding program will be wrinen by a
professional C programmer like this:
'include <stdio.h>
#include <stdlib.h>

int main(int argc, char *argv[})

{
double pounds;

i f (arge! ;2) {
printf (·Usage: CONVERT <ounces:>\n-); "
printf ("Try Again-) ;

)
•
ex~t(l); /* stop the program *'
pounds; atof(argv[l]) / 16.0;
printf ( • \f pounds', pounds);
 .,
220 WlCH YOURSlI.f
c

return -0;
)

When some condition necessary for a program's execution has

not been met, most C programmers call the standard library
function exit( ) to terminate the program. The exit( ) function
has this prototype:

lloid exit(int return-code);

and uses the STDLlB.H header file. When exit( ) terminates

the program, it returns the value of return-code to the operating
system . By convention, most operating systems use a return
code of zero to mean that a program has terminated normally.
Nonzero values indicate abnormal termination.

1. Write a program that accepts two command-line arguments.

Have the program compare them and report which is
lexicographically greater than the other.
2. Write a program that takes two numeric arguments and display
their sum.
3. Expand the program in Exercise 2 so that it takes three
arguments. The first argument must be one of these words: add
subtract, multiply, or divide. Based on the value of the first
argument, perform the requested operation on.the remaining
two numeric arguments.

COMPARE. OLD-STYLE TO MODERN

FUNCTION PARAMETER DECLARA TlONS
/

Early versions of C used a different parameter declaration method

than has been shown in this book. This original declaration method is
 ACIose.Look .. _ 221
7.5 COMPARE OW-STYI..E TO MODERN FUNCTION PAA'METER DEClARATIONS "
.
now called the old-style or classic form. The form used in this book is
the modern form. It was introduced when the ANSI C standard was
created. While the modem form should be used for all new programs,
you will still find examples of old-style parameter declarations in older
programs and you need to be familiar with it.
The general form of the old-style parameter declaration is shown
here:

type function-name(parameler I, parameter2,._parameterN)

type parameter I ;
type parameler2;

type parameterN;
(
function-code
I
Notice that the declaration is divided into two parts. Within the
parentheses, only the names of the parameters are specified. Outside
the parentheses, the types and names are specified. For example ,
given the following modem declaration
float f(char ch, long size. double rnaK)
(

the equivalent old-style declaration is

float f(ch, size. max)
char ch;
long size;
double max,
(

)
222 TEACH YOURSELF

'" C

One other aspect of the old-style declaration is that you can specify
more than one parameter after the type name. For example, this is
perfectly valid:
myfunc (i. j. k)
int i, j, k;
{

The ANSI C standard specifies that either the old-style or the

modern declaration form may be used. The reason for this is to
maintain compatibility with older C programs. (There are literally
millions oflines of C code still in existence that use the old-style
form.) So, if you see programs in books or magazines that use the
classic form, don't worry; your compiler will be able to compile
them. However for all new programs, you should definitely use the
modern form.

EXAMPLE

I. This program uses the. old declaration fonn;

#include <stdio.h>

int area{int 1. int w)

int main(voidl
(
printfC-area is %d-, area(lO. 13»);
return 0;
}

int area(l. w)
int 1. w;
(
return 1 .. w.
)
 A CIo_1.Dok at FunctIons 223
7.5 COMPARE OLD·STYl.E TO MODERN FUNCT10N PARAMETER DECLARATKJNS '"
Notice that even though the old form of parameter declaration is
used to define the function, it is still possible to prototype the
function.

EXERCISE
1. Convert this program so that Cto_m( ) uses the old-style
declaration form.
#include <stdio.h>

int main (void)

(
double feet;

printf (-Enter feet: .. );

scanf("'lf", &feet);
printf("Meters: %f", f_to_m(feetl l;

return 0;
)

double f_to_m{double fl
(
return f I 3.28;
)

At this point you should be able to answer these questions and

perfonn these exercises:

1. How do you prototype a function that does not have parameters?

2. What is a function prototype, and what are the benefits of it?
3. How do command-line arguments get passed to a C program?
224
., TEACH YOURSlLf
c

4. Write a program that uses a recursive function to display the

letters of the alphabet.
s. Write a program that takes a string as a command-line
argument. Have it output the string in coded form . To code the
string, add 1 to each character.
6. What is the prototype for this function?
double myfunc(int x, int y, char ch)
(

7. Show how the function in Exercise 6 would be coded using the

old-style function declaration.
8. What does the exit( ) function do?
9. What does the atoi( ) function do?

This section.checks how wen you have integrated the material in

this chapter with that from earlier chapters.

1. Write a program that allows access only if the user enters the
correct password as a command-line parameter. If the user
enters the right word, print Access Pennitted; otherwise print
. Access Denied.
2. Create a function caned 8triniLup( ) that transforms the string
it is caned with into uppercase characters. Demonstrate its use
in a program. (Hint. use the toupper( ) function to convert
lowercase characters into uppercase.)
 A Closer Look at Functions
,.
225
7.5 COMPARE OLD-STYLE TO MODERN FUNCT10NPARAMETfR DEClAAAnONS

3. Write a function called avg( ) that averages a list of floating-point

values. The function will have twO arguments. The first is a
pointer to the array containing the numbers: the second is an
integer value, which specifies the size of the array. Demonstrate
its use in a program.
4. Explain how pointers allow C to construct a call-by-reference
parameter. .
8
Console 110

chapter objectives

8.1 Learn another preprocessor directive

B.2 Examine character and string input and

output

B.3 Examine some non-standard console

functions

8.4 Take a closer look at gels( ) and pUIS()

8.5 Master printfO

8.6 Master scanlO

227
~
228 TEACH YOURSELF

• C

N this chapter you will learn about.C's console I/O functions.

These are the functions that read or write information to and fro
the console. You have already been using some ·ofthese functions.
Here we will look at them in detail. This chapter begins with a
short but necessary digression that introduces another of C's
preprocessor directives: #define.

Before proceeding, you should be able to answer these questions

and perform these exercises:

1. What. must you do to enable the compiler to check that a

function is being called correctly?
2. What are the principal advantages of using function prototypes?
3. Write a program that uses a function called hypot( ) that
returns the length of the hypotenuse of a right triangle when
passed the length of the two opposing sides. Have the function
return a double value. The type of the parameters must be
double as well. Demonstrate the function in a program. (The
Pythagorean theorem states that the sum of the squares of the ·
two opposing sides equals the square of the hypotenuse..)
4. What return type should you use for a function that returns
no value?
5. Write a recursive function called rstrlen( ) that uses recursion
to compute the length of a string. Demonstrate it in a program.
6. Write a program that reports how many command line
arguments it has been called with. Also, have it display the
contents of the last one.
7. How is this declaration coded using the old-style function
declaration form?
void func(int a, char ch. double d)
 CONSOU 110 22B
a, LEARN ANOTHER PREPROCESSOR DIRECTIVE '"

EARN ANOTHER PREPROCESSOR

DIRECTIVE

As you recall, the C preprocessor performs various manipulations on

the source code of your program before it is actually compiled. A
preprocessor directive is simply an instruction to the preprocessor. Up
to this point, you have learned about and have used one preprocessor ·
directive, #include. Before proceeding, you need to learn about
another: #define.
(The #define directive tells the preprocessor to perform a text
substitution throughout your entire program. That is, it causes one
sequence of characters to be replaced by another. This process is
generally referred to as macro substitution. The general form of the
#define statement is shown here:

4Fdefine macro-name character-sequence

Notice that this line does not end in a semicolon. Each time the
macro-name is encountered in the progranl, the associated
character-sequence is substituted for il) For example, consider
this program:
*include <stdio.h>

#define MAX 100

int main(void)
(
int i;

for(i=O; i<MAX; i++) printf("%d i);

return 0;
)

When the identifier MAX is encountered by the preprocessor, 100 is

automatically substituted. Thus, the for loop will actually look like this
to the compiler;
for(i=O; i<100; i++) printf(-'d ., i);

(Keep one thing clearly in mind: At the time of the substitution, 100 is
simply a string of characters composed of a I and two as. The
230 TfACH YOURSELF
... C

preprocessor does not convert a numeric string into its internal binary
format. This is left to the compiler.)
(The macro name can be any valid C identifier. Thus, macro names
must follow the same naming rules as do variables. Although macro
names can appear in either upper- or lowercase letters, most
programmers have adopted the convention of using uppercase for
macro names. This makes it easy for anyone reading your program to
know when a macro name is being used.)
(There must be one or more spaces between the macro name and
the character sequence. The character sequence can contain any type
of character, including spaces. It is terminated by the end of the line.)
Preprocessor directives in general and #define in particular are not
affected by C's code blocks. That is(whether you define a macro name
outside of all functions or within a function, once it is defined, all code
after that point may have access to it. For example, this program
prints 186000 on the screen.
#include <stdio.h>

void f (void) ;

int mpin(void)
(
#define LIGHTS PEED 186000

f!) ;

return 0;
)

void f (void)

printf("%ld", LIGHTSPEED);

There is one important point you must remember: Each

preprocessor directive must appear on its own line.
Macro substitutions are useful for two main reasons. First, many C
library functions use certain predefined values to indicate special
conditions or results. Your programs will need access to these values
when they use· one of these functions. However, many times the
actual value will vary between programming environments. For this
 CONSOLE 110 231
'II'
8.] LEARN ANOTHER PREPROCESSOR DIREcnVE

reason, these values are usually specified using macro names. The
macro names are defined inside the header file that relates to each
specific function. You will see an example of this in the next section.
The second reason macro substitution is important is that it can
help make it easier to maintain programs. For example, if you know
that a value, such as an array size, is going to be used several places in
your program, it is better to create a macro for this value. Then if you
ever need to shange this value, you simply change the macro
definition. A'll references to it will be changed automatically when the
program is recompiled.

EXAMPLES
1. Since a macro substitution is simply a text replacement, you can
use a macro name in place of a quoted string. For example, the
following program prints Macro Substitutions are Fun.
#include <stdio.h>

#define FUN "Macro Substitutions are Fun"

int main(void )
{
printf (FUN);

return 0;
)

To the compiler, the printfC ) statement looks like this:

printf("Macro Substitutions a~e Fun");

2. Once a macro name has been defined, it can be used to help

define another macro name. For example, consider this program:
#include <stdio.h>

#define SMALL 1
#define MEDIUM SMALL.!
#define LARGE MEDIUM+!

int main (void) .

{
printf("%d %d %d", SMALL, MEDIUM, LARGE);
232 TEACH YOURSElf
c

return 0;
)

As you might expect, it prints} 2 3 on the screen.

3. If a macro name appears inside a quoted string, no substitution
will take place. For example, given this definition
#define ERROR ·catastrophic error occurred"

the following statement will not be affected.

printf (·ERROR: Try again");

EXERCISES

1. Create a program that defines two' macro names. MAX and

COUNTBY. Have the program count from zero to MAX-} by
whatever value COUNTBY is defined as. (Give COUNTBY the
value 3 for demonstration purposes.)
2. Is this fragment correct?
#define MAX MIN+I00
#define MIN 10

3. Is this frilgment correct?

#define STR this is a test

printf(STR) ;

4. Is this program correct?

#define STDIO <stdio.h>
#include STOIC

int main(void)
(
printf("This is a test.");

return 0;
)
 CONSOLE va 233
8.2 EXAMfNE CHARACTER AND STRING INPUT AND O/.JTPUT
"

XAMINE CHARACTER AND STRING

INPUT AND OUTPUT
Although you have already learned how to input and output characters
and strings, this section looks at these processes more formally.
The ANSI C standard defines these two functions that perform
character input and output, respectively;

int getchar(void);
int pUti:har(int ch);

They both use the header file STDIO.H. As mentioned earlier in

this book, many compilers implement getchar( ) in a line-buffered
manner, '\vhich makes its use limited in an interactive environment.
Most compilers contain a non-standard function called getche( ),
which operates like getchar( ) , except that it is interactive. Discussion
of getche( ) and other non-standard functions will occur in a later
section.
The getchar( ) function returns the next character t"ped on the
ke'yboard. This character is read as an unsigned char convened
to an into Ho\\'c\'er, most commonly, your program will assign this
value to a char variable, even though getchar( ) is declared as
returning an intoIf you do this, the high-order byte(s) of the integer
is simply discarded.
The reason that getchar( ) returns an integer is that when an error
occurs while reading mput, getchar( ) returns the macro EOF, which
is a negat've integer (usually -1). The EOF macro, defined in
STDIO.H, stands for end-of-file. Since EOF is an integer value, to allow
it to be returned, getchar( ) must return an integer. In the vast
majority of circumstances, if an ~rror occurs when reading from the
keyboard, it means that the computer has ceased to function.
Therefore, most programmers don't usually bother checking for EOF
when using getchar( ). They just assume a valid character has been
returned. Of course, there are circumstances in which this is not
appropriate-for example, when lIO is redirected, as explained in
Chapter 9. But most of the time you wi11 not need to worry about
getchar( ) encountering an error.
The putchar( ) function outputs a single character to the screen .
Although its parameter is declared to be of type int, this value is
converted into an unsigned char by the funr.rion Thus, only the
234 TfACH VOURSElF
c

low-order byte of ch is actually displayed. If the output operation is

successful, putchar() returns the character written. If an output
error occurs, EOF is returned. For reasons similar to those givert for
getchar( ), if output to the screen fails, the computer has probably
crashed anyway, so most programmers don't bother checking the
return value of putchar( ) for errors.
The reason you might want to use putchar( ) rather than printf( )
with the %c specifier to output a character is that putchar( ) is faster
~nd more efficient. Because printf( ) is more powerful and flexible, a
call to iirintf( ) geti'er,iltes greater oyerhead than a call to putchar().

EXAMPLES

1. As stated earlier, getchar( ) is generally implemented using

line buffering When input is line buffered, no characters are
actually passed back to the calling P' ugram until the user
presses E~TER. The follOWing program demonstrates this:
#include <stdio.h>

int main(voidl
(
char chi
do (
ch = getchar ( ) ;
pu tchar ( , . ' ) ;
while(ch != '\n');

return 0;
}

Instead of printing a period between each character, what you

will see on the screen is all the letters you typed before pressing
ENTER, followed by a string of periods.
One other point: When entering characters using getchar( ),
pressing ENTER will cause the newline character (\n) to be
returned. However, when using one of the alternative
non-standard functions, pressing ENTER will cause the carriage
return character (\r) to be returned. Keep this difference in mind.
2. The following program illustrates the fact that vou can use C's
backslash character constants with putchar( )-.
 c:08sllu I/O 235
43 EXAMINE SOME NON-STANONID CONSOLE FiJNCnONS
'"
'include <stdio.h>

int main (void)

(
putchar( 'A');
putchar(' \n');
putchar( 'B');

return 0;
)

This program displays

A
B

on the screen.

EXERCISES

1. Rewrite the program shown in the first example so that it

checks for errors on both input and output operations.
2. What is wTOng with this fragment?
char str[a01 = -this is a test";

putchar(str) :
•

AMINE SOME NON-STANDARD

CONSOLE FUNCTIONS

Because character input using getchar( ) is usually Jine-buffel'ed, many

compilers supply additional input routines that provide interactive
character input. You have already been introduced to one of these:
. getche( ). Here is its prototype and that of it~ close relative getch( ):
int getche(void);
int getch(void);
238 TEACH YOURSELF

• C

Both functions use the header file CONIO.H. The getche( ) function
waits until the next keystroke is entered at the keyboard. When a key
is pressed, getche( ) echoes it to the screen and then immediately
returns the character. The character is read as an unsigned char and
elevated to into However, your routines can simply assign this value to
a char value. The getch( ) function is the same as getche( ), except
that the keystroke is not echoed to the screen.
Another very useful non-ANSI-standard function commonly
supplied with a C compiler is kbhit( ). It has this prototype:

int kbhit(void);

The kbhit( ) function also requires the header file CONlO.H. This
function is used to determine whether a key has been pressed or not.
If the user has pressed a key, this function returns true (nonzero), but
does not read the character. If a keystroke is waiting, you may read it
with getche( ) or getch(). If no keystroke is pending, khhit( )
returns false (zero).
For some compilers. the non-standard i/O functions such as
getche( ) are not compatible with the standard i/O functions such
as printf( ) or seanf( ). When this is the case, mixing the two can
cause unu.sui\1 program behavior. Most troubles caused by this
incompatibility occur when inputting information (although problems
could occur on output). If the standard and non-standard 110
functions are not compatible in your compiler, you may need to use
non-standard versions of scanf( ) and/or printf( ), too. These ar~
called cprintf( ) and cscanf( ) .
The cprintf( ) function works like printf( ) except that it does not
translate the newline character (\n) into the carriage return, Iinefeed
pai~ as does the printf( ) function. Therefore, It is necessary to
explicitly output the carriage return (\r) where desired. The cscanf( )
function works like the seanf( ) functiOn. Both cprintf( ) aDd .
cscanf( ) use the CONIO.H header file. The cpdJltf( ) and C8C8nf( )
Junctions are expressly designed to be compatible with .,ach( ) and
getche( ), as well as other non-standard I/O functions.

Microsoft C++ suppons the functions just dasctfbed. In addifion. it prrNides

altemative names for the functions that begin with an undetscote. For ~
when using Visual C++, you can specify getdre() as ...fJflIChe(), too.
 COlli DIUO D'J
• •

One last point: Even for compilers that have Incompatibilities

between the standard and non-standard 1/0 functions, such
incompatibilities sometimes only apply in one case and not another. If
you encounter a problem, just try substituting a different function.

EXAMPLES

1. The getc:h( ) function lets you take greater control of the

screen because you can determine what is displayed each time a
key is struck. For example, this program reads characters until a
'q' is typed. All characters are displayed in uppercase using the
cprintf( ) function.
'include <stdio.h>
'include <conio.h>
'include <ctype.h>

int mainevoid)
{
char Chi

do (
ch = getch();
cprintf(·%c". toupper(ch»;
} while(ch != ' q ' };

return 0;
)

2. The kbbit( ) function is very useful when you want to let a

user interrupt a routine without actually forcing the user to
continually respond to a prompt like ·Continue?'. For example,
this program prints a 5-percent sales-tax table in increments of
20 cents. The program continues to print the table until either
the user strikes a key or the maximum value is printed.
'include <8tdio.h>
'include <conio.h>

int main (void)

{
double amount;
238 TEACHYOURSW
... C

amount = 0.20;
cprintf( "Printing 5-percent tax table\n\r");
cprintf("Press a key to stop.\n\n\r");
do (
cprintf(-amount: 'f. tax: %f \n\r-, amount.
amount-O.OS);
if(kbhit(» break;
amount = amount + 0.20;
} while(amount < 100.0);

return 0:
)

In the calls to cprintf( ), notice how both the carriage return

(\r) and the newline (\n) must be output. As explained,
cprintf( ) does not automatically convert newlines into
carriage return, Iinefeed pairs.

DEBe. .

I. Write a program that displays the ASCII code of each character

typed. Do not display the actual character, however.
2. Write a program that prints periods on the screen until you
press a key.

KE A CLOSER LOOK AT AND

puts( )

Although both gets( ) and puts( ) were introduced earlier, leI's take a
closer look at them now. Their function prototypes are

char "gets(char ·slr);

int puts(char "51";

These functions use the header file STDlO.H-. The gets( ) function
reads characters entered at the keyboard until a carriage return is read
(i.e., until the user presses ENTER). It stores the characters in the array
 CON8OL£ va 238
'"

pointed to by str. The carriage return is not added to the string.

Instead, it is converted into the null terminator. If successful. gets( )
returns a pointer to the start of SfT. If an error occurs, a null pointer
is returned.
The puts( ) function outputs the string pointed to by str to the
screen. It automatically appends a carriage return. line-feed sequence.
If successful, puts( ) returns a non-negative value. If an error occurs,
EOF is returned.
The main reasol1you may want to use puts( ) instead of printf( ).,
to output a string is that puts( ) is much smaller and faster. WhtJe this
is not important in the example programs shown in this book. it may
be in some applications.

EXAMPLES

I. This program shows how you can use the return value of gets( )
to access the string holding the input information. Notice that
this program also confirms that no error has occurred before
attempting to use the string.
'include <stdio.h>

int main(void)
(
char "p. str[BO];

printf (6Enter a string: ");

p = gets(str);
if{p) IIr if not null */
printf("%s %5", p. str);

return 0;
}

2. If you simply want to make sure that gets( ) did not encounter
an error before proceeding, you can place gets( ) directly inside
an if statement, as illustrated by the following program:
'include <stdio.h>

int main(void)
(
char str[80];
240
•
printf (-Enter a string: .);
if(gets(str)) /* if not null */
printf(~Here is your string: %5·, str);

return 0;
)

Because a null pointer is false, there is no need for the

intermediary variable p, and the gets( ) statement can be put
directly inside the if. .•
3. It is important to understand that even though gets( ) returns a
pointer to the start of the string, it still must be called with a
pointer to an actual array. For example, the following is wrong:
char .p; -
p = gets(pl; /"" wrong!!! */

Here, there is no array defined into which gets( ) can put the
string.,This will result in a program failure.
'4. This program outputs the words one, two, and three on three
separate Jines. using puts( ).
#include <stdio.h>

int main(void)
{
puts Vone");
puts(ntwo") ;
puts(ftthree") ;

return 0;
)

EXERCIRS

1. Compile the program shown in Example 2, above. Note the size

of the compiled code. Next, convert it so that it uses pl'intf( )
statements, instead ofputs( ). You wiJI find that the printf{ :'
version is several byte, larger.
 CONSOLE 1/0

8.5 MASTER printfO

,.
241

2. Is this program correct? If not, why not?

#include <stdio.h>

int main (void l

(
char "p, *q;

printf (" Enter a string: ~);

p = gets(q);
printf (p) ;

return 0;
}

STER printf()

Although you already know many things about printf( ) , y ou will be

surprised by how many more features it has. In this section you wi1l
Jearn about some morc of them . To begin, let's revicv." what you
know so far.
The printf( ) function has this prototype:

int printf(char 'control-string, ... );

The periods indicate a variable-length argument list. The printf( )
function returns the number of characters output. If an error occurs, it
returns a negative number. Frankly, few programmers bother with the
return value of printf( ) because, as mentioned earlier, if the console
is not working, the computer is probably not functional anyway .
The control string may contain two types of items: characters ,to be
output and format speCifiers. All format speCifiers begin with % . A
[annat specifier, also referred to as a format cat/e, determines lmw Its
matching argument will be displayed . Fe!l'mat specifiers and their
arguments are matched from left to right , and there must be as many
arguments as there are specifiers.
The format specifiers accepted by printf( ) are shown in Table 8-1 .
You have already learned about the %c, %d, %5, %u , %p, and %f
specifiers. The others will be examined now.
242 TfACH YOURSELF
." C

Code Fonnat
%c Character
%d Signed decimal integers
<¥oi Signed decimal integers
%e Scientific notation Oowercase 'e')
OfoE Scientific notation (uppercase 'F)
%f Decimal floating point
%g Uses Ofoe or Ofof, whichever is shorter
%G Uses %E or 0/01, whichever is shorter
0100 Unsigned octal
%s String of characters
%u Unsigned decimal integers
%x Unsigned hexadecimal Oowercase letters)
OfoX Unsigned hexadecimal (uppercase letters)
Ofop Displays a pointer
Ofon The associated argument is a pointer to an integer into which the
number of characters written so far is placed.
010010 Prints a % sign
The prinlf() Formal Specifiers ."

The %i command is the same as %d and is redundant.

You can display numbers of type float or double using scientific
notation by using either %e or %E. The only difference between the
two is that %e uses a lowercase 'e' and %E uses an uppercase 'E'.
These specifiers may have the L modifier applied to them to allow
them to display values of type long double.
The %g and %G specifiers cause output to be in either normal or
scientJfic notation, depending upon which is shorter. The dIfference
between the %g and the %G is whether a lower- or uppercase 'e' IS
used in cases where scientific notation is shorter. These speCifiers may
have the L modifier applied to them to allow them to display values of
type long double.
You can display an integer in octal format using %0 or in
hexadecimal using %x or %X. Using %x causes the letters 'a' through
'f' to be displayed in lowercase. Using %X causes them to be displayed
in uppercase. These specifiers may have the h and I modifiers applied
to allow them to display short and long data types, respectively .
 CONSOLE 110 243
."
a 5 MAS1l'R prind( )

The argument that matches the %n specifier must be a pointer to an

integer. When the %n is encountered, printf( ) assigns the integer
pointed to by the associated argument the number of characters
output so far.
Since all format commands begin with a percent sign, you must use
%% to output a percent sign.
All but the %%, %p, and %e specifiers may have a minimum-
field-width specifier and / or a precision specifier associated with them.
Both of these are integer quantities. If the item to output is shorter
than the specified minimum field width, the output is padded with
spaces, so that it equals the minimum width. However, if the
output is longer than the minilTIUm, output is nor truncated. The
minimum-field-width specifier is placed after the % sign and before
the format specifier.
The precision speCifier follows the minimum-field-width speCifier.
The two are se parated by a period. The precision specifier affects
different types of format specifiers differently . If it is applied to the
%d, %i, %0 , %u or %x specifiers, it determines how many digits are to
be shown . Leading zeros are added if needed. When applied to %f, %e,
or %E, it determines how many digits will be displayed after the
decimal point . for %g or %G, it determines the number of significant
digits. When applied to the %s, it specifies a maximum field width. If a
string is longer than the maximum-field-width specifier, it will be
truncated.
By default, all numeric output is right justified. To leftjustify
output , put a minus sign directly after the % sign.
The general form of a format speCifier is shown here. Optional
items are shown between brackets,

% [-] [minimum-field-width] [.] [precision] format -specifier

for example , this format specifier tells printf( ) to output a double

value using a field width of 15, with 2 digits after the decimal point.
%15.2f

.'. . . EXAMPLES '

I. If you don't want to specify a minimum field width, you can

still specify the precision . Simply put a period in front of the
precision value, as illustrated by the following program:
244 TEACH YOURSELF

• C

#include , <stdio.h>

int main{V01!tt
(
l>rintfl"%.5d\n", 10);
printfl"S%.2f\n", 99.95);
printf("%.lOs", -Not all of this will be printed\n-j;

return 0;
}

The output from this program looks like this:

00010
$99.95
Not all of

Notice the effect of the precision specifier as appli~d to each

data type.
2. The minimum-field-width specificr is especially useful lor
creating tables that contain columns of numbers that must line
up. For example, this program prints 1000 random numbers in
three columns. It uses another of C's standard lib'rary functions,
rand( ), to generate the random numbers. The raJ:ld( )
function returns a randolll integer value each time it is called. It
uses the header STDLlB.H.
#include <stdio.h>
#include <stdlib.h>

int main (void)

int ii

for(i=O; i<1000; i++)

printf!"%10d %lOd %10d\n", randt) , rand!), rand!));

return 0;
}

Part of the output from this program is shown here. Notice how
the columns are aligned. (Remember, if you try the program,
you will probably see different numbers.)
 10982 130 346
7117 11656 1090
22948 6415 17595
14558 9004 31126
18492 22879 3571
26721 5412 1360
27119 25047 22463
13985 7190 31441
30252 27509 31214
19816 14779 26571
17995 19651 21681
13310 3734 23593
15561 21995 3979
11288 18489 16092
5892 8664 28466
5364 22766 13863
20427 21151 17639
8812 25795 100
12347 12666 15108

3. This program prints the value 90 four different way.: decimal,

octal, lowercase hexadecimal, and uppercase hexadecimal. It
also prints a floating-point number using scientific notation with
a lowercase 'e' and an uppercase 'E' .
• include <stdio.h>

int main(voidl
{
printf ("%d %0 %x %X\n" , 90, 90. 90, 90);
printf("%e U\n", 99.231, 99.231);

return. 0;
)

The output from this program is shown here:

90132 Sa 5A
9.92310e+01 9.92310E+01
4. The following program demonstrates the %n specifier:
.include <stdio.h>

int main(void)
{
246 TtACH YOURSW'

'" C

-
int i;

printf(-%d %f.n%n-, 100, 123.23, &i);

printf (-%d characters 01.J.tput so far-. i);

return 0;
)

Its output looks like this:

100 123.230000
15 characters output so far

The fifteenth character is the newline.

EJlERCISU

1. Write a program that prints a table of numbers, each line

consisting of a llunlber, its square, and 1[5 cuhe. Have the table
begin at 2 and end at 100. Make the columns line up, and left
justify each column.
2, How would 'you output this line using printf( F
Clearance price: 40% off as marked

3. Show how to display 1023,03 so that only two decimal places

are printed.

. ' 8_6 . MASTER scanf()

Like printf( ), seanf( ) has many more features than we ha,'e used
so far. In this section, several of these additional features are explored
Let's begin by reviewing what you have already learned.
The prototype for seanf( ) is shown here:

int scanf(char 'control-string, ... );

The col1trol-sml1g consists mostly of format specifiers. However, it can

contain other characters. (You will learn about the effect of other
characters in the control string soon.) The format specifiers determine
 CONSOLE 110 247
8.6 MASTER scanf()

how seanf( ) reads information into the variables pointed to by the

arguments that follow the control string. The specifiers are matched in
order, from left to right, with the arguments. There must be as many
arguments as there are specifiers. The format specifiers are shown in
Table 8-2. As you can see, the seanf( ) specifiers are very much like
the printf( ) specifiers.
The seanf( ) function returns the number of fields assigned values.
If an error occurs before any assignments are made, EOF is returned.
The specifiers %x and %0 are u~ed to read an unsigned integer
using hexadecimal and octal number bases, respectively.
The specifiers %d, %i, %u, %x, and %0 may be modified by the h
when inputting into a short variable and by 1 when inputting into a
long variable.
The specifiers %e. %f, and %g are equivalent. They all read
floating-point numbers represented in either scientific notation or
8tandard decimal notation . Unmodified, they input information into a
float vanable . You can modify them using an 1 when inputting into a
double. To read a long double, modify them with an L.
You can use seanf( ) to read a string using the %s specifier, but you
probably won 't want to . Here's why: When seanf( ) inputs a string, it
stops reading that string when the first whitespace character is
encountered. A whitespace character is either a space, a tab, or a

.ode Meaning
c Read a single character
d Read a decimal integer
Read a decimal mteger
e Read a floating-point number
f Read a floating-point number
g Read a floating-point number
Read an octal number
Read a string
Read a hexadecimal number
Read a pointer
n Receives an integer value equal to the number of characters read so far
%u Read an unsigned integer
%[ 1 Scan for a set of characters

The scanf( ) Format Specifiers T

248 ltACH YOU/lSElf
c

newline, This means that you cannot easily use seanf( ) to read input
like this into a string:
this is one string

Because there is a space after "this," seanf( ) will ,stop inputting

the string at that point. This is why gcts( ) is generally used to
input strings.
The %p specifier inputs a memory address using the format
determined by the host environment, The %n specifier assigns the
number of characters input up to the point the %n is encounte~ed to
the integer variable pointed to by its matching argument, The %n may
be modified bv "
either I or h so that it may
. assign
.
its value to either a
long or short variable,
A very interesting feature of seanf( ) is called a scanser , A scanset
specifier is created by putting a list of characters inside square
brackets, for example , here is a scanset specifier containing the
letters 'A Be.'
% [,>,BC]

When s"anf( ) encounters a scanset, it begins reading input into

the character array pointed to by the scanset's matching argument. It
will only continue reading characters as long as the next character is
part of the scanset. As soon as a character that IS not part of the
scanset is found, seanf( ) stops reading input for this specifier and
moves on to any others in the control string.
You can specify a range in a scanset using the· (hyphen), for
example, this scanset specifies the characters 'A' through 'Z',
\; [A-Z]

Technicaliy, the use of the hyphen to specify a range is not specified

by the ANSI C standard, but it is nearly universally accepted,
When the scanset is very large, sometimes it is easier to specify
what is not part of a scanset, To do this, precede the set with a " ,
for example,
%[-0123456789]

When seanf( ) encounters this scanset , it will read any characters

except the digits 0 through H.
 CON~IIO 249
46 WlS7fRICMIt) •

You can suppress the assignment of a field by putting an asterisk

immediately after the % sign. This can be very useful when inputting
information tha~contains needless characters. For example, given this
seanf( ) statement
int first , second;
scanf( "%d'*c%d", &first. &second);

this input
555-2345

will cause seanf( ) to assign 555 to first, discard the ., and assign
2345 to second. Since the hyphen is not needed, there is no reason to
assign it to anything. Hence, no associated argument is supplied.
You can specify a maximum field width for all specifiers except %e,
for which a field is always one character, and %n, to which the
concept does not apply. The maximum field width is speCified as an
unsigned integer, and it immediately precedes the format specifier
character. For example , this limits the maximum length of a string
assigned to str to 20 characters:
sca n f ( - %2 0 s· , str);

If a space appears in the control string, then seanf( ) will begin

reading and discarding whitespace characters until the first
non-whitespace character is encountered. If any other character
appe ars in the control string, seanf( ) reads and discards all matching
characters until it reads the first character that does not match
that character.
One other point: As seanf( ) is generally implemented, it
line-buffers input in the same way that getehar( ) often does. While.
this makes lit~le difference when inputting numbers, its lack of
interactivity tends to make seanf( ) of limited value for other
types of input.

EXAMPLES

1. To see the effect of the %s specifier, try this program. When

prompted, type this is a test and press ENTER. You will see
only this redisplayed on the screen. This is because, when
reading strings, seanf( ) stops when it encounters the first
whitespace character.
250 TEACH YOURSW
c

#include <stdio.h>

int main (void)

(
char str [aD);

'*printfC"Enter
Enter "this is a test" */
a string: ");
scanf (. %5·, str);
printf (str);

return 0;
)

2. Here's an example of a scanset that accepts both the upper- and

lowercase characters. Try entering some letters, then any other
character, and then some more letters. After you press ENTER,
only the letters that you entered before pressing the non-letter
key will be contained in str.
#include <stdio.h>

int main(voidl
(
char str(80];

printf("Enter letters. anything else to stop\n");

scanfl"%[a-zA-Z]", str);

printf (str);

return 0:
}

3. If you want to read a string containing spaces using scanf( ),

you can do so using the scanset shown in this slight variation of
the previous program.
'include <stdio.h>

int main(void)
(
char strlaO);

printfC"Enter letters and spaces\n");

 CONSOLEVa 2&1

scanf(-%(a-zA-Z ]-, str}:

printf (str) ;

return 0;
)

You could als,o specify punctuation symbols and digits, so that

you can read virtually any type of string. However, this is a
fairly cumbersome way of doing things.
4. This program lets the user enter a number followed by an
operator followed by a second number, such as 12 + 4. It then
performs the specified operation on the two numbers and
displays the results .
• include <stdio.h>

int main (void)

(
int. i. j;
char op;

printf(-Enter operation: "I;

scanf(-%d%c%d", &i. &OP, &j);

s ..... itch(op) {
case '.': printf("%d-, i+j);
break;
case '-': printf("'d-, i-j);
break;
case 'I': if{j) printf(-'d-, i/j):
break;
case '.' .: printf(-'d-, i*j);
)

return 0;
)

Notice that the format for entering the information is somewhat

restricted because no spaces are allowed between the first
number and the operator. It is possible to remove this
restriction. As you know, seanf{ ) automatically discards
leading whitespace characters except when you use the '!be
specifier. However, since you know that the operator will not be
252 TEACH YOURSW
c

a whitespace character, you can modify the scanf( ) command

to look like this:
scanfC"'d 'c'd" , &i. &op, &j);

Whenever there is a space in the control string, scanf( ) will

match and discard whitespace characters until the first
non-whitespace character is found. This includes matching zero
whitespace characters. With this change in place, you can enter
the information into the program using one or more spaces
between the first number and the operator.
5. This program illustrates the maximum-field-width specifier:
.include <stdio.h>

int main (void)

C
int i, j;

printf (- Enter an integer: .);

scanfC"'3d'd", &i, &j);
printfC "'d 'do, i. j);

return 0;
)

If you run this program and enter the number 12345, i will be
assigned 123, and j will have the value 45 . The reason for this is
that scanf( ) is told that i's field is only three characters long.
The remainder of the input is then sent to j.
6. This program illustrates the effect of having non-whitespace
characters in the control string. It allows you to enter a decimal
value, but it assigns the digits to the left of the decimal point to
one integer and those to the right of the decimal to another. The
decimal point between the two %d specifiers causes the decimal
point in the number to be matched and discarded.
'include <stdio.h>

int rnain(void)
(
int i. j;

printf(-Enter a decimal number: .);

 CONSOLE 110 253
a6 MAS1fR IeOnf()

scanf ("%d. %d", &i. &j);

printf (-left part: %d, right part: %d ~, i, j);

return 0;
)

EXERCISES
1. Write a program that prompts for your name and then inputs
your first, middle, and last names. Have the program read no
more than 20 characters for each part of your name. Finally,
have the program redisplay your name.
2. Write a program that reads a floating-point number as a string
using a scanset.
3. Is this fragment correct? If not why not?
char ch;

scanf ( -'t2c·, &ch);

4. Write a program that inputs a string, a double, and an integer.

After these items have been read, have the program display how
many characters were input. (Hint: use the %n specifier.)
5. Write a program that converts a hexadecimal number entered
by the user into its corresponding decimal and octal equivalents.

Before proceeding you should be able to answer these questions and

perform these exercises:

1. What is the difference between getchar( ), getche( ), and

getch( )?
2. What is the difference between·the %e and the %E printf( )
format specifiers?
254 TEACH YOURSELF
y
C

3. What is a scanset?
4. Write a program, using scanf( ), that inputs your first name,
birth date (using the format mml dd/yy), and telephone
number. Redisplay the information on the screen to verify that
it was input correctly.
5. What is one advantage to using puts( ) over printf( ) when
you only need to output a string? What is one disadvantage to
puts( )?
6. Write a program that defines a macro called COUNT as the
value 100. Have the program then use this macro to control a
for loop that displays the numbers a through 99.
7. What is EOF, and where is it defined?

Cumulative
Check

This section checks how well you have integrated the material in
this chapter with that from earlier chapters.
1. Write a program that allows you to enter the batting averages for
the players on a little league team. (Assume there are exactly 9
players.) Have the user enter the first name and batting average
of each player. Use a two-dimensional character arrav to hold
the names and a one-dimensional double array to hold the
batting averages. Once all the names are entered, have the
program report the name and average of the players with the
highest and lowest averages. Also, have the program display the
team average.
2. Write a program that is a simple electronic library card catalog.
Have the program display this menu:

Card Catalog:
1. Enter
2. Search by Author
3. Search by Title
4. Quit
Choose your selection:
 CONSOLE 110
8.6 MASTER scanf()
.,
255

If you choose Enter, have the program repeatedly input the name,
author, and publisher of a book. Have this process continue until the
user enters a blank line for the name of the book.
For searches, prompt the user for the specified author or title and
then, if a match is found, display the rest of the information. After you
finish this program, keep your file, because in the next chapter you
wi11learn how to save the catalog to a disk file.

'.
'.
9
File I/O

chapter objectives

9.1 Understand streams

9.2 Master file-system basics

9.3 Understand leol() and lelTor()

9.4 Learn some higher-level text functions
9.5 Learn to read and write binary data

9.G Understand random access

9.7 Learn about various file-system functions

9.8 Learn about the standard streams

257
'"
 .
258 TfACH YOURSELF

L THO UGH C does not define any keywords that perform

file 110, the C standard library contains a very rich set of
110 functions. As you will see in this chapter, C's approach
to 110 is efficient, powerful, and flexible.

Most C compilers supply two complete sets of file 110 functions. One is called
the ANSI file system (sometimes called the buffered file system). This file
system is defined by the ANSI C standard. The second file system is based on
the original UNIX operating environment and is called the UNIX-like file
system (sometimes called the unbuffered file system). This file system is not
defined by the ANSI C standard. The ANSI standard only defines one file
system because the two file systems are redundant. Further, not all
environments may be able to adapt to the UNIX-like system. For these
reasons, this book only discusses the ANSI file system. For a discussion of the
UNIX-like file system, see my book C: The Complete Reference (Berkeley,
CA, Osborne/McGraw-HilI).

Before proceeding you should be able to perform these exercises

and ans',,'er these questions:

I . What is the difference between getchar( ) and gctche( F

2. Gwe one ·reason why you probably won't use 8canf( )'s %8
option to read strings from the keyboard.
3. Write a program that prints a four-column table of the prime
numbers between 2 and 1000. Make sure that the columns
are aligned.
4. Write a program that inputs a double, a character, and a string
not longer than 20 characters. Redisplay the values to confirm
that they were input correctly.
5. Write a program that reads and discards leading digits and then
reads a string. (Hint: Use a scanset to read past any leading digits.)

•
 File I/O 259
9.1 UNDERSTAND STREAMS

NDERSTAND STREAM S

Before we can begin our discussion of file 1/0, you must understand
two very important concepts: the stream and the file. The C I/O system
supplies a consistent interface to the programmer, independent of the
actual I/O device being used. To accomplish this, C provides a level of
abstraction between the programmer and the hardware. This abstraction
is called a strean1':(The actual device providing I/O is called a file.
Thus, a stream is a logical interface to a file. As C defines the term file,
it can refer to a disk file, the screen, the keyboard, memory, a port, a
file on tape, and various other types of I/O devices. The most common
form of file is, of course, the disk file. Although files differ in form and
capabilities, all streams are the same. The advantage to this approach
is that to you, the programmer, one hardware device will look much
like any other The stream automatically handles the differences.
A stream is linked to a file using an open operatton. A stream is
disassociated from a file using a close operation)
There are two types of streams: text and binary. A text stream
contains ASClI characters. When a text stream is being used, some
character translations may take place. For example, when the newline
character is output, it is usually converted into a carriage return, !inefeed
pair. For this reason, there may not be a one-to-one correspondence
between what is sent to the stream and what is written to the file.
A binary stream may be used with any type of data. No character
translations will occur, and there is a one-to-one correspondence
between what is sent to the stream and what is actually contained
in the file.
One final concept you need to understand is that of the CUlTent
location. The current location, also referred to as the CUlTent position,
is the location in a file where the next file access will occur. For
example, if a file is 100 bytes long and half the file has been read, the
next read operation will occur at byte 50, which is the current location.
To summarize: In C, disk I/O (like certain other types of I/O) is
performed through a logical interface called a stream. All streams have
similar properties, and all are operated on by the same I/O functions,
no matter what type of file the stream is associated with. A file is the
••
260 TEACH YOURSElF

• C

actual physical entity that receives or supplies the data. Even though
files differ, streams do not. (Of course, some devices may not support
random-access operations, for example, so their associated streams
will not support such operations either.) .
Now that you are familiar with the theory behind C's file system, it
is time to begin learning about it in practice.

MASTER FILE-SYSTEM BASICS

In this section you will learn how to open and close a file. You will also
learn how to read characters from and write characters to a file.
(TO open a file and associate it with a stream, use fopen( ) . Its
prototype is shown here:

FILE 'fopen(char '(name, char 'mode); )

The fopen( ) fpnction, like all the file-system functions, uses the
header STDIO.H:(rhe name of the file to open is pointed to by fname.
It must be a valid file name, as defin~d by the operating system. The
string pointed to by mode determines how the file may be accessed.
The legal values for mode as defined by the ANSI C standard are shown
in Table 9-1. Your compiler may allow additional modes)
.../ (If the open operation is successful, fopen( ) returns a valid file
pointer. The type FILE is defined in STDIO.H. It IS a structure that
holds various kinds of information about the file, such as its size, the
current location of the file, and its access modes. It essentially
identifies the file)(A structure is a group of variables accessed under
one name. You will learn about structures in the next chapter, but you
do not need to know anything about them to learn and fully use C's
file system.)(The fopen( ) function rerurns a pointer to the structure
associated with the file by the open process. You will use this pointer
with all other functions that operate on the file. However, you must
never alter It or the object it pomts to)
...... G!fthe fopen( ) function fails, it returns a null pointer. The header
STDlO.H defines the macro NULL, which is defined to be a null pointer.
It is very important to ensure that a valid file pointer has been
remrned . To do so, check the value returned by fopen( ) to make
sure that it IS not NULL)for example, the proper way to open a file
called my file for text input is shown in this fragment:
 9,2
-
AiellO
MASTER FILE-SYSTEM BASICS
.
261

FILE Tfp;

ifllfp = fopenl "myfile" , Or") I == NULL) (

printf(MError opening file.\n~);
exit(l); / . or substitute your own error handler */
)

--/(Although most of the file modes are self-explanatory, a few

comments are in order. If, when opening a file for read-only
operations, the file does not exist, fopen( ) will fail. When opening a
file using append mode, if the file does not exist, it will be created.
Further, when a file is opened for append all new data written to the
file will be written to the end of the file. The original contents will
remain unchanged. If, when a file is opened for writing, the file does
not exist, it will be created. If it does exist, the contents of the original
file will be destroyed and a new file created. The difference between
modes r + and w + is that r + will not create a file if it does not exist;
however, w + will. Further, if the file already exists, opening it with
w + destroys its contents; opening it with r + does not)

I
Mode Meaning
I'r" Open a text file for readIng.
It
I'W Create a text file for writing.
I'a" Append to next file.
Open a binary file for reading.
"wb lt
Create a binary file for writing.
"ab" Append to a binary file.
"r+1t Open a text file for read/write
"w+" Create a text file for read/write
"a+" Append or create a text file for read/write.
IIr+b" Open a binary file for read/write. You may-also use "rb+".
I'w+b ll
Create a binary file for readiwrite.'(b~-niily '~iso u~e "wb+".
"a+b" Append or create a binary file fo~ read/write. You may ~Iso use "ab+".
, .....
The Legal Values for Mode :y. " " . ..... '.
TABLE 9-1
, ' "
. '.'.
-'
262 TEACH YOURSELF

" C

/ ( To close a file, use fcJose( ), whose prototype is

int fcloseCFILE 'fp):

The fcJose( ) function closes the file associated with fp, which must
be a valid file pointer previously obtained using fopen( ), and
disassociates the stream from the file. In order to improve efficiency,
most file system implementations write data to disk one sector at a
time. Therefore, data is buffered until a sector's worth of information
has been output before the buffer is physically written to disk. When
you call fcJose( ), it automatically writes any information remaining
in a partially full buffer to disk. This is often referred to as flushing the
buffer.
You must never call fcJose( ) with an invalid argument. Doing so
will damage the file system and possibly cause irretrievable data loss.
The fcJose( ) function returns zero if successful. If an error occurs,
EOF is returned)
J'(Once a file has been opened, depending upon its mode, you may
read and/or write bytes (i.e., characters) using these two functions:

int fgetc(FILE 'fp):

int fputcQnt eh, FILE' fp):

The fgetc( ) function reads the next byte from the file described by
fp as an unsigned ehar and returns it as an integer. (The character is
returned in the. low-order byte.) If an error occurs, fgete( ) returns
EOF. As you should recall from Chapter 8, EOF is a negative integer
(usually -1). The fgete( ) function also returns EOF when the end of
the file is reached. Although fgete( ) returns an integer value, your
program can assign it to a ehar variable since the low-order byte
contains the character read from the file.
The fputc( ) function writes the byte contained in the low-order
byte of eh to the file associated with fp as an unsigned char. Although
ell is defined as an int, you may cal1 it using a char, which is the
common procedure. The fpute( ) function returns the character
written if successful or EOF if an error occurs)
Historical note: The traditional names for fgete( ) and fpute( ) are
getc( ) and pute( ). The ANSI C standard still defines these names,
and they are essential1y interchangeable with fgete( ) and fputee ).
One reason the new names were added was for consistency. Al1 other
ANSI file system functio" names begin with 'f: so 'r was added to
 FlI.1I0 263
9.2
-
MASTER RLE-SYSTEM BASICS
~

getc( ) and putc( ). The ANSI standard still supports the traditional
names, however, because there are so many existing programs that
use them. Ifyoll see programs that use getc( ) and putc( ), don't
worry. They are essentially different names for fgetc( ) and fputc( ).

~. EXAMPLES '

1. This program demonstrates the four file-system functions you

have learned about so far. First, it opens a file called MYFILE for
output. Next, it writes the string "This is a file system test." to
the file. Then, it closes the file and reopens it for read
operations. Finally, it displays the contents of the file on the
screen and closes the file.
#incliJde <stdio.h>
_include <stdlib.h>

int main(void)
{
char str(801 = "This is a file system test. \n";
FILE "'fp;
char .p;
int i:

/- open myfile for output ... /

H(lfp = fopen{"rnyfile", "w"JJ==NULLJ (
printf(nCannot open file.\nR);
exit(l):
)

,- write str to disk ~I

p = str;
while l"pJ {
if(fputcl"p, fpJ==EOFJ (
printf(MError writing file. \n");
exit(!);

p++ ;

fclose(fp) ;
264 TEACH YOURSW

" C

/* open myfile for input */

ifllfp = fopenl"myfile" , "r"))==NULL) (
printf(-Cannot open file.\nM);
exitll) ;
)

/* read back the file */

forI;;) I
i = fgetc Ifp);
i£(i == EOF) break;
putchar{il;
)
fclose (fp) ;

return 0;
)

In this version, when reading from the file, the return value of
fgetc( ) is assigned to an integer variable called i. The value of
this integer is then checked to see if the end of the file has been
reached_ For most compilers, ho,,-ever, you can simply assign
the value returned by fgetc( ) to a char and still check for EOF,
as IS shown in the follo\,·.. ing version:

#include <s~dio.h>

#include <std~ib.h>

int main (void)

(
char str[BO] = -This is a file system test. \ n-;
FILE 'fp;
char ch. "p;

; * open myfile for output * /

if«fp = fopen(-myfile-, ·w·»==N'tJ.:....L) {
printf / ·Cannet oper. file. \ n-) i
exit (1);
)

It write str to disk */

p = str;
while ('p) {
if (fputc ('p, fp) ==EOF) (
printf! ~ Errcr writing file.\n·)j
exit(1 ) .
 File 110 265
'I'
9.2 MASTER FILE-SySTEM BASICS

}
p++
}
fcloselfp} ;

/* open myfile for input */

if«fp = fopen{"myfile", "r-»==NtlLL)
printf{~Cannot open file.\n");
exit(l);
}

/* read back the file */

fori;;} I
ch = fgetc (fp) ;
if(ch == EOF) break;
putchar(ch) ;

fclose (fp) ;

return 0;
}

The reason this approach works is that when a char is being

compared to an int, the char value is automatically elevated to
an equivalent int value.
There is, however, an even better way to code this program.
For example, there is no need for a separate comparison step
because the assignment and the comparison or> be performed
dL cne ~alne nme, \vithln the if, as shown here:
#include <stdio.h>
'include <stdlib.h>

int main(void}
(
char str(801 = "This is a file system test. \n~;
FILE *fp;
char ch, *p;

/* open myfile for output */

if«fp = fopen("myfile", "w"»==NULL) {
printf("Cannot open file.\n");
exitU);
}
 ..
266 nACH YOURSELF

/* write str to disk *j

P str;
=
while(*p) {
iflfputcl'p, fp)==EOF) (
printf(ftError writing file.\n-);
exit{l) ;
)
p-t-+ ;
)
fclose (fp) ;

/* open myfile for input */~

if«fp = fopen("myfile", "r"~)==N'uLL) {
printf("Cannot open file.\n");
exit(1);
)

/* read back the file ~ /

for!;;) {
if«ch = fgetc'fpl I -- EOF) break;
putchar(ch) ;
',.
fclose (fp) ;

return 0;

Don't let the statemem

if«ch = fgetc(fp)) == EOF) break;

fDol you. Here's what is happening. First, inside the if, the
return value of fgetc( ) is assigned to ch. As you may recall, the
assignment operation in C is an expression. The entire value of
(ch = fgetc(fp)) is equal to the return vaiue offgetc().
Therefore, it is this integer value that is tested against EOF.
Expanding upon this approach, you will normally see this
program written by a professional C programmer as follows:
#include <stdio.h>
#include <stdlib.h>

int main(void)
{
char strl8D] = "This is a file system test.\n";
 File 1/0 267
~
9.2 MASTER FILE-SYSTEM BASICS

FILE "'fp;
char ch, "'p;

/'" open myfile for output "'/

H( (fp = fopen( "myfile", "w") )==NULL)
printf("Cannot open file. \n") ;
exit(l) ;

/'" write str to disk * /

p = str;
while (*p)
if(fputc("'p++, / fp)==EOF)
printf("Error writing file.\n");
exi t (l) ;

fclose (fp) ;

/ * open myfi!e for input */

if{(fp = fopen("myfile", "r"))==NULL) (
printf("Cannot open file.\n");
exit{l);

/'" read back the file */

while( (ch = fgetc(fp)) != EOF) putchar(ch);
fclose(fp) ;

return 0;
)

Notice that now, each character is read, assigned to ch, and

tested against EOF, all within the expression of the while loop
that controls the input process. If you compare this with the
original version, you can see how much more efficient this one
is. In fact, the ability to integrate such operations is one reason
C is so powerful. It is important that you get used to the kind of
approach just shown. Later on in this book we will explore such
assignment statements more fully.
2. The following program takes two command-line arguments. The
first is the name of a file, the second is a character. The program
searches the specified file, looking for the character. If the me
26B TIACH YOURSElF
'I' C

contains at least one of these characters, it reports this fact.

Notice how it uses argv to access the file name and the
character for which to search.
/* Search speGified file for specified character. */
~include <stdio.h>
#include <stdlib.h>

inL mainline argc, char ·argv[])

FILE: *fp;
char chi

/* see if correct number of command line arguments */

if(argc!=3) {
printf ("Usage: find <file name> <ch>\n");
exi t (1) ;
}

... open file for input */

if( (fp = fopen(arg.v(ll, Or") )==NULL) (
printf(MCannot open file.\nM);
exit(l) ;
)

/* look for character */

while( (eh = fgetc{fp» != EOF)
if (ch==*argv[2) {
printf("%c found-, ch};
break;
}
fclose (fp) ;

return 0;
)

1. Write a program that displays the contents of the text file

specified on the command line.
 Filei/O 269
9.3 -
UNDERSTAND feof( ) AND tenor() "

2. Write a program that reads a text file and counts how many
times each letter from 'A' to 'Z' occurs. Have it display the results.
(Do not differentiate between upper- and lowercase letters.)
3. Write a program that copies the contents of one text file to
another. Have the program accept three command-line arguments.
The first is the name of the source file, the second is the name
of the destination file, the third is optional. If present and ifit
equals 'watch," have the program display each character as it
copies the files; otherwise, do not have the program display any
screen output. If the destination file does not exist, create it.

NDERSTAND feof() AND fe rr or

As you know, when fgetc( ) returns EOF, either an error has occurred
or the end of the file has been reached , but how do "vou know which
event has taken place? Further if you are operating on a binary file, all
values are valid. This means it is possible that a byte will have the
same value (when elevated to an int) as EOF, so how do you know if
valid data has been returned or if the end of the file has been reached?
The solution to these problems are the functions feof( ) and ferror( ),
whose prototypes are shown here:

int feof(FILE 'fp);

int ferror(FILE 'fp);

The feof( ) function returns nonzero if the file associated with fp has
reached the end of the file. Otherwise it returns zero. This function
works for both binary files and text files. The ferror( ) function
returhs nonzero if the file associated with fp has experienced an error;
othefW'ise, it returns zero.
Using the feof( ) function, this code fragment shows how to read to
the end of a file:
FILE *fp;

while ( ! feaf (fp)) ch = fgetc (fp) ;

270 TEACH YOURSELF

'" C

This code works for any type of file and is better in general than
checking for EOF. However, it still does not provide any error
checking. Error checking is added here:
FILE *fp;

while ( ! feof (fp» (

ch = fgetc (fp) ;
if(ferror(fp» (
printf(~File Error \ n-)i
break;
)
)

Keep in mind that ferrore ) only reports the status of the file system
relative to the last file access. Therefore, to provide the fullest error
checking, you must call it after each file operation.
The most damaging file errors occur at the operating-system level.
Frequently, it is the operating system that intercepts these errors and
displays its own error messages. For example, if a bad sector is found
on the disk, most operating systems will, themselves, stop the
execution of the program and report the error. Often the only types of
errors that actually get passed back to your program are those caused
by mistakes on your part, such as accessing a file in a way inconsistent
with the mode used to open it or when you cause an out-of-range
condition. Usually these types of errors can be trapped by checking
the return type of the other file system functions rather than by
calling ferrore ). For this reason, you will frequently see examples of
C code in which there are relatively few (if any) calls to ferrore ). One
last point: Not all of the file system examples in this book will provide
full error checking, mostly in the interest of keeping the programs
short and easy to understand. However, if you are writing programs
for actual use, you should pay special attention to error checking.
 File I/O 271
9.3 UNDERSTANDfeof() ANDfenor()

EXAMPLES

1. This program copies any type of file, binary or text. It takes two
command-line arguments. The first is the name of the source
file, the second is the name of the destination file. If the
destination file does not exist, it is created. It includes full error
checking. (You might want to compare this version with the
copy program you wrote for text files in the preceding section.)
/* Copy a file. */
#include <stdio.h>
#include <stdlib.h>

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

(
FILE *from, ·to;
char Chi

/* see if correct number of command line arguments */

if (argc ! = 3) {
printf("Usage: copy <source> <destination>\n");
exi t (1) ;
)

/ * open source file */

if{(from = fopen{argv[l]. "rb")::::::NULL)
printfC-Cannot open source file.\n");
exi t (1) ;
)

J* open destination file */

if«to = fopen (argv [2]' "wb"')==NULLI (
printf("Cannot open destination file.\n"j;
exit(1);
)

/* copy the file * j

272 TEACH YOURSELF

• C

while ( ! feof (from» {

ch = fgetc(from);
i f I ferror I from)) (
printf(-Error reading source file.\nM);
exit (1);
)
i f I! feof I from)) fputc Ich, to);
if(ferror(to)} (
printf(-Error writing dest~nation file.\n-);
exit(l);
)
)

if(fclose(from)==EOF) (
printf{-Error closing source file.\n·);
exit(1);
)

iflfcloseltoi==EOFi (
printf(-Error closing destination file.\nM);
exit(lJ;
)

return .0;
)

2. This program compares the two files whose names are specified
on the command line. It either prints Files are the same, or it
displays the byte of the first mismatch. It also uses full error
checking.
'* Compare files. *j
#include <stdio.h>
#include <stdlib.h>

int main(int argc, char *argv[t)

(
FILE *fpl, *fp2;
char chI. ch2, same;
unsigned long 1;

1* see if correct number of command line arguments */

if(argc!=3) {
printf(-Usage: compare <file 1> <file 2>\n-);
exit (1) ;
 273
•

/* open first file *'

if((fpl = fopen(argv[lJ, "rb"}}=2NULL} (
printf("Cannot open first file.\n");
exit(l);
}

/* open second file *'

if((fp2 = fopen(argv [2]. "rb"»&=NULL) (
printf(-Cannot open second file.\n-);
exH (1) ;
}

1 = 0;
same = 1;
'* compare the fi1es'*,
while(!feof(fpl» (
chl = fgetc(fpl);
if (ferror ( fpl» (
printf '-Error reading first file.\nW);
exit(l); •

ch2 = fgetc(fp2);
if(ferror(fp2» (
printf(-Error reading second file.\n-);
exit(l) ;
}
if(chl!=ch2) .(
printf("Files differ at byte number 'lu", l);
same 0:
'II:

break;
}
1++;
}
if (same) printf("Files are the same.\n");

if(fclose(fpl)==EOF} (
printf(-Error closing first file.\ne):
exit(l):
}

if (fclose(fp2)==EOF) (
printf (-Error closiJ"".g second file. \n-) :
exit(l);
274 lJACH YOURSEl.F
. ,.,-
c

return 0;
}

EXERCIIES
1. Write a program that counts the number of bytes in a file (text
or binary) and displays the result. Have the user specify the file
to count on the command line.
2. Write a program that exchanges the contents of the two files
whose names are specified on the command line . That IS , given
two files called FILE! and FlLE2, after the program has run ,
F!LE! will contain the contents that originally were in FlLE2,
and FILE2 will contam FILE! 's original contents. (Hint: Use a
temporary file to aid in the exchange process.)

EARN SOME HIGHER-LEVEL TEXT

FUNCTIONS

../ (When working with text files, C provides four functIOns that make file
operations easier. The first two are called fputs( ) and fgets( ) , which
write a string to and read a string from a file, respectively . The ir
pmtotypes are

int fputs(char 'sir. FILE 'fp);

char 'fgets(char 'sir. int nurn, FILE 'fp);

The fputs( ) function writes the string pointed to by SIT to the file
associated with [p. It returns EOF if an error occurs and a non-negative
value if successful. The null that terminates str is not written. Also,
unlike its related function puts( ) it does not automatically append a
carriage return, linefeed pair.
 RIoIiO 27&
9.4 LEARN SOME HIGHER· LEVEL TEXT FUNCnONS
'"

The fgets( ) function reads characters from the file associated with
fp into the string pointed to by str until num-l characters have been
read , a newline character is encountered, or the end of the file is
reached. In any case, the string is null-terminated. Unlike its related
function gets( ), the newline character is retained. The function
returns str if successful and a null pointer if an error occurs.
The C file system contains two very powerful functions similar to
two you already know. They are fprintf( ) and fscanf( ). These
functions operate exactly like printf( ) and seanf( ) except that they
work with files. Their prototypes are:

int fprintf(FILE '(p, char 'control-string, ...J;

.
int fscanf(FILE '(p, char 'control-string, ...J;

Instead of directing their I/O operations to the console , these

functions operate on the file specified by fl'. Otherwise their
operations are the same as their console-based relatives. The
advantage to fprintf( ) and fscanfe ) is that they make it very easy to
",nte a_wide variety of data to a file using a text format)

EXAMPLES

1 This program demonstrates fputse ) and fgets( ). It reads lines

entered by the user and writes them to the file specified on the
command line. When the user enters a blank line, the input
phase terminates, and the file is closed. Next, the file is
reopened for input, and the program uses fgets( ) to display
the contents of the file.
#include <stdio.h>
tinclude <stdlib.h>
#include <string.h>

int main(int argc. char *argv(])

{
FILE *fPi
char str (80];

j . check for command line arg * /

 i f (argc!=2) {
printf("Specify file name.\n");
exit(l);
)

j* open file for output

if«fp = fopen(argv[l].
*'
"w"))==NlJLL) (
printf(-Cannot open file.\n-);
exit(l);
)

printf(-Enter a blank line to stop.\n-);

do {
printf(", ");
gets(str);
strcat(str, -\n-); '*
add newline */
iff*str != '\n') fputs{str. fpl;
} while(*str != '\n');
fclose (fp) ;

/* open file for input

i f ( (fp = fopen (argv [1].
*'"r")) ==NULL) (
printf(-Cannot open file.\n-):
exitll);
)

j* read back the file */

do ( .
fgets(str. 79. fp);
if(!feof(fp)) printf(str);
) while ( ! feof (fp)) ;
fclose (fp) ;

return 0;
)

2. This program demonstrates fprintf( ) and facanf( ). It first

writes a double. an into and a string to the file specified on the.
command line. Next. it reads them back and displays their
values as verification. If you examine the file created by this
program. you will see that it contains human-readable text. This
 277
9..
-
RleVQ

LEARN SOME HIGHER-LEVEL TfXT FUNCTIONS •

is because fprintf( ) writes to a disk file what printf( ) would

':"rite to the screen. No internal data formats are used.
'include <stdio.h>
#include <stdlib.h>
'include <string.h>

int main(int argc, char ·argv[))

{
FILE *fp;
double Id;
int d;
char str(aO};

/ * check for command line arg */

if (argc! =2) {
printf(-Specify file name. \ n ~ l i
exit (1) ;
)

'* open file for output * /

if ( (fp = fopen(argv[l). "w"»==NULL) {
printf{"Cannot open file. \ n" ) ;
exit(l) ;

fprintf(fp, "%f %d %s", 12345 . 342, 1908, "hello"

fclose(fp) ;

1 * open file for input * /

if((fp = fopen(argv[lj. "r"»==NULL) {
printf (·Cannot open file. \n");
exit(l);
)

fscanf(fp, "%If%d%s", &ld, &d, str);

printf("%f %d %s", ld, d, str);
fclose(fp) ;

return 0;
)
278 TEACH YOURSELF
'I' C

EXERCISES

1. In Chapter 6 you wrote a very simple telephone-directory

program. Write a program that expands on this concept by
allowing the directory to be saved to a disk fIle. Have the
program present a menu that looks like this:

1. Enter the names and numbers

2. Find numbers
3. Save directory to disk
4. Load directory from disk
5. Quit
The program should be capable of storing 100 names and
numbers. (Use only first names if you like.) Use fprintf( ) to
save the directory to disk and fscanf( ) to read it back into
memorv.
2. Write a program that uses fgets( ) to display the contents of a
text file , one screenful at a time. After each screen is displayed,
have the program prompt the user for more.
3. Write a program that copies a text file . Specify both the source
and destination file names on the command line. Use fgets( )
and fputs( ) to copy the file. Include full error checking.

EARN TO READ AND WRITE BINARY

DATA
AS useful and convenient as fprintf( ) and fscanf( ) are, they are not
necessarily the most efficient way to read and write numeric dala. The
reason for this is that both functions perform conversions on the data.
For examp.le, when you output a number using fprintf( )'the number
is converted from its binary format into ASCII text. Conversely, when
you read a number using fseanf( ) , it must be converted back into its
binary representation. For many applications, this conversion time
will not be meaningful; for others, it will be a severe limitation. further,
for some types of data, a file created by fprintf( ) will also be larger
than one that contains a mirror image of the data using its binary
 AI.1I0 279
."
9.5 LEARN TO READ AND WRI1f BINARY DATA

fonnat. For these reasons, the C me system includes two important

functions: fread( ) and fwrite( ). These functions can read and write
any type of data, using its binary representation. Their prototypes are

size_t fread(void 'buffer, size_t size, size_t num, FILE '(p);

size_t tw'rite(void 'buffer, size_t size, size_t num. FILE '(p);

As you can see, these prototypes introduce some unfamiliar elements.

However, before discussing them, a brief description of each function
IS necessary.
The fread( ) function reads from the file associated with [p, num
number of objects, each object sIZe bytes long, into the buffer pointed
to by buffer. It returns the number of objects actually read. If this value
is less than nUtn, either the end of the file has been encountered or an
error has occurr-ed. You can use feof( ) or ferror( ) to find out which.
The fwrite( ) function is the opposite of fread( ). It writes to the
me associated with [p, num number of objects, each object size bytes
long, from the buffer pointed to by buffer. It returns the number of
objects written . This value will be less than ntlm only if an output
error has occurred.
Before looking at any examples, let's examine the new concepts
introduced by the functions' prototypes.
The first concept is that of the void pointer. A void pointer is a
pointer that can point to any type of data without the use of a type
cast. This is generally referred to as a generic poinrer. In C, void
pointers are u~d for two primary purposes. First, as illustrated by
fread( ) and fwrite( ), they are a way for a function to receive a
pointer to any type of data without causing a type mismatch error. As
stated earlier, fread( ) and fwrite( ) can be used to read or write any
type of data. Therefore, the functions must be capable of receiving any
sort of data pointed to by buffer. void pointers make this possible. A
second purpose they serve is to allow a function to return a generic
pointer. You will see an example of this later in this book.
The second new item is the type size_to This type is defined in the
STDlO.H header file. (You will learn how to define types later in this
book.) A variable of this type is defined by the ANSI C standard as
being able to hold a value equal to the size of the largest object
supported by the compiler. For our purposes, you can think of size_t
as being the same as unsigned or unsigned long. The reason that
size_t is used instead of its equivalent built-in type is to allow C
2aO TEACH YOURSELF
... C

compilers running in different environments to accommodate the

needs and confines of those environments.
When using fread{ ) or fwrite{ ) to input or output binary data,
the file must be opened for binary operations. Forgetting this can
cause hard-to-find problems.
To understand the operation of fread{ ) and fwrite{ ), let's begin
with a simple example. The following program writes an integer to a
file called MYFILE using its internal, binary representation and then
reads it back. (The program assumes that integers are 2 bytes long.)
#include <stdio.h>
#include <stdlib.h>

int main (void)

(
FILE *fp;
int i;

1* open file for output -;

H«fp = fopen("myfile", "wb"))==NULL) (
printf(ftCannot open file.\nR):
exit(!) ;
)

i = 100;
i f ( fwri te ( .. i.
2. 1. fp) ! = 1) (
printf(·~Tite - error occurred.\n M ) ;
exit(1) ;
)
fclose(fp) ;

/* open file for input */

H«fp = fopen("myfile", "rb"))==NULL) (
printf (·Cannot open file. \nR);
exit(l);
)

H(fread("i. 2. 1, fp) != 1) (
printf (RRead error occurred. \nR);
exit(l);
)
printfC-i is %d-, i);
fclose (fp) ;
 RIoVO 281
T
9.5 LEARN ro READ AND WRITE BlNAJlY £lATA

return 0;
}

Notice how error checking is easily performed in this program by

simply comparing the number of items written or read with that
requested. In some situations, however, you will still need to use
feof( ) or ferror( ) to determine if the end of the file has been
reached or if an error has occurred.
One thing wrong with the preceding example is that an assumption
about the size of an integer has been made and this size is hardcoded
into the program. Therefore', the program will not work properly with
compilers that use 4-byte integers, for example. More generally, the
size of many types of data changes
, between systems or is difficult to
determine manually. For this reason, C includes the keyword sizeof,
which is a compile-time operator that returns the size , 'in bytes, of a
data type or variable. It takes the general forms '.

sizeol(type)

or

sizeol var_name;
For example, if floats are four bytes long and f is a float variable, both
of the following expressions evaluate to 4:
sizeof f
sizeof(float)

When using sizeof with a type, the type must be enclosed between
parentheses. No parentheses are needed when using a variable name,
although the use of parentheses in this context is not an error.
By using sizeof, not only do you save yourself the drudgery of
computing the size of some object by hand, but you also ensure the
portability of your code to new environments. An improved version of
the preceding program is shown here, using sizeof.
#include <s tdio.h>
#include <stdlib.h>

int main(void)
{
FILE *fp;
282 TtACH YOIlRSW
~ c

int i;

/* open file for ~utput -,

if ( (fp - fopen ("myfile", "wb")) ==NULL) (
printf(-Cannot open file.\n-);
exit(l);
)

1. ::: 1 00 ;
--
if(fwrite(&i. sizeof(intl, 1, fp) != 1) {
print! (MWrite error occurred. \n-);
exit(1) ;
)
fclose (fp) ;

'* open flie for input */

if«fp = fopen("myfile", "rb"))==NULL) (
printf (·Car..not opel !ile. \ n-);
exit(1);
)

if { fread (&i, sizeo! i, 1. fp ) != 1) {

printf ( ftRead error occurred. \ n-);
exit ( l);
)
printf{-i is %d-,i);
fclose (fp) ;

return 0;
-' )

EXAMPLES

I. This program fills a ten-element array with floating-point

numbers, writes them to a file, and then reads them back. This
program writes each element of the array separately. Because
binary data is being written using its internal format , the file
must be opened for binary I/O operations.
#include 's::stdio.h>
.include <stdlib.h>
 File 1/0 283
9.5 LEARN TO RE.MJ·AND WRITE BINARY DATA 't'

double d(10] • (
10.23. 19.87. 1002.23. 12.9. 0.897.
11.45. 75.34. 0.0. 1 . 01. 87 5 .875
) ;

int main(void)
(
int ii
FILE *fPi

if «fp • fopen ("rnyfile", "wb" II ==NULLI (

printf("Cannot open file.\n");
exi t (1) :
)

for(i=O; i<10; i++)

if{fwrite(&d(ij, sizeof(doublel. 1, fp) t= 1)
princf{-Write error.\nn);
exit(l) ;
)
fclose{fp} ;

if«fp = fopen("rnyfile", "rb"II==NULLI (.

printf ("Cannot open file. \n");
exit(l);
)

/* clear the" array */

for(i=O; i<10; i++) d[iJ = -1.0;

for(i=O; i<10; i++)

if(fread(&d[i), sizeof(doub1e) , I, fp) != 1) {
printf(ftRead error.\n~);
exit(l) ;
)
fclose (fp) ;

/* display the array */

for(i=O; i<10; i++) printf("%f" d(i]):

return 0;
)
2 ••
y

,
The array is cleared between the write and read operations only
to 'prove' that it is being filled by the &eIId( ) statement.
2. The following progIam does the same thing as the first, but here
only one call to fwrite( ) and fread( ) is used because the
entire array is written in one step. which is much more
efficient. This example helps illustrate how powerful these
functions are.
'include <stdio.b>
'include <stdlib.b>

double d[lOJ = {
10.23. 19.87. 1002.23. 12.9. 0.897.
11.45, 75.34. 0 . 0. 1.01. 875.875
);

int mainevoid)
{
int i;
FILE 'fp;'

it«fp = fopen("myfile", "wb"I)==N\JLL ) {

printf("Cannot open file.\n");
exit(l);
)

/* write the entire array in one step

it (fwrite(d, s~zeof d, I, fp) != 1) {
*'
printf ( -Wri te error. \n· ) ;
exit(l);
)
fclose(fp);

if«fp = fopen("myfile", "rb"))==N\JLL) {

printf("Cannot open file,\n");
exit(l);
)

/* clear the array *'

for(i=O; i<lO; i++) d[iJ = -1.0;

, '* read the entire array in one step

if(fread(d, sizeof d, 1, fp) != 11 {
*'
printf (-Read error. \n-):
 Rlel/O 2U
9.6 -
UNDERSTAND RANDOM ACCESS "

exit(l) ;
)
~close(fp) ;

/* display the array *1

for(i=O; i<lO; i++) printf("%f" d[i]);

return 0;
.)

EXERCISES

1. Write a program that allows a user to input as many double

values as desired (up to 32.767) and writes them to a disk file as
they are entered. Call this file VALUES. Keep a count of the
number of values entered. and write this number to a file called
COUNT.
2. Using the file you created in Exercise 1. write a program that
first reads the number of items in VALUES from COUNT. Next.
read the values in VALUES and display them.

DERSTAND RANDOM ACCESS

So far. the examples have either written or read a file sequentially

from its beginning to its end. However. using another of C's file
system functions, you can access any point in a file at any time. The
function that lets you do this is called fseek( ), and its prototype is

in! fseek(FILE "fp, long offset in! origin);

Here, fp is associated with the file being accessed. The value of offset
determines the number of bytes from origin to make the new current
188 lEACH YOURSELF
.., C

position. ongin must be one of these macros, shown here with their
meanings:

Origin Meaning
SEEKJ)ET Seek from start of file
SEEK_CUR Seek from current location
SEEK_END Seek from end of file

These macro~. are defined in STDlO.H. For example, if you wanted to

set the current location 100 bytes from the start of the file, thlln ·origin
will be SEEK_SET and offset will be 100.
The fseek( ) function returns zero when successful and nonzero if
c·
a failure occurs. In most implementations, you may seek past the end
of the file:but you may never seek to a point before the start of the file .
You can determine the current location of a file using fwU( ),
another of C's file system functions. Its prototype is

long ftell(FILE *fp):

It returns the location of the current position of the file associated with
[p. If a failure occurs, it returns -1.
In general, you will want to use random access only on binary files.
The reason for this is simple. Because text files may have character
translations performed on them, there may not be a direct
correspondence between what is in the file and the byte to which it
would appear that you want to seek. The only time you should use
fseek( ) with a text file is when seeking to a position previously
determined by fiell( ), USing SEEK_SET as the origin.
Remember one important point: Even a file that contains only texl
- can be opened as a binary file, if you like. ·There ~ no inherent
restriction about random access on files containing text. The
restriction applies only to files opened as text fillls.

EXAMPLES

I. The following program uses fseek( ) to report the value of any

byte within the file specified on the command line.
 287
-
FIoI/O
9.' UNDERSTMD RANDOM ACCESS

#include <stdio.h>
#include <stdlih.h>

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

(
long lOCi
FILE -fPi

/ * see if file name is specified */

H(argc!=2) (
print~{·File name missing.\n-);
exit(!) ;
)

if( ( fp = fopen(argv(lJ , "rb"»==N1JLL) (

printf(-Cannot open file . \n~);
exitCl);
)

printf(~Enter byce to seek to: "I;

scanf("%ld~, &loc);
if ( f seek ( fp, 10c, SEEK_SET» (
printf("Seek error. \ n~);
exi t (1) ;
)

printf("Va1ue at 10c 'ld is 'do, 10c, getc(fp);

fclose (fp) ;

returr. 0;
)

2. The following program uses ftell( ) and fseek( ) to copy the

contents of one file into another in reverse order. Pay special
attention to how the end of the input file is found. Since the
program has sought to the end of the file, the program backs up
one byte so that the current location of the file associated with
in is at the last actual character in the file.
/* Copy a file in reverse order */
.include <stdio.h>
#include <stdlib.h>

iot main(int argc. char *argv[)

 288 1tACH YOURSW

'" C

long loci
FILE • in, ·out;
char chi
•
/ * see if correct number of command line arguments */
i f large! =3 l I
printf(-Usage: revcopy <source> <destination>.\n-);
exit(1);
}

ifllin = fopenlargv[l], "rb")}==NULL) (

printf(·Cannot open input file.\n·);
exit (1);
}
ifllout = fopen(argv[2], "",b"))==NULL) (
printf(-Cannot open output file.\n-);
exit(1) ;
}

I T find end of source file */

fseeklin, OL, SEEK_END);
lac = ftell(in);

1* copy file in reverse order *'

lac = lac-I; /* back up past end-af-file mark */
while Iloc >= OLl (
fseeklin, loc, SEEK_SET);
ch = fgetc(in);
fputc{ch. out);
loc--;
}
fclose(in) ;
fclose (out) ;

return 0;
}

3. This program writes ten double values to disk. It then asks you
which one you want to see. This example shows how you can
randomly access data of any type. You simply need to multiply
the size of the base data type by its index in the file.
iinclude <stdio.h>
iinclude <stdlib.h>
 FIoIJO 2118
-
9.6 UNDERSTAND IWIDOM ACCESS '"

double d[101 = {
10.23, 19.87, 1002.23, 12.9, 0.897,
11.45, 75.34, 0.0, 1.01, 875.875
} ,
int main(void}
{
long Icc;
double value;
FILE "'fp;

if«fp = fopen("myfile", "wb"))==NULL) {

printf ("Cannot open file. \n·) ;
exit(l) ,
}

/ * write the entire array in one step

H(fwrite(d, sizeof d, 1, fp} != 1) {
*'
printf("Write error.\n"l i
exit (1);

fclose (fp) ,

if«fp = fopen("myfile", "rb"»==NULL) {

printf("Cannot open file.\n");
exit(l) ,
}

printf("Which element? .);

scanf("'ld", &loc),
if (fseek(fp, loc*sizeof(double) , SEEK-SET)) {
printf(·Seek error.\n");
exit(l) ,
}

fread(&value, sizeof(double) , 1, fp),

~rintf("Element 'ld is 'f", loc, value),

fclose(fp) ,

return 0;
}
290 T£ACH YOURSELF

'" C

EXERCIIU
1. Write a program that uses fseek( ) to display every other byte
in a text file. (Remember, you must open the text file as a
binary file in order for fseek( ) to work properly.) Have the
user specify the file on the command line.
2. Write a program that searches a file, specified on the command
line, for a specific integer value (also speCified on the command
line). If this value is found, have the program display its
location, in bytes, relative to the start of the file.

EARN ABOUT VARIOUS FILE-SYSTEM

FUNCTIONS

YOll can rename a file using rename( ), shown here:

int rename(char 'oldname, char 'newname);

Here, oldname points to the original name of the file and newname
points to its new name. The function returns zero if successful and
nonzero if an error occurs.
You can erase a file using remove( ). Its prototype is

int remove(char 'file-name);

This function will erase the file whose name matches that pointed to
by file-name. It returns zero if successful and nonzero if an error occurs.
You can position a file's current location to the start of the file using
rewind( ). Its prototype is

void rewind (FILE '(p);

It rewinds the file associated with [p. The rewind( ) function has no
return value, because any file that has been successfully opened can
be rewound.
 RleIlO 281
.., .
9.7 LEARN ABOUT VARIOUS FlU-SYSTEM FUNCnONS

Although seldom necessary because of the way Cs file system

works, you can cause a file's disk buffer to be flushed using fflush( ).
Its prototype is

int fflush(FILE "fp);

It flushes the buffer of the file associated with {p. The function returns
zero if successful, EOF if a failure occurs. If you call fflush( ) using a
NULL for {p, all existing disk buffers are flushed.

EXAMPLES

,. This program demonstrates remove( ). It prompts the user for

the file to erase and also provides a safety check in case the user
entered the wfong name.
#include <stdio.h>
~inc:ude <stdlib.h>
_include <ctype.h>

int main(void)
{
char fname[BO};

printf("Enter name of file to erase: ");

gets (fname) ;
printf("Are you sure? (YIN) ");
if(toupper(getchar(»=='Y') remove (fname) ;

return 0;
)

2. The following program demonstrates rewind( ) by displaying

the contents of the file specified on the command line twice.
#include <stdio.h>
#include cstdlih.h>

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

292 TEACH VOURSW •

• C

(
FILE *fp;

1* see if file name is specified */"

if(argc!=2) {
printf (-File name missing. \ n");
exitll) ;
)

if II fp = fopen largv [1J, "r"» ==NULL) (

printfC"Cannot open file. \ n-);
exit(1) ;
)

/* show it once * 1
whil,.l! feof I fp) )
putcharlgetclfp»;

rewind ( £p) ;

1 * show it twice • I

while I! feof I fp) )

putchar(getc(fp» ;

fcloselfp) ;

return 0;
)

3. This fragment causes the buffer associated with fp to be flushed

to disk.
FILE *fp;

fflush(fp) ;

4. This program renames a file called MYFILE.TXT to

YOURFILE.TXT.
#include <stdio.h>

int main(void)
(
if (rename ( "myfi Ie. txt", ·yourf ile. txt") )
- - ---
file 1/0 293
9.8 LEARN ABOUT THE STANDARD STREAMS
'"
printf ("Rename failed. \n") ;
else
printf (" Rename successful. \n ~ ) ;

return 0;
}

EXERCISES

I. Improve the erase program so that it notifies the user ifhe or

she tries to renlove a nonexistent file.
2. On your own, think of wavs that rewind( ) and fflush( ) could
be useful in real applications.

EARN ABOUT THE STANDARD

STREAMS

When a C program begins exccmion, three streams are automatically

opened and available for use. These streams are called standard inpw
(stdin), standard output (stdout), and stand{l)'d error (stderr). By
default, they refer to the console, but in environments that SUppOTt
redirectable ]/0, they can be redirected bJ' the operating system to
some other device.
Normally, stdin inputs frol11 the keyboard; stdout and stderr write
to the screen. These standard streams are FILE pointers and may be
used with any function that requires a variable of type FILE '. For
example, you can use fprintf( ) to print formatted output to the
screen by specifying stdou~ as its output stream. The following two
statements are functionally the same:
fprintf(stdout, "%d %c %5", 100. 'c', "this is a string");
printf("%d %c %5-, 100, te'. ~this is a string");

In actuality, C makes little distinction between console ]/0 and file

1/ 0. Asjust shown, it is possible to perform console 1/ 0 using several
 .,
294 TEACH YOURSELF

of the file-system functions. Although it may come as a bit of a

surprise, it is also possible to perform disk file JlO using console JlO
functions, such as printf( ) . Here's why.
All of the functions described in Chapter 8 and referred to as
"console JlO functions" are actually special-case file-system functions
that automatically operate on stdin and stdout. Thus, the console JlO
functions are just conveniences for you, the programmer. As far as C
is concerned, the console is simply another hardware device. You
don't actually need the console functions to access the console. Any
me-system function can access it. (Of course, non-standard JlO
functions like getche( ) are differentiated from the standard
file-system functions and do, in fact, operate only on the console.) In
environments that allow redirection of I/O, stdin and stdout could
refer to devices other than the keyboard and screen. Since the console
functions operate on stdin and stdout. if these streams are redirected,
the "console" functions can be made to operate on other devices. For
example, by redirecting the stdout to a disk file. you can use a
"console" I ' 0 function to write to a disk file.
One important point: stdin, stdout , and stderr are not variables.
They may not be assigned a value using fopen( ) , nor should you
attempt to close them using fclose( ) . These streams are maintained
internally by the compiler. You are free to use them, but not to
cnange them.

EXAMPLES

1. Consider this program:

#include <stdio.h>

int main(voidl

printf("This is an example of redirection.\n M

):

return 0;
)

Assume that this program is called TEST. If you execute TEST

normally, it displays the string on the screen. However, if an
 AiellO 295
."
9.8 LEARN ABOlfT THE STANDARD STREAMS

environment supports redirection of [/0, stdout can be

redirected to a file. For example, in a DOS, OS/2, Windows, or
UNIX environment, executing TEST like this
TEST > OUTPUT

causes the output of TEST to be written to a file called OUTPUT.

You might want to try this now for yourself.
2. Input can also be redirected. For example, consider the
following program:
#include <stdio.h>

int main (void)

(
int i;

scanf ("%d", &i);

printf ("%d", i);

return 0;
}

Assuming it is called TEST, executing it as

TEST < INPUT

causes stdin to be directed to the file called INPUT. Assuming

that INPUT contained the ASCII representation for an integer,
the value of this integer will be read from the file and printed on
the screen.
3. As mentioned earlier in this book, when using gets( ) it is
possible to overrun the array that is being used to receive the
characters entered by the user because gets( ) provides no
bounds checking. One way around this problem is to use
fgets( ), specifying stdin for the input stream. Since fgets( )
requires you to specify a maximum length, it is possible to
pre'Vent an array overrun. The only trouble is that fgets( ) does
not remove the newline character and gets( ) does. This means
that you will have to manually remove it, as shown in the
following program:
#include <stdio.h>
#include <string.h>
296 TEACH YOURSELF

" C

int main(void)
(
char str[lO);
int i;

printf (-Enter a string: .. );

fgets(str, 10, stdinl;

/~ remove newline, if present */

i = strlen(str)-l;
if(str[i]=='ln'l str[i] = '10';

printf ("This is your string: %s", str);

return 0;
}

EXERCISES

1. Write a program that copies the contents of one text file to

another. However, use only "console" 110 functions and
redirection to accomplish the file copy.
2. On your .own, experiment using fgets( ) to read strings entered
from the keyboard.

Skills Check

Before continuing, you should be able to answer these questions

and complete these exercises:

1. Write a program that displays the contents of a text file (specified

on the command line), one line at a time. After each line is
displayed, ask the user ifhe or she wants to sec another line.
 RleIlO 2S7
9.8 -
LEARN ABOUT THE STIWDAJ/D SIRfAMS '"

2. Write a program that copies a text file. Have the user specify
both file na~es on the command line. Have the copy program
convert all lowercase letters into uppercase ones.
3. What do fprintf( ) and fscanf( ) do?
4. Write a program that uses fwrite( ) to write 100 randomly
generated integers to a file called RAND.
5. Write a program that uses fread( ) to display the integers stored
in the file called RAND, created in Exercise 4.
6. Using the file called RAND, write a program that uses fseek( )
to allow the user to access and display the value· of any integer
in the file.
7. How do the "console' I/O functions relate to the file system?

This section checks how well you have integrated the material in
this chapter with that from earlier chapters.

1. Enhance the card-catalog program you wrote in Chapter 8 so

that it stores its information in a disk file called CATALOG.
When the program begins, have it read the catalog into memory.
Also, add an option to save the information to disk.
2. Write a program that copies a file. Have the user specify both
the source and destination files on the command line. Have the
program remove tab characters, substituting the appropriate
number of spaces.
3. On your own, create a small database to keep track of anything
you desire-your CD collectiorl;" for el<ample:
 10
Structures and Unions

ch....... ableed...

10.1 Master structure basics

10.2 Declare pointers to structures

10.3 Work with nested structures

10.4 Understand bit-fields

10.5 Create unions

299
"
300 TtACIf YOUIISfI.f

'" C

N
. ~

this chapter you wil11earn about two of C's most important

user·defined types: the structure and the union.

Review
Skills Check

Before proceeding you should be able to answer these questions and

perform these exercises:

1. Write a program that copies a file. Have the user specifY both
the source and destination file names on the command line.
Include full error checking.
2. Write a program using fprintf( ) to create a file that contains
this information:
this is a string 1230.23 IFFF A

Use a string, a double, a hexadecimal integer, and character

format specifiers and values.
3. Write a program that contains a 20·element integer array.
Initialize the array so that it contains the numbers 1 through 20.
Using 0Tlly one fwrite( ) statement, save this array to a file
called TEMP.
4. Write a program that reads the TEMP file created in Exercise 3
into an integer array using only one fread( ) statement. Display
the contents of the array.
5. What are stdin, stdout, and stderr?
6. How do functions like printf( ) and scanf( ) relate to the C
file system?

10.1 MASTER STRUCTURE BASIC.S

(A structure is an aggregate (or conglomerate) data type that is composed
of two or more related variables called members. Unlike an array in
 S1RUC1URiS AND UNIOIIS 301
,.., MASTfR SrRUCTURE BASiCS •

which each element is of the same type, each member of a structure

can have its own type, which may differ from the types of the other
members. Structures are defined in C using this general form:

struet tag-name {
type member I ;
type member2;
type f!lember3;

I¥pe memberN;
} variable-list;

The keyword struct tells the compiler that a structure type is being
defined. Each type is a valid C type . The tag-name is essentially the
type name of the structure, and the variable-list is where actual
instances of the structure are declared. Either the tag-name or the
variable-list is optional, but one must be present (you will see why
shortly) . The members of a structure are also commonly referred to as
fields or elements) This book will use these terms interchangeably.
( Generally, the information contained in a structure is logically
related. For example, you might use a structure to hold a person's
address. Another structure might be used to support an inventory
program in which each item's name, retail and wholesale cost, and the
quantity on hand are stored. The structure shown here defines fields
that can hold card-catalog information:
struct catalog (
c har name [ 40] ; / ' author name ' /
char title[40] ; / ' title ' /
char pub[40] ; / ' publisher ' /
unsigned datei / ' copyright date ' /
unsigned char edi / ' edition ' /
) card;

Here, catalog is the type name of the structure. It is not the name
of a variable. The only variable defined by this fragment is card. It is
important to understand that a structure declaration defines only a
logical entity, which is a new data type . It is not until variables of that
type are declared than an object of that type actually exists. Thus,
catalog is a logical template; card has physical realitYJigure 10-1
 fiGURE 10-1
~_n~rum~e_____~~bc.Y=~=-___________?~~~______~
How the etmI
structure variable ~~ti~·tl7e~____~40~b~Yt~~__________--(~~
appears in ~~p_u_b____~4~O~b~Y_t~ ______- ,__-J\ \~________~
memory date 2 bytes I
(assuming 2-byre ed Ibte I
integers)
T

shows how this structure will appear in memory (using 2-byte

integers).
(To access a member of a structure, you must specifY both the
structure variable name and the member name, separated by a period.
For example, using card, the following statement assigns the date
field the value'I776:
card.date = 1776;

C programmers often refer to the period as the dot operator. To print

the copyright date, you can use a statement such as:
printf(~Copyright date: %U", card.date);

To input the date , use a scanfe ) statement such as:

scanf{~%u~ . &Card.datel :)

Notice that the & goes before the structure name, not before the
member name.0n a similar fashion, these statements input the
author'S name and output the title:
gets{card.name);
printf(~%s·, card.~itle);

To access an individual character in the title field, simply index

title. For example, the following statement prints the third letter:
printf(~%c·, card.title[2]);)

(once you have defined a structure type, you can create additional
variables of that type using this general form:
 STRUCTUAES 1IMo'- 303
~
10. I WoSTfR STRUCTURE BASICS

Assuming, for example, that catalog has been defined as shown

earlier in this section, this statement declares three variables of type
struct catalog:
struct catalog varl, var2, var3i

This is why it is not necessary to declare any variables when the

structure type is defined. You can declare them separately, as needed.)
A key concept to understand is that each instance of a structure
contains its own copy of the members of the structure. For example,
given the preceding declaration, the title field of varl is completely
separate gom the title field of var2.0n fa~t, the only relationship that
varl, var2, and var3 have with one another is that they are all
variables of the same type of structure. There is no other linkage
among the three.
If you know you only need a fixed number of structure variables,
you do not need to specify the tag name. For example, this code
creates two structure variables, but the structure itself is unnamed:
struct {
int aj
char Chi
} varl, var2 i

In actual practice, however, you will usually want to specify the tag name)
(Structures can be arrayed in the same fashion as other data types.
For example, the following structure definition creates a ! DO-element
array of structures of type catalog:
struct catalog cat[lOO];

To access an individual structure of the array, you must index the

array name. For example, the· following accesses the first structure:
cat [O!

To access a member within a specified structure, follow the index

with a period and the name of the member you want. For example,
the following$tatement loads the ed field of structure 33 with the
value of2:
cat[33! .ed = 2;)
 ( Structures may be passed as parameters to functions just like any
other type of value. A function may also return a structure. )
(You may assign the contents of one instance of a structure to
another as long as they are both of the same type. For example, this
fragment is perfectly valid:
struct s_type {
int a;
float f:
} varl, var2:

varl.a = 10:
varl.f = 100.23;

var2 = varl;

After this fragment executes, var2 will contain exactly the same thing
as varl)

EXAMPLES

I. This program demonstrates some ways to access structure

members:
'include <stdio.h>

struct s_type {
int i;
char chi
double d;
char str[BO}:
} 5;

int main(void)
(
printf(-Enter an integer: .);
scanf ( • %d: ., &s . i) ;
printf(-Enter a character: .);
scanf(- %c~. &s.ch):
printf(·Enter a flo~ting point numbe=: .);
scanf("'lf", &s.d);
printf("Enter a string: .);
 STRUCTURES AND UNIONS 305
10.1 MASTER STRUCTURE BASICS .,

printf(-'d %c %f %s", s.i, s.ch. s.d, s.str);

return 0;
)

2. When you need to know the size of a structure, you should use
the sizeof compile-time operator. Do not try to manually add'up
the number of bytes in each field. There are three good reasons
for this. First, as you learned in the preceding chapter, using
sizeof ensures that your code is portable to different
environments. Second, in some situations, the compiler may
need to align certain types of data on even word boundaries. In
this case, the size of the structure will be larger than the sum of
its individual elements. Finally, for computers based on the
8086 family of CPU. (such as the 80486 or the Pentium), there
are several different ways the compiler can organize memory.
Some of these ways cause pointers to take up twice the space
they do when memory is arranged differently.
When using sizeof with a structure type, you must precede
the tag name with the keyword 8truct, as shown in this program:
tinclude <stdio.h>

struct s_type {
int i;
char ch:
int *p;
double d;
} S;

int rnain{void)
(

return 0;
)

3. To see how useful arrays of structures are, examine an

improved version of the card-catalog program developed in ,he
preceding two chapters. Notice how using a structure makes it
easier to organize the information about each book. Also notice
306 TEACH YOUASl1F

" C

how the entire structure array is written and read from disk in a
single operation.
/* An electronic card catalog. */
.include <stdio.h>
#include <string.h>
#include <stdlib.h>

#define MAX 100

int menu (void) i

void display(int i):
void author_search (void) ;
void title_search (void) ;
v01d enter(void);
void save(void);
void 1oadCvoidJ ,

struct catalog (
char name (80 J ; ,.,. .,
author name .,
char title[801, title
," publisher
char pub[80] ,
unsigned date; ," copyright "' "'
date
," edition
]
unsigned char ed,
cat [1dAX] , "'
int top = 0; /* last location used */

int main (void)

(
int choice;

load() ; 1* read in catalog * /

do (
choice =. menu();
switch (choice) (
case 1: enter();
break,
'* enter books */

CAse 2: author_search();
break;
'* search by author */

case 3: title_search(); /* search by ti~e *;

break;
case 4: save();
J
 I1IIUC1URES lIND IINIONS 307
10.1 MASTER STRUCTURE 8/ISICS y

} while(choicel=S);

return 0;
)

/* Return a menu selection. */

menu (void)
(
int i;
char str[80];

printf(NCard catalog:\n");
printE(" 1. Enter\n~);
printE(" 2. Search by Author\n");
printE(" 3. Search by Title\n");
printE(" 4. Save catalog\n");
printE(" 5. Quit\n");

do (
printf{~Choose your selection: ");
gets (str);
i = atai (str);
printE("\n") ;
) whUeli<l II i>5);

return i;
)

/* Enter books into database. */

void enter(void)
(
int i;
char temp[a01;

for{i=top; i<MAX; i++) {

printf("Enter author name (ENTER to quit): It);
gets(cat(i] .name);
if(!*cat[ij . name) break;
printf("Enter title: ");
gets{cat[i).tit!e):
printfC"Enter publisher: ");
getslcat(iJ .pub);
printi ("Enter copyright date: ");
gets I temp) ;
cat[iJ . date = (unsigned) atOi(temp);
308 WlCH YOURSELF
'If C

printf("Enter edition: ·)i

gets (temp) ;
catlij.ed = (unsigned char) atoi{temp);
)
top = i;
)

/* Search by author. */
void author_search(voidl
{
char name[BO);
int i, found;

printf ("Name: ~);

gets (narne) ;

found = D.
for(i=O; i<top; i++)
if(!strcmp.{narne. cat[ij.name») {
display (i) ;
found =
1;
printf("\n") ;

if (! found) printf ("Not Found\n") i

;* search by title. */
void title_search(voidl
{
char title[8D];
int i, found;

printf("Title: 10);
gets (title);

found = 0;
for(i=O; i<top; i++)
if{!strcrnp(title. cat{i}.title) {
display (i) ;
found = 1;
printf ( .. \n");
)
if{!found} printf("Not Found\n");
)
 STRIJCllJRfS AND_
'0. , MASTER STRUCTURE BASICS

/* Display catalog entry. */

void display(int i)
(
printf{"%s\n", cat(i].title);
priPtf("by %s\n", cat[i].name);
printf("published by %s\n", cat[i] .pub);
prlntf("Copyright: %u, %u edition\n", cat(i] .date,
cat(ij.ed) ;
)

/* Load the catalog file. */

void load(void)
(
FILE *fp;

if( (fp = fopen("catalog", "rb") )==NULL) (

printf{"Catalog file not on disk.\n");
return;
)

if(fread(&top, sizeof top, 1, fp) != 1) { /* read count */

printf("Error reading count.\nn);
exit(1) ;
)

if(fread(cat, sizeof cat, 1. fp) != 1) { /* read data */

printf("Error reading catalog data.\n");
exit(l) ;

fclose (fp) ;
}

/* Save the catalog file. */

void save(void)
(
FILE *fp;

if((fp = fopen("catalog", "wb"))==NULL) (

printf("Cannot open catalog file.\n");
exit (1);
)
310 TEACH YOUASElf
... C

if (fwrite(&top. sizeof top. 1. fp) != l} ( /. write CO\mt */

printf(-Error writing count.\n-);
exit 11);
)
if(fwrite(cat, sizeof cat, 1, fp) != 1) { j* write data */
printf(~Error writing catalog data.\n-);
exit(1) ;
)

fclose (fp) ;
)

4. In the preceding example, the entire catalog array is stored on

disk, even if the array is not full. If you like, you can change the
load( ) and save( ) routines as follows, SO that only structures
actually holding data are stored on disk:
.
/* Load the catalog file.
v .o id load (void)
*'
I
FILE ·fpi
int i;

i f l l f p = fopenl"catalog", Orb") )==NULL) {

printf (·Catalog file not on disk. \n·) ;
return;
)

if(fr~ad(&top. sizeof top, 1, fp) != 1) { /* read count */

printf(-Error reading count.\n M ) ;
exit(1) ;
)
for(i=O; i<=top; i++) /* read data */
if(fread(&cat[i}. sizeof(struct catalog), 1, fp)!= 1) (
printf(-Error reading catalog data.\n-);
exit(!) ;
)

fclose (fp) ;
)
 10.1 MASTER STRUCTURE BASJCS

1* Save the catalog file. */

void save (void)
(
FILE *fp;
int i:

if«fp = fopen("catalog", "wb"»==NULL) (

printf(MCannot open catalog file.\n-) i
exit(l) ;
}

if (fwrite(&top, sizeof top, 1, fp) != 1) ( 1* write count *j

printf(ftError writing count.\nM);
exit(1);
}
for(i=O; i<=tcp: i++) 1* write data *j
if(fwrite(&cat{il. sizeof{struct catalog), 1. fp)!= 1) {
prlntf(MError writing catalog data.\n-);
exit(l);
}

fclose ( fp) ;
}

5. The names of structure members will not conflict with other

variables usmg the same names. Because the member name is
linked with the structure name, it is separate from other
variables of the same name. For example, this program prints
10 100 101 on the screen .
• include <stdio.h>

int main (void)

(
struct s_type {
int i;
int j;
} 5;

int i;

i = 10;
312 _'fUP f

... C

B.i = 100;
B.j = 101;

printf("d 'd 'd', i, B.i, B.j);

return 0;
)

The variable i and the structure member i have no relationship

to each other.
6. As stated earlier, a: function may return a structure to the calling
procedure. The following program, for example, loads the
members of ...arl with the values 100 and 123.23 and then
displays them on the screen:
'include <stdio.h>

at .....:t a_type (
int i;
double d;
);

st ....ct s_type f(void);

int main (void)

(
struct s_type varl;

varl ~ fO;
printf("d 'f', varl.i, varl.d);

return 0;
)

.truct s_type f(void)

(
struct s_type temp;

temp. i = 100;
temp.d = 123.23; •
return temp;
)
 STRUCTURES AND UNIONS 313
"I'
lal MASTER STRUCTURE BASICS

7. This program passes a structure to a function:

'include <stdio.h>

struct s_type {
int i;
double d:
};

void f(struct s_type temp);

int main(void)
{
struct s_type varl;

varl.i = 99;
varl.d = 98.6;
f (varl);

return 0;
}

void f(struct s_type temp)

•
{
printf(M%d tf M
, temp.i. temp.d);
)

.:1', . . . . . .

1. In Chapter 9, you wrote a program that created a telephone

directory that was stored on disk. Improve the program so that it
uses an array of structures, each containing a person's name,
area code, and telephone number. Store the area code as an
integer. Store the name and telephone number as strings. Make
the array MAX elements long, where MAX is any convenient
value that you choose.
2. What is wrong with this fragment?
struct s_type
int i;
314 TEACH YOURSELF
'I' C

long 1;
char str[aO];
} 5;

~ = 10:

3. On your own, examine the header file STDIO.H and look at how
the FILE structure is defined.

DECLARE POINTERS TO STRUCTURES

It is very common to access J structure through a pOInter.(you declare
a pointer to a structure in the same way that you declare a pointer to
any other type of variable. For example, the following fragment
defines a structure called s_type and declares two variables. The first,
S, is an actual structure variable. The second, p, is a pointer to
structures of type s_type.
struct s_type {
int i;
char str[aO];
S. *Pi

Given this definition, the following statement assigns to p the

address of s:
p = &5:

Now that p points to s you can access s through p. However, to access

an individual element of s using p you cannot use the dot operator.
Instead, you must use the mTOtI' opel"{ltor, as shown in the following
example:
p->i =. 1;)
 STRUCTURES AND UNIDNS 315
10.2 DEClARE POINTERS TO STRUCTURES

This statement assigns the value 1 to element i of s through p. The

arrow operator is formed using a minus sign followed by a greater-than
sign. There must be no spaces between the two.
CC passes structures to functions in their entirety. However, if the
structure is very large, the passing of a structure can cause a
considerable reduction in a program's execution speed. For this
reason, when working with large structures, you might want to pass a
pointer to a structure in situations that allow it instead of passing the
structure itself.)

Remember ( When accessing a member usirig a structure variable, use the dot operator.
When accessing a member using a pointer, use the arrow operator. )

EXAMPLES

1. The following program illustrates how to use a pointer to a

structure:
#include <stdio.h>
#include <string.h>

struct s_type {
int it
char str[aO];
s. *p;

int main(void)
(
p = &5;

s.l = 10; / * this is functionally the same */

p->i = 10; / * as this */
strcpy(p->str, "I like structures.");

printf("%d %d %s", 5.i, p->i, p->str);

return 0;
316 TEACH YOURSELF

" C

2. One very useful application of structure pointers is found in C's

time and date functions. Several of these functions use a pointer
to the current time and date of the system. The time and date
functions require the header file TIME.H, in which a structure
called tm is defined. This structure can hold the date and time
broken down into its elements. This is called the bmken-down
time. The tm structure is defined as follows:
struct tm {
int tm _sec; I' seconds, 0-61 ' /
int tm _min; I' minutes, 0-59; 'I
int tm_hour; I ' hours. 0-23 ' I
int tm_rnday; I' day of the month. 1-31'/;
int tm _mon; / ' months since Jan, 0-11 ' I
int tm_year; / ' years from 1900 ' I
int tm_wday; I' days since Sunday, 0-6' 1 •
int tm_yday; • days s~nce Jan 1, 0-365 ' /
int tm_isdst: I ' Daylight Saving Time indicator . /
};

The value of tm_isdst ",ill be positive if Daylight Saving Time is

in effect, zero if it is not in effect, and negative if there is no
information available. Also defined in TIME.H is the type
time_to It is essentially a long integer capable of representing
the time and date of the system in an encoded implementation-
specific internal format. This is referred to as the calendar time.
To obtain the calendar time of the system, you must use the
timer ) function, whose prototype is:

time_t time(time_t 'systime) ;

The timer ) function returns the encoded calendar time of the

system or -I ifno system time is available. It also places this
encoded form of the time into the variable pointed to by systimc.
However, if systime is null, the argument is ignored.
Since the calendar time is represented using an implementation-
speCified internal format, you must use another of C's time and
date functions to convert it into a form that is easier to use. One
of these functions is called localtime( ). Its prototype is

struet tm 'Iocaltime(time_t 'systime);

 STRUClURES AND UNIONS 317
10.2 DECLARE POINTERS TO STRUCTURES

The localtime( ) function returns a pointer to th e broken-down

form of systime. The structure that h olds the broken-down time
is internally allocated by the compiler and will be overwritten
by each subsequent call.
This program demonstrates time( ) and localtime( ) by
displaying the current time of the system:
#include <stdio.h>
#include <time.h>

int main(void)
(
struct tm *systime;
time_t 't;

t = time (NULL) ;
systime = localtime(&tl;

printf("Time is %.2d:%.2d:%.2d\n", systime->tm_hour,

systime->trn_min, systime->tm_sec);
printf("Date: %.2d/%.2d/%.2d", systime->trn_mon+l,
systime->tm_rnday, systime->tm_year);

return 0;
J

Here is sample output produced by this program:

Time is 10:32:49
Date: 03115/97

EXERCISES

1. Is this program fragment correct'

struct s_type {
int a;
int bi
 318 TfACH YOURSELF
'I' e

int main (void)

(
p = &5;

p.a = 100;

2. Another of C's time and date functions is called grntime( ) . Its

prototype is

struet tm 'gmtime(time_t 'time);

The grntime( ) function works exactly like localtime( ) ,

except that it returns the Coordinated Universal Time (which is,
essentially, Greenwich Mean Time) of the system. Change the
program in Example 2 so that it displays both local time and
Coordinated Universal Time. (Note: Coordinated Universal
Time may not be available on your system.)

10.3 WORK WITH NESTED STRUCTURES

So far, we have only been working with structures whose members

consist solely of C's basic types. However, members can also be other
structures. These are referred to as nested structures. Here is an
example that uses nested structures to hold information on the
• performance of two assembly lines, each with ten workers:

struct worker (
char narne[80];
int aV9_uni ts-per_hour: •
int aV9_errs-per_hour;
) ;

struct asm_line {
int product_code;
double material_cost;
 STRUCTURES AND UNIONS 319
10.3 WORK WITH NESTED STRUCTURES

struct worker wkers[NUM_ON_ LINE];

} line!. line2;

To assign the value 12 to the av~units_pechour of the second

wkers structure of line I , use this statement:
linel.wkers[l] .av9_units-per_hour = 12;

As you see, the structures are accessed from the outer to the inner.
This is also the general case. Whenever you have nested structures,
you begin with the outermost and end with the innermost.

, EXAMPLE

I. A nested structure can be used to improve the card catalog

program. Here, the mechanical information about each book is
stored in its own structure, which, in turn, is part of the catalog
structure. The entire catalog program using this approach is
shown here. Notice how the program now stores the length of
the book in pages.
/ - An electronic card catalog--3rd Improvement. */
#include <stdic.h>
#include <string.h>
#i~clude <stdlib.h>

#define MAX 100

int menu (void) ;

void display(int i);
void author_search (void);
void title~search(void);
void enter(void);
void save(void);
void load (void) i

struct book_type (
unsigned date; /. copyright date ./
unsigned char ed; /. edition ' f
unsigned pages; f' length of book ' f
)
320 TEACH YOURSElf
~ C

struct catalog {
char name[80] ; /* author name *j
char title[80]; /- title */
char pub[80]; /- publisher -/
struct book_type book; /* mechanical info wi
) cat [MAX] ;

int tOP = 0; '* last location used */

int main(voidl
{
int choice;

load(); /* read in catalog */

do {
choice = menu () ;
5wi tch (choice) (
case 1: enter(); '* enter books */
break;
case 2: author_search(); /* search by author */
break;
case 3: title_search(); /* search by title */
break;
case 4.: save();
)
} while(choice!=51;

returr. 0;
)

'* Return a menu selection. */

menu (void)
{
int i;
char str(80};

printf("Card catalog:\n");
printf(" 1. Enter\n");
printf{" 2. Search by Author\n");
printf{" 3. Search by Title\n");
printf'" 4.. Save catalog\n");
printf'" 5. Quit'n");

do {
 10.3
STRUCnJRES AND UNIONS
WORK WITH NESTED STRUCTURES
,.
321

printf(-Choose your selection: R);

gets(str) ;
i = atoi(str);
pnntf("'n") ;
) while(i<l II i>5);

return i;
)

/* Enter books into database. */

void enter (void)
(
int i;
char temp(8D]:

for (i=top; i<MAX; i++) {

printf(NEnter author name (ENTER to quit): ~)
gets(cat(i] .name):
if(!*cat[i] .name) break;
printf{"Enter title: ");
gets(cat[i] .title):
printf("Enter publisher: ");
gets(cat(i] .pub) i
printf("Enter copyright date: ");
gets (temp) ;
cat[i] .book.date = (unsigned) atoi(temp);
printf("Enter edition: ");
gets (temp) ;
cat[i].book.ed = (unsigned char) atoi(temp);
printf("Enter number of pages: ");
gets (temp) ;
cat[i].book.pages = (unsigned) atoi(temp);
)
top = i;
)

/* Search by author. */
void author_search (void)
{
char name [ 80] ;
int i, found;

printf ("Na.'lle: ");

gets (name) ;
322 TEACH YOURSELF
c,

found = 0;
for(i=O; i<top; i++)
if(!strcmp(name. cat{i] .narne» {
display(i) ;
found = 1;
printf,-'n M
) ;

if(!foundl printf(MNot Found\n-);

)

/'" Search by title. "'I

void title_search(voidl
{
char titleIBO];
int i. found;

printftMTitle: .);
gets{':.ltle);

fCl1.:.nd = 0;

for{i=O; i<top; i++)

if(!strcmp(title. cat[ij.title» {
(hsplay (i) ;
found = 1;
printf!"\n-) ;
)
if(!foundl printf(-Not Founc\n*);

,- Display catalog entry. */

void display(int il
{
printf(-'s\n-, catli] .title);
printf(-by %5\n-, cat[~J .name);
printf(*Published by %s\n"', cat[i] .pub);
printf(·Copyright: %u. edition: %u\n-,
catli1.book.date. cat[i] .book.edl;
p~intf(·Pages: %u\n-, catli] .book.pages);
)

/* Load the catalog file. */

 STRUCTURES AND UNIDNS 323
1a3 WORK WITH NESTED STRUCTURES

void load(void)
(
FILE ·fp;

if«fp = fopen("catalog", "rb"))==NULL) (

printf("Catalog file not on disk.\n");

return;
}

if(fread(&top, sizeof top, 1, fp) != 1) ( /* read count */

printf("Error reading count.\n");
exit{l);
}

if(fread(cat, sizeof cat, 1, fp) != 1) { /* read data */

printi (~Error reading catalog data. \n") ;
exi t (I) ;

fclose (fp) ;
}

/* Save the catalog file. */

void save(void)
(
FILE "fp;

if({fp = fopen{"catalog", "wb."»==NULL) (

printf("Cannot open catalog file.\n");
exit(l) ;
}

if(fwrite(&top. sizeof top, 1, fp) != 1) { /* write count */

printf{"Error writing count.\n");
exit(l);
}

if(fwrite(cat, sizeof cat, 1. fp) != 1) ( /* write data */

printf(·Error writing catalog data.\n");
exit(!);
324 TEACH YOURSELF

" C

EXERCISE

1. Improve the telephorle-directory program you wrote earlier in

this chapter so that it includes each person's mailing address.
Store the address in its own structure, called address, which is
nested inside the directory structure.

- . , 10.4 . ' U N DERSTAND BIT-FIELDS

(c allows a variation on a structure member called a blt-jield A blt-jicld

is composed of one or more bits. Using a bit-field, you can access by
name one or more bits within a byte or word'f[o define a bit-field, use
this general form:

type name: size;

Here, type is either int or unsigned. If you specify a signed bit-field,

then the high-order bit is treated as a sign bit, if possible. The number
of bits in the field is specified by size. Notice that a colon separates the
name of the bit-field from its size in bits')
(Bit-fields are useful when you want to pack information into the
smallest possible space. For example, here is a structure that uses
bit-fields to hold inventory information.
struct b_type {
unsigned department: 3 ; f' up to 7 departments of
unsigned instock: 1 ; /0 1 if in stock, 0 if out Of
unsigned backordered: 1 ; fo 1 i f backordered, 0 i f not 'f
 STRUCTURES AND UNIONS 325
10.4 UNDERSTAND BIT-RELDS
"
unsigned lead_time: 3; /* order lead time in months */
) inv [MAJeITEM] ;

In this case one byte can be used to store information on an inventory

item that would normally have taken four bytes without the use of
bit-fields)(Xou refer to a bit-field just like any other member of a
structure. The following statement, for example, assigns the value 3 to
the department field of item 10:
inv[9] .department = 3:

The following statement determines whether item 5 is out of stock:

if(!inv[4].instock) printf("Out of Stock");
else printf ( ~ In Stock"): )

(It is not necessary to completely define all bits within a byte or

word . For example, this is perfectly valid:
struct b_type (
int a: 2;
int b: 3 i •

The C compiler is free to store bit-fields as it sees fit. However,

usually the compiler will automatically store bit-fields in the smallest
unit of memory that will hold them. Whether the bit-fields are stored
high-order to low-order or the other way around is implementation-
dependent. However, many compilers use high-order to low-order.)
(You can mix bit-fields with other types of members in a structure's
definition. For example, this version of the inventory structljre also
includes room for the name of each item:
struct b_type {
char name[40]: /* name of item */
unsigned department: 3; /* up to 7 departments */
unsigned instock: 1: /* 1 if in stock, 0 if not */
unsigned backordered: 1; /* 1 if backordered, 0 if not */
unsigned lead_time: 3; /* order lead time in months */
} inv(MAX_ITEMJ: }

(Because the smallest addressable- unit of memory is a byte , you

cannot obtain the address of a bit-field variable.)
326 TEACH YOURSElf
,. c

(Bit-fields are often used to store Boolean (true/false) data because

they allow the efficient use of memory-remember, you can pack ·
eight Boole~n values into a single byte)

EXAMPLES

1.( It is not nece~sary to name every bit when using bit-fields. Here,
for example, is a structure that uses bit-fields to access the first
and last bit in a byte.
struct b_type {
unsigned first: 1;
int : 6;
unsigned last: 1;
};

The use ·of unnamed bit-fields makes it easy to reach the bits
you are interested in.)
2. To see how useful bit-fields can be when working with Boolean
data, here is a crude simulation of a spaceship flight recorder.
By packing all the relevant information into one byte,
comparatively little disk space is used to record a flight.
/* Simulation of a 100 minute spaceship

. /
flight recorder .

#include <stdlib.h>
#include <stdio.h>

/* all fields indicate OK if 1,

malfunctioning or low if 0 */
• struct telemetry {
unsigned fuel: 1;
unsigned radio: 1;
unsigned tv: 1;
unsigned' water: 1;
unsigned food: 1;
unsigned waste: 1;
} flt_reed;

void display(struct telemetry i);

 STIIUCTURES AND UNIONS 327
10.4 UNDERSTAND Bff-FIELDS

int main (void)

(
FILE *fPi
int i;

H{lfp = fopen{" flight" , "wb"»==NULL) (

printf (~Cannot open file. ~n")';
exit(l) ;
)

/ * Imagine that each minute a status report of

the spaceship is recorded on disk.
*1
for(i=O; i<100: i++) {
fIt_reed,fuel = rand()%2:
f1 t_recd. radio = rand ( ) %2;
flt_reed.tv =
rand(1% 2 ;
flc_reed.water = rand(1%2;
flt~recd.food = rand()%2;
flt_reed.waste = rand( ) %2 ;

display(flt_recd) :
fwrite(&flt_recd, sizeof flt_recd, 1, fp) i
)

fcloselfp) :

return 0:
)

void display(struct telemetry i)

(
if(i.fuel) printf(-Fuel OK\n-);
else printf(MFuel low\n-):
if(i.radio) printf("Radio OK\n-);
else printf(-Radio failure\n M
);

if(i.tv) printf("TV system OK\n-);

else printf(,"TV malfunction\n") i
if(i.water) printf("Water supply OK\n");
else printf("Water supply low\n");
if(i.food) printf(qFood supply OK\n");
else printf("Food supply low\n"):
if(i.waste) printf("Waste containment OK\nM);
328 TEACH YOURSELF

" C

else printf(~Waste containment failure\n-);

printf(~\n·l ;
}

Depending on how your compiler packs the bit-fields, after you

run this program, the file on disk may be as short as 100 bytes
long. Now try the program after modifying the telemetry
structure as shown here:
struct telemetry {
char fuel;
char radio;
char tv;
char water;
char food;
char waste;
} fIt_reed;

In this version, no bit-fields are used and the resulting file is at

least 600 bytes long. As you can see, using bit-fie ids can provide
substantial space savings.

1. Write a program that creates a structure that contains three

bit-fields called a, b, and c. Make a and b three bits long and
make c two bits long. Next, assign each a value and display the
values.
2. Many compilers supply library functions that return the status
of various hardware devices, such as a serial port or the
keyboard, by encoding information in a bit-by-bit fashion. On
your own, consult the user's manual for your compiler to see if
it supports such functions. If it does, write some programs that
read and decode the status of one or more devices.
 STlIUCTURES AND UNIONS 329
10.5 CREATE UNIONS

EATE UNIONS

(In C, a union is a single piece of memory that i; shared by two or more

variables. The variables that share the memory may be of different
types. However, only one variable may be in use at anyone time'{A
union ,is defined much like a structure. Its general form is

union tag-name {
type member 1;
type member2;
type member3;

Like d structure, either the tag-name or the vanable-nan1cs may be

·mssmg. Members may be of any valid C data type) For example, here
is a union that contains three elements: an integer, a character array,
~nd a double:

(union u_type (
int i;
charc{2];
double d;
} samplei

This union will appear in memory as shown in Figure 10-2. ')

t-I- - - - - - - - d
t-c[O) + c[I)-1
in
I I I
330 TEACH YOURSElf

'" C

( To access a member of a union, use the dot and arrow operators just
as you do for structures. For example, this statement assigns 123.098
to d of sample:

samp1e.d = 123.098;
If you are accessing a union through a pointer, you must use the arrow
operator. For example, assume that p points to sample. The following
statement assigns i the value 101:
p->i = 101;)

(It is important to understand that the size of a union is fIxed at

compile time and is large enough to accommodate the largest member
of the union. Assuming 8·byte doubles. this means that sample will
be 8 bytes long. Even if sample is currently used to hold an int value,
it wrtrstill occupy 8 bytes of memory. As is the case with structures,
you should use the sizeof compile·time operator to determine the size
of a union. You should not simply assume that 11 will be the size of the
largest element, because in some environments, the compiler may pad
the union so that it aligns on a word boundary)

EXAMPLES.

I. Unions are very useful when you need to interpret data in two
or more different ways. For example, the encode( ) function
shown below uses a union to encode an integer by swapping its
two low-order bytes. The same function can also be used to
decode an encoded integer by swapping the already exchanged
bytes back to their original positions.
#include <stdio.h>

int encode(int i);

int main(void)
(
int i;

i = encode (10) ; f* encode it */

 STRUCTURES AND UNIONS 331
10.5 CREATE UNIONS

printf("lO encoded is %d\n", il;

i = encode(i); /* decode it *1
printf("i decoded is %d-, i);

return 0;
)

/* Encode an integer, decode an encoded integer. *1

int encode{int i)
{
union crypt_type {
int num;
char c (2];
} crypt j
unsigned char Chi

crypt.num = i;

j *swap bytes */
ch = crypt.e(O];
crypt.e[O] = crypt.e[l];
crypt.e(l] = chi

/ * return encoded integer *1

return crypt.num;
)

The program displays the following:

10 encoded is 2560
i decoded is 10

2. The following program uses the union of a structure containing

bit-fields and a character to display the binary represe ntation of
a character typed at the keyboard:
/ * This program displays the binary code for a
character entered at the keyboard .
' /
#include <stdio.h>
#include <conio . h>

struct sample (
unsigned a: 1;
unsigned b: 1;
332 nACH YOURSELF
." C

unsigned c, 1 ;
unsigned d, 1;
unsigned e, 1 ;
unsigned f' 1;
unsigned g' 1;
unsigned h, 1;
) ;

union key_type {
char chi
struct sample bits;
key; ,

int main(void~
(
printf("Strike a key: ~);

key.ch = getche();
printf (~\nBinary code is: "l;

if(key.bits.h) printf("l H);

else printf("O "};
if(key.bits.g) printf("l ~);

else printf("O to);

if(key.bits.f) printf("! ");
else printf ("0 ");
if(key.bits.e) printf("l ");
else printf ("0 .. );
if{key.bits.d) printf{"l ");
else printf{"O ");
if(key.bits.c) printf("l "J;
else printf ("0 .. );
if{key.bits.b) printf("l n);
else printf("O M);
if(key.bits.a) printf(~l ");
else printf{"O ");

return 0;

When a key is pressed, its ASCII code is assigneq to !<ey.ch,

which is a char. This data is reinterpreted as a serjf's ofbit-fields,
which allow the binary representation of the key to be
displayed. Sample output is shown here:
 STRUCTURES AND UNIONS 333
10.5 CREATE UNfONS
'"

Strike a key: X
Binary code is: 0 1 0 1 1 0 0 0

EXERCISES
1. Using a union composed of a double and an 8-byte character
array, write a function that writes a double to a disk file, a
character at a time. Write another function that reads this value
from the file and reconstructs the value using the same union.
(Note: If the length of a double for your compiler is not 8 bytes,
use an appropriately sized character array.)
2. Write a program that uses a union to convert an int into a long.
Demonstrate that it works.

At this point you should be able to answer these questions and

perform these exercises:

1. In general terms what is a structure, and what is a union?

2. Show how to create a structure type called s_type that contains
these five members:
char chi
float d:
int i;
char str[80];
double balance:

Also, define one variable called s_ var using this structure.

3. What is wrong with this fragment>
struct s_type {
int a;
334 TtACH YOURSW
't' C

char b;
float bali
} myvar, .p;

p = &myvar;

p.a = 10;

4. Write a program that uses an array of structures to store

employee names, telephone numbers, hours worked, and
hourly wages. Allow for 10 employees. Have the program input
the information and save it to a disk file. Call the file EMP.
5. Write a program that reads the EMP file created in Exercise 4
and displays the information on the screen.
6. What is a bit-field?
7. Write a program that displays individually the values of the
high- and low-order bytes of a short integer. (Hint: Use a union
that contains as its two elements a short integer and a two-byte
character array.)

This section checks how well you have integrated the material in
this chapter with that from earlier chapters.

1. Write a program that contains two structure variables defined as:

struct s_type {
int i;
char chi
double d;
} varl, var2;

Have the program give each member of both structures initial

values, but make sure that the values differ between the two
structures. Using a function called struct_swap( ), have the
program swap the contents ofvarl and var2.
2. As you know from Chapter 9, fgetc( ) returns an integer value,
even though it only reads a character from a file. Write a
 335
STRUCTURES AND UNIONS
10.5 CREATE UNIONS
.,

program that copies one file to another. Assign the return value
of fgetc( ) to a union that contains an integer and character
member. Use the integer element to check for EOF. Write the
character element to the destination file. Have the user specify
both the source and destination tile names on the command line.
3. What is wrong with this fragment?
Sl:.ruct s_type {
int a;
int b: 2;
int c: 6;
} vari

scanf ("%d", &var);

4. In C, as you know, you cannot pass an array to a function as a

parameter. (Only a pointer to an array can be passed.) However,
there is one
, way around this restriction. If you enclose the array
within a structure, the array is passed using the standard
call-by-value convention. Write a program that demonstrates
this by passing a string inside a structure to a function, altering
its contents inside the function and demonstrating that the
original string is not altered after the function returns.
11
Advanced Data Types
and Operators

11.1 Use the storage class specifiers

11.2 Use the access modifiers
11.3 Define enumerations
11.4 Underllla~d """"'"
11.5 Use C's bitwise operators
11.8 Master the shift operators
11.7 Underllland the ?·oPerator
11.8 Do more with the assignment operator
11.9 Understand the comma operator
11.10 Know the precedence summary

337
••
338 TEACH YOURSELF

'" ' C

-lIB E C language includes a rich set of data type modifiers that

allow you to better fit the type of a vari ..ble to the information
it will be storing. AlstJ, C includes a number of special
operators that permit the creation of very efficient routines.
Both of these items are the subject of this chapter.

Before proceeding, you should be able to answer these questions

and perform these exercises.

I. Write a program that uses an array of structures (0 hold the

squares and cubes of the numbers I through 10. Display the
contents of the array.
2. Write a program that uses a union to display as a character the
individual bytes that make up a short integer entered by the
user.
3. What does this fragment display? (Assume two-byte ints and
eight-byte doubles.)
union {
int i;
double d;
} uvar;

printf("%d", sizeof uvar) i

4. What IS wrong with this fragment?

struct {
int i;
char str[80J;
double balance;
} svar;

svar->i = 100;

5. What is a bit-field?
 ADVANCED DATA TYP£S AND OPERATORS 339
11. J USE THE STORAGE ClASS SPECJRERS
"

n.1 SE THE STORAGE CLASS SPECIFIERS

( C defines four type modifiers that affect how a variable is stored. They are

auto
extern
register
static
These spe cifiers precede the type name) Let's look at each now.
( The specifier auto is completely unnecessary)It is provided in C to
allow compatibility with its predecessor, B.Qts use is to declare
automatic variables. Automatic variables are simply local variables,
which are auto by default. You will almost never see auto used in any
C program')
Although the programs we have been working with in this book are
fairly short, programs in the real world tend to be quite long. As the
size of a program grows, it takes longer to compile. For this reason,~
allows you to break a program into two or more files. You can
separately compile these files and then link them together. This saves
compilation time and makes your projects easier to work with)CThe
actual method of separate compilation and linking will be explained in
the instructions that accompany your compiler.) When working with
multiple source files there is, however, one issue that needs to be
addressed. As a general rule, global data can only be defined once.
However,~lobi\l data may need to be accessed by two or more files
that form a program. In this case, each source file must inform the
compiler about the global data it uses. To accomplish this you will
need to use the keyword extern. To understand why, consider the
following program, which is split between two files :

FILE #1:
#include <stdio.h>

int count;

void f1 (void) ;

int main(void)
{
int i;

•
rou:H YOUASB.F
C

fl(); /* set count's value */

.... for(i=O; i<count: i++)

printf("%d ", i):

return 0;
}

FILE #2:

'include <stdlib.h>

void fl(void)
(
count = rand () ;
}

If.you try to compile the second file, an error will be reported because
count is 'not defined. However, you cannot change fILE #2 as follows:
.include <stdlib.h>

int count;

void fl(void)
(
' count = rand();
J

If you declare Count a second time, many linkers will report a

duplicate-symbol error, which means that count is defined twice, and
the linker doesn't know which to use)
(The solution to this problem is C's extern speCifier. By placing
extern in front of count's declaration in FILE #2, you are telling the
compiler that count is an integer defined elsewhere. In other words,
using extern informs the compiler about the existence and the type of
the variable it precedes, but it does not cause storage for that variable
to be allocated. The correct version of FILE #2 is
#include <st~~ib.h>

extern int count;

void fl(vclidJ
 ADVANCED DIITA TYPES AND _TORS 341
11.1 USE TIlt; STOAAGE ClASS SPfClRERS •

(
count ='rand();
) )
( Although rarely done, it is not incorrect to use extern inside a
function to declare a global variable defined elsewhere 10 the same
file . For example, the following is valid:
#include <stdio.h>

int count;

lnt main(void)
(
extern int count; /* this refers to global count *1

count = 10;
printf(-%d 8
, count);

return 0;
)

The reason you will rarely see this use of extern is that it is .
redundant. Whenever the compiler encounters a variable name not
defined by the function as a local variable, it assumes that it is global.)
(One very important storage-class specifier is register. When you
specify a register variable you are telling the compiler that you want
access to that variable to be as fast as possible)In early versions of C,
register could only be applied to local variables (including formal
parameters) of types int or char, or to a pointer type. It caused the
variables to be held in a register of the CPU. (This is how the name
register came about.) By using a register of the CPU, extremely fast
access times are achieved.(ln modern versions of C, the definition of
register has been broadened to include all types of variables and the
requirement that register variables must be held in a CPU register
was removed. Instead, the ANSI C standard stipulates that a register
variable will be stored in such a way as to minimize access time. In
practice, however, this means that register variables of type int and
char continue to be held in a CPU register-this is s~iI1 the fastest way
to access them.
No matter what storage method is used, only so many variables
can be granted the fastest possible access time. For example, the CPU
II
has a limited number of registers. When fast-access locations are
342 TEACH YOURSElf

" C

exhausted, the compiler is free to make register variables into regular

variables. For this reason, you must choose carefully which variables
you modifY with register)
(One good choice is to make a frequently used variable, such as the
variable that controls a loop, into a register variable . The more times a
variable is accessed, the greater the increase in performance when its
access time is decreased. Generally, you can assume that at least two
variables per function can be truly optimized for access speed)
Important:(j3ecause a register variable may be stored in a register of
the CPU, it may not have a memory address. This means that you
cannot use the £5 to find the address of a register variable)
( When you use the static modifier, you cause the contents of a local
variable to be preserved between function calls. Also, unlike normal
local variables, which are initialized each time a function is entered, a
static local variable is initialized only once. For example, take a look
at
this program,
#include <stdio.h>

void f (void) ;

int main(void)
(
int i;

for(i=O; i<;10; i++) f();

return 0;
)

void f(void)
(
static int count = 0;
count++;
printf("count is %d\nM, count);
)

which displays the following output:

count is 1
count is 2
 ADVANCBl OATA TYPES AND DP£RATORS 343
",
//./ USE THE STORAGE CLASS SPECIFIERS

count is 3
count is 4
count is 5
count is 6
collnt is 7
\
count is B
count is 9
count is 10
As you can see, count retains its value between function caHs. The
advantage to using a static local variable over a global one is that the
static local variable is still known to and accessible by only the
function in which it is declared.)
<'!he static modifier may also be used on global variables. When it is,
it causes the global variable to be known to and accessible by only the
functions in the same file in which it is declared. Not only is a function
not declared in the same file as a static global variable unable to
access that global variable, it does not even know its name. This
means that there are no name conflicts if a static global variable in
one file has the same name as another global variable in a different file
of the same program. For example, consider these two fragments,
which are parts of the same program:

FilE #1 FilE #2
int count; static int count;

count = 10; count = 5:

printf("%d-, count); printf(M%d M , ccu~t);

Because count is declared as static in FILE #2, no name conflicts

arise. The printf( ) statement in FILE #\ displays 10 and the printf( )
statement in FILE #2 displays 5 because the two counts are
different variables)

EXAMPLES

I. To get an idea about how much faster access to a register

variable is, try the foHowing program. It makes use of another of
C's standard library functions caHed c1ock( ), which returns the
 number of system clock ticks since the program began
execution. It has this prototype:

clock_t clock(void);

It uses the TIME.H header. TIME.H also defines the c1ock_t

type, which is more or less the same as long. To time an event
using c1ock{ ), call it immediately before the event you wish to
time and save its return value. Next, call it a second time after
the event finishes and subtract the starting value from the
ending value. This is the approach used by the program to time
how long it takes two loops to execute. One set of loops is
controlled by a register variable, 'the other is controlled by a
non-register variable.
'include <stdio.h>
.include <time.h>

int i; /* This will not be transformed into a

register variable because it is global.* /

int main(void)
(
~ register int j:

int k;
clock_t start, finish;

start '= clock();

for(k=Oi k <100; k++)
for(i~O; i<32000; i+.)
finish = clock();
pr~ntf( · Non-register loop: tId ticKs \ n- , finish - start);

start = clock();
for l k=O; k<lOO; k++l
for(j=O; j<32000 ; j++l;
finish = clock() ;
printf (-Register 'loop: tId ticKs \ n-, finish - start );

return 0;
I
 ADVANCW DATA TYPES AND OPERATDRS 345

For most compilers, the register-controlled loop will execute

about twice as fast as the non-register controlled loop.
The non-register variable is global because, when feasible,
virtually all compilers will automatically convert local variables
not specified as register types into register types as an
automatic optimization. If you do not see the predicted results,
it may mean that the compiler has automatically optimized i
into a register variable, too. Although you can't declare global
variables as register, there is notbing that prevents a compiler
from optimizing your program to this effeet. If you don't see
much difference between the two loops, try creating extra global
variables prior to i so that it will not be automatically optimized.
2. As you know, the compiler can optimize access speed for
only a limited number of register variables in anyone function
(perhaps as few as two). However, this does not mean that your
program can only have a few register variables. Because of the
way a C program executes, each function may utilize the
maximum number of register variables. For example, for the
average compiler, all the variables shown in the next program
will be optimized for speed:
#include <stdio.h>

void f2 (void) :
void f (void) ;

int main (void)

(
register int a , b;

void f (void)
(
register int i, j;
c

void f2(voidl
(
register int j. k;

3. Local static variables have several uses. One is to al1o-;'" a

function to perform various ininalizations only once, when it is
first called. For example, consider this function :
void myfunc(void)
(
static inc first = 1;
if{f~rst) { j* initialize the system */
rewH~d (fp) ;

a = 0;
lee = 0,
fprintf(~System IniLialized~);

first = 0;
)

Because first is static, it will hold its value between calls. Thus,
the initialization code will be executed only the first time the
function is called.
4. Another interesting use for a local static variable is to control a
recursive function. For example. this program prints the
numbers 1 through 9 on the screen:
#include <stdio.h>

void f (void) ;

int main (void)

(
t();

re turn 0;
}
 ADVANCUI DATA TYPES AND OPERATORS 347
~
11.' USE THE STORAGE ClASS SPEC/FifRS

void f (void)
(
static int stop=O:

stop++;

if(stop==10) return;
printf("%d ", stop);
f(); /* recursive call *1
)

Notice how stop is used to prevent a recursive call to C( ) when

it equals 10.
5. Here is another example of using extern to allow global data to
be accessed by two files:

FILE #1:
fiinclude <stdio.h>

char str[80];

void getname(void);

int main (void)

(
getname () ;
printf"{ -Hello %s·. str);

return 0;
)

FILE #Z,
'include <stdio.h>

extern char str[BO]:

void getname(void)
(
printf("Enter your first name: .. ,; .
gets (str);
)
TEACH YOURSElF
C

EXERCISES

1. Assume that your compiler will actually optimize access time of

only two register variables per function. In this program, which
two variables are the best ones to be made into register variables?
#include <stdio.h>
#include <conio.h>

int main (void)

{
int i. j, k, m;

do ("
printf t -Enter a value: .);
scanf ("M', &i);

m = 0;
for(j=O; j<i; j++)
for (k=Oi k<100i k++ )
m = k + mi
while t i>O) ;

return 0;
)

2. Write a program that contains a function called sum_it( ) that

has this .prototype:
void sum_it (int value);

Have this function use a local static integer variable to maintain

and display a running total of the values of the parameters iUs
called with. for example, if sum_it( ) is called three times with
the values 3,6, 4, then sum_it( ) will display 3, 9, and 13.
3. Try the program descnbed in Example :.. Be sure to actually use
two files. If you are unsure how to compile and link a program
consisting of two files, check your compiler's user manual.
4. What is wrong with this fragment?
register int i;
int .p;

p = &i;
 ADVIINGlD DATA TYPES AND OPERATORS 348
1 1.2 USE THE ACCESS MOD/FlEAS
'"

SE THE ACCESS MODIFIERS

.
C includes two type modifiers that affect the way variables are
accessed by both your program and the compiler. These modifiers are
const and volatile. This section examines these type modifiers.
If you precede a variable's type with const, you prevent that
variable from being modified by your program. The variable may
be given an initial value, however, through the use of an initialization
when it is declared. The compiler is free to locate const variables
in ROM (read-only memory) in environments that support it. A
const variable may also have its value changed by hardware-
dependent means.
The const modifier has a second use. It can prevent a function
from modifying the object that a parameter points to. That is, when a
pointer parameter is preceded by const, no statement in the function
can modify the variable pointed to by that parameter.
When you precede a variable's type with volatile, you are telling
the compiler that the value of the variable may be changed in ways
not explicitly defined in the program. For example, a variable's
address might be given to an interrupt service routine, and its value
changed each time an interrupt occurs. The reason that volatile is
important is that most C compilers apply complex and sophisticated
optimizations to your program to create faster and more efficient
execu table programs. If the compiler does not know that the contents
of a variable may change in ways not explicitly specified by the
program, it may not actually examine the contents of the variable
each time it is referenced. (Unless it occurs on the left side of an
assignment statement, of course.)

EXAMPLES

J. The following short program shows how a con8t variable can be

given an initial value and be used in the program, as long as it is
not on the left side of an assignment statement.
• include <stdio.h>

int main(void)
{
const int i = 10;
350 TEACH YOURSELf
... c

printf ( "%d·, i); / * this is OK ./

return 0;
}

The following program tries to assign i another value. This

program will not compile because i cannot be modified by
the program.
#include <stdio.h>

int main(void)
{
Canst int i : 10;

i = 20; /* this is wrong */

return 0;
}

2. The next program shows how a pointer parameter can be

declared as const to prevent the object it points to from
being modified.

'include <stdio.h>

void pr_str{const char *p);

int main (void)

{
char strr80];

printf(·Enter a string: .};

gets{str~ ;

pr..... str(str) ;

return 0;
}

void pr_str(const char *p)

{

while(*p) putchar(*p++); /* this is ok */

}
 IIDVANC£D DATA rtP£S AND DPERATDRS 351
".2 USE THE ACCESS MODIFIERS

Ii you change the program as shown below, it will not

compile because this version attempts to alter the string pointed
to by p.
#include <stdio.h>
Jinclude <ctype.h>

void pr_str(const char *p);

int main(void)
(
char str [80 1 ;

printf{"Enter a string: ");

gets(str) i
pr_str(str) i

return 0;
)

void pr_str\const char .p)

(
.
while ('p) (
.p = toupper(~p); /. this will not compile */
putchar ( .. p++) ;
)
)

3. Perhaps the most important feature of const pointer parameters

is that they guarantee that many standard library functions will
not modify the variables pointed to by their parameters. For
example, here is the actual prototype to strlen( ) specified by
the ANSI standard:

size_t sll1en(const char ·str);

Since sIT is specified as conat, the string it points to cannot be

changed.
4. While' short examples ofvoJatUe are hard to find, the following
fragment gives you the flavor of its use:
volatile unsigned u;
3&2 ltACH YOURSELF
.., C

for(;;) { j* watch value of u -,

printf (-%d-, u);

In this example, if u had not been declared as volatile, the

compiler could have optimized the repeated calls to printf( } in
such a way that u was not reexamined each time. The use of
volatile forces the compiler to actually obtain the value of u
whenever it is used.

I. One good time to use const is when you want to embed a

version control number into a program. By using a const
variable to hold the version, you prevent it from aCCidentally
being changed. Write a short program that illustrates how this
can be done. Use 6.01 as the version number.
2. Write your own version of strcpy( ) called mystrcpy( ), which
has the prototype

char "mystrcpy (char "to, canst char 'from);

The function returns a pointer to 10. Demonstrate your version

of mystrcPY( ) in a program.
3. On your own, see if you can think of any ways to use volatile.

11.3 DEFINE ENUMERATIONS

In C you can define a list of named integer constants called an
enumeration. These constants can then be used any place an integer
can. To define an enumeration, use this general form:

enum tag-name { enumeration list) variable-list;

Either the tag· name or the variable-list is optional. The tag-name is
essentially the t ype name of the enumeration. For example ,
enum color_type {red. green, yellow} color ;
 ADVANCED DATA TYPES AND OP£RATDRS 353
11.3 DfFlNE fNUMERA nollS
"
Here, an enumeration consisting ofthe constants red, green, and
yellow is created. The enume ration tag is color_type and one
variable , called color, has been created.
By default, the compiler assigns integer values to enumeration
constants, beginning with 0 at the far left side of the list. Each constant
to the right is one greater than the constant that precedes it.
Therefore, in the color enumeration, red is 0, green is 1, and yellow '
is 2. However, you can override the compiler's default values by
explicitly giving a constant a value. For example, in this statement
enum color_type {red • . green=9. yellow} color;

red is stm 0, but green is 9, and yellow is 10.

Once you have defined an enumeration, you can use its tag name to
declare enumeration variables at other points in the program. For
example. assuming the color_type enumeration, this statement is
perfectly valid and declares mycolor as a color_type variable:
enum c o lor_type mycolori

An enumeration is essentially an integer type and an enumeration

variable can hold any integer value-not just those defined by the
enumeration . But for clarity and structure, you should use
enumeration variables to hold only values that are defined by their
enumeration type.
Two of the main uses of an enumeration are to help provide
self-documenting code and to clarify the structure of your program.

EXAMPLES
1. This short program creates an enumeration consisting of the
parts of a computer. It assigns comp the value CPU and then
displays its value (which is 1). Notice how the enumeration tag
name is used to declare comp as an enumeration variable
. separately from the actual declaration of computer.

iin clude <stdio . h>

e num compu ter {keyboard , CPU, screen , printer);

i n t ma i n (v oi d )
{
enum compu ter c a mp ;
3&4 TfACH YOURSnf

'" C

comp = CPU;
p.r intf(-'d-, comp);

return 0;
}

2. It takes a little work to display the string equivalent of an

enumerated constant. Remember, enumerated constants are not
strings; they are named integer constants, The following
program uses a sWitch statement to output the string equivalent
of an enumerated value. The program useS C's random-number
generator to choose a means of transportation. It then displays
the means on the screen. (This program is for people who can't
make up their minds!)
tinclude <stdio.h>
*include <stdlib.h>
.include <conio.h>

enum transport (car, train, airplane. bus) tPi

int main (void)

{
printf(-Press a key to select transport: .)j

/* generate a new random number each time

th~ program is run
"
while ( ! kbhit (II rand () ;
getch(): /* read and discard character */

tp = rand () % 4;
switch (tp I (
case car: printf(·car*):
break;
case train: printf(·train*);
break;
case airplane: printf(*airplane-);
break;
case bus: printf{*bus·);
}

return 0;
}
 AOVANC£D DATA TVPfS AND _TORS 355
..,
11.3 DEFINEENUMERATIONS

In some cases, there is an easier way to obtain a string

equivalent of an enumerated value. As long as you do not initialize
any of the constants, you can create a two-dimensional string array
that contains the string equivalents of the enumerated values in
the same order that the constants appear in the enumeration. You
can then index the array using
. an enumeration .value to obtain its
corresponding string. The following version of the transportation-
choosing program, for example, uses this approach:
#include <stdio.h>
#include <stdlib.h>
#include <conio.h>

enum transport {car. train. airplane. bus} tp;

char trans() (20) = (

·car~, "train-. -airplane", "bus"
) ;

int main(void)
{
printf ("Press a key to select transport: ");

j * Generate a new random number each time

., the program is run

while(!kbhit(» rand();
getch(); j * read and discard character */
•
tp = rand!) % 4;
printf(~%s·, trans {tp] ):

return 0;
)

3. Remember, the names of enumerated constants are known only

to the program, not to any library functions. For example, given
the fragment
enurn numbers {zero, one, two, ·three} nurn;
.,
printfC"Enter a number: to);
scanf (" %d", &num):

you cannot respond to seanf( ) by entering one.

356 T£ACH YOURSELF
c

EXERCISES

I. Compile and run the example programs.

2. Create an enumeration of the coins of the U.S. from penny
to dollar.
3. Is this fragment correct? If not, why not'
enurn cars {Ford, Chrysler, GM} make;

make = GM;
printf{"car is %s", make);

11.4 UNDERSTAND typedef

( In C you can create a new name for an existing type using typedef.
The general form of typedef is

typedef old-name new-name;

This new name can be used to declare variables)~or example., in the

following program, smallint is a new name for a signed char and is
used to declare i.
#include <stdio.h>

typedef signed char smallint:

int main{void)

smallint i;

for(i=O; i<lO; i++)

pr int f ( "%d ., i);

return 0;
J)
CKeep two points firmly in mind: First, a typedef does not cause the
original name to be deactivated. For example, in the program, signed
char is still a valid type. Second, you can use several typedef
statements to create many different, new names for the same type)
 ADVANCED DATA TYPU AND DPEAATORS 317
y
11.' UNDERSTAND typOdeI

. ( T~ere are basically two reasons to use typedef. The first is to create
portable programs. For example, if you know that you will be writing a
program that will be executed on computers using 16-bit integers as
well as on computers using 32-bit integers, and you want to ensure
that certain variables are 16 bits long in both environments, you might
want to use a typedef when compiling the program for the 16-bit
machines as follows:
typedef int myint;

Then, before compiling the code for a 32-bit computer, you can change
_the typedef statement like this:
typedef short int myint;

This works because on computers using 32-bit integers. a short int

will be 16 bits long. Assuming that you used myint·to declare all
integer values that you wanted to be 16 bits long, you need change
only one statement to change the type of all variables declared
using myint.
The second reason you might want to use typedef is to help
provide self-documenting code. For example, if you are writing an
inventory program, you might usc this typedef statement.
typedef double subtotal;

Now, when anyone reading your program sees a variable declared as

subtotal, he or she will know that it is used to hold a subtotal.)

EXAMPLES

I . (The new name created by one typedef can be used in a

subsequent typedef to create another name. For example,
consider this fragment:
typedef int height;
typedef height length;
typedef length depth;

depth d;

Here, d is still an integer)

2. In addition to the the basic types, you can use typedef on more
complicated types. For example, the followil1~ is perfectly valid:
368 nACH VOURSlLF

enum e_type {one, two, three }

typedef enurn e_type mynums;

mynums num; /* declare a variable */

Here, nurn is a variable of type e_type.

EXERCISES

1. Show how to make UL a new name for unsigned long.

Show that it works by writing a short program that declares a
variable using UL, assigns it a value, and displays the value
on the screen.
2. What is wrong with this fragment?
typedef balance float;

11.5 USE C'S BITWISE OPERA TORS

C contains four special operators that perform their operations on a
bit·by-bit level. These operators are

& bitwise AND

I bitwise OR
bitwise XOR (eXclusive OR)
1's complement

(These operators work with character and integer types; they cannot be
used with floating-point types.)
The AND, OR, and XOR operators produce a result based on a
comparison of corresponding bits in each operand. The AND operator
sets a bit if both bits being compared are set. The OR sets a bit if either
of the bits being compared is set. The XOR operation sets a bit when
either of the two bits involved is 1, but not when both are 1 or both are
O. Here is an example of a bitwise AND:
 ADVANCED DATA TYPES AND OPERATORS
11.5 USE C'S BfTWISf OPERATORS
,.
359

10100110
& 0011 1011

00100010
Notice how the resulting bit is se{, based on the outcome of the
operation being applied to the corresponding bits in each operand.
The 1's complement operator is a unary operator that reverses the
state of each bit within an integer or character.

EXAMPLES

I. The XOR operation has one interesting property. Given two

values A and B, when the outcome of A XOR B is XORed with B
a second time, A is produced. For example, thi& output

initial value of i: 100

i after first XOR: 21895
i after second XOR: 100

is produced by the following program:

#include <stdio.h>

int main(voidl
(
int i;

i = 101);
printf("initial value of i: %d\n-, il;

i = i ;, 21987;
printf(-i after first XOR: %d\n", i);

i = 21987;
i A

printf{"i after second XOR: %d\n", il;

return 0;
)

2. The following program uses a bitwise AND to display, in binary,

the ASCII value of a character typed at the keyboard:
#include <stdio.h>
#include <conio.h>
380 TEACH YOURSELF
c

int main (void)

(
char chi
lnt i;

printf(MEnter a character: ");

ch = getche();
printf{"\n") ;

j* display bin~ry representation *;

for(i=128; i>oi .i=i/2)
if(l & chI printf("l ");
else printf("O ");

return 0;
)

'f.he program works by adjusting the value of i so that only one

bit is set each time a comparison is made. Since the high-order
bit in a byte represents 12B. this value is used as a starting point.
Each time through the loop. i is halved. This causes the next
bit position to be set and all others cleared. Thus. each time
through the loop. a bit in eh is tested. If it is I. the comparison
produces a true result and a I is output. Otherwise a 0 is
displayed. This process continues until all bits have been tested.
3. By modifying the program from Example 2. it can be used to
show the effect of the 1's complement operator.
#include <stdio.h>
#include <conio.h>

int main (void)

(
char Chi
int i;

ch = 'a';

/*. ?isplay binary representation ~ I

for!i=128; i>O; i=i/2)
ifli & ch) printfl"l ");
else printf("Q ");

j * reverse bit patt~rn */

 ADVANCED DATA TYPES AND OPERATORS 361
, 1.5 USE C'S BITWISE OPERATORS

ch = -Chi
printf("\n") ; •

/ * display binary representation */

for(i=128: i>Oi i=i/2)
if(i & chI printf("l .);
else printf{"O .);

return 0;
}

When you run this program, you will see that the state of bits in
ch are reversed after the operation has occurred.
N

4. The following program shows how to use the [5 operator to

determine if a signed integer is positive or negative. (The
program assumes short integers are 16 bits long.) Since negative
numbers are represented with their high-order bit set, the
comparison will be true only if i is negative. (The value 32768 is
the value of an unsigned short integer when only its high-order
bit is set. This value is \000 0000 in binary.)
#include <stdio.h>

iot main (void)

(
short i; •
printf("Enter a number: "I;
scanf("%hd", &i);

if(i & 32768) printf("Number is negative.\n M ) :

return 0;
}

5. The following program makes i into a negative number

by setting its high-order bit. (Again, 16-bit short integers
are assumed.)
#include <stdio.h>

int rnain(void)
(
short i;
.'
.'
362 TEACH YOURSElF
.,. C

i :: 1;
i=iI32768;
printf("%hd", il;

return 0;
)

It displays - 32 ,767.

EXERCISES

1. One very easy way to encode a file is to reverse the state of each
bit using the N operator. Write a program that encodes a file
using this method. (To decode the file, simply run the program
a second time.) Have the user specify the name of the file on
_ -the command line.
2. A better method of coding a file uses the XOR operation
combmed wah a user-defined key. Write a program that
encodes a file using this method. Have the user speCJfy the file
to code as "'ell as a single character key on the command line.
(To decode the me, run the program a second time using the
Silme key.)

• 3. What is the outcome of these operations?

A 1010 001 J [< OIOJ J JOJ
B OIOJ 1I0J I J Il J lOll
C. OIOJ 01 JO A IOJO 101 J
4. Sometimes, the high-order bit of a byte is used as a panty bIt by
modem programs. It is used to verify the integrity of each byte
transferred. There are two types of parity: even and odd. If even
parity is used, the parity bit is used to ensure that each byte has
an even number of 1 bits. If odd parity is used, the parity bit is
used to ensure that each byte has an odd number of 1 bits. Since
the parity bit is not part of the information being transferred,
show how you can clear the high-order bit of a character value
 ADVANCED DATA TYPES AND OPERATORS 363
..,
11.6 MASTER THE SHIfT OPERA TORS

ST E R THE SHIFT OPERATORS

C includes two operators not commonly found in other computer

languages: the left and right bit-shift operators. The left shift operator
is «, and the right shift operator is ». These operators may
be applied only to character or integer operands. They take these
general forms:

value« number-ot-bits
value» number-ot-bits
The integer expression specified by """,ber-oJ-bits determines how
many places to the left or right the bits within value are shifted. Each
left-shift causes all bits within the specified value to be shifted left one
position and a zero is brought in on the right. A right-shift shifts all bits
to the right one position and brings a zero in on the left. (Unless the
number is negative, in which case a one is brought in.) When bits are
shifted off an end. they are lost.
A right shift is equivalent to di\'idmg a number by 2, and a left shift
is the same as multiplying the number by 2. Because of the internal
operation of virtually all CPUs, shift operations are usually faster than
their equivalent arithmetic operations.

EXAMPLES

1. This program demonstrates the right anq left shift operators:

#include <stdio.h>

void show_binary(unsigned ul ;

int main(void)
{
unsigned short u;

u = 45678;

show_binary (ul ;
u =U « 1;
show_binary (ul ;
u = U » 1;
show_binary (ul ;
 return 0;
}

void show_binary(unsigned u}
(
unsigned n;

for (n=32768, n>O, n=n/21

if(u & nl printf("l "I,
else printfC·O .);

pri.ntf("\n"I,
}

The output from this program is

101100100110tll0
01 1.001 001 1 011 1 00
0011001001101110
Notice that after the leti shift. a bit of information has been lost.
When the right shift occurs, a zero is brought in. As stated
"arlier, bits that are shiftl'd off one end are lost.
2. Since a right shift is the same as a division by two, but faster,
the sbow_binary( ) function can be made more efficient as
shown here:
void sh~_binary(unsigned u)
(
unsigned n;

for(n=32768: n: n=n»l)
if(u & nl printf("l "I,
else printf"(" 0 "I,

printf("\n"l;
}
 ADVANCED DATA TYPES AND OPERATORS 365
117 UNDERSTAND THE? OPERATOR
"

ElWICISES
1. Write a program that uses the shift operators to multiply and
divide an integer. Have the user enter the initial value. Display
the result of each operation.
2. C does not have a rotate operator. A rotate is similar to a shift,
except that the bit shifted off one end is inserted onto the other.
For example, 1010 0000 rotated left one place is 0100 0001.
Write a function called rotate( ) that rotates a byte left one
position each time it is called. (Hint, you will need to use a
union so that you can have access to the bit shifted off the end
of the byte.) Demonstrate the function in a program.

11.7 UNPERSTAND THE? OPERATOR

( C contains one ternary operator: the ?'. A ternary operator requires

three operands . The? operator is used to replace statements such as:
if (co nd it ion ) var = expli
el s e ~ ar = exp2;

The general form of the? operator is

var = condition? exp I: exp2 ..

Here, emu/awH lS an expression that evaluates to true or false. If it is

true, liar is assigned the ,'alue of c.vpl . If it is false , ,'W ' is aSSigned the
value of exp2't!he reason for the? operator is that a C compiler
can produce more efficient code using it instead of the equivalent
if/else statement)

EXAMPLES

1. The following I?rogram illustrates the? operator. It inputs a

number and then converts the number into 1 if the number is
positive and -1 if it is negative.
 .,
366 TEACH YOURSELF
C

#include <stdio.h>

int main(void)
(
int i;

printf("Enter a number: ");

scanf (" %d", &i) i

i=i>0?1:-1;

printf{"Outcome: %d M
, i);

return 0;

2.(The next program is a computerized coin toss. It waits for you to

press a key and then prints either Heads or Tails.
#include <stdio.h>
#include <stdlib.h>
tiinclude <conio.h>

int main(void)

int i;

while(!kbhit(» rand();

i = rand() %2 ? 1: 0;

it (i) printf{"Heads");
else printf ("Tails");

return 0;

The coin-toss program can be written in a more efficient

way. There is no technical reason that the? operator need
assign its value to any variable. Therefore, the coin toss program
can be written as:
#include <stdio.h>
#include <stdlib.h>
#include'<conio.h>
 ".8 DO MORE WITH THE ASSIGNMENT OPERATOR

int main(void)
(
while(!kbhit(» rand();

rand()%2 ? printf(-Heads") printf(-Tails") ;

return 0;
)

Remember, since a call to a function is a valid C expression, it is

perfectly valid to call printf( ) in the? statement)

1. One particularly good use for the? operator is to provide a

means of preventing a division-by-zero error. Write a program
that inputs two integers from the user and displays the result of
dividing the first by the second. Use? to avoid division by zero.
2. Convert the following statement into its equivalent? statement.
if(a>b) count = laO;
else count = 0;

o MORE WITH THE ASSIGNMENT

OPERATOR
The assignment operator is more powerful in C than in most other
computer languages. In this section, you will learn some new things
about it.
(You can assign several variables the same value using the
general form

var I = var2 = var3 = ... = yarN = value: )

For example, this statement

i = j = k = 100;
388 WICII YOURSW
c

assigns i, j, and k the value 100. In professionally written C code, it is

common to see such multiple-variable assignments.
Another variation on the assignment statement is sometimes called
C shorthand. In C, you can transform a statement like
a =a + 3;

into a statement like

a += 3;

In general, any time you have a statement of the form

var = var op expression;

you can write it in shorthand form as

var op = expression;

Here, op is one of the following operators.

+ - • / % « » & I
There must be no space between the operator and the equal sign. The
reason you will want to use the shorthand form is not that it saves you
a little typing effort, but because the C compiler can create more
efficient executable code.

EXAMPLES

1. The following program illustrates the multiple-assignment

statement:
#include <stdio.h>

int main(voidl
(
int i. j, k;

i = j =k = 99;

printf("%d %d %d". i. j. kl;

return 0;
)
2. The next program counts to 98 by twos. Notice that it uses C
shorthand to increment the loop-control variable by two each
iteration.
#include <stdio.h>

int main(void)
(
int i:

/* count by 25 */
for(i=O: i<100: i+=2)
printf(-%d -, i);

return 0;
}

3. The following program uses the left-shift operator in shorthand

form to multiply the value ofi by 2, three times. (The resulting
value is 8.)
#include <stdio.h>

int main(void)
(
int i = 1;

i «= 3; /* multiply by 2, 3 times *j

printf("%d", i);

return 0;
}

1. Compile and run the program in Example 1 to prove to yourself

that the multiple-assignment statement works.
2. How is the following statement written using C shorthand?
x = x & y;
370 TEACH YOURSnf
c

3. Write a program that displays all the even multiples of 17 from

17 to 1000. Use C shorthand.

11.9 UNDERSTA~D THE COMMA OPERATOR

The last operator we will examine is the comma. It has a very unique
function: it tells the compiler to "do this and this and this: That is,
the comma is used to string together several operations. The most
common use of the comma is in the for loop. In the following loop,
the comma is used in the initialization portion to initialize two
loop-control variables, and in the increment portion to increment
i and j.
for(i=Q, j=O: i+j<count: i++, j++) ...

The value of a comma-separated list of expressions is the rightmost

expression. For example, the following statement assigns 100 to value:
value = (coi..lnt, 99, 33, 100):

The parentheses are necessary because the comma operator is lower

in precedence than the assignment operator.

EXAMPLES

I. This program displays the numbers a through 49. It uses the

comma operator to maintain two loop-control variables.
#include <stdio.h>

int main(void)
(
int i, j;

/* count to 49 */
for(i=O, j=100: i<j; i++. j--)
printfl"%d", ii,

return 0;
)
 ADVANCED DATA TVl'£S AND OPERATORS 371
/1.9 UNDERSTAND THE COMMA OPERATOR

2. In many places in C, it is actually syntactically correct to use the

comma in place of the semicolon. For example, examine the
following short program:
iinclude <stdio.h>

int main (voidl

{
char chi

ch = getchar(), /* notice the comma here *1

putchar (ch+l) ;

return 0;
}

Because the comma tells the compiler to 'do this and this: the
program runs the same with the comma after getchar( ) as it
would had a semicolon been used. Using a comma in this way is
considered extremely bad form, however. It is possible that an
unwanted side effect could occur. (This use of the comma
operator does make interesting coffee-break conversation,
however! Many C programmers are not aware of this interesting
twist in the C syntax.)

EXERCISES

I . Write a program that uses the comma operator to maintain

three for loop-control variables. Have one variable run from 0
to 99, the second run from -50 to 49, and have the third set to
the sum of the first two, both initially and each time the loop
iterates. Have the loop stop when the first variable reaches 100.
Have the program display the value of the third variable each
time the loop repeats.
2. What is the value of i after the following statement executes?
i=(l,2,3);
372 1EACH YOURSElF
c •

NOW THE PRECEDENCE SUMMARY

The following table shows the precedence of all the C ~perators.

Highest . ( [ 1 ->

•
- + - ++ (type cast) • & sizeof
/ %
+ -
« »
< <= > >=
-- .-
1-

&

&&
II
?:
; += -= *= /= etc.
Lowest

At this point you should be able to answer these questions and

perform these exercises:

J. What does the register specifier do?

2. What do the const and volatile modifiers do?
3. Write a program that sums the numbers J to J 00. Make the
program execute as fast as possible.
4. Is this statement valid? If so, what does it do?
typedef long double bigfloat;

S. Write a program that inputs two characters and compares

corresponding bits. Have the program display the number of
each bit in which a match Occurs. For example, if the two
integers are
 ,I. ro KNOW THE PRECEDENCE SUMMARY

1001 0110
11101010

tre program will report that bits 7, 1, and 0 match. (Use the
bitwise operators to solve this problem.)
6. What do the « and » operators do?
7. Show how this statement can be rewritten:
c=c+l0i

8.· Rewrite this statement using the? operator:

if(!done) count = 100;
else count = 0;

9. What is an enumeration? Show an example that enumerates

the planets.

This section checks how well you have integrated the material in
this chapter with that from earlier chapters.

1. Write a program that swaps the low-order four bits of a byte

with the high-order four bits. Demonstrate that your routine
works by displaying the contents of the byte before and after,
using the show_binary( ) function developed earlier.
(Change show_hinary( ) so that it works on an eight-bit
quantity, however.)
2. Earlier you wrote a program that encoded files using the 1's
complement operator. Write a program that reads a text file
encoded using this method and displays its decoded contents.
Leave the. actual file encoded, however.
3. Is this fragment correct?
register FILE *fPi

4. Using the program you developed for Chapter 10, Section 10.3,
ElIerCise 1, optimize the program by selecting appropriate local
variables to become register types.
 12
The ·C Preprocessor
and Some Advanced
Topics

ell ..... obIecd . .

12.1 Learn more about .define and .include

12.2 Understand conditional compilation

12.3 Learn about ."""', .uncle!, .Iine, and
.pragma
12.4 Examine C's built-in macros
12.5 Use the. and •• operators

c 12.6 Understand function pointers

12.7 Master dynamic allocation

375
....
376 TIACH YOURSELF
c

ONG RATU LAT! ON S! If you have worked your way

through all the preceding chapters, you can definitely call
yourself a C programmer. This chapter examines three
topics: the C preprocessor, pointers to functions, and C's
dynamic allocation system. All of the features discussed in
this chapter are important, and you need to be aware of their existence.
However, you won't use many of them right away. This is not because
any of the features discussed in this chapter are particularly difficult,
but because some features are more applicable to large programming
efforts and the management of sophisticated systems. As your
proficiency in C increases, however, you will find these features
quite valuable.

c· Before proceeding you should be able to answer these questions and

perform these exercises:

I. What is the major advantage gained when a variable is declared

using register?
2. What is wrong with this function?
void myfunc(const int *i)
{
*i = *i / 2;
)

3. What is the outcome of these operations?

a. 11011101 & 1110 0110
b. 1101110111110 OllD
c. 11011101" 1110 0110
4. Write a program that uses the left and right shift operators to
double and halve a number entered by the user.
5. How can these statements be written differently?
a = 1;
b = 1·,
c = 1;
 THE C PREPROCESSOR AND SOME ADVANCED TOPICS 377
12.1 /.EARN MORE ABout fdllflne A N D _
"
if(a<b) max; 100;
else max = 0;

i = i * 2;

6. What is the extern type specifier for?

EARN MORE ABOUT #define AND

#include

Although you have been using #define and #inc1ude for some time,
both have more features than you've read about so far. Each is
examined here in detail.
In addition to using #define to define a macro name .that will be
substituted by the character sequence associated with that macro, you
can use #define to create [zmctio>1-1ike macros. In a function-like
,.....
~,.j,)d...
macro, arguments can be passed to the macro when it is expanded by
the preprocessor. For example, consider this program:
#include <stdio.h>
,,; ,,'
,,",,,,,

#define smnL j) i+j

int main{void)
(
int sum;

sum; SUM(lO, 20);

printf{-%d-, sum);

return 0;
}

The line

10 sum; SUM(lO, 20);

is transformed into
sum = 10+20;
by the preprocessor. As you can see, the values 10 and 20 are
automatically substituted for the parameters i and j.
378 TfACH YOURSELF
y
C

A more practical example is RANGE( ), illustrated in the following

simple program. It is used to confirm that parameter i is within the
range specified by parameters min and max. You can imagine how
useful a macro like RANGE( ) can be in programs that must perform
several range checks. This program uses it to display random numbers
between 1 and 100.
#include <stdio.h>
#include <conio.h>
#include <stdlib.h>

#deflne RANGE(i, min, max) (i<min) 11 (i>max) ? 1 0

int main{void)
(
int r;

/* print random numbers between 1 and 100 */

do (
do (
r = rand();
) while.(RANGE(r, 1, 100));
printf (~%d ", r);
) while ( !kbhitll);

return 0;
)

The advantage to using function-like macros instead of functions is

that in-line code is generated by the macro, thus avoiding the time it
takes to call and return from a function. Of course, only relatively
simple operations can be made into function-like macros. Also,
because code is duplicated, the resulting program might be longer than
it would be if a function were used.
The #include directive has these two general forms:

#include <filename>
#include "filename"

So far, all the example programs have used the first form. TI1e
reason for this will become apparent after you read the following
descriptions.
 THE C PREPROCESSOR AND SOME ADVANCED TOPICS 379
12.' LEARN MORE ABOLR fdeflne AND *include

If you specif'y the file name between angle brackets, you are
instructing the compiler to search for the file in some implementation-
•
defined manner. For most compilers, this means searching a special
directory devoted to the standard header files. This is why the sample
programs have been using this form to include the header files
required by the standard library functions. If you enclose the file
name between quotation marks, the compiler searches for the file in
another implementation-defined manner. If that search fails, the
search is restarted as if you had specified the file name between angle
brackets. For the majority of compilers, enclosing the name between
quotation marks causes the current working directory to be searched
first. Typically, you will use quotation marks to include header files
that you create.

EXAMPLES

J. Here is a program that uses the function-like macro MAX( ) to

compute which argument is larger. Pay close attention to the
last printf( ) statement.
#include <stdio.h>

#define MAX{i , j) i>j ? i j

int main( vo i d)
{
l
printf("%d\n", MAX(l, 2));
printf("%d\n", MAX(l, -1));

/ * this statement does not work correctly */

printf("%d \ n", MAX(100 && -1, a));

return 0;
)

When the preprocessor expands the final printf( ) statement,

the MAX( ) macro is transformed into this expression:
100 && -1 > a ? 100 && -1 , a
380
., TtACH YOURSElf
c

Because of C's precedence rules, however, this expression is

executed as if parentheses had been added like this:
100 && (-1 > 0) ? 100 && -1 , 0

As you can see, this causes the wrong answer to be computed.

To fIx this problem, the macro needs to be rewritten as:
#de fine MAX (i. j) « i ) > (j » ? (i) , (j )

Now the macro works in all possible situations. In general,

you will need to fully parenthesize all parameters to a
function-like macro.

The RANGE() macro discussed earlier will need similar

parenthesization as well if it is to work in all possible
situations. This is left as an exercise.

2. The next program uses quotes in the #include directive.

#include "stdio.h"

int main(void)
(
printf("This is a test~);

return 0;
}

While not as efficient as using the angle brackets, the #include

statement will still find and include the STDlO.H header fIle.
3. It is permissible to use both forms of the #include directive in
the same program. For example,
#include <stdio.h>
#include ~stdlib.h~

int rnain(voidl
(
printf("This is a random number: %d-, rand();

return 0;
}
 381
TIll C PAEPROClSSORANO SOME ADVIIHCEO TOPICS
122 UNDERSTIWOCONDmONAI. COMPILAnON
..,

1. Correct the RANGE( ) macro by adding parentheses in the

proper locations.
-... ;
'r- .• 2. Write a program that uses a parameterized macro to compute
the absolute value of an integer. and demonstrate its use in a
program.
3. Compile Example 2. If your compiler does not find STOIO.H.
recheck the installation instructions that came with your
compiler.

12.2 UNDERSTAND CONDITIONAL

COMPILA TlON

The C preprocessor includes several directives that allow parts of the

source code of a progra,m to be selectively compiled. This is called
cOl1ditiona/ compi/aHol1. These directives are

#if
#else
#elif
#endif
#ifdef
#ifndef
-.-, This section examines these directives.
'.'
The general form of #if is shown here:

#if constant-expression
statement-sequence
#endif

If the value of the constant-expression is true. the statement or

statements between #if and #endif are compiled. If the

-
382 TEACH YOURSW

constant-expression is false, the compiler skips the statement or

statements. Keep in mind that the preprocessing stage is the
first stage of compilation, so the constant-expression means exactly that.
No variables may be used.
You can use the #else to form an alternative to the #if. Its general
form is shown here:

#if constant-expression
statement-sequence
#else
statement-sequence
#endif
Notice that there is only one #endif. The #else automatically
terminates the #ifblock of statements. If the constant-expression is
false, the statement or statements associated with the #else are
compiled.
You can create an if-else-if ladder using the #elif directive,
as shown here:

# if constant-expression-!
statement-sequence
#elif constant-expression-2
statement-sequence
#elif constant-expression-3
statement-sequence

#endif

As soon as the first expression is true, the lines of code associated with
that expression are compiled, and the rest of the code is,skipped.
Another approach to conditional compilation is the #ifdef directive.
It has this general form:

#ifdef macro-name
statement-sequence
#endif

If the macro-name is currently defined, then the statement-sequence

associated with the #ifdcf directive will be compiled. Otherwise, it is
 TIlE C PAlPIIOCESSOR AND SOME ADVANCED TOPICS 383
y
'21 UNDERSTAND CONDmONAL COMP/IJ'.TION

skipped. The #else may also be used with #ifdef to provide an

alternative.
The complement of #ifdef is #ifndef. It has the same general form
as #ifdef. The only difference is that the state.ment sequence
.,-.' associated with an #ifndef directive is compiled only if the .
macro-name is not defined.
In addition to #ifdef, there is a second way to determine if a macro
name is defined. You can use the #if directive in conjunction with
the defined compile-time operator. The defined operator has this
general form:

defined macro-name
If macro-name is defined, then the outcome is true. Otherwise, it is
I false. FOT example, the following two preprocessor directives are
equivalent:
# ifdef WIN) 2
#if defined WIN32

You can also apply the I .operator to defined to reverse the condition.

EXAMPLES

1. Sometimes you will want a program's behavior to depend on

a value ciefined within the program. Although examples that
are both short and meaningful are hard to find, the following
program gives the flavor of it. This program can be compiled
to display either the ASCII character set by itself, or the full
I
extended set, depending on the value of CHAR_SET. As you
know, the ASCII character set defines characters for the values 0
'f
through 127. However, most computers reserve the values 128
through 255 for foreign-language characters and mathematical
and other special symbols. (You might want to try this program
with CHAR_SET set to 256. You will see some very interesting
characters')
#include <stdio.h>

/* define CHAR_SET as either 256 or 128 */

 'define CHAR-SET 256

int main (void)

I
int i;
.if CHAR-SET ==256
printf I ·Displaying ASCII character set plus extensions. \n· ) ;
'else
printfl·Displaying only ASCII character set.\n·):
.endif

for(i=O: i<CHAR-SET: i++)

printf(-'c·, i);

return 0:
I

2. A good use of jJjfclef is for imbedding debugging information

into your programs. For example, here is a program that copies
the contents of one file into another:
/* Copy a file. */
'include <stdio.h>
'include <stdlib.h>

'define DEBUG

int main(int argc, char ""argv[])

I
FILE ·from. ""to:
char Chi

'*iflargc!=3)
see if correct number of command line arguments
I
*/

printf(-Usage: copy <source> <destination>\n-);

exitt1):
I

'* open source file *'

if«from = fopen(argv[11. ·rb·»==NULL) I
printf(·Cannot open source fi1e.\n:):
exit(1):
I
 _ C PREPIIOCU8OO MIl SOM£ IIIlVANCm TOPICS 385
'21 UNDERSTAND CONDfTIONAi. COMP/U4T1ON •

/*open destination file */

i f I Ito = fopen largv[2], "wb"» ==NULL) {
printE (~Cannot open destination file. \n");
exit(!);
)

/* copy the file */

while ( ! feaf (from» (
ch = fgetc (from) ;
if (ferror (from» (
princf("Error reading source file.\n");
exit(!);
)
if(!feof(from» {
."
.., fpute (ch, to);
#ifdef DEBUG
putchar (ch);
#endif
)
if(ferror(to»)- {
printE ("Error writing destination file. \0") ;
exit(!):
)
)
fclose (trom) ;
fclose(to) ;

return 0;
,,

If DEBUG is defined, the program displays each byte as it is

"
transferred. This can be helpful during the development phase.
Once the program is finished, the statement defining DEBUG is
removed, and the output is not displayed. However, if the
program ever misbehaves in the future, DEBUG can be defin~d
again, and output will again be shown on the screen. While this
might seem like a lot of work for such a simple program, in
actual practice programs may have many debugging statements,
and this procedure can greatly facilitate the development and
testing cycle.
386
.,. T~ACH YOURSElf

As shown in this program, to simply define a macro name,

you do not have /0 associate any character sequence with it

3. Continuing with the debugging th~me, it is possible to use the

Hif to allow several levels of debugging code to be easily managed.
For·example, here is one of the encryption programs from the
answers to Chapter II that supports three debugging levels:
#include <stdio.h>
#include <std!ib.h>

/ * DEBUG leve ls:

0: no debug
1. d".splay byte read flom source file
2. d~.s play byte wr~tten to destination file
3: display bytr::s read and hyt.es written
0/
ijdefin-::) DEBUG ::

int mainline arqc, char *argv[] I

(
FILE *in, "o~r:;
unsigned char chi

j* see if correct number of command line arguments */

if(argc!.::4) [
printf (·Usage: code <in> <out> <key> ~) ;
exit (1);
)

I~ open input file */

if{(in = fopen{arg'l[l] "rb"»==NULL) {
printf("Cannot open ~pput file.\n")i
exit(!);
)

/w open output file */

if«out = fopen(argv[2}, "wh") )==NULL) (
printf("Cannot open output file.\n~ ) ;
exit(l);
)

while(!feof(in» {
 THE C PREPROCESSOR AND SOME ADVANCED TOPICS 387
12,2 UNDERSTAND CONDITIONAL COMPILAnON •
ch = fgetc (in) ;
DEBUG == 1 II DEBUG == 3
putchar feb) ;
Itendif
ch = *argv(31 A chi
/tif DEBUG >:: 2
'IO~fr, putchar (eh) ;
illy ntaIii1 #endif
miroml, if(!feof(in» fputc(ch, out);
,glelll )

fclose (in) ;
fclose(outl;

return 0;
)

••
..-."" ...:
;
4. The following fragment illustrates the "elif. It displays NUM IS
.:::; 2 on the screen .
#define NUM 2

tiE NOM == 1
printf("NUM is 1");
#ellf NUM == 2
printf ("NUM is 2");
_€~'.~ t

#elif NOM == 3
printf{"NUM is )N);
#eliE NOM == 4
printf(~NUM is 4");
#endif
•
5. Here, the defined operator is used to determine if
TESTPROJECT is defined .
'include <stdio.h>

#define TESTPROJECT 29

iif defined TESTPROJECT

int main (void)
(
printf(~This is a test.\n");
388 ltACH YOUAS£Lf

" C

I'"eturn .0 :- __
}
#end i f

EXERCISES

1. Write a program that defines three macros called INT, FLOAT,

and PWR_TYPE. Define INT as 0, FLOAT as J , and
PWR_TYPE as either INT or FLOAT. Have the program
request two numbers from the user and display the result of
the first number raised to the second number. Using #if and
depending upon the value of PWR_TYPE, have both numbers
be integers, or allow the first number to be a double.
2. Is this fragment correct' If not , show one ,;'ay to fix it .
#define MIKE

#ifdef !MIKE

#endif

EARN ABOUT #error, #undef, #Iine,

AND #pragma

C's preprocessor supports four special-use directives: #error, #undef,

#line, and #pragma. Each will be examined in tum here.
The #error directive has this general form:

#error effor-message
 _ C PIIEPROCUSOR AND SOME IIDVANCED TOPICS 381
12.3 LEARN ABOUT terror, fundef. tHne. AND #pragma
"

It causes the 'compiler to stop compilation and issue the error-message

along with other implementation-specific information, which will
generally include the number of the line the #error directive is in and
the name of the file. Note that the error-message is not enclosed
between quotes . The principal use of the #error directive is in
debugging.
The #undef directive undefines a macro name. Its general form is

#undef macro-name
r- If the macro-name is currently undefined, #undef has no effect. The
fIC·
principal use for #undef is to localize macro names.
When a C compiler compiles a source file , it maintains two pieces of
" information: the number of the line currently being compiled and the
l.
name of the source file currently being compiled. The #line directive
is used to change these values. Its general form is

#Iine line-num "filename"

Here , line-nUll1 becomes the number of the next line of source code ,
and filename becomes the name the compiler will associate with the
source file. The value ofline-num must be between 1 and 32,767. The
filename may be a string consisting of any valid file name. The principal
use for #line is for debugging and for managing large projects.
The #pragma directive allows a compiler's implementor to define
other preprocessing instructions to be given to the compiler. It has this
general form:

- #pragma instructions

If a compiler encounters a #pragma statement that it does not

recognize, it ignores it. Whether your compiler supports any

-
lill #pragmas depends on how your compiler was implemented.

EXAMPLES

1. This program demonstrates the #error directive .

#include <stdio.h>
380--
" c

int main (void)

(
int i:

i :Ie 10 ;
.error This is an error message.
printf( - 'd-, i, : /* this line will not be compiled * 1

return 0;
)

As soon as the #error directive is encountered, compilation

stops.
2. The next program demonstrates the #undef directive. As the
.,program states, only the first printf( ) statement is compiled.
'include <stdio.h>

Idefine DOG

int main (void)

(
.ifdef DOG
printf("DOG is defined . \n");
.endif

.undef cpo
'ifdef DOG
printfC MThis line is not compiled . 'n-);
.endif

return 0:
}

3. The following program demonstrates the #1ine directive. Since

virtually all implementations of #error display the line number
and name of the file, it is used here to verify that -uae did, in
fact, perform its function correctly. (In the next section, you will
see how a C program can directly access the line number and
file name).
'include <stdio.h>

int main (void)

 THE C PREPROCESSOR AND SOME ADVANCED TOPICS 391
12.4 EXAMINE C'S BUlL T·IN MACROS •
(
int i;

/* reset line number to 1000 and file name to

myprog.c

"'
*line 1000 -myprog.c·
#error Check the line number and fi~e name.

return 0;
)

4. Although the ANSI C standard does not specifY any #pragma

directives, on your own check your compiler's user manual and
learn about any supported by your system.

EXERCISE

J. Try the example programs. See how these directives work on

your system.

XAMINE C'S BUILT-IN MACROS

If your C compiler complies with the ANSI C standard. it will have

at least five predefined macro names that your program may lise.
They are
LINE__
FllE__
__ DATE__
__TIME__
__STDC__

Each of these is explained here.

392
., TEACH YOURSElF
c

The __ LINE__ macro defines an integer value that is equivalent to

the line number of the source line currently being compiled.
The __ FILE__ macro defines a string that is the name of the file
currently being compiled,
The __ DATE__ macro defines a string that holds the current
system date, The string has this general form:

month/day/year

The __TIME__ macro defin~s a string that contains the tim~ the
compilatlOn of a program began. The string has this general form:

hours:minutes:seconds

The __STDC__ macro is defined as the value 1 if the compiler

conforms to the ANSI standard.

'EXAMPLES

J. This program demonstrates the macros __ LINE __ . __ FILE__ .

__ DATE__ . and _ -'rIME__ ,
#include <stdio.h>

int main(void)
(
printf(PCompil i ng %5, line: %d, o n %5, at ~s·,
__ FILE __ • __ LINE __ • __ DATE__ •
__ TIME __ I;

return 0;
)

It is important to understand that the values of the macros are

fixed at compile time. For example. if the above program is
called T.C. and it is compiled on March 18. 1997. at 10 A.M..
it will always display this output no matter when the
program 1S run,

Compiling T.C, line: 6. on Mar ~8 1997. at 10:00:00

 _ C PREPIIOCESSORAND SOME _ _ TOPICS 393
125 USE THE f AND f. OPfAATORS
"

The main use of these macros is to create a time and date stamp,
which shows when the program was compiled.
2. As you learned in the previous section, you can use the #line
directive to change the number of the current line of source
code and the name of the file. When you do this, you are
actually changing the values of __LlNE__ and __FILE__ . For
example, this program sets __ LlNE__ to 100 and _ "':FILE__ to
myprog.c:
#include <st~io_h>

int main/void)
(
#line 100 ~myprog.c~
printf(-Compiling %s, line, %d, on %5, at %5",
__ FILE __ • __ LINE __ . __ DATE __ •
___ TIME __ I:

return 0;
)

The program displays the following output, assuming it was

[ compiled on March 18, 1997, at 10 AM.

Compiling myprog.c. line: 101. on Mar 18 199i at 10:00:00

1. Compile and run the example programs.

12.5 USE THE # AND ## OPERATORS

The C preprocessor contains two little-used but potentially valuable
operators: # and ##. The # operator turns the argument of a
 ~-===

..,
3M

function-like macro into a quoted string. The ## operator concatenates

two identifiers.

EXAMPLES

1. This program demonstrates the # operator.

#include <stdio.h>

#define MKSTRING(str) # str

int main(vqid)
(
int value:

value = 10;

printf(-%s is %d", MKSTRING(value), value);

return 0;
)

The program displays value is 10. This output occurs because

MKSTRING( ) causes the identifier value to be made into a
quoted string.
2. The following program demonstrates the ## operator It creates
the output( ) macro, which translates into a call to printf( ).
The value of two
, variables, which end in 1 or 2, is displayed.
'include <stdio.h>

'define output(i) printf("'d %d\n", i #t 1, i It 2)

int main (void)

{
int countl, count2;
int i1. i2;

count! = 10;
count2 = 20;
i l = 99;
i2 = -10;
 126 UNDEIISTIWD FUNCT10N POIN1fRS

output(count);
output (i) ;

return 0:
)

The program displays 10 20 99 -10. In the calls to OUtput( ),

count and i are concatenated with 1 and 2 to form the variable
names countl , rounU, i1 and i2 in the printf( ) statements.

1. Compile and run the example programs.

2. What does this program display?
.include <stdio.h>

#define JOIN(a, b) a •• b

int main(void)
{
pr i ntf (J'OIN( • one • -two·»;

return 0;
)

3. On your own , experiment with the # and ## operators. Try to

think of ways they can be useful to you in your own
programming projects.

I!

12.6 UNDERSTAND FUNCTION POINTERS

This section introduces one of C's most important advanced features:

the function pointer. Although it is beyond the scope of this book to
ofthei( value. Like the program in Example 1, this program
prompts the user for two numbers. Next, it asks the user to
enter the number of the operation to perform. This numbet is
then used to index the function-pointer array to execute the
proper funCtion. Finally, the result is displayed.
'include <stdio.h>

int sum(int a, int b) ;

int subtract (int a, int b) ;
int mul (int a, int b) ;
int div(int a, int b) ;

int (*p[4}) (int x. int y);

int main (void)

(
int result:
int i, j. op;

p [0 J = sum: f' get address of sum() 'f

p [1 J = subtract; f' get address of subtract (I 'f
p [2 J = mul, f' get address of mul I) 'f
p [3 J = div; f' Qet address of divl) 'f

printf{DEnter two numbers: ");

scanfl"%d%d" , &i, &j);
printf("O: Add, 1: Subtract, 2: Multiply, 3: Oivide\n-);
do I
printf(nEnter number of operation: .);
scanf(·'d-, &op);
) whilelop<O II op>3);

result = I'p[op») Ii, i);

printf(-'d-, result);

return 0;
)

int sum(int a, int b)

(
return a+b;
)

int subtract(int a, int bl

 TIIlC~AllDSOME-Mc:m- 388
126 UNOERSTNoIO RJNCTION POIN1fIIS •
{
i return a-b;
}

int mul{int a, int b)

{
return a*bi
}

int div(int a, lnt b)

(
if(b) return a l b:
else return 0:
)

When you study this code, it becomes clear that using a

function-pointer array to call the appropriate function is more
efficient than using a switch( ) statement.
Before leaving this example, we can use it to illustrate
one more point: function-pointer arrays can be initialized, just
like any other array. The following version of the program
shows this.
'include <stdio . h>

int sum (int a, int b) :

int subtract (int a , int b) :
int mul ( int a, int b) :
int div(int a, int b) :

/ * initialize the pointer array */

int ( 'p[4]) (int x, int y) = (.
- sum, subtract. mul , div
) :

int main (void)

(
int result;
int i. j, op;

printf(-Enter two numbers: .);

scanf(-'d%d·, &i. &j);
printf("O: Add, 1: Subtract, 2: Multiply, 3: Divide\n"),
do (
printf(-Enter number of operation: .);
400
., 1UCH yOUASW
c

scanf("%d", &Op);
) while(op<O II op>3 ) ;

resul t = ('p [op)) (i. j);

printf("%d~, result);

return 0;
)

int sum(int a, int b)

(
return a+b;
)

i nt subtract(int a, int b)
(
return a-b;
)

int mul(in L a, int b)

(
. return a·bi
)

int div(int a, int b)

(
iftb) return a l b:
else return 0;
)

3. One of the most common uses of a function pointer occurs

when utilizing another of C's standard library functions, qsort{ ).
The q80rt( ) function is a generic sort routine that can sort any
type of singly dimensioned array, using the Quicksort algorithm.
Its prototype is

void qsort(void 'array, size_t number, size_t size,

int (,comp)(const void '8. const void 'b));
Here, array is a pointe no the first element in the array to be
sorted. The number of elements in the array is specified by
number, and the size of each element of the array is specified by
 THE C PREPROCESSOR AND SOME ADVANCED TOPICS 401
~
12.6 UNDERSTAND FUNCnON POINTERS

size. (Remember, size_t is defined by the C compiler and is

loosely the same as unsigned.) The final parameter is a pointer
to a function (which you create) that compares twO elements of
the array and returns the following results:

·a <·b returns a negative value

·a ==·b returns a zero
·a >·b returns a positive value

The qsort( ) function has no return value. It uses the STDLlB.H

header file.
The following program loads a J DO-element integer· array
with random numbers, sorts it, and displays the sorted form.
Notice the necessary type casts within the comp( ) function.
#include <stdio.h>
;include <stdlib.h>

int comp(const void *i, const void *j);

int main(void)
(
int sort[lOO], i;

fOl"(j:=O; i<100; i+ .... )

sort[ij := randr);

qsort(sort, 100, sizeof(int), comp):

for i<100; i++)

(i=l);
printf{'"%d\n",sort[i]};

return 0;
)

int comp(const void canst void *j)

(
return *(int*)i - *iint*)j;
)
402 TEACH VOURSt.LF

'" C

EXlRCISfS

1. Compile and run all of the example programs. Experiment with

them, making 1111nor changes.
2. Another of C's standard library functions is called bsearc h ( ) .
This functlon searches a sorted a rray, gIven d key. It returns a
pointer to the first entry in the array thot matches the key If no
match is found, a null pointer is returned. Its prowtype is

void 'bsearch (consl VOid ' key, const void ' array, slze_t number. size_t Size,
Inl (, comp)(const void ' a, const void ' b));
•
All the pa rameters to b search( ) arc the same as for q sort ( )
except the first. \\'hieh is a pOInter to kelj, the object bc:ing
,,,ught. The comp( ) fun, tlon operates the same for bscarch( )
", it does for qsort( ) .
.\lnd;fy the prugraln In L... dlr'pl~ 3 so that anf"r the iiff<-ly is
sOIlPd the tl~er IS prompted to cntpr a number :\'C\t. using
hscarch( ). search the s()rH~d arr.JY and repan ji a nla~ch is
found
3. Add a function call~d modulus( ) to the tinal verSion of the
.nithmctic progranl In Example 2. Ha\'e the functIOn return the
result of a % h . Ada tll1S optIOn to the menu dnd fully integrate
it lnW the p rogcam.

MASTER DYNAMIC A L LOCA TION

this final section at the book mtroduces you to C's dynanlic-alioration
SY5tem. Dyl1mmc allocatIOn is the p;o(.cs~ by whIch memory is
allocated as needed during runtime. ThIS allocated memory can be
used fOT a \"aricry ot purposes l\.lost commonly, memory is allocJted
bvap!}l1 r l[1f'1r (1Ft np.('.i i-, ta' e fu:] advantage of all tile nlenHJry In
(h(" r l.TJ.! 'T( r r(n examplC a \"Old processOT \\·m want to let tht> user
edit docuJJ1l:nt~ that arc as large as pOSSible Ht}\'o.'c\'er, if the word
proC{~ss()r u:-,es a nOTmal character array. It must th its siu' at (.(JmpiJe
time. Thl1~, It ""ollld have [Q hp compiled to run in computers WIth the
m-inlJlHlnl 11l10U 'f nt memory not allowing u . . ers \,!1[h mUTe Jnf'!l1()JT
 THE C PREPROCESSOR AND SOME ADVAHCfO TOPICS 403
12.7 MASTER DYNAMIC AUOCA nON
...

to edit larger documents. If memory is allocated dynamically (as

needed umil memory is exhausted), however, any user may make full
use of the memorv in the system. Other uses for dvnamic
- " "
allocation
include linked lists and binary trees.
The core of C's dynamic-allocation functions are malloc( ), which
d~ociltes memory, and free( ), which releases previously allocated
·,nory. Their prototypes are

void 'malloc(size_1 numbytes);

void free(void •plr);

Here, nwnhytcs is the number of bytes ofnlemory you wish to

allocate. The malloc( ) function returns a pointer to the start of the
allocated piece of memory. If malloc( ) cannot fulfill the memory
request-for example, lhere may be insufficient memory availdble-lt
n'Wl ns a null pointer. To free memory, call frce( ) \\-irh a pointer to
tl'" start of the block of memor\' (previously allocated using malloc( ))
you "'ish to free. Both functions usc the header file STDLlB.H.
:-''''mor), is allocated from a region called the heap. Although the
,l( mal physical layout of memory may differ, conceptually the heap

lies hetv,"een your program and the stack. Since this is a finite area, an
illloc.1t1On request can fail when memory is exhausted.
When a program terminates, all allocated memory is automatically
released.

EXAMPLES

1. You must confirm that a call to malloc( ) is successful hetore

you use the pointer it returns. If you perform an operation on a
null pomter, you could crash your program and maybe even the
entire computer. The easiest way to check for a valid pointer is
shown in this fragment:
p = malloc(SIZE1;
if ( ! pl{
printft"Allocation Error");
404 TEACH YOURSELF
... C

exit(l) ;
)

2. The following program allocates 80 bytes and assigns a

character pointer to it. This creates a dynamic character array.
It then uses the allocated memory to input a stri~ using gets( ).
Finally, the string is redisplayed and the pointe,",s freed. (As
stated earlier, all memory is freed when the program ends, so
the call to free( ) is included in this program simply to
demonstrate its use.)
#include <stdio.h>
#include <stdlib.h>

int main(voidl
(
char *p;

p = malloe (80);

ifl!p) (
printf ("Allocation Failed");
exit{l);
)

printfl-Enter a string: ~);

gets(p) ;
printf (PI;
free(p) ;

ret.urn 0;
)

3. The next program tells you approximately hoI\' much free

memory is available to your program.
~inc:ude <stdio.h>
#include <st.dlib.h>

int ma::.n(·,fOid)
(
char .~.
~,

long 1;

1 = 0;
 THE C PREPROCESSOR AND SOME ADVANCED TOPICS 401
12. 7 MASTER DYNAMfC ALLOC4 nON
"
do (
p = rnalloc{lOOO);
iElp) 1 += 1000;
} while{p);
:~

printf("Approximately 'tld bytes of free memory.", 1);

return 0;
}

The program works by allocating IOOO-byte-iong chun ks of

memory until an allocation request fails. When malloc( )
returns null, the heap is exhausted . Hence, the value on
"cpresents (within 1000 bytes) the amount of free memory
available to the program
4. One good use for dvnamic allocation is to create buffe rs for file
I 0 when you are using fread( ) and , or fwritc( ) . Often, vou
only n""d a buffer for a short perioe! of time, so it makes sense
to allocate it whe n needed and free it \\'hen done. T he fo llowi ng
program shows how dynamic allocation can be used to crc<ltc
a butkr. The program allocates enough space to holcl te n
floating-point values. It then assigns ten ranclom n u mbe rs to
the allocated memory, indexing the pointer as a n array. Next,
it writes the values to disk and frees the memory. Fi nally, it
reallocates memory, reads the tile and displays the ranclom
numbers. Although there is no need to free and then reallocate
the memory that serves as a file buffer in this short exa m ple, it
illustrates the hasic idea.
#include <stdio.h>
#include <stdlib.h >

i nt main(vo id )
{

int i;
d ouble *p;
FILE '"fp;

/ '" get memory * /

p = malloctlO * sizeof(double));
iEl:p) {
printf("Allocation Error");
406 TEACH YOURSfLF
... C

exit(l) ;
}

/* generate 10 random numbers */

for(i=O; i<10; i++)
p[i] = (double) rand();

if«fp = fopen("myfile", "wb"))==NULL)

printf(MCannot open file.\n")i
exit(l);
}

/* write the entire array in one step */

if(fwrite(p, lO*sizeof(double) , I, fp) != 1) (
printf("Write Error.\n");
exit(l) ;
}
fclose (fp) i

free(p); /'" memory not needed now " I

I '

imagine something transpires here

'f

I~ get memory again */

p = malloc(lO .. sizeof(double»
if(!p) {
print f (M Allocat ion Error");
exit(1) ;

if«fp = fopen("myfile", "rb"))==NULL) (

printf("Cannot open file. \n");
exit(l) ;
}

/* read the entire array in one step */

if{fread(p, lO*sizeof(double) , 1,' fp) != 1)
printf("Read Error.\n");
 THE C PREPROCESSOR AND SOME ADVANCED TOPICS 407
127 MASTER DYNAMIC ALLOCAnON •

exit (1);

fclose (fp) ;

/* display the array */

•
for(i=O; i<10; i++) printf{ft%f" p[ii);
free (p) ;

return 0;

5. Just as array boundaries can he overrun, so can the boundaries

of allocated memory. For example, this fragment is syntactically
valid, but wrong.
p = malloc(lO);
for(i=O; i<100; i) plil = i;

EXERCISES

I. Compile and run the example programs.

2. Vlritc a program that creates a tenRelernent dynamic integer
array. Next, using pointer arithmetIc or array indexing, c1ssign
the values I through 10 to the integers that wmprisc the arra\'.
Finally, display the values and free the memory.
3. What's wrong \·" ith this fragment?
char "p;

'p = mallocllO);
gets (p) ;
 ,.
408 TEACH YOURSELF

At this point yoti should be able to answer these qttestions and

perform these exercises:

1. What is the difference between using quotes and angle brackets

\vith the #include directive?
2. Using an #ifdef, show how to conditionally compile this
fragment of code hased upon whether DEBUG is defined or not.
if l ! (j %2)) {
print.f("J = %d\n", J);
J = 0;
)

3 L'smg the tragmcm from ExerCIse 2, show how you can

conditionallv compile the cncie when DEBUG is defined as 1.
(Hint: Use #if).
-I. Hov,' do VOll undefine a macro?
5. \"lhat is __ FILE__ and what does it repre!:.f'nt:'
6. What do the # and ## preprocessor operators do:'
7. Write a prugr<lm that sorts the string "this is a test ofqsort".
Display the sorted output.
8. Write a program that dynamically allocates mem(lfY for one
double. Have the program assign that locanon the value ~9.01,
display the value, and then free the memory.

Skills Check

This section checks how well you have integrated the material in
this ch<lptcr with that from "<lrlier ch<lpters.

]. Section 10.1, Example 3, presents .1 computerized card-catalog

program tllat liSt'S an array \)f structures to hold infornlation on
books. Change this program so that only an array of structure
 THE C PREPROCUSOR AND SOME ADVANCEII TOPICS . . . .
~

pointers is created, and use dynamically allocated memory to

actually hold the information for each book as it is entered. This
way, less memory is used when information on only a few
books is stored.
2. Show the macro eqUivalent of this function;
char code_it(char c)
hraci.m (
return -Ci
)

Demonstrate
, that your macro version works in a program.
3. On your own, look over the programs that you have written in
the course of working through this book. Try to find places
where you can;
.. Use conditional compilation.
• Replace a short function with a f~tion.like macro.
.. Replace statically allocated arrays with dynamic arrays.
... Use function pointers.
4. On your own, study the user's manual or online documentation
for your C compiler, paying special attention to the description
of its standard library functions. The C standard library contains
several hundred library functions that can make your
programming tasks easier. Also, Appendix A in this book
discusses some of the most common library functions.
5. Now that you have finished this book, go back and skim through
each chapter, thinking about how each aspect of C relates to the
rest of it. As you will see, C is a highly integrated language, in
which one feature complements another. The connection
between pointers ahd arrays, for example, is pure elegance.
6. C is a language best learned by doing! Continue to write
programs in C and to study other programmers' programs. You
will be surprised at how quickly C will become second nature!
7. Finally, you now have the necessary foundation in C to a1\ow
you to move on to C++;C'sobject-orientedextension. If C++ .
programming is in your future, proceed to Teach Yoursel{C++,
11~
(Berkeley, CA, Osborne/McGraw-Hill). It picks up where this
book leaves off.
.-
A
Some Common C
Library Functions

411
'f'
412 1tACII YOURSElF
c

IS appendix discusses a number of the more frequently used

ANSI C library functions. If you have looked through the
library section in your C/ C++ compiler's documentation, you
are no doubt aware that there are a great many library
functions. It is far beyond the scope of this book to cover each
one. However, the ones you will most commonly need are discussed here.
The library functions can be grouped into the following categories:
... i/O functions
... String and character functions

... Mathematics functions

... Time and date functions
... Dynamic allocation functions
... Miscellaneous functions
The I/ O functions were thoroughly covered in Chapters 8 and 9 and
will not be expanded upon here. .
Each function's description begins with the header file required by
the function followed by its prototype. The prototype provides you
with a quick way of knowing what types of arguments and how many
of them the function takes and what type of value it returns.
Keep in mind that ANSI C specifies many data types, which are
defined in the header files used by the functions. New type names will
be discussed as they are introduced.

TRING AND CHA RA C TER FUNCTIONS

The C standard library has a rich and varied set of string- and
character-handling functions. In C, a string is a null-terminated array
of characters. The declarations for the string functions are found in th
header file STRlNG.H. The character functions use CTYPE.H as their
header file.
Because C has no bounds-checking on array operations, it is the
•
programmer's responsibility to prevent an array overflow
 _ _ cL_fUIIClIONI .13
AI STRJNG ANa aW!.ocmI FUNCTIONS
"
-......
len~Uli: The character functions are declared with an integer parameter.
~the While this is true, only the low-order byte is used by the function.
~~ Generally, you are free to use a character argument because it will
il.'j automatically be elevated to int at the time of the call.
ttru:\
lil'!dft! *Include <ctype.h>
~tfJ1ntt Int isalnum(i"t ch)i

Description The isalnum( ) function returns nonzero if its argument

is either a letter or a digit. If the character is not alphanumeric, then 0
is returned. .

Example This program checks each character read from stdin and
reports all alphanumeric ones:
.include <ctype.h>
.include <stdio.h>

int main(void)
{
char chi

for ( ; ;) (
ch = getchar()i
if(cn==' ') break;
ifCisalnumCch» printf(-%c is alphanumeric\n-, chI;
)

return OJ
)

*Include <ctype.h>
Int ISJtIpha(int ch)i

DaaiptIoq The isalpba( ) function returns nonzero if ch is a letter of

the alphabet; oth~rwise 0 is returned.

Exsmple This program checks each character read fromstdin and

reports all those that are letters of the alphabet:
414 TIACIt YOUASnf
c

'include <ctype.h>
#include <stdio.h>

int main(void)
(
char ch;

fori;;) (
ch = getchar();
if(ch==' ') break;
if(isalpha(ch» printf(-%c is a letter\n-, ch);
)

return 0;
)

#include <ctype.h>
int iscntrl(int ch);

Description The iscntrl( ) function returns nonzero if ch is between 0

and Ox] F or is equal to Ox7F (DEL); otherwise 0 is returned.

Example This program checks each character read from stdin and
reports all control characters:
#include <ctype.h>
#include <stdio.h>

int main(void)
(
char chi

for!;;) (
ch = getchar();
iflch==' ') break;
if(iscntrl{ch) )
printf(·~c is a control character\n-, ch):
)

return 0;
)
 SOME COMMON C ~ RINCYIONS 415
AI STRING AND ClWlACWI FUNCTIONS •
.,
#include <ctype.h>
int isdigit(int ch);

Description The isdigit( ) function returns nonzero if ch is a digit (0

through 9); otherwise 0 is returned.

Example This program checks each character read from stdin and
reports all those that are digits:
#include <ctype.h>"
#include <stdio.h>

int main (void)

{
char Chi

fori;;) (
ch = getchar () ;
if (ch==' .) break;
if(isdigit(ch» printf(M%c is a digit\n-, ch);
)

return 0;
)

#include <ctype.h>
Int Isgraph(int ch);

Da.cripfJoo The isgraph( ) function returns nonzero if ch is any

printable, character other than a space; otherwise 0 is returned.
Printable characters are in the range Ox21 through Ox7E .
.
Example This program checks each character read from stdin and
reports all printing characters: •
• include <ctype.h>
.include <stdio.h>

int main(void)
{
char chi
418 TEACH YOURSW
c

for (;;) (
ch = getchar();
if(ch==' ') break;
if(isgraph(ch» ,
. printf (·'c is a printing character\n-, ch);
)

return 0;
)

.Include <ctype.h>
int lalower(lnt ch);

DeM:tfptIon The lalower( ) function returns nonzero if ch is a

lowercase letter (a through z); otherwise 0 is returned.

EnmpIe This program checks each character read from stdin and
reports all those that are lowercase letters:
'include <ctype.h>
'include <stdio.h>

int main (void)

(
char Chi

for (;;) {
ch c getchar () ;
if (c:h=-' .) break;
H(hlO'j>er(chl) printf(·'c is lowercase\n· ,. chI;
)

return 0;
)

jJlndude <ctype.h>
Int IsprintOnt ch);

Oo.ett>fIon The iaprint( ) function returns nonzero if ch is a

printable character, including a space; otherwise 0 is returned.
Printable characters are often in the range Ox20 through Ox7E.
EnmpIe This program checks each character read from atdin and
reports all those that are printable:
'include <ctype.h>
#include <stdio.h>

int main (void)

{
char chi

for ( ; ;) (
ch = getchar () ;
if(ch=='Q') break;
if(isprint(ch» printf(-'c is printable\n-, ch);
)

return 0;
)

.include <ctype.h>
Int Ispunct(lnt 00);

Description The ispunct( ) function returns nonzero if ch is a

punctuation character, excluding the space; otherwise 0 is returned.
The term 'punctuation: as defined by this function, includes all
printing characters that are neither alphanumeric nor a space.

Example This program checks each character read from stdin and
reports all those that are punctuation:
.include <ctype.h>
'include <Std10.h>

int main (void)

(
char chj

forI;;) (
ch = getchar();
if(ch==' ') break;
if(ispunct(ch» printf{-'c is punetuation\n-, ch);
}
.1.., TtACHYOURSW
c

return 0;
)

#include <ctype.h>
int isspace(int 00);

Description The i88pace( ) function returns nonzero if ch is either a

space, tab, vertical tab, form feed, carnage return, or newline
character; otherwise 0 is returned.

bample This program checks each character read from stdin and
reports all those that are whitespace characters:
_include <ctype.h>
.include <stdio.h>

int main(void)
(
char chi

for I; ; I (
ch = getchar();
if{isspace(ch)) printf("\c is . . .1hitespace \n ... chI;
if(ch==' 'J breakj
)

return 0:
)

#include <ctype.h>
int isupper(Jnt ch);

DescriptiOll The isupper( ). function returns nonzero if ell is an

uppercase letter (A through Z); otherwise 0 is returned.

bample This program checks each character read from stdin and
reports all those that are uppercase letters:
#include <ctype.h>
#include <stdio.h>
 SOME COMMON C UBRARY FUNcnONS

A 1 STRING AND CHARACTER FUNCTIONS

.1.'"
int main(void)
{
char Chi

for I ;; I I
eut,ri ch ;; getchar():
if(ch==' ') break;
if(isupper(ch)) printf("%c is uppercase\n", ch);
}

,·,
1••• return 0;
}

#include <ctype.h>
int isxdigit(int ch);

Description The isxdigit( ) fUllcrion returns nonzero if ch is d

hexadecimal digit; otherwise 0 is returned. A hexadecimal digit ,,·i11 be
in nne of these ranges: A through F, a through f, or 0 through 9.

Example This program checks each character read from stdin and
reports all those (hat are hc,,-adenma\ digits.
•
~include <ctype.h>
#lr".ciude <stdio.h>

int main{vo1d)
(
char eh;

for I ; ; I [
ch ~ getchar () i
if (ch=;;:' ') break;
if(isxdigit(ch» printf("~c is hexadecimal \n-, ch);
}

return 0;
}
420
.., TEACH VOURSELF
C

mnclude <string.b>
char ·stn:8t(char ·""1, const char .str2);
DeIu..,., The strcat{ ) function concatenates a copy of str2 to strl
and tenninates strl with a null. The null terminator originally e nding
slrl is overwritten by the first character of str2. The string str2 is
untouched by the operation. The atrcat( ) func.tion returns strl .

No bounds-checking takes place, so it is the programmer's responsibility

to ensure that strl is large enough to hold both its original contents and
those of str2.

Example Th is program append, the Illst StI mg lead from stdin to the
second . For exampl e. assum ing the user p-nters hello and the re . the
progra m will pr int there hello .
• inc l ude <string.h>
' i nc l u de <stdio.h>

in t ma i n (VOl d )
{
char 51 801. s21801;

printf("Enter two strings: ~);

g e t s(sl,;
ge t s (52) ;

strcat(s2. 51);
print f (52) ;

return 0 ;
)

.incIude <string.b>
char °strchr(const char 'str, int ch);

Description Th e strchr( ) functio n re turns a pointe r to the first

occurre nce of the low-order byte of cl1 In th e stlin~ pointed to by str If
no match is found, a null poi nter is re turned.
 SOME COMMON C LIBRARY fUNCTIONS 421
A.I srmNG AND G'HAR4crfR FUNCTIONS •

Example This prints the string is a test:

#include <string.h>
#include <stdio.h>

int main(void)

char .p;

p =
strchr( "th~s is a test", ' ');
print f (p) ;

rE>tur-n )

#include <String.h>
int strcmp(const char 'strl, const char 'str2);

Description strCJnp( ) tUllL.t1011 lexicographlcally compdrl's two

r\
null-terminated sIrings md returns .1Il intc!:!,cr bd~cd 011 -lv' Ollt(nl11t'
lS shO\\'1l here:

Result Meaning
less than a SlrT is less than sIr2
a SlrT is equal to sIr2
greater than a slrT is greater than SIr2

Example The following function can be used as a password

verification routine. It will return a on failure and I on success.
#include <string.h>

int password (void)

(
char s[801;

printf("Enter password: ");

gets(s);
wll
if (strcmp{s, "pass » { H

printf (" Invalid Password\n");

return 0;
422 TEACH YOURSELF

'" C

lE-!.:.urn _.
,\

#include <String.h>
char 'strcpy(char 'str1, const char 'str2);

Description The strcpy( ) function is used to copy the contents of str2

into strl; str2 must be a pointer to a null-terminated string. The
slrcpy( ) function returns a pointer to str!.
If strl and str:! overlap. the behavior of slrcpy( ) is undefined

Example The following code fragment will copy 'hello' into string sir·
C!1ar stt" [80: ;
s!.:.!:"cpy(str, -hello");

#include <string.h>
size_t strlen(const char 'str);

Description The strIen( ) function rerums the length of the

null-terminated string pointed to by .'fr. The null is not counted. The
size_1 tvpe is defined in STRING. H

Example The following code fragment \\ill print 5 on the screen:

strcpy(s, "~elloM I ;
printf (M %d" , strlen(s));

#include <Stdio.h>
char 'strstr(const char 'str1, const char 'str2);

Description The strslr( ) function returns a pointer to the first

occurrence of the strlng pointed to by str2 in the string pointed to by
str] (except st/2's null terminator). It returns a null pointer ifno match
is found .

Example This program displays the message is a lesl:

 _ ~ C IJIIRAAYfUNClIONS 423
A 1 STRING AND CHARACTER FUNCTIONS •

*include <strinq.h>
.include <stdio.h>

lnt main(void)
(
char *Pi

p.; strstr(-this is a t~st". "is~);

printf (pi:

return 0;
)

.include <string.h>
char °strtok(char ·slrl, con .. char 0str2);

Dat:rlptJoII The .trtok( ) function returns a pointer to the next token

in the string pointed to by MrI. The characters making up the strin,
pointed to by SII':! are the delimiters that &cparate cach token. ,-\ null
pointer is returned when thernu'e no more tokens
The first time strtok( ) is called, strl is actually used in till' ('all.
Subsequent calls use a null pointer for the first argument In this "'.l"
the entire string can be reduced to its tokens.
• It is possible to usc a different set of delimiters for each call to strtok( ) ,

~ This program tokenizes the string "The sUl11mer soldier, the

sunshine patriot' Ivith spaces and commas as the delimiters. The
output will be The I Aummer I soldier I the I sunshine I patriot.
'include <strinq . h>
#include <stdio.h>

int main (vo id)

(
char "'p;

p = strtok.("The summer soldier, the sunshine patriot", " ,"J:

printf(p) ;
do (
P = strtok( '\0', .. Of);
if(pl printf('i's·, pi;
424 TIACII VOURSRF
... C

}whilelp};

return 0;
}

#include <ctype.h>
int tolower(int ch);

Oe5cription The tolower( ) function returns the lowercase equivalent

of ell if el> is a letter; otherwise ell is retur.n ed unchanged.
~. - .

Example This fragment displays q:

putchar(to!ower('Q')) ;

#include <ctype.h>
int toupper(int ch);

Description The toupper( ) function returns the uppercase

equivalent of eil if ell is a letter; otherwise ell is returned unchanged.

Example This displays A:

put char (toupper ( 'a' ) ) .

E MA THEMATICS FUNCTIONS
ANSI C defines several mathematics functions that take double
arguments and return double values. These functions fall into the
following categories:

" Trigonometric functions

" Hyperbolic functions

" Exponential and logarithmic functions

" Miscellaneous functions

 SOME COMMON C UBIWIY FUNtnONS 425
y
A.2 THE MATHEMAnCS FUNcnONS

All the math functions require that the header MATH.H be included
in any program that uses them. In addition to declaring the math
functions, this header defines a macro called HUGE_VAL. If an
operation produces a result that is too large to be represented by a
double, an overflow occurs, which causes the routine to return
HUGE_VAL. This is called a ra>1ge error. For all the mathematics
functions, if the input value is not in the domain for which the
function is defined, a domai>1 error occurs .
." All angles are specified in radians.

#include <math.h>
double acos(double arg);
•

Description The acos( ) function returns the arc cosine of argo The
argument to acos( ) must be in the range -1 through 1; otherwise a
domain error will occur. '.

Example This program prints the arc cosines, in one-tenth increments,

." of the values -1 through 1:
"
#include <math.h>
#include <stdio.h>

int rnain(void)
(
double val = -1.0;

do (
printf(~arc cosine of %f is %f\n". val, acos(val);
va l +:-; 0.1;
} while(val<=l,O);

return 0;
}

#inclUde <math.h>
double asin(double arg);

Description The asin( ) function returns the arc sine of argo The
argument to asin( ) must be in the range -1 through 1; otherwise a
domain error will occur.
426 TEACH YOURSELF
." C

Example This program prints the arc sines, in one-tenth increments,

of the values -1 through 1:
#include <math.h>
#include <stdio.h>

int main(void)

double va!=-l.O:

do (
printf("arc sine of %f is %f\n", val, asin(val»:
val += 0.1:
} while(val<=1.0);

return 0;
}

#include <math.h>
double atan(double arg);

Description The atan( ) function returns the arc tangent of a.-g.

Example This program prints the arc tangents, in one-tenth

increments, of the values -I through 1:
#include <math.h>
#include <stdio.h>

int main(void)
(
double val=-l.O;

do (
printf("arc tangent of %f is %f\n-, val. atan(val» i
val += 0.1;
} while(val<=1.0);

return 0;
 SOME COMMON C LIBRARY FUNCTIONS 427
A2 THE MATHEMATICS FUNCTIONS
"

#include <math.h>
""
~,
....
.~
double atan2(double y, double x);

Descriplion The atan2( ) function returns the arc tangent of ylx. It

uses the signs of its arguments to compute the quadrant of the return
value.

Example This program prints the arc tangents, in one-tenth

increments ofy, from -1 through 1:
#include <math.h>
#include <stdio.h>

int main(void)
(
double y=-l.O;

do (
printf(".;I,tan2 of %f is %f \ n", y, atan2(y, 1.0»;
y += 0.1;
} while{y<=1.0);

, -. return 0;
I ~
}
•

#include <math.h>
double ceil (double num);

Description The ceil( ) function returns the smallest integer

(represented as a double) that is not less than I1wn. For example, given
1.02, ceil() would return 2.0; given -1.02, ceil( ) would return-I.

Example This fragment prints 10.0 on the screen:

printf ("tf", ceil (9.9» i
428 1fACH YOUIISII.F
~ C

ftlnclude <math.h>
double cos(double erg);

Dat:IfptIon The cos( ) function returns the cosine of argo The value
of arg must be in radians.

&IImpIe This program prints the cosines, in one-tenth increments, of

the values -1 through 1:
'include <math.h>
.
'include <stdio . h>

int main (void)

(
double val=-l.O;

do (
printf( · cosine of %f is %£ \ n-, val. cos{val»:
val += 0.1;
) while(val<=l.O);

return 0;
}

•
ftinc:lude <math.1I:;>
double cosh(double erg);

DescripdM The cosh( ) function returns the hyperbolic cosine of arg

This program prints the hyperbolic cosines, in one-tenth

EJtIImpIe
increments, of the values -I through I :
'include <math.h>
.include <stdio . h>

int main (void)

(
double val=-l.O;

do (
print£(-hyperbolic cosine of \f is %£\n-, val . cosh{val
val += 0 . 1;
 /
SOME COMMON C UBAARY FUNcnONS 429
~
A2 THE MATHEMAnCS FUNGnONS

) whilelval<=l.O);

return 0;
)

#include <math.1I>
•
double exp(double arg);

Oescrlptlon The exp( ) function returns the natural logarithm e raised

to the arg power.

Example This fragment displays the value ofe (rounded to 2.718282):

printf(-Value of e to the first: %f-, exp(l.O»;

#include <math.1I>
double fabs(double nurn);

Oescrlpt/on The fabs( ) function returns the absolute value of HWH.

Example This program prints the numbers 1.0 1.0 on the screen:
.include <math.h>
.include <stdio.h>

int main{void)
(
printfl·%1.lf %1.lf·, fabsl1.0) , ' fabsl-1.0»;

return 0;
)

#include <math.1I>
double floor(double nurn);

Oescrlptlon The floor( ) function returns the largest integer

(represented as a double) not greater than nurn. For example, given
1.02, floor( ) would return 1.0; given -1.02, floor( ) would return
-2.0.
430 TEACH YOURSELF
.,. C

Example This fragment prints 10.0 on the screen:

printf{"%f", floor{lO.9));

#include <math.1I>
double log (double num);

Description The loge ) function returns the natural logarithm for

I1UJ11. A domain error occurs if nUJ11 is negative and a range error
occurs if the argument is O.

Example This program prints the natural logarithms for the numbers
I through 10:
#include <math.h>
#include <stdio.h>

int main(void)
{
double val=l.O;

do {
printf ("%f %f\n", val. log (val» ;
val++;
} while(val<11.0);

return 0:
)

#include <math.1I>
double log10(double num);

Description The loglO( ) function returns the base 10 logarithm for

the variable nurn. A domain error occurs if nurn is negative and a
range error occurs if the argument is O.

Example This program prints the base 10 logarithms' for the numbers
I through 10:
#include <math.h>
#include <stdio.h>
 ,
SOME COMMON C UBRARY FUNCTIONS 431
A2 THE MATHEMAnCS FUNcnONS

int main (void)

(
double val=l.O:

do (
printf(·%f %fln·, val, loglO(val»;
val++;
} while(val<11.0);

return 0;
)

#include <math.h>
double pow(double base, double exp);
•
Description The pow( ) function returns b(lSe raised to the exp power
e
(base ,/). A domain error may occur if base is 0 and e,<p is less than or
equal to O. A domain error will occur if base is negative and exp is not
an integer. An overflow produces a range error.

Example This program prints the tlrst ten powers of 10:

#include <math.h>
.include <stdio.h>

int main (void)

(
double x=lO.D. y=O.O;

do (
printf{-%f· pow(x, y»);
y++:
} while{y<ll);

return 0;
)

#include <math.h>
double sin(double arg);

Description The sine ) function returns the sine of argo The value of
arg must be in radians.
432 1IACH YOURSELF

" c

Example This program prints the sines, in one-tenth increments, of

the values -1 through 1:
.include <math.h>
'include <stdio.h>

int rnain(void)
{
double val=-l.O;

do (
printf(-sine of %f is %£\n-, val, sin(val»;
val += 0.1;
} while(val<=l.O):

return 0:
}

#include <rnath.h>
double sinh(double arg);

Description The sinh( ) function returns the hyperbolic sine of argo

ExBmple The following program prints the hyperbolic sines, in

one-tenth increments, of the values -1 through 1:
.include <math.h>
'include <stdio.h>

int main (void)

[
double val=-l.O:

do [
printf(-hyperbolic sine of %f is %£\n-, val, sinh(val»i
val += 0.1;
} while(val<=l.O):

return 0;
}
#include <math.h>
double sqrt(double num);

Description The sqrt( ) function returns the square root of num. If

called with a negative argument, a domain error will occur.

Example This fragment prints 4.0 on the screen:

printf("%f", sqrt(16.011;

#include <math.h>
double tan(double 8rg);

Description The tan( ) function returns the tangent of argo The value
of arg must be in radians.

Example This program prints the tangents, in one-tenth incremen'ts,

of the values -1 through 1:
#include <math.h>
#include <stdio.h>

int main(void)
{
double va!=-l.O;

do (
printf(-tangent of %f is %f\n H , val, tan(val»:
val += 0.1;
} while(val<=1.0);

return 0;
}

#include <math.h>
double tanh (double arg);

Description The tanh( ) function returns the hyperbolic tangent of argo

434 ltACII YOURSElF
c
-

Example This program prints the hyperbolic tangents, in one-tenth

increments, of the values -1 through 1:

*include <math.h>
*include <stdio.h>

int main (void)

{
double val=-l.O;

do (
print f("tanh of %f is %f\n", val, tanh(val»;
-,?-1 += 0.1;
j '""hi Ie (val<=l.0) i

return 0;
}

ME AND DATE FUNCTIONS

The time and date functions require the header TIME. H for their
prototypes. This header file also defines four types and two macros.
The type time_t is able to represent the system time and date as a
long integer. This is called the calendar time. The structure type tm
holds date and time broken down into its elements. The tm structure
is defined as shown here:
struct tm {
int tm_sec; /- seconds, 0-61 -/
int tm_min; /- minutes. 0-59 -/
int trn_houri/- hours, 0-23 "/
int tm_mday; /- day of the month, 1-31-/
int tm_man; /- months since Jan, 0-11 -/
int tm-year; /- years from 1900 -/
int tm_wday; /- days since Sunday, 0-6 -/
int
int
tm-yday; /- days since Jan 1, 0-365 ,.-
tm_isdst; /- Daylight Saving Time indicator -/
};

The value of tm_isdst will be positive if Daylight Saving Time is in

effect, 0 if it is not in effect, and negative if there is no information
 A3 nME AND DATr FUNCTIONS

.',..
'-
available. When the date and time are represented in this way, they
are referred tD as broken-down time.
The type c1ock_t is defined the same as time_to The header file
also defines size_to
The macros defined are NULL and CWCKS_PER_SEG.

#include <time.h>
char *asctime{const struct tm ·ptr);

Description The asctime( ) function returns a pointer to a string that

contains the time and date stored in the structure pointed to by ptr
after it has been converted into the following form:

day month date hours:minutes:seconds year\n\O

For example:

Wed Jun 19 12:05:34 1999

-
The structure pointer passed to asctime( ) is generally obtained from
either localtime( ) or gmtime( ).
m: The buffer used by asctime( ) to hold the formatted output string is
ro- a statically allocated chatacter array and is overwritten each time the
Idtli function is called. If you want to save the contents of the string, you
~ ryl need to copy it elsewhere.
!me
Example This program displays the local time defined by the system:
.include <time.h>
#include <stdio.h>

int main(void)
{
struct tm "ptri
time_t It;

It = time(NULL);
ptr = !ocaltime{&ltl;
printf(asctime{ptr»;

return 0 i
)
.
438~_F

.include <time.h>
clock_t clock(void);

Dacliptlon The c1ock{ ) function returns the number of system

clock cycles that have occurred since the program began execution.
To compute the number of seconds, divide this value by the
CLOCKS_PER_SEC macro.

Exemple The following program displays the number of system clock

cycles occurring since it began:
.include <stdio.h>
*include <time.h>

int main(void)
(
int i;

for(i=O: i<10000; i++);

printf (. %u·. clock () I ;

return 0;
I

.include <time.h>
char ·ctime(const time_t ~me);

DescrIptIon The ctime{ ) function returns a pointer to a string of th

form .

day month date hours:minutes:seconds year\n\O

given a pointer to the calendar time. The calendar time is generally

obtained through a call to timet ). The ctime( ) function is
equivalent to:
asctime(localtime{time»)

The buffer used by ctime( ) to hold the formatted output string is a

statically allocated character array and ii overwritten each time the
 SOM£ COMMON C UBRAII't FIlNClWNS 437
A3 nMf AND CMTE FUNCnONS

function is called. If you wish to save the contents of the string, you
need to copy i~ elsewhere.

&ample This program displays the local time defined by the system:

.include <time.h>
.includ~ <stdio.h>

int main (void)

{

It = tirne(NULLl;
printf {.ctime (&1 t) ) ;

return ai
)

#include <time.h>
double difftime(time_t time2, time_t time1);

OescriprJon The difftime( ) function returns the difference, in

seconds, between timel and time2. That is, time2 - timel.

Example This program times the number of seconds that it takes for
the empty for loop to go from 0 to 500000.
#include <time.h>
.include <stdio.h>

int main (void)

(
time_t start, end;
long unsigned int t;

start = time (NULL) i

for(t=O; t<500000L; t++);
end = time{NULL);
printf("Loop required %f seconds.\n", difftime(end, start»;

;tring lll return 0;

)
ime ttl
438 1fACII YOORSW'
c

#include <time.h>
strut tm 'gmtime(const time_t *time);

Description The gmtime( ) function returns a pointer to the

broken-down form of time in the form of a tm structure. The time is
represented in Coordinated Universal Time (i.e., Greenwich Mean
Time). The time value is generally obtained through a call to time( ).
The structure used by grntirne( ) to hold the broken-down time is
statically allocated and is overwritten each time the function is called.
If you wish to save the contents of the structure, you need to copy it
elsewhere.

Example Thi~
program prints both the local time and the Coordinated
Universal Time of the system:
#include <time.h>
#include <stdio.h>

; - print local and Coordinated Universal time * /

int main (voidl
{
struct. tm *local, ·coordinated:
time_t t;

t time (NULL) ;
=
local = localtime(&t);
printf(~Local time and date: %5·, asctime(local»);
coordinated = gmtime(&t l ;
priotf(·Coordinated Universal time and date: %5·,
asctime(coordinated» ;

return 0:
}

#include <time.h>
struct tm 'Iocaltime(const time_t 'time);

Description The loca1tirne( ) function returns a pointer to the

broken-down form of time in the form of a tm structure. The time is
represented in local time. The hme value is generally obtained through
a call to the time( ) function .
 SOME COMMON C UBRARY FUNCTKINS
A,3 nME AND DATE FUNCnONS
,.
43.

The structure used by localtime( ) to hold the broken-down time is

statically allocated and is overwritten each time the function is called.
If you wish to save the contents of the structure, you need to copy it
elsewhere.

Example This program prints both the local time and the Coordinated
Universal time of the system:
#include <tirne.h>
#include <stdio.h>

1* print local and Coordinated Universal time */

int rnain(void)
,
(
struct tm *local;
time_t t;

t = time (NULL) ;
local = loca!time(&t) i
printf(~Local'time and date: %s~, asctime(local));
local = gmtime(&t);
printf(~Coordinated Universal time and date: %5".
asctime(local)) ;

return 0:
}

#include <time.h>
time_t time(time_t ·systime);

Description The time( ) function returns the current calendar time of

the system. If the system has no time-keeping mechanism, then -1 is
returned. _.
The time( ) function can be called either with a null pointer or
with a pointer to a variable of type time_to If the latter is used, then
the argument will also be assigned the calendar time.

Example This program displays the local time defined by the system:
#include <time.h>
#include <stdio.h>
440 TEACH YOURSElF
c

int main{void)
(
struct tm *ptri
time_t It;

It = time(NULL);
ptr = !ocaltime(&lt);
printf(asctime(ptr» ;

return 0,
)

YNA MIC ALL OCA TION

There are two primary ways a C program can store information in the
main memory of the computer. The first uses global and local
variables-including arrays and structures. In the case of global and
static local variables, the storage is fixed throughout the runtime of
your program. For dynamic local variables, storage is allocated on the
stack. Although these variables are efficiently implemented in C, they
require the programmer to know in advance the amount of storage
needed for every situation. The second way information can be stored
is with C's dynamic allocation system. In this method, storage for
information is allocated from the free memory area (called the heap)
as it is needed. '
The ANSI C standard specifies that the header information
necessary to the dynamic allocation system is in STOLIB.H. In ~is
file, the type size_t is defined. This type is used extensively by the
allocation functions and is essentially the equivalent of unsigned.

#include <stdlib.h>
void 'calloc(size_t num, size_~ size);

Description The calloc( ) function returns a pointer to the allocated

memory. The amount of memory allocated is equal to num • size. That
is, calloc( ) allocates sufficient memory for an array of num objects of
size size.
 SOME COMMON C UBRARV fU~S 1141
A' O"YNAMIC ALLOCATION '"

The canoc( ) function returns a pointer to the first byte of the

allocated region. If there is not enough memory to satisfy the request,
a null pointer is returned.
It is always important to verify that the return value is not a null
pointer before attempting to use it.

Example This function returns a pointer to a dynamically allocated

array of 100 floats: .
#include <stdlib.h>
#include <stdio.h>

float "'get_rnem(void)
(
float "'Pi

p = calloc(lOQ, sizeof(float)};
iflJp) (
p.rintf{-Allocation error - aborting.\n~);
exi t ( 1 ) ;
001 )

C:. return p;
)

#include <stdlib.h>
'iIi1/' void free(void 'ptr);

Description The free( ) function de allocates the memory pointed to

by prr. This makes the memory available for future allocation.
It is imperative that the free( ) function be called only with a
pointer that was previously allocated using one of the dynamic
allocation system's functions, such as manoc( ) or canoc( ). Using
an invalid pointer in the call will probably destroy the memory
management mechanism and cause a system crash.

Example This program first allocates room for 100 user-entered

strings and then frees them:
#include <stdlib.h>
#include <stdio.h>

int main (void)

TEACH YOUASW
C

{
char ·str[lOO];
int i;

for (i=O; i<100; i++) (

ifllstr(i) = malloc(128))"NULL) (
printf(-Allocation error ~ aborting.\n~);
exitlO) ;
)
getslstr(i)) ;
)

j *now free the memory */

for{i=O; i<100; i++) free(str[il);

return 0;
)

linclude <stdlib.h>
void *malloc(size_t size);

Description The maUoc( ) function returns a pointer to the first byte

of a region of memory of size size that has been allocated from the
heap. (Remember, the heap is a region of free memory managed by
C's dynamic allocation subsystem.) If there is insufficient memory in
the heap to satisfy the request, malloc( ) returns a null pointer. It is
always important to verify that the return value is not a null pointer
before attempting to use it. Attempting to use a null pointer will
usually result in a system crash.

Example This function allocates sufficient memory to hold structures

of type addr:
#include <stdlih.h>
#include <stdio.h>

struct addr (
char name (40);
char street (40);
char city(40) ;
char state(3);
r.har zip(lO) ;
 SOM£ COMMON C ~ FUNCOONI ..:I
A< DYNAMIC ALLOCATION •

) :

struct addr *get_struct(void)

(
struct addr *p;

if{{p = malloc(sizeof(struct addr»)==NULL)

(
printf(KAllocation error - aborting.\n-);
exit(O) ;
)
return in
)

#include <Stdlib.h>
void ·realloc(void ·ptr, size_t size);

Description The realloc( ) function changes the size of the allocated

memory pointed to by plr to that specified by size. The value of size
~.
may be greater or less than the original. A point~r to the memory
block is returned since it may be necessary for realloc( ) to move the
block to increase its size. If this occurs, the contents of the old block
are copied into the new block - no information is lost.
If there is not enough free memory in the heap to allocate size
bytes, a null pointer is returned. This means it is important to verify
the success of a call to realloc( ).

Example This program first allocates 17 characters, copies the string

"this is 16 chars" into the space, and then uses realloc( ) to increase
the size to 18 in order to place a period at the end.
'include <stdlib.h>
#include <stdio.h>
*incluge <string.h>

int main(void)
(
char *p;

p = malloc(17);
444 _eH VOURSlLf
c

if(!p) (
printf(-Allocation error - aborting.\n");
exit(l);
}

strcpy(p, "this is 1."6 chars");

p = realloc(p,lB);
if(!p) (
printf(·Allocation error - aborting.\n");
exit(1) ;
}

strcat (P, ". ") ;

printf (p) ;

free (p) ;

return 0;
)

ISCELLANEOUS FUNCTIONS

The functions discussed in this section are all standard functions that
don't easily fit in any other category.

#include <stdlib.h>
void abort(void);

Description The abort( ) function causes immediate termination

of a program. Whether it closes any open files is defined by the
implementation, but generally it won't.

Example In this program, if the user enters A, the program will

terminate:
.include <stdlib.h>
#include <conio.h>
 OUMt COMMON C UBRARV fUNcnONS 445
'f'
AS MISCELLANEOUS FUNCnONS

int main (void)

(
for (; ; )
if(getche()=='A') abort();

return 0;
}

#include <stdlib.h>
int abs(int num);

Description The abs( ) function returns the absolute value of the

integer >1urn

Example This function converts the user-entered numbers into their

absolute values:
#include <stdlib.h>
#include <stdio.h>

int get_abs(void)
(
char num[80];

gets (num) ;

return abs(atoi(num»;
}

#include <stdlib.h>
double atof(const char • str);

-
,"
Description The atof( ) function converts the string pointed to by str
into a double value. The string must contain a valid floating-point
number. If this is not the case, the returned value is O.
The number may be terminated by any character that cannot be
part of a valid floating-point number. This includes whitespace
characters, punctuation (other than periods), and characters other
than 'E' or 'e'. Thus, if atof( ) is called with "lOO.OOHELLO", the value
100.00 will be returned.
 .,
448 TIACII YOURSElf

Example This program reads two floating'point numbers and displays

their sum:
#include <stdlib~h>
#include <stdio.h>

int main(void)
(
char nurnl(80) , nurn2[80);

printf("Enter first: ");

gets(numl);
printf("Enter second: ");
gets (num2) ;
printf("The sum is: %f", atof(numl) + atof(num2));

return 0;
)

#include <stdlib.h>
int atoi(const char ·str);

Description The atoi( ) function converts the string pointed to by str

into an int value. The string must contain a valid integer number. If
this is not the case, the returned v~lue is o.
The number may be terminated by any character that cannot be
part of a integer number. This includes whitespace characters,
punctuation, and other characters. Thus, if atoi( ) is called with
123.23, the integer value 123 will be returned, and the 0,23 ignored,

Example This program reads two integer numbers and displays

their sum:
#include <stdlib.h>
#include <stdio.h>

int main (void)

(
char nurnl[BO], nurn2[80];

printf ("Enter first: ");

gets (nurnl) ;
printf("Enter second: ");
 A5 MISCELLANEOUS FUNCTIONS

gets (num2l :
prin~f("The sum is: %d", atoi(numl) + atoi(num2»);

return 0;
)

#include <stdlib.h>
long atol(const char ·str)j

Description The atol( ) function converts the string pointed to by sty

into a long int value , The string must contain a valid long integer
number. If this is 110t the case, the returned value is o.
The number may be terminated by any character that cannot be
part of an integet number. This includes whitespace characters,
punctuation, and Other characters. Thus, if atol( ) is called with
123.23, the integer value 123 will be returned, and the 0.23 ignored.

Exemple This program reads two long integer numbers and displays
their sum:
~include <stdlib.h>
#include <stdio.h>

int main(void)
(
char numl{BO), num2{BO);

printf("Enter first: ") ;

gets (nurn! ) ;
printf(~Enter second: ");
gets (num2) ;
printfC"The sum is: %1d", atol(numl) + atol(num2»;

return 0 ;
)
448 TfACH YOURS£LJ'
c

#include <stdlib.h>
void ·bsearch(const void ·key, const void ·base,
size_t nurn, size_t size,
int(*cornpare)(const void ., const void .));

Description The bsearch( ) function performs a binary search on' the

sorted array pointed to by base and returns a pointer to the first
member that matches the key pointed to by key. The number of
elements in the array is specified by num and the size (in bytes) of
each element is described by size. (The size_t type is defined in
STDLlB.H and is essentially the equivalent of unsigned .)
The function pointed to by compare is used to compare an element
of the array with the key. The form of compare must be

int function_name(const void 'argl, const void 'arg2J

It must return the following values:

Less than 0 If arg I is less than arg2

o If arg I is equal to arg2
Greater than 0 If arg I is greater than arg2

The array must be sorted in ascending order, with the lowest

address containing the lowest element.
If the array does not contain the key, then a null pointer is returned.

Example This program reads characters entered at the keyboard and

deteT]nines whether they belong to the alphabet.
#include <stdlib.h>
#incl ude <ctype.h>
#include <stdio.h>

char *alpha = -abcdefghijklrnnopqrstuvwxyz - ;

int comp (c on s t void *eh , const voi d * 5);

int main {v o id )
{
char Chi
char *P i
 SOME COMMON C UIIWIY FUNCTIOIII
A5 MISCELlANEOUS RJNC7IONS

do (
printfC-Enter a character: -);
scanfC-%c%*c-,&ch);
ch = tolower CchI ;
p = bsearch(&ch, alpha, 26, 1, comp);
if(p) printf('is in alphabet.\n');
else printfC-is not in alphabet.\n");
) while (p) ;

return 0;
)

/ * compare two characters */

int comp(const void *ch, const void ·s)
(
return ·(char *)ch - *(char *)s;
)

#include <Stdlib,h>
void exit(int status);

De&crlpUon The exit( ) function causes immediate normal

termination of a program.
The value of status is passed to the calling process, usually the
operating system, if the environment supports it. By convention, if
the value of status is 0, normal program termination is assumed. A
nonzero value may be used to indicate an error.
You may also use the predefined macros EXIT_SUCCESS and
EXIT_FAILURE as arguments to exit( ) .

Ez8mp1e This function performs menu selection for a mailing list

program. If Q is selected, the program is terminated.
char menu (void)
(
char chi

do (
printf (-Enter names (El \n");
printfC"Delete name (D)\n·);
printf(·Print (P)\n");
 printf('Ouit (Q)"\n');
) whilellstrchr('EDPQ',toupper(ch)));
if(ch=='Q') exit(O);
return ch:
}

#include <stdlib.h>
long labs(long num);

DescriptionThe labe( ) function returns the absolute value of the

long int num.

Example This function converts the user-entered numbers into their

absolute values:
#include <stdlib.h>
#include <stdio.h>

long int get_labs (voidl

(
char nurn(80];

gets (numl ;

return labs(ato!(num»;
}

#include <setjmp.h>
void longjmp(jmp_buf envbuf, int val);

Description The longjrnp( ) function causes program execution to

resume at the point of the last call to setjrnp( ). These two functions
are the way ANSI C provides for ajump between functions. Notice
that the header SETJMP.H is required.
The longjrnp( ) function operates by reset"iiIlg the stack as
descnbed in enubuf, which must have been set by a prior call to
setjmp( ). This causes program execution to resume at the statement
following the setjrnp( ) invocation-the computer is 'tricked' into
thinking that it never left the function that called setjrnp( ). (As a
somewhat graphic explanation, the longjrnp( ) function 'warps'
 AS MISCELLANEOUS FUNCnONS

across time and (memory) space to a previous point in your program,

without having to perform the normal function-return process.)
The buffer envbufis of type jrnp_buf, which is defined in the
header SETJMP.H. The buffer must have been set through a call to
setjrnp( ) prior to calling iongjrnp( ). .
The value of val becomes the return value of setjrnp( ) and may be
interrogated to determine where the long jump came from. The only
value not allowed is O.
It is important to understand that the iongjrnp{ ) function must be
called before the function that called setjrnp{ ) returns. If not, the
result is technically undefined. In actuality, a crash will almost
certainly occur.
By far the most common use ofiongjrnp( ) is to return from a
deeply nested set of routines when a catastrophic error occurs.

Example This program prints 1 2 3:

#include <setjmp.h>
.include <stdio.h>

void f2 (void) ;

int main(void)
(
char first=!;
int i;

printf("! ");
i = setjmp(ebuf);
i f (first) (
first = ! firpt:
f2 () ;
printf{"this will not be printed");
}
printf("%d-, i);

return 0;
}

void f2(void)
{
 printf('2 .);
longjmp(ebuf, 3);
}

.include <atdlib.h>
void qsot1(void ·base, size_t num, size_t size,
int("compare)(const void·, const. void"»;

Description The q80rt( ) function sorts the array pointed to by base

using a Quicksort (which was developed by C.A.R. Hoare). The
Quicksort is generally considered the best general-purpose sorting
algorithm. Upon termination, the array will be sorted. The number of
elements in the array is specified by 11um and the size (in bytes) of
each el~ment is described by size. (The size_t type is defined in
STDLIB.H and is essentially the equivalent of unsigned.)
The function pointed to by compare is used to compare two
elements in the array. The form of compare must be

int function_name(const void *argl, const void *arg2)

It must return the following values:

Less than 0
•
If erg I is less than arg2
o If argl is equal to tRg2
Greater than il . If erg I is greater than erg2

The array is sorted in ascending order, with the lowest address

containing the lowest element.

&ample This program sorts a list of integers and displays the results.
'include <stdlib.h>
'include <stdio.h>

int comp(const void *i, const void .j ) ;

int num[lO}' {
1,3, 6, 5. 8, 7, 9, 6, 2, 0
};

int main (void )

 SOM£ COMMON C UIIIWIY RlNClIONS 4U
A5 MISCELLANEOUS FUNCnONS
"

int i:

printf(-Original array: .);

for(i=O; i<10; i++) printf(-'d· num(i]);
printf ( • \ n·) ;

qsort(num, 10, sizeof{int), comp);

p,rintf (. Sorted array: .);

forli=O: i<10: i++) printf(-'d· num[i});

return 0;
}

" ' j'

/ * compare thl integers */
int comp(const void *i, const void *j)
{
return *(int *)i - ·(int .)j;
)

*include <stdIIb.h>
Int rand(void);

0......... The rand( ) function generates a sequence of

pseuda..random num~rs. Each time it is called, an integer between 0
and RAND_MAX is returned. RAND_MAX is defined in STDLIB.H.
The ANSI standard stipulates that the macro RAND_MAX will have a
value of at least 32,767.

Example This program displays ten pseudo-random numbers:

_include <stdlib.h>
.include <stdio.h>

int main (void)

(
~nt i;

for(i=O; i<10; i++)

printf{"id ", randl));
454 _CIt YOURSa.f
c

return 0;
)

#include <Setjmp0h>
int setjmpOmp_buf envbuO;

Description The setjrnp( ) function saves the contents of the system

stack in the buffer envbuf for later use by longjrnp( ).
The setjrnp( ) function returns 0 upon invocation. However,
longjrnp( ) passes an argument to setjrnp{ ) when it executes, and it
is this value (always nonzero) that will appear to be the value of
setjmp( ) after a call to longjrnp( ).
See the longjrnp( ) section for more information.

Example This program prints 1 2 3:

#include <setjmp.h>
.include <stdio.h>

void t"2 (void) ;

int main(void)
(
char first=!;
int i;

printflol °1;
i = setjrnp(ebuf);
iflfirstl (
first = !first;
f2 I I;
printf(-this will not be printed-);
)
printfIO%dO,il;

return 0;
)

void f2(void)
(
printfl02 °1;
 SOME COMMON C UBIWIY FUNCTIONS
A5 ftl/SCELIANEOUS FUNGnONS

longjmp(ebuf, 3);
)

#include <stdlib.h>
void srand(unsigned seed);

DescrlpIJon The srand( ) function is used to set a starting point for

the sequence generated by rand( ), which returns pseudo-random
numbers.
Generally srand( ) is used to allow multiple program runs to use
different sequences of pseudo-random numbers.

Example This program uses the system time to randomly initialize the
rand( ) function using srand( ):
#include <stdio.h>
#include <stdlib.h>
#include <time.h>

/* Seed rand with the system time

and display the first 100 numbers.
"/
lnt main{void)
(
int i, utime;
long Itime;

/* get the current calendar time *1

Itime = time(NULL);
utime = (unsigned intI Itime/2:
srand(utime) ;

for(i=O; i<10; i++) printf("%d" rand()l;

return 0;
)
B
C Keyword Summary

457
•
 . . . TEACH YOURSELF
" . C

- ' N ERE are 32 keywords that, when combined with the formal
C syntax, form the C language as defined by the ANSI C
standard. These keywords are shown in Table B-1.
All C keywords use lowercase letters. In C, uppercase
and lowercase are different; for instance, else is a keyword,
ELSE is not.
An alphabetical summary of each of the keywords follows:

auto
auto is used to create temporary variables that are created upon entry
into a block and destroyed upon exit. For example:
#include <stdio.h>
'include <conio.h>
.. ~ .
int mainlvoid)
{
fori;;} {
iflgetche(}=='a'} (
auto int t;

... for(t=O; t<'a'; t++)

printfl·'d· t};
break;
l
l

return 0;
l

auto double int struct

break else long switch
case enum register typedef
char extern return union
• const float short unsigned
continue for signed void
default goto sizeof volatile
do if static while

'''BU fI 1 Keyword Ust .,

 CK~-'"
•

In this example, the variable t is created only if the user strikes an a.

Outside the ifblock, t is completely unknown; and any reference to it
would generate a compile-time syntax error. The use of auto is
completely optional since all local variables are auto by default.

break
break is used to exit from a do, for, or while loop, bypassing the
normal loop condition. It is also used to exit from a switch statement.
An example of break in a loop is shown here :
while Ix<lOO I (
x = get_oew_x ( );
if (kbhit( )) break ; / * key hit on keyboard * /
process (x) i
}

Here, if a key is presse~, the loop will terminate no matter what the
value of x is .
In a switch statement, break effectively keeps program execution
from "falling through" to the next case. (Refer to the switch section for
details.)

case

case is covered in conjunction with switch.

char

char is a data type used to declare character variables. For example, to

declare ch to be a character type, you would write:
char Chi
•
In C, a character is one byte long.

const

The const modifier tells the compiler that the contents of a variable
cannot be changed. It is also used to prevent a function from
modifying the object pointed to by one of its arguments.
460 TlACH YOURSELF
... C

continue
continue is used to bypass portions of code in a loop and forces the
conditional expression to be evaluated. For example, the following
while loop will simply read characters from the keyboard until an sis
typed:
while(ch=getche())
if(ch != ' 5 ' ) continue; j * read another char */
process (chl;
)

The call to process( ) will not occur until ch contains the character s.

default
default is used in the switch statement to signal a default block of
code to be executed If no matches are found in the switch. See the
switch se ction .

do
Th" do loop is one of three loop constructs i!\'ailahJe in C. The geJl('ral
form of the do loop is

do (
. statement block
) while(condition);

If only one statement is repeated, the braces are not necessary, but
they add clarity to the statement. The do loop repeats as long as the
condihon is true.
The do loop is the only loop in C that will always have at least one
iteration because the condition rs tested at the bottom of the loop.
A common use of the do loop is to read disk files. This code will
read a file until an EOF is encol1nte,rcd.
do (
ch = getc (fp) ;
if(!feof(fp)) printf("%c", chl:
while(!feof(fp));
 c~-4&l
...

double
double is a data type specifier used to declare double-precision
floating-point variaDles. To declare d to be of type double you would
write the following statement:
double di

else

See the if section.

enum
The enum type specifier is used to create enumeration types. An
enumeration is simply a list of named integer constants. For example,
the following code declares an enumeration called color that consists
of three constants: red, green, and yellow.
#include <stdio.h>

enum color {red, green, yellow};

enum color Ci

int main(void)
(
c = red;
if (c==red) printf(~is red\n");

return 0;
}

extern

The extern data type modifier tells the compiler that a variable is
defined elsewhere in the program. This is often used in conjunction
with separately compiled files that share the same global data and are
l'nked together. In essence, it notifies the compiler of a variable
without redefining it.
As an example, if first were declared in another file as an integer,
the following declaration would be used in subsequent files:
extern int first;
TEACH YOURSa.F

float
float is a data type specifier used to declare floating-point variables. To
declare f to be of type float you would write:
f loat f;

for

The for loop allows automatic initialization and incrementation of a

counter variable. The general form is

for (initialization; condition; increment) {

statement block
}

If the statement block is only one statement, the braces are not
necessary .
Although the for allows a number of variations, generally the
l11itialization is used to set a counter variable to its starting value . The
condition is generally a relational statement that checks the counter
variable against a termination value , and the increment increments (or
decrements) the counter value. The loop repeats until the condition
be comes false.
The following code will print hello ten times.
for ( t=O; t<10; t++) printf { ~Hello \ n·);

goto
The goto causes program execution to jump to the label specified in
the goto statement. The general form of the goto is

goto label;

label:

All labels must end in a colon and must not conflict with keywords or
function names. Furthermore, a goto can branch only within the
current function, and not from one function to another.
The following example will prim the message right but not the
message wrong:
goto labl;
printf(~wrong·l ;
labl,
printf ("right");

if
The general form of the if statement is

if(condition) (
statement block 1
) .
else {
statement block 2
}

If single statements are used, the braces are not needed. The else is
optional.
The condition may be any expression. If that expression evaluates to
any value other than 0, then statement block 1 will be executed;
otherwise, if it exists, statement block 2 will be executed.
The following code fragment can be used for keyboard input and to
look for a 'q' which signifies 'quit.'
ch = getche () :
i f (ch=='q') (
printf(~Progr~~ Terminated-}i
exit (0);
)
else proceed( ) ;

int
int is the type specifier used to declare integer variables. For example,
to declare count as an integer you would write
int count;
4&4 T£ACII VOURSEU
c

long
long is a data type modifier used to declare long integer and long
double variables. For example,·to declare count as a long integer you
would write
long int count;

register
The register modifier requests that a variable be stored in the way
that allows the fastest possible access. In the case of characters or
integers, this usually means a register of the cpu. To declare i to be a
register integer, you would write
register int i;

return
The return statement forces a return from a function and can be used
to transfer a value back to the calling routine. For example, the
following function returns the product of its two integer arguments.
int mul(int a, int b)
(
retunJ. a '*b;
)

Keep in mind that as soon as a return is encountered, the function

will return, skipping any other code in the function.

short

short is a data type modifier used ·to declare small integers. For
example, to declare sh to be a short integer you would write
short int sh;

signed

The signed type modifier is most commonly used to specifY a signed

char data type. .
 C KEYWORD SUMMARY 485
y

sizeof
The sizeof keyword is a compile-time operator that returns the length
of the variable or type it precedes. If it precedes a type, the type must
be enclosed in parentheses. For example,
printf (~%d·, sizeof (short int»;

will print 2 for most C implementations.

The sizeof statement's principal use is in helping to generate
portable code when that code depends on the size of the C built-in data
types.

static
The static keyword is a data type modifier that causes the compiler to
create permanent storage for the local variable that it precedes. This
enables the specified variable to maintain its value between function
calls. For example, to declare lase time as a static integer, you would
write
static int last_time;

static can also be used on global variables to limit their scope to the
file in which they are declared.

struct
The struct statement is used to create aggregate data types, called
structures, that are made up of one or more members. The general .
form of a structure is

strue! struct-name {
type member' :
type member2 ;

type member N ;
} variable-list;

The individual members are referenced using the dot or arrow

operators.
466 TtACH YOURSELF

" c

switch
The switch statement is C's mUlti-path branch statement. It is used to
route execution in one of several ways. The general form of the
statement is

switchCint-expression) {
case constant I: statement-set I;
break;
case constant2: statement-set 2 ;
break;

case constantN: stBtement-set N ;

break;
default: default-statements;
}

Each statement-set may be one or many statements long. The default

portion is optional. The expression controlling the switch and all case
constants must be of integral or character types.
The switch works by checking the value of int-expression against
the constants. As soon as a match is found, that set of statements is
executed. If the break statement is omitted, execution will continue
into the next case. You can think of the cases as labels. Execution will
continue until a break statement is found or the switch ends.
The following example can be used to process a menu selection:
ch = getche () ; •
switch(ch) {
case 'e': enter();
break;
case '1': list();
break:
case's': sort'();
break;
case 'q': exit(O);
break;
default; printf(·Unknown Command\n-);
printf(-Try Again \ n-);
)
 CKEYWOADSUMMARY 487
...

typedef

The typedef statement allows you to create a new name for an

existing data type. The general form of typedef is

typedef type-specifier new-name;

For example, to use the word 'balance' in place of 'float,' you would
write
typedef float balance:

union

The union keyword creates an aggregate type in which two or more

variables share the same memory location. The form of the
declaration and the way a member is accessed are the same as for
struet. The general form is

union union-name {
type memberT :
type member2 :

type member N ;
: } variable-list;
••• ..,
unsigned

The unsigned type modifier tells the compiler to create a variable that
holds only unsigned (Le ., positive) values. For example, to declare big
to be an unsigned integer you would write
unsigned int big;

void

The void type specifier is primarily used to declare void functions

(functions that do not return values). It is also used to create void
pointers (pointers to void) that are generic pointers capable of
pointing to any type of object and to specify an empty parameter list.
468 1lACH YOURSElf
... •
c

volatile
The volatile modifier tells the compiler that a variable may have its
contents altered in ways not explicitly defined by the program.
Variables that are c}!anged by the hardware, such as real-time clocks,
interrupts, or other inputs are examples.

while

The while loop has the general form:

while(condition) {
statement block
}

If a single statement is the <lbject of the while, the braces may be

omitted. The loop will repeat as long as the condition is true.
The while tests its condition at the top of the loop. Therefore, if the
condition is false to begin with, the loop will not execute at all. The
condition may be any expression.
An example of a while follows. It reads characters until end-of-file
is encountered.
t = 0;

While ( ! feof (fp» (

sIt] =
getc(fp);
t++;
]
c
Bu.ilding a Windows
Skeleton

469
~
470 TEACH YOURSELF
c

is a popular language for Windows programming. As such,

it makes sense that some coverage of this important topic
be included in this book . But be forewarned : Programming
for Windows requires a thorough knowledge of both C ana
Windows. Frankly, before you can write useful Windows
programs, you will need to hone your C programming skills and then
invest substantial time in learning the ins and outs of the Windows
operating system. Keep in mind that just a description of the functions
available within Windows requires approximately 2,000 printed pages'
The preceding notwithstanding, if you will be moving on to
Windows programming, you are probably anxious to begin. The
purpose of this appendix is to give you a brief overview of Windows
programming and to explain a few of its most fundamental element~ .
In essence, the information presented here is designed to give you a
"jump start" into the world of Windows programming.
This appendix discusses in a general way what Windo\I's is, how a
program must interact with it , and what rull's must be followed by
every Windows application . It also develops an application skeleton
that you can use as a basis for your own Windows programs. As you
will see, all Windows programs share several common traits. It is these
shared attributes that will be contained in the application skeleton.

ICH VERSION OF WINDOWS?

,
At the time of this writing, there are three versions of !he Windows
operating system in common use : Windows 3.1, Windows 95, and
Windows NT. The skeleton developed in this appendix is designed for
32-bit versions of Windows, such as Windows 95 or Windows NT, since
these are the most Widely used versions. However, the basic principles
apply to all versions of Windows.

W'NDOWS PROGRAMMING
PERSPfCTlVE

The' goal of Windows is to enable a person who has basic familiarity

with the system to sit down and run Virtually any application without
prior training. To accomplish this end, Windows provides a consistent
interface to the user. In theory, if you can run one Windows-based
 BUILDING" WINDOWS SK£L£TON 471
WINDOWS PROGRAMMING PERSPfCT1VE •

program, you can run them all. Of course, in actuality, most useful
programs will still require some sort of training in order to be used
effectively, but at least this instruction can be restricted to what the
program does, not how the user must interact with it. In fact, much of
the code in a Windows application is there just to support the user
interface.
Before continuing, it must be stated that not every program that
runs under Windows will necessarily present the user with a Windows-
. style interface. It is possible to write Windows programs that do not
take advantage of the Windows interface elements. To create a
Windows-style program, you must purposely do so. Only those
programs written to take advantage of Windows wi1llook and feel like
Windows programs. While you can override the basic Windows design
philosophy, you had better have a good reason to do so, because the
users of your programs will, most likely, be very disappointed. In
general, any application programs you- are writing for Windows should
utilize the normal Windows interface and conform to the standard
Windows design practices.
Windows is graphics-oriented, which means that it provides a
Graphical User Interface (GUI). While graphics hardware and video
modes are quite diverse, many of the differences are handled by
Windows. This means that, for the most part, your program does not
need to worry about what type of graphics hardware or video mode is
being used.
Let's look at a few of the more important features of Windows.

rIlE DESIITOP MODEL

With few exceptions, the point of a window-based user interface is

to provide the equivalent of a desktop on the screen. On a desk you
might find several different pieces of paper, one on top of another,
often with fragments of different pages visible beneath the top page.
The eqUivalent of the desktop in Windows is the screen. The pieces of
paper are represented by windows on the screen. On a desk you may
move pieces of paper about, maybe switching which piece of paper is
on top, or how much of another is exposed to view. Windows allows
the same type of operations on its windows. By selecting a window,
472 T£AClfYOU~

you can make it current, which means putting it on top of all the other
open windows. You can enlarge or shrink a window, or move it about
on the screen. In short, Windows lets you control the surface of the
screen the way you control the items on your desk.
While the'ldesktop' model forms the foundation of the Windows user
interface, Windows is not limited by it. In fact, several Windows
interface elements emulate other types of familiar devices, such as
slider controls, spin controls, property sheets, and toolbars. Windows
gives you, the programmer, a large array of features from which you
may choose those most appropriate to your specific application.

TIlE I/IIOUSE

Windows allows the use of the mouse for almost all control,
selection, and drawing operations. Of course, to say that it allows the
use of the mouse is an understatement. The fact is that the Windows
interface was designed for tile mot/se-it allows the use of the keyboard!
Although it is certainly possible for an application program to ignore
the mouse, it does so only in violation of a basic Windows design
principle.

Windows encourages the use of icons and bitmaps (graphics

. images). The theory behind the use of icons and bitmaps is found in
the old adage 'a picture is worth a thousand words.'
An icon is a small symbol that represents some operation or
program. Generally, the operation or program can be activated by
selecting the icon. A bitmap is often used to convey information
quickly and simply to the user. However, bitmaps can also be used as
menu elements.

MElIUS /UID DlllLOG BOXES

Aside from standard windows, Windows also provides several

special-purpose windows. The most common of these are the menu
and the dialog box. A menu is, as you would expect, a special wi:1dow
that contains choices from which the user makes a selection. The
 BUILDING" WINDOWS SKEUTON 473
'II'

thing that makes menus valuable is that they are largely automated.
Instead of having to manage menu selection manually in your
program, you simply create a standard menu-Windows will handle
the details for you.
A dialog box is a special window that allows more complex
interaction with the application than that allowed by a menu. For
example, your application might use a dialog box to request a file
name. With few exceptions, non-menu input is accomplished via a
dialog box.

OW WINDOWS AND YOUR PROGRAM

INTERACT

When you write a program for many operating systems, it is your

program that initiates interaction with the operating system. For example,
in a DOS program, it is the program that requests such things as input
and output. Put differently, programs written in the "traditional way"
call the operating system. The operating system does not call your
program. However, Windows generally works in the opposite way. It is
Windows that calls your program. The process works like this: Your
program waits until it is sent a message by Windows. The message is
passed to your program through a special function that is called by
Windows. Once a message is received, your program is expected to
take an appropriate action. While your program may call Windows
when responding to a message, it is still Windows that initiates the
activity. More than anything else, it is the message-based interaction
with Windows that dictates the general form of all Windows programs.
There are many different types of messages that Windows may send
your program. For example, each time the mouse is clicked on a
window belonging to your program, a mouse-clicked message will be
sent to your program. Another type of message is sent each time a
window belonging to your program must be redrawn. Still another
message is sent each time the user presses a key when your program
is the focus of input. Keep one fact firmly in mind: As far as your
program is concerned, messages arrive randomly. This is :why
Windows programs resemble interrupt-driven programs. You can't
know what message will be next.
One final point: Messages sent to your program are stored in a
message queue associated with your program. Therefore, no message
474 TIACH YOURSELf

will be lost because your program is busy processing another message.

The message will simply wait in the queue until your program is ready
for it.

NDOWS IS MYL TlTASKING

Since the start, Windows has been a multitasking operating system.

This means that it can run two or more programs concurrently. All
32-bit versions of Windows (such as Windows NT and Windows 95)
use preemptive multitasking. Using this approach, each active
application receives a slice of CPU time. It is during its time slice that
an application actually executes. When the application's ti~ slice
runs out, the next application begins executing. (The previously
executing application enters a suspended state in which it awaits
another time slice.) In this fashion, each application if) the system
receives a portion of CPU time. Although the application skeleton
developed in this appendix is not concerned with the multitasking
aspects of Windows, they will be an important part of any application
you create.

Older, /6-bit versions of Windows used a form of multitasking called non-

preemptive multitasking. With this approach, an application retained the CPU
until it explicitly released it This aI/owed applications to monopolize the CPU
and effectively "lock out" other programs. Preemptive multitasking eliminates
this problem.

E WIN32 API

In general, the Windows environment is accessed through a call-based

interface called the Application Program Interface (API). The API
consists of sevedl hundred functions that your program calls as
needed. The API functions provide all the system services performed
by Windows. There is a subset to the API called the Graphics Device
Interface (GDI), which is the pal'! of Windows that provides device-
. independent graphics support. It is the GDI fiInctions that make it
possible for a Windows application to TUn on a variety of hardware .
Programs designed for use by 32-bit versions of Windows, such as
Windows 95 and Windo"fs NT, use the Win32 AP.I. For the most part,
Win32 is a superset of the older Windows 3.1 API (WinI6). Indeed, for
 BUILDING A WINDOWS SKEUTON 475
THE COMPONENTS OF A WINDOW
'"

the most part, the functions are called by the same name and are used
in the same way. However, even though similar in spirit and purpose,
the two APIs differ because Win32 supports 32-bit addressing while
Winl6 supports only the 16-bit, segmented-memory model. Because of
this difference, several of the older API functions have been widened
to accept 32-bit arguments and return 32-bit values. A few API
functions have had to be altered to accommodate the 32-bit archi-
tecture . API functions have also been added to support preemptive
multitasking, new interface elements, and other enhanced features.
Because modern versions of Windows support 32-bit addressing, it
makes sense that integers are also 32 bits long. This means that types
int and unsigned are 32 bits long, not 16 bits, as is the case for
Windows 3.1. If you want to use a I6-bit integer, it must be declared as
short. Windows provides portable typedef names for these types, as
you will see shortly.

THE COMPONENTS OF A WINDOW

Before moving on to specific aspects of Windows programming, a few

important terms need to be defined. Figure C-I shows a standard
window with each of its elements pointed out.

fiGURE C-l System menu icon Title 1\1 inim ize box
The elements of a Maximize box
standard window <;:\o,sebox
...

Border _ _-oj
Vertical
scroll bar

Client area Horizontal scroll bar

478 WlCH YOURSUF
c

All windows have a border that defines the limits of the window; the
borders are also used when resizing the window. At the top of the
window are several items. On the far left is the system menu icon
(also called the title bar icon). Clicking on this box displays the system
menu. To the right of the system menu icon is the window's title . At
the far right are the minimize, maximize, and close boxes. The client
area is the part of the window in which your program activity takes
place. Most windows also have horizontal and vertical scroll bars that
are used to move information through the window.

ME WINDOWS APPLICA TlON BASICS

Before developing the Windows a pplication skeleton , some basic

concepts common to all Windows programs need to be discussed .

WlnMaln(J

All Windows programs begin execution with a call to WinMain( ) .

(Windows programs do not have a maine ) function.) WinMain( )
has some special properties that differentiate it from other functions
in your application. First, it must be compiled using the WIN API
calling convention. (You will see APIENTRY used as well. They both
mea~ the same thing.) By default, functions in your C programs use
the C calling convention. However, it is possible to compile a function
so that it uses a different calling convention; Pascal is a common
alternative . For various technical reasons, the calling convention
Windows uses to call WinMain( ) is WINAPI. The return type of
..
WinMain( ) should be into

THE WINDOW FUNCTION

All Windows programs must contain a special function that is not

called by your program, but is called by Windows. This function is
generally referred to as the window function or the window procedure.
The window function is called by Windows when it needs to pass a
message to your program. It is through this function that Windows
communicates with your program. The window function receives the
message in its parameters. All window functions must be declared as
 BUILDING A WINDOWS SKELETON 477
SOME IMNDOWS APPUCAnON BASICS

returning type LRESULT CALLBACK. The type LRESULT is a

typedef that, at the time of this writing, is another name for a long
integer. The CALLBACK calling convention is used with those.
functions that will be called by Windows. In Windows terminology,
any function that is called by Windows is referred to as a callback
function.
In addition to receiving the messages sent by Windows, the window
function must"'initiate any actions indicated by a message. Typically, a
window function's body consists of a switch statement that links a
specific response to each message that the program will respond to.
Your program need not respond to every message that Windows sends.
For messages that your program doesn't care about, you can let
Windows provide default processing. Since there are hundreds of
different messages that Windows can generate, it is common for most
messages simply to be processed by Windows and not by your
program.
All messages are 32-bit integer values. Furthermore, all messages
are linked with any additional information that the messages require .

WINDOW ClASSES

When you'r Windows program first begins execution, it will need to

define and register a window class. When you register a window class,
you are telling Windows about the form and function of the window.
However, registering the window class does not cause a window
to come into existence. To actually create a window requires
additional steps.

THE MESSAGE LOOP

As explained earlier, Windows communicates with your program by

sending it messages. All Windows applications must establish a
message loop inside the WinMain( ) function. This loop reads any
pending message from the application's message queue and dispatches
that message back to Windows, which then calls your program's
window function with that message as a parameter. This may seem to
be an overly complex way of passing messages, but it is, nevertheless,
the way all Windows programs must function. (Part of the reason for
I
this scheme is to return control to Windows so that the scheduler can
 .,
478 TEACH YOURSELf

allocate CPU time as .it sees fit rather than waiting for your
application's time slice to end.)

WINDOWS DATA TYPES

As you will soon see, Windows programs do not make extensive use
of standard C data types, such as int or char'. Instead, all data types
used by Windows have been typedefed within the WINDOWS.H file
and/or its related files. The WINDOWS.H file is supplied by your
Windows-compatible compiler and must be included in all Windows
programs. Some of the most common types are HANDLE, HWND,
BYTE, WORD, DWORD, UINT, LONG, BaaL, LPSTR, and
LPCSTR. HANDLE is a 32-bit integer that is used as a handle. As you
will see, there are a number of handle types, but they are all the same
size as HANDLE. A handle is simply a value that identifies some
resource. Also, all handle types begin with an H. For exampl.e, HWND
is a 32-bit integer used as a window handle. BYTE is an 8-bit unsigned
character. WORD is a 16--bit unsigned short integer. DWORD is an
unsigned long integer. UINT is a 32-bit unsigned integer. LONG is
another name for long. BOOL is an integer; this type is used to
indicate values that are either true or false. LPSTR is a pointer to a
string, and LPCSTR is a const pointer to a string.
In addition to the basic types described above, Windows defines
several structures. The two that are needed by the skeleton program
are MSG and WNDCLASSEX. The MSG structure holds a Windows
message, and WNDCLASSEX is a struchlre that defines a window
class. These structures will be discussed later in this appendix.

A WINDOWS SKELETON

Now that the necessary background information has been covered,

it's time to develop a minimal Windows application. As stated, all
Windows programs have certain things in common. This section
develops a Windows skeleton that provides these necessary features.
In the world of Windows programming, application skeletons are
commonly used because there is a substantial "price of admission"
when creating a Windows program. For instance, the short example
programs shown in this book are designed for a command-line
interface (such as DOS), in which a minimal program is about 5 lines
 BUILDING A WINDOWS SKELETON 479
SOME WlNOOWS APPUCATION BASICS
...

long. A minimal Windows program, however, is approximately 50

lines long.
A minimal Windows program contains two functions: WinMain.( )
and the window function. The WinMain( ) function must perform
the following general steps:

1. Define a window class.

2. Register that class with Windows.
3. Create a window of that class.
4. Display the window
5. Begin running the message loop.

The window function must respond to all relevant messages. Since

the skeleton program does nothing but display its window, the only
message that it must respond to is the one telling the application that'
the user has terminated the program.
Before considering the specifics, examine the following program,
which is a minimal Windows skeleton. It creates a standard window
that includes a title. The window also contains the system menu and
is, therefore, capable of being minimized, maximized, moved, resized,
and closed. It also contains the standard minimize, maximize, and
close boxes.
1* A minimal 32-bit Windows skeleton. *1

#include <windows . h>

LRESULT CALLBACK WindowFunc(HWND, UINT, WPARAM, LPARAM);

char szWinName[] = "MyWin"; 1* name of window class *1

int WINAPI WinMain(HINSTANCE hThislnst, HINSTANCE hPrevlnst,

LPSTR IpszArgs, int nWinMode)
(
HWND hwnd;
MSG msg;
WNDCLASSEX wcl;

1* Define a window class. *1

wcl.cbSize = sizeof(WNDCLASSEX); 1* size of WNDCLASSEX *1

wcl.hlnstance = hThislnst; 1* handle to this instance */

480 TEACH YOURSElf

'" C

wcl.lpszClassName = szWinNamej ;* window class name */

wcl.lpfnWndProc = WindowFunc; ;* window function *'
wel.style = 0; ;* default style *'
wcl.hlcon
weI. hlconSm
= Loadlcon(NULL.
= Loadlcon (NULL,
lDI_APPLICATION); /* icon style *'
IDI_W""INLCX:;O); /* small icon style */

wcl.hCursor = LoadCursor(NULL, rOC_ARROW); /* cursor style */

wcl.lpszMenuName = NULL; /* no menu */

weI. cbClsExtra = 0; /* no extra */

wcl.cbWndExtra = 0; ,
/* information needed */

j- Make the window background white. */

wcl.hbrBackground = (HBRUSH) GetStockO~ject(WHITE_BRUSH)j

1* Register the window class. *j

if(!RegisterClassEx(&wcl» return 0;

/* Now that a window class has been registered, a window

can be created. */
hwnd = CreateWindow(
szWinName, 1* name of window class *1
~Windows Skeleton /* title * I
M
,

WS_OVERLAPPEDWINDOW, /* window style - normal *1

.CW_USEDEFAULT, /* X coordinate - let Windows decide */
CW_USEDEFAULT, 1* Y coordinate - let Windows decide */
CW_USEDEFAULT, 1* wfdth - let Windows decide * I
CW_USEDEFAULT, /* height - let Windows decide */
HWND_DESKTOP, 1* no parent window *1
NULL, 1* no menu *1
hThisInst, 1* handle of this instance of the program */
NULL 1* no additional arguments *1
};

I * Display the window. */

ShowWindow{hwnd, nWinMode);
UpdateWindow{hwnd) i

/* Create the message loop. *1

while {GetMessage {&msg, WJLL. 0, 0)
(
 tIUILDIIIG,,-~ 48'
SOME WlNllDM N'PUCATIDN MSICS
"
TranslateMessage(&msg)i I~ translate keyboard messages */
DispatchMessage(&msg); /* return control to Windows */
}
return msg~wParam:

/ * This function is called by Windows and is passed

messages from the message queue.
Of

LRESULT CALLBACK WindowFunc(HWND hwnd. UI~ message.

WPARAM wParam. LPARAM IParam)
(
switch(message) (
case WM_DESTROY: 1* terminate the program *1
PostQuitMessage(O);
break;
default:
1* Let Windows , process any messages not specified in
the preceding switch statement. */
return OefWindowProc(hwnd. message, wParam, IParam);
}
return 0;
}

The window produced by this program is shown in Figure C-2. Now

let's go through this program step by step.
First, all Windows programs must include the header file
WINDOWS.H. As stated, this file (along with its support files) contains
the API function prototypes and various types, macros, and definitions
used by Windows. For example, the data types HWND and
WNDCLASSEX are defined in WINDOWS.H.
,
• The window function used by the program is called WindowFunc( ).
It is declared as a callback function, because this is the function that
Windows calls to communicate with the program.
Program execution begins with WinMain( ). which is passed four
parameters. hThislnst and hPrevlnst are handles. hThislnst refers
to the current instance of the program. Remember, Windows is a
multitasking system, so more than one instance of your program may
be running at the same time. hPrevlnst will always be NULL. (In
482 TEACH YOURSW

.
FIGURE C-2 Window, ')~ .. I..t".. j " f - j-

The window
. produced by the
Windows skeleton
T

Windows 3.1 programs, hPrevInst would be non-zero if there were

other instances of the program currently executing, but this doesn't
apply to 32-bit versions of Windows.) The IpszArgs parameter is a
pointer to a string that holds any command line arguments specified
when the application was begun. The nWinMode parameter contains
a value that determines how the window will be displayed when your
program begins execution .
Inside the function, three variables are created.' The hwnd variable
will hold the handle to the program's window. The msg structure
variable will hold window messages, and the weI structure variable
will be used to define the window class.

The first two actions that WinMain( ) takes are to define a

window class and then register it. A window class is defined by filling
in the fields defined by the WNDCLASSEX structure. Its fields are
shown here:
UINT cbSize: / * size of the WNDCLASSEX structure * /
UINT style; /* type of window */
WNDPROC IpfnWndProc; / * address to window tunc * /
 BUILDING A WINDOWS SIU1£1'ON 483
y .
SOME WINDOWS APPI./CAnON BASICS

int cbClsExtra: /* extra class info *j

lnt cbWndExtra: j* extra window info */
HINSTANCE hlnstance; /* handle of this instance */
HICON hleon; /* handle of standard icon */
HICON hlconSm; / * handle of small icon
HCURSOR hCursor; /* handle of mouse cursor */
*'
HBRUSH hbrBackground; /* background color *j
LPCSTR IpszMenuName: /* name of main menu */
LPCSTR IpszClassName; /* name of window class */

As you can see by looking at the program, cbSize is assigned the size
of the WNDCLASSEX structure. The hlnstance field is assigned the
current instance handle as specified by hThislnst. The name of the
window class is pointed to by IpszClassName, which points to the
string "MyWin' in this case. The address of the window function is
.' assigned to IpfnWndProc. No default style is specified, and no extra
information is needed.
All Windows applications need to define a default shape for the
mouse cursor and for the application's icons. An application can define
its own custOln version of these resources or it may use one of the
built-in styles, as the skeleton does. In either case, handles to these
resources must be assigned to the appropriate members of the
WNDCLASSEX structure. To see how this is done, let's begin
with icons.
A modern Windows application has at least two icons associated
with it: one standard size and one smalL The small icon is used when
the application is minimized and it is also the icon that is used for the
system menu. The standard icon is displayed when you move or copy
an application to the desktop. Typically, standard icons are 32 by 32
bitmaps and small icons are 16 by 16 bitmaps. The style of each
icon is loaded by the API function Loadlcon( ), whose prototype is
shown here:

HICON Loadlcono-IINSTANCE hlnst, LPCSTR IpszName);

This function returns a handle to an icon. Here, hlnst specifies the

handle of the module that contains the icon and the icon's name is
specified in lpszName. However, to use one of the built in icons, you
484 TEACH YOURSELf
c

must use NULL for the first parameter and specify one of the
following macros far the second:

Icon Macro Shape

IDI_APPLICATION Default icon
IDI_ASTERISK Information icon
IDI_EXCLAMATION Exclamation point icon
IDI_HAND Stop sign
IDI_QUESTION Question mark icon
IDI_WINLOGO Windows Logo

In the skeleton , IDI - APPLICATION is used for the standard icon

and IDCWINLOGO is used for the small icon.
To load the mouse cursor, use the LoadCursar( ) API function .
This function has the following prototype:

HCURSOR LoadCursor(HINSTANCE hlnst, LPCSTR IpszName);

This function returns a handle to a cursor resource . Here, hI11se

specifies the handle of the module that contains the mouse cursor, and
the name of the mouse cursor is specified in IpszNa111e. However, to
use one of the built-in cursors, you must use NULL for the first
parameter and specify one of the built-in cursors, using its macro, for
the second parameter. Some of the most common built-in cursors are
shown here:

Cursor Macro Shape

IDC_ARROW Default arrow pointer
IDC_CROSS Cross hairs
IDCJBEAM Vertical I-begm
IDC_WAIT Hourglass

The background color of the window created by the skeleton is

specified as white, and a handle to this brush is obtained using the API
function GetStackObject( ). A brush is a resource that paints the
screen using a predetermined size, color, and pattern. The function
GetStockObject( ) is used to obtain a handle to a number of standard
 BUILDING'" WINDOWS SKELETON 485
'"

display objects, including brushes, pens (which draw lines), and

character fonts. It has this prototype:

HGDIOBJ GetStockObjectCint object);

The function returns a handle to the object specified by object. (The
type HGDIOBJ is a GDI handle.) Here are some of the built-in
brushes available to your program:

Brush Macro Background T~l!e

BLACK_BRUSH Black
DKGRAY_BRUSH Dark gray
HOLLOW_BRUSH See-through window
LTGRAY_BRUSH Light gray
...
WHITE_BRUSH White

You can use these macros as parameters to GetStockObject( ) to

obtain a brush.
Once the window class has been fully specified, it is registered with
Windows using the API function RegisterClassEx( ), whose prototype
is shown here:

ATOM RegisterClassEx(CONST WNDCLASS 'lpWClass);

The function returns a value that identifies the window class. A TOM
is a typedef that means WORD. Each window class is given a unique
value. /pWClass must be the address of the WNDCLASSEX structure.

CREATIIIG A ""'''DOW

Once a window class has been defined and registered, your

application can actually create a window of that class using the API
function CreateWindow( ), whose prototype is shown here:

HWND CreateWindow(
LPCSTR IpClassName, 1* name of window class '/
LPCSTR IpWinName,1* title of window '/
DWORD dwStyle, /' type of window '/
int X. int Y.I* upper-left coordinates '/
488 TtACH YOURSElF

'" C

int Width, int Height, I' dimensions of window '/

HWND hParent, I' handle of parent window '/
HMENU hMenu, I' handle of main menu '/
HINSTANCE hThislnst,1' handle of creator '/
LPVOID IpszAdditionall' pointer to additional info '/
);

As you can see by looking at the skeleton program , many of the

parameters to CreateWindow( ) may be defaulted or specified as
NULL. In fact, most often the X, Y, Width, and Height parameters will
simply use the macro CW_USEDEFAULT, which tens Windows to
select an appropriate size and location for the window . If the window
has no parent, which is the case in the skeleton, then hParent must be
specified as HWND_DESKTOP. (You may also use NULL for this
parameter.) If the window does not contain a main menu, then hMenu
must be NULL. Also, if no additional information is required, as is
most often the case, then lpSzAddlMol1al is NULL. (The type LPVOID
is typedefed as void' . Historically, LPVOID stands for 'long pointer
{o void .')

The remaining four parameters must be set explicitly by your

program. First, lpszClassNClJne must point to the name of the window
class. (This is the name you gave it when it was registered.) The title
of the window is a string pointed to by lpszWinName. This can be a
null string, but usually a window will be given a title. The style (or
type) of window actually created is determined by the value of
dwStyle. The macro WS_OVERLAPPEDWINDOW specifies a standard
window that has a system menu, a border, and minimize, maximize,
and close boxes. While this style of window is the most common, you
can construct one to your own specifications. To accomplish this,
simply OR together the various style macros that you want. Some
other common styles are shown here :

Style Macros Window Feature

WS_OVERLAPPED Overlapped window with border
WS_MAXIMIZEBOX Maximize box
•
WS_MINIMIZEBOX Minimize box
WS_SYSMENU System menu
WS_HSCROLL Horizontal scroll bar
WS_VSCROLL Vertical scroll bar
 BUILDING A WINDOWS SKElETON 487
SOME WINDOWS APPt.ICAnON BASICS

The I1Thislnst parameter must contain the Cllrrent instance handle of

the application.
The CreateWindow( ) function returns the handle of the window it
creates or NULL if the window cannot be created.
Once the window has been created, it still is not displayed on the
screen. To cause the window to be displayed, call the ShowWindow( )
API function. This function has the following prototype:

I BOOl ShowWindow(HWND hwnd, int nHow);

The handle of the window to display is specified in hwnd. The display

mode is specified in nHoW. The first time the window is displayed, you
will want to pass WinMain( )'8 nWinMode as the nHow parameter.
Remember, the value of nWinMode determines how the window will
be displayed when the program begins execution. Subsequent calls can
display (or remove) the window as necessary. Some common values
for nHow are shown here:

Display Macros Effect

SW_HIDE Removes the window
SW_MINIMIZE Minimizes the window into an icon
SW_MAXIMIZE Maximizes the window
~i
SW_RESTORE Returns a window to normal size
I!
The ShowWindow( ) function returns the previous display status
of the window. If the window was displayed, then nonzero is returned.
If the window was not displayed, zero is returned.
Although not technically necessary for the skeleton, a call to
UpdateWindow( ) is included because it is needed by Virtually every
Windows application that you will create. It essentially tells Windows
to send a message to your application that the main window needs to
be updated.

The final part of the skeletal WinMain( ) is the message loop. The
message loop is a part of all Windows applications. Its purpose is to
receive and process messages sent by Windows. When an application
is running, it is continually being sent messages. These messages are
488 TEACH YOURSElF
~
C

stored in the application's message queue until they can be read and
processed. Each time your application is ready to read another message,
it must call the API function GetMessage( ), which has this prototype:

BOOL GetMessage(lPMSG msg, HWND hwnd, UINT min, UINT max);

The message will be received by the structure pointed to by msg.

All Windows messages are contained in a structure of type MSG,
shown here:
/* Message structure */
typedef struct tagMSG
~
HWND hwnd; /. window that message is for ./
UINT message; /. message ./
WPARAM wParam; /. message-dependent info ./
LPARAM IParam; /. more message-dependent info ./
DI,ORD time; /. time message posted ./
POINT pt; /. X,Y location of mouse "I'
t1SG;

111 MSG, the handle of the window for which the message is intended
is contained in hwnd. All Win32 messages are 32-bit integers, and the
message is contained in message. Additional information relating to
each message is passed in wParam and IParam. The type WPARAM
is a typedeffor UINT, and LPARAM is a typedeffor LONG.
The time the message was sent (posted) is speCified in milliseconds
in the time field.
The pt member will contain the coordinates of the mouse when the
message was sent. The coordinates are held in a POINT structure,
which is defined like this:
typedef struct tagPOINT {
LONG x, y;
) POINT;

If there are no messages in the application's message queue, then a

call to GetMessage( ) will pass control back to Windows.
The I,Wl1d parameter to Get!l1essage( ) specifies the window for
which messages will be obtained. It is possible, and even likely, that an
application will contain several windows, but you only want to receive
messages for a specific window. If you want to receive all messages
directed at your application, this parameter must be NULL.
 The remaining two parameters to GetMetllBge( ) specify a range of
messages that will be received. Generally, you want your application
to receive all messages. To accomplish this, specify both min and max
as 0, as the skeleton does.
GetMeaaage( ) returns zero' when the user terminates the program,
causing the message loop to terminate. Otherwise it returns nonzero.
Inside the message loop, two functions are called. The first is the
API function TranalateMeaaage( ). This function translates raw ,
keyboard input into character messages. Although it is not necessary
for all applications, most applications call TranalateMeMage( )
because it is needed to allow full integrationofthe keyboard into your
flPplication program.
Once the message has been read and translated, it is dispatched
back to Windows using the DispatcbMetluge( ) API function.
Windows then holds this message-until it can be passed to the
program's window function.
Once the message loop terminates, the WinMain( ) function ends
by returning the value of msg. wParam to Windows. This value
.
contains the return code generated when your program' terminates.
,

E WINDOW FUNCTION

The second function in the application skeleton ,is its window function.
In this case, the function is called WindowFunc( ). but it could have
any name you like. The window function is passed the first four
members of the' MSG structure as parameters. For the skeleton,
the only parameter used is the message itself. However, actual
applications will use the other parameters to this function.
The skeleton'S window function responds to only one message
explicitly: WM_DESTROY. This message is sent when the USer
terminates the program. When this message is received, your program
must execute a call to the API function PostQuitMessage( ). The
argument to this function is an exit code that is returned in
msg.wParam inside WinMain( ). Calling PostQuitMessage()
Ci!uses a WM_QUIT me.ssage to be sent to your application, which
causes GetMessage( ) to return false. thus stopping your program.
Any other messages received by WindowFunc( ) are passed to
Windows, via a call to DefWindowProc( ), for default processing.
This step is necessary because all messages must be dealt with in one
fashion or another.
490 TfACH YIlUASElS

'" C

SHORT WORD ABOUT DEFINITION

FILES

You may have heard or read about definition files. For 16-bit versions
ofWjndows, such as 3.1, progtams need to have a definition file
associated with them. A definition file is simply a text file that
specifies certain information and settings required by a Windows 3.1
program. However, because of the 32-bit architecture (and other
improvements) of modern versions of Windows, definition files are no
longer needed.

MING CONVENTIONS
-'
Before concluding this appendix, a shan comment on the naming of
functions and variables needs to be made. Several of the variable and
parameter names in the skeleton program and its description probably
seemed rather unusual. This is because they follow a set of naming
con~ntions that was invented for Windows programming by
Microsoft. For functions , the name consists of a verb followed by a
noun . The first character of the verb and noun is capitalized.
For variable names, Microsoft chose to use a rather complex system
of embedding the data type into the name. To accomplish this, a
lowercase type prefix is added to the start of the variable's name. The
name itselfbegins with a capital letter. The type prefixes are 6hown in
Table C-l. Frankly, the use of type prefixes is controversial·and is'not
universally supported. Many Windows programmers use this method,
but many do not . You are free to use any naming convention you like.

LEARN MORE

The foregoing overview of Windows programming just scratches the

surface. In order to write Windows programs that are useful, you must
learn mueh more about Windows programming. To learn more about
Windows 9S programs you will want to read the following books:

Schildt's Windows 95 Programming in C and C++

Schildt's Advanced Windows 95 Programming in C and C++

 BUILDING A WINDOWS SKELETON 491
TO LEARN MORE "
,

Prefix Data Type

b Boolean (one byte)
c Character (one byte)
dw Long unsigned integer
I 16-bit bit-field (flags)
In Function
h Handle
Long integer
Ip Long pointer
n Short integer
p Pointer
pt Long integer holding screen coordinates
w Short unsigned integer
52 Pointer to null-terminated string
Ipsz Long pointer to null-terminated st[ing
rgb Long integer holding RGB color values

'ABU C-l Variable Type Prefix Characters "

To learn more about Windows NT programming, you will find

Windows NT 4 Programming From the Ground Up

especially useful. These books are written by Herbert Schildt and

published by Osborne/ McGraw-Hili.
-. ,
 o
Answe~

c
..." lIACIIYOURSaF
C

APTER 1

ERCISES

2 . • include <stdio.h>

int main (void)

{
int num;

num • 1000;
printf(-'d is the value of num-, num};

return 0;
}

ERCISES

2. 'include <stdio.h>

ine main (void)

(
float 4, b;

printf(-Enter two numbers: .);

scanf(-'f-, &4);
scanf("\f", &b);
printf(-Their sum is 'f.-, a+b);

return 0;
}

£RCISES

1. linclude <stdio.h>

int main (void)

(
int len, width, height;

printf("Enter length: ");

 ,
"

scanf("\d", &len);

printf(-Enter width: .);

scanf("'d", &width);

printf("Enter height: "); •

scanf("'d", &height);

printf("Voiume is 'd.", len· width· height);

return 0;
}

2. linclude <stdio.h>

int main(voidl
(
printf(·Number of seconds in a year: .):
printf("\f",' 60.0 • 60.0 • 24.0 • 365.0);

return 0;
}

2. Yes, a comment can contain nothing.

,
3. Yes, you can temporarily remove a line of code from your
program by making it into a comment. This is sometimes called
'commenting out' a line of code.

2. #include <stdio.h>

void one (void) ;

void two (void) ;

int main(void)
(
oneIl:
two () ;
·1. --
c

return 0;
)

void one (void)

(
, printf{"The summer soldier, "); •
)

void two (void)

(
printf("the sunshine patriot.");
)

3. The compiler will report an error. The prototype is needed in

order for the compiler to properly call funcl( ).

ERCISES
2. linclude <stdio. h>

int convert (void) ;

int main (void)

(
printf("%d", convert());

return Or
)

int convert(void)
(
int dollar.:

printfC-Enter number of dollars: .);

scanf(~%d·, &dollars);
return dollars I 2;
}

3. There is nothing technically wrong with the program. However,

function fl( ) returns an integer value, but it is being assigned
to a variable of type double. This would lead one to suspect
that perhaps the programmer has misunderstood the purpose of
the fl ( ) function.
 N .... _

MAS7lRY SlaUS atECI( •

4. A function declared with a void return type cannot return

a value.

ERCISES
.
1. 'include <stdio.h>

void outnum(int num):

int mAin (void)

{
outnum(lO) ;

return 0;
}

void outnum(int num)

(
printf (10 %d 10. num);
}

2. The sqr_it( ) function requires an integer argument, but it is

called with a floating-point value.

STERY SKILLS CHECK

1. tinclude <stdio.h>

int main (void) .

{
float weight;

printf(IOEnter your weight: 10);

scanf(IO%f &weight):
lO
,

printf("Effective moon weight: 'f-, weight· 0.17);

return 0;
}

2. The comment is not terminated with a */.

3. 'include <stdio.h>
. . . 1IACH YOURSW'

" C

int main (void)

{
int ouncesr
int cups;

printf(MEnter ounces: -):

scanf ( • %d M. &tounces):

cups = o_to_c{ounces):
printf{"'d cupS", cups)/

return 0;
}

int o_to_c(int 0)
{
return 0 I 8;
}

4. char, int, float, double, and void.

5. The variable names are wrong because
a. A dash may not be used in a variable name.
b. A dollar sign may not be used in a variable name.
c. A + sign may not be used in a variable name.
d. 'A digit may not begin a variable name .

. . . . . CHAPTER 2

. . . . . REVIEW 'SKILLS CHECK

1. All prograljl1s must have a maine ) function . This is the first
function called when your program begins executing.
2: fHnclude <stdio. h>

int main (void)

{
printf(-This is the number %d M
, 100);

I
 -498
21 EXERCISES ~

return 0;
)

3. To include a header file, use the #inc1ude compiler directive.

For example,
'include <stdio.h>

includes the STDIO.H header.

4. The five basic data types are char, int, f101lt, double, and void.
S. The invalid variable names are b, c, and e.
6. The acanf( ) function is used to input information from the
keyboard.
7 .• include <8tdio.h>

int main (void)

(
int i;

printf(-Enter a number: .);

scanfl"d", &i):
printf(-'d-, i*i);

return 0;
)

8. Comments must be surrounded by the /. and' / comment

symbols. For example, this is a valid C comment.
/* This is a comment. */

9. A function returns a value to the calling routine using return.

10. void Myfunc(int count, float-balance, char chl;

XERCISES

1. b, d, and e are true.

2. 'include <stdio.h>

int main (void)

{
int i;
&00
...

print£(*Enter a number: .);

scanf("%d", &i);
if( (i%2)==0) printf("Even");
if((i%2)==1) printf("Odd");

return 0;
)

ERCISES

1. 'include <stdio.h>

int main(void)
(
int a, b. op;

printf(·Enter first number: .);

scanf("'d", &a);

printf{*Enter second number: .);

scanf (·'d*. &b);

printf("Enter 0 to add, 1 to multiply: ");

scanf("'d", &op);

if(op==O) printf("'d", a+b);

else print£ (·'d·. a-b);

return 0:
)

2. 'include <atdie. h>

int main (void)

(
int i;

printf(-Enter a number: .);

scanf (-'d-, &i);
if((i%2)==0) printf("Even");
else printf("Odd");
'.
 2..
-
EXERCISES
1101
•

return 0;
)

ERCISES
1. linclude <stdio.h>

int main(void)
(
int a, b. op;

printf(·Ertter 0 to add, 1 to subtract: .);

scanf("'d", &Op);

if(op==O) ( J* add *J
printf("Enter first number: .);
scanf ("\d", &a);
printf(·Enter second.number: ·)i
scanf("\d", &b),
printf(·'d·, a+b);
)
else { ,- subtract -,
printf,"Enter first number: .);
scanf("'d", &a);
printf("Enter second number: ~);
scanf ("'d", &b);
printf("'d", a-b);
)

return 0;
)

2. No, the opening curly brace is missing.

fRelSES
,

1. 'include cstdio.~>

int lIIain(yoid)
(
int i; .

for(i=l; i<101; i=i+1) printf("'d" i);

102 WICII YOURSELf

" C

return 0;
}

2. 'include <stdio.h>

int main (void)

{
int i;

for(i=l7: i<101; i=i+1)

i£«i'171==01 print£("'d" il;

return 0:
}

3. iinclude <stdio.h>

int mainevoid)
( :
t
int num, i;

printfC-Enter the numPer to test: .);

scanf ( • %d·, &num);

for(i=2; i«num/2)+1: i=i+1)

if«num%iI==OI print£("'d" iI,
•
return 0:
}

ERCISES

1. 'include <stdio.h>

int main (void)

{
int i;

for(i=l; i<101; i++1 printf("'d" il;

return 0:
}
 MIlt •

• include <stdio.h>

int main (void)

(
int i;

forti=17; i<101; i++)

if«i\17)==0) printf("\d" i

return 0;
}

'include <stdio.h>

int main (void)

{
int nwn. i;

printf(-Enter the number to test: ·)i

scanf ( - %d -, &nwn);

for(i=2; i«num/2)+1; i++)

if«numh)==O) J?!intf("'d" i);

return 0;
}

2.• include <stdio.h>

int main(void)
{

a := 1;
a++;
b := ai
b--;
printf ("'d \d", a, b);

return 0;
}
..."
ERCISES

1. .include <stdio.h>

int main (void)

(
int i;

for(i=l; i<11; i++)

printf("'d 'd 'd\n", i, i"'1. i*i*i):

return 0;
)

2. • include <stdio.h>

int main (void)

{
int i. j; .
,
printf (-Enter a number: .);
scanf("\d", &i);

for(j=i; j>O; j--) printf("'d\n '

printf(·\a-);

return 0;
)

ISES
1. The loop prints the numbers 0 through 99.
2. Yes .
•
3. No, the first is true, the second is false.

STERY S ILLS CHECK

1 .• include <stdio.h>

int 'main(void)
{
int magic; /* mag~c number *'
 int guess: ,- user's que •• -,
int i:

magic = 1325:
guess = 0;

for(i=O; i<10 && guess!=magic; i++) {

printf(-Enter your guess: -):
scanf ( - %d -, &guess);

if(guess == magic) {
printf(-RIGHT!-):
printf(- %d is the magic number.\n-, magic);
)
else (
printf(- ... Sorry, you're wrong ... -);
if(guess > magic)
printf (- Your guess i . too high. \n-) ;
else printf(- Your guess is too low.\n-);
)
)
return 0;
)

2. 'include <stdio.h>

int main (void)

(
int rooms, len, width, total:
int i;

printf{-Number of rooms? -);

scanf ( - 'd -, "rooms);

total = 0:
for (i=rooma: i>O; i--) {
printf('Enter length: '):
Bcanf(-'d-, &len):

printf('Enter width: .,:

scanf ("d', "width):

total = total • len - width:

I
prtntf('Total square footage: 'd', total):
 508 TfACH YOURSElJ'

" c

return 0;
}

3. The increment operator increases a variable by one and the

decrement operator decreases a variable by one.
4. #include <stdio. h>

int main(void)
(
int answer, count;
int right, wrong:

right = 0;
wrong = 0:

• for(count=l; count < 11; count=count+l" (

printf("What is 'd + 'd? ", count, count);
scanf(*'d-, &Anawer):

if(answer == count+count) (
printf ( " Rillht ! " ) ;
..,right++1
}
else (
printf(*Sorry. you're wrong . • );
printf(-The answer is 'd. ., count+count);
wrong.+;
}
}

printf(*you got %d right and %d wrong.-, right. wrong);

re.turn 0;
}

5. 'include <stdi6.h>

int main (void)

(
int i:

for(i=l; i<=100; i •• ) (
printf("'d\t", i);
 -1107
REVIEW SJ(JLLS CHECK ~

iflli%5)==0) printfl"'n");
)

return 0:
)

APTER 3

W , SKILLS, CHECK
1, C's relational and logical operators are <, >, < =, > =, ! =, = =,
I, &&, and II,
2, A block of code is a group oflogically connected statements, To
make a block, surround the statements with curly braces,
3, To output a newline, use the \n backslash character code,
4. #include <stdio.h>

int main(void)
1
int i;

for(i=-lOO; i<101; i++) printf(M%d" i);

return 0:
)

5 . • include <stdio.h>

int main (void)

(
int i;

printf(-Enter proverb number: .);

scanf ( • %d·, &i);

Hli==l) printfl"A bird in the hand" ,");

i£(i==2) printfC"A rolling stone .. 0·);
if(i==3) printfC·Once burned, twice shy.-);
i£(i==4) printfC-Early to bed, early to rise ... ·);
i£(i==5) printf("A penny saved is a penny earned."};
IS08 TEACltY~LI

'" C

return 0;
}

6. count++;
/* or */
++count;

7. In C, true is any nonzero value. False is zero.

ERCISES
1. 'include <stdio.h>
'include <conio.h>

int main (void)

(
int i;
char ch. smallest;

printf(·Enter 10 letters.\n-);

smallest = 'z' /* make largest to begin with */

for(i=Q; i<10; i++) {

ch = getche ( ) ;
if(ch < smallest) smallest = chi
}
printf(-'nThe smallest character is 'c.-, smallest);

return 0;
)

2. 'include <stdio.h>

iDt main (void)

(
char chi

for(ch='A', ch<c'Z'; ch++)

printf("'d ", chI;

printf(~'n");

for'cb='a'; ch<='z'; ch++'

 #If'_? I" 1M
*#

printf ("M" ch);

return 0;
}

The codes differ by 32.

ERCISES
1. The else relates to the first if; it is not in the same block as
the second.
2 .• include <stdio.h>

int main (void)

(
char ch:
int sl.. s2:
float radius;

printf'-Compute area of Circle. Square, or Triangle? .);

ch = getchar();
printf (" In") ;

if(ch=='C') (
printf(-Enter radius of circle: -);
scanf(-'f-, .radius);
print£(-Area is: 'f-, 3.1416*radius*radius);
}
else it(ch.~'S') (
printf("Enter length of first ~ide: ");
scanf(-'d-, &sl);
printf(-Enter length of •• cond side: -);
Icanf(-'d-, &82);
printf("Area i.: 'do, al-.2);
}
elaeif(ch=='T') (
printf("Enter length of baa.; ");
acanf{-'d-, &al);
printf("Enter height: ");
scanf("'d", ~s2);
printf("Area is: 'do, (al-.2}/2);
}
 ~-,
c

return 0;
}

XERCISES

,
1. #include <stdio.h>

int main (void)

(
float dist. speed;
int num;

printf(-Enter number of drive time computations: .);

scant ("%d", &nwn);.

for (; nwn; num-- ) (

printf(-'nEnter distance: .);
scant ("%t", &di.t);

printf(-Enter average speed: .);

seanfeM'f-, &speed);

printf(-Drive time is %£\n-, dist/speed);

}

• return 0;
}

2. 'include <stdio.h>

int main (void)

(
int i;

printf(-Enter a number: .);

scant ("%d", &i);.

fore ; i: i--)

printt (" \a") ;

 MllJdERS 811
"
return 0;
)

3 .• include <stdio.h>

int main(void)
(
int i;

for(i-l; i<lOOl; i-i+i) printf("%d" i);

return 0;
}
•

1. 'include <stdio.h>

int main (void)

(
float dist. speed;
int nwn;

printf(-Enter number of drive time computations: -);

scanf(-'d-, &num);

while (nurn) (
printf(-\nEnter distance: .);
scanf ("\t", &dist) I.
•
printf(~Enter average speed: -)~
scanf (-'f-, &speed);

printf("Drive time is 'f'n", dist/speed);

num--;
}

return 0;
}

2. tinclude c'stdio. h.
'include <conio.h>
·,2
.,

int main (void)

{
char ch.

printf(-Enter your encoded measage.\n-};

ch = getche();
while(ch!='\r') (
printf(-'t-, ch-l);
ch = getche () ;
I

return 0;
} •

ERCISES
1. UDelude <stdio. h>

int main (void)

(
float gallons;

printf("\nEnter gallons: ");

scanf("'f", &gallons);

dol
printf("Liters: 'f\n", gallons·3.7854);

printf(-Enter gallons or 0 to quit . • );

scanf (··'f-. ~CJallon.);

} while(gallons!=O);

return 0:
I

2 .• include <atdio. ,,>

int main('lOid)
{
int choice;

printf("Kailing list =enu:\n\n"';

 u -
EXERCISES
1513
~.

printf(- l. Enter addresses\n-);

printf (" 2. Delete addresses\n-);
printf(" 3. Search the list\n-):
printf(- 4. Print the list\n");
printf(- 5. Quit\n") ;

do (
printf(-Enter the number of the choice (1-5): .);
scanf(·'d-, &choice);
} while(choice<1 I I choice>5);

return 0;
)

s
1. /* This program finds the prime numbers from
2 to 1000.
"I

tinclude <stdio.h>

int main(void)
(
int i, j, prime;

for(i=2j i<1000; i++) (

prime = 1;
for(j=2; j <= i/2; j++)
if(!(i\j» prime=O;
if (prime) printf{-'d is prime.\n-, il;
)

return 0;
)

2. 'include <stdio.h>
'include <conio.h>

int rnain(void)
(
int i:
char chi
 .,
&14 TEACH YOURSElF

for(i=O; i<10; i++) (

printf(·\nEnter a letter: .);
ch = getche () ;
printf ( • \n·) ;
fore ; chi ch--) prinlf'-'c·, ','):
)

return 0;
}

XERCISES

2. #include <stdio. h>

#include <conio.h>

int main (void)

(
float i;
char chi

printf(ftTip Computer \ n");

for(i=1.0; i<101.0; i=i+1.0) (

printf(ft%f %f %f Itf\n-, i, i+i*,1. i+i*.15, 1+1'*.2);
printt("l1ore? (YIN) .);
ch = getche ( ) ;
printf(- ' n-) ;
if(ch=='N') break;
}

return 0;
}

ERCISES

1. .include <8tdio.h>

int main (void)

(
int i;

forti;:!; i<101; i++) {

if(! (1%2» continue;
 .1.'"
printfC-%:d - i);
)

return 0;
)

XERCISES

1. Floating point values may not be used to control-switch..

2. jinclude <stdio.h>
'include <conio.h>

int main(void)
(
char ch;
int digit, punc, letter;

printfC-Enter characters, ENTER to stop.\n-);

digit = 0;
punc = 0;
letter = 0;

do (
ch = getche () ;
switchCch) (
case '1' !
case • 2' !
case • 3' :
case '4' :
case ' 5 • :
case '6' :
case • 7' !
case • 8' :
case '9' :
case • O· :
digit++;
break;
case , , .
case , , , ..
case .? :
case • ! • :
·
case ...
 - ._- -"'"

51. TlACIf YOURSElf

c

case 'i':
punc+-+;
break;
default:
letter+"';
)
while(ch!=' \ r') ;
printf(- \ nDigits: %d \ n-, digit) i
printf(-Punctuation: %d\n-, punc);
printf{-Letters: %d\n-, let~er);

return 0;
)

ERCISES

I. #include <stdio.h>

int main(void)

int i;

i = 1;

jump_label,
if(i>=11) goto done_label:
printf(-'d -, il;
i+-+;
goto jump_label;
done_label: printf{-Done-);

return 0;
)

ASTERY SKILLS CHECK

1. #include <stdio.h>
#include <conio.h>

int main(void)
{
char Chi
 - .&17
MAS1!RY SKJLLS CHECK '"

printf(-Enter lowercase letters . • );

printf(" (Press ENTER to Quit.)\n");
do (
ch = getche () :
iflch!='\r') printf("'c", ch-32);
} whilelch!='\r'):

return 0;
}

2. #include <stdio.h>

int main(voidl
(
int i:

printf("Enter a number: "J;

scanf(-'d", &i);

if(!i) printf("zero");
else i£(i<O) printf("negative");
else printf ("positive") i

return 0;
}

3. The for loop is valid. C allows any of its expressions to be empty.

4. for ( : : )

5. /* for *1
far(i=1: i<11: i++) printf("td" i):

1* do *1
i = 1:
do {
printf ("M ", i):
i++:
} whileCi<11);

/* while */
i=1:
while (i<11) {
printf("%d" i);
1518 TfACM YOIlRIInF
c

i++;
)

6 The break statement causes immediate te, · lination of the loop.

7. Yes.
8. No, the label is missing the colon.

UMULATIVE SKILLS CHECK

1. #include <stdio.h>
#include <conio.h>

int main(void)
{
char chi

printf(ftEnter characters (q to q~it): \n");

do {
ch = getche ( ) i
swi tch (ch) {
case ' \ t ' : printf("tab\n");
break;
case '\b': printf{"backspace\n");
break;
case '\r': printf("Enter\n");
)
} while(ch!='q');

return 0;
)

2. include <stdio. h>

int rnain(void)
(
int i, j, k;

for{k=O: k<lO; k++) { j* use increment operator *;

printi ("Enter first number: .. );
scanf("%d", &i);

printfC"Enter second number: ");

scanfC"'d", &j);
 ANSWERS 51.
REVIEW SKiUS CHECK Y

if(j) printf("'d\n", i/j); /. simplify condition ./

else printf(-Cannot divide by zero. \n"); ,' .. use else 'Ill
)

return 0;
)

APTER ..

VIEW SKILLS CHECK

1. int i;
forei=1; i<11: i++) printf(-'d to il; -
i = 1;
do {
printfl"%d" i);
i++i
) while ( i<ll);

i ::!: 1 :
while I i<l1l {
printfl"\d" il.
i++;
)

2. switchlchl {
case 'LI! load();
breaK;
case ·s· : save () ;
break;
case • E' ~ enter();
break;
case '0' : display II ;
break;
case 'Q' : quit ();
break;
)

3. 'include <stdio.h>
linclude <conio.h>
&20 T£ACII YOURSELF
~
C

int main (void)

{
char Chi

do (
ch = getche{);
} while(ch!='\r');

return 0;
)

4. The break statement causes immediate termination of the loop

that contains it. It also terminates a statement sequence in a
switch.
5. The continue statement causes the next iteration of a loop
to occur.
6. #include <stdio . h>

int main (void)

{
int i;
f.loat feet, meters, ounces. pounds;

do (
printf("Convert \ n\n R ) ;
printf ( -l. feet to meters\n");
printf("2. meters to feet\n"):
printf'-3. ounces to pounds\n - ):
printf(-4. pounds to ounces\n-);
printf("5. Quit\n \ n");
do (
printf("Enter the number of your choice: .);
scanf("'d", &i);
) while(i<D II i>5);

swi tch (i) {

case 1:
printf("Enter feet : .);
seani("tf n , &feet);
printf("Meters: %£\n", feet I 3.28):
break;
case 2:
printf( "Enter meters: .);
scant'"'f-, &meters);
 •. I -
EXERCISES
01
•

print£t"Feet: '£\n", rnet er£ • 1.28);

break;
case 3:
print£("Enter ounces: ");
scant("'t", &ounces);
printf ( - Pounds: %f \n". ouncp':: 16) i
break:
case 4:
printf("Enter pounds: "):
scanf(-'f", &pounds);
printf(-ounces: %f\n~. pounds * 16);
break;
}
_ } while(i!=5);

return 0;
}

XERCISES
1. unsigned short int lac_counter ;

2. #include <stdio.h>

int main (vaid)

(
Unsigned long int distance:

printft-Enter distance: .) ;

scanf("'lu·, &distance}:

printt("%ld seconds·, distance I lr'G!'(00);

,return 0;
}

- 3. The statement can be recoded using L's shortlland as follows:

short i:
 ,.
522 T£ACIt YOURSILF
C

XER·CISES

1. Local variables are known only to the function in which they

are declared. Global variables are known to and accessible by all
functions. Further, local variables are created when the function
is entered and destroyed when the function is exited. Thus they
cannot maintain their values between function calls. However,
global variables stay in existence during the entire lifetime of
the pJogram and maintain their values.
2. Here is the non-generalized version .
• include <stdio.h>

void soundspeed(void);

double distance;

int main (void)

(
printfC-Enter distance in feet: -,;
scanf(~%lf", &distance);

soundspeed ( ) ;

return 0;
)

void soundspeed(void)
(
printf(·Travel time: 'f", distance I 1129);
)

Here is the parameterized version.

'include <stdio.h>

void soundspeed (double distance);

int main (void)

(
double distance:

printfC-Enter distance in feet: .);

scanf(-'lf-, &discance);
soundspeed{distance) ;
 •.•
ANSWERS
EXERCISES
.,
523

return 0;
)

void soundspeed(double distance)

,. (
printf(-Travel time: %f·, distance / 1129);
)

ERCISES

1. To cause a constant to be recognized by the compiler explicitly

as a float, follow the value with an F.
2 . #include <stdio.h> -

int main (void )

(
long int i;

printf (- Enter a number: .) j

scanf (. %ld·. &i);

printf ( ·'~d·. i);

return 0 ;
)

3 . #include <stdio.h>

int main (void)

(

return OJ
)

ERCISES

1. .include <stdio. h>

int main (void)

{
int i=100j
TEACH YOURSElF
C

[',r( ; i>O; i--) printf(-'d· i);

return 0;

2 ")0. You cannot initialize a global variable using another variable.

3 y ,5. A local variable can be initialized using any expression
'"lhd at the time of the initialization.

1 [I,e entire expression is float.

2. " ',ubexpression is unsigned long.

4.6 .-.,'-- .

1 : h program displays 10.

2. n-.o program displays 3.0.

]. :bn~lu de <stdio.h>

: ". llai n (void)

{
.l.·')at f;

n l f =1.0; fint) f<=9; f=f + 0.1)

):;-j;.tf(-%f -, f ) ;

~ - .rn 0;

2. ;, O,T!, is the corrected statement.

/ = (int)123.23 % 3; j* now fixed */
 MASTERY SJ(JUS CHECK

STERY SKILLS CHECK

1. The data-type modifiers are

unsigned
long
short
signed

They are used to modifY the base type so that you can obtain
variables that best fit the needs of your program.
2. To define an unsigned constant, follow the value with a U. To
define a long constant, follow the value with an L. To specifY a
long double, follow the value with an L.
3 . floa.t balance = o. 0;
4. When the C compiler evaluates an expression, it automatically
converts all chars and shorts to into
5. A signed integer uses the high-order bit as a sign flag. When the
bit is set, the ilUm»er is negative, when it is cleared, the number
is positive. An unsigned integer uses all bits as part of the
number and can represent only positive values.
6. Global variables maintain their values throughout the lifetime of
the program. They are also accessible by all functions in the
program.
7. #include <stdio.h>

int series(void);

int num = 21;

int main (void)
(
int ii

for(i=O: i<10; i++)

printf ( .. %d ., series () ) ;

return 0;
)
 int series(void)
{
num = (num'146B) % 467;
return num;
}

8. A type cast temporarily changes the type of a variable. For

example, here the int i is temporarily changed into a double.
(double) i

MULA TIVESKIL LS CHECK

1. The fragment is not valid because to C, both 'A' and 65 are the
same thing, and no two case constants can be' the same.
2. The reason that the return value of getc\tar( ) or getche( ) can
be assigned to a char is because C automatically removes the
high-order byte.
3. No. Because i is a signed integer, its maximum value is 32,767.
Therefore, it will never exceed 33,000,

APTER 5

EVIEW SKILLS CHECK

1. A local variable is known only to the function in which it is

declared. Further, it is created when the function is entered and
destroyed when the function returns. A global variable is known
throughout the entire program and remains in existence the
entire time the program is executing.
2. C compiler will assign the following types:
a. int
b. int
c. double
d. IO!lg
e. long .

•
3. 'include <stdio.h>

int main (void)

(
long 1;
short s;
double d;

printf(WEnter a long value: ~);

scanf (·%ld~, &1);

printf(-Enter a short value: .);

scant ("\hd", &s);

printf(-Enter a double value: ~);

scanf('Uf", &d); "

printf (:%ld\n", 1);

printf("'hd\n", 5);
printf ( "'f\n", d);

return 0:
}

4. A type caSt temporarily changes"the type of a value.

5. The else is associated with the if(j) statement, contrary to what
the (incorrect) indentation would have you believe.
6. When i is I, a is 2. When i is 4, a is 5.

XERCISES
J. The array count is being overrun. It is only 10 elements long,
but the program requires one that is 100 elementi long.
2 .• include <stdio.h>

int main (void)

(
int i [101, j, k, match;

printf(-Enter 10 numbers:\n·);
for(j=O; j<lO; j++) 5canf("\d", &i[j]);
,.
U8 TEACH YOUIISllf

/* see if any match

for(j=O; j<lO; jH) (
*'
match = i(jJ;
for(k=j+l; k<lO; k++)
i f (match==i (kJ )
printf("'d is duplicated\n", match);
}

return 0;
}

3. *include <stdio.h>

int main (void)

(
float item(lOOJ, t;
int a, bi
lnt count;

/* read in numbers */
printf(-How many numbers? ·)i
scanf(-'d-, &count);
for{a=O; a<count; a++) scanf{-'f-, &item[a]);

/* now sort them using a bubble sort */

for(a=l; a<count; ++a)
for (b=count-l; b>=a; --b) (
/* compare adja:ent element,
i f (item[b-lJ > item[bJ) (
*'
/* exchange elements */
t = item[b-lJ;
item(b-lJ = item[bJ;
item[bJ = t;
)
}

/* display sorted list */

for(a=O; a<counti a++) printfC-'f· item{a))i

return 0;
}
 ANSWERS 529
5.2 EXERCISES

;~ ' 5.2 .. - . EXERCISES

1. /* Reverse a string.
#include <stdio.h>
*'
#include <string.h>

int main(void)
(
char str[BO];
int i;

printf(-Enter a string: .);

gets (str) ;
•
for(i=strlen(str)-l; i>=O: i- )
printf("tc", str[i]);

return 0;
)

2. The string str is not long enough to hold the string 'this is a test".
3. #include <stdio.h>
#include <string.h>

int main (voidj

(
char bigstr[lOOOJ =" str[BO];

for ( ; ; ) (
printf("Enter a string: ~);

gets (str) ;
if(!strcmp(str, -quit"» break;
strcat (str, "'n") j
/* prevent an array overrun */
if(strlen(bigstr}+strlen(str) >= 1000) break;
strcat(bigstr. str);
)

printf (bigstr) ;

return 0;
)
&eo T£ACH YOURSElf

" C

ERCISES

1. 'include <stdio.h>

int mainevoid)
(
int three_d(3J [3J [3J;
int i. j, k. x;

x = 1;
for(i=O; i<3; i++l
for(j=O; j<3; j++)
for(k=O; k<3; k++) (
three_d[iJ [j) [kJ = Y.;
x++;
printf("%d ", three_d[iJ [jJ [kJ) ;
)

return 0;
J

2. #include cstdio.h>

int main(void)
(
int three_d[3J [3J [3J;
int i, j, k. sum;

for(i=O; i<3; i~~)

for{j=O; j<3; j++)
for(k=O; kc3; k++) { •
three_d[iJ [jJ [kJ = (i+1) * (j+l) • lk+l);
printf("%d .. , three_d[i} [jJ ;kJ);
J

/* sum all elements */

sum = 0;
for(i=O; ie3; i++)
for(j=O; jel; j++)
for(k=O; k<3; k++)
sum =
sum + three_d[iJ [jJ [k);
 ANSWERS 531
5.4 EXERCISES ,
"
return 0;
)

ERCISES

I. No. The list must be enclosed between curly braces.

°
2. No. The array name is only 4 characters long. The attempted
caB to strcpy( ) will cause the array to be overrun.
3. #include <stdio. h>

int main(vcid)
(
int cube[) [3) = (
1, 1. 1,
2, 4, 8,
3, 9, 27,
4, 16, 64,
5, 25, 125,
6, 36, 216,
7, 49, 343,
8, 64, 512,
9, 81, 729,
10, 100, 1000
) ;
int num, i;

printfC-Enter cube: .);

scanfC-%d-, &num);

for(i=O; i<10; i++)

if{cube[i) [2)==nurn) (
printf{ORoot; %d\n°, cube[i) [0));
printf{OSquare; %dO, cube[i) [1));
break;
)

if(i==10) printf(~Cube not found.\n");

return 0;
)
532 TEACH YOURSElF

• C

EXERCISE S
1 #include <stdio.h>
#include <conio.h>

int rnain(void)
(
char digits[10J [10J = (
Rzero", "one", "two", "three",
"four" "five", "six", "seven",
"eight", "nine"
) ;
char nurn;

prlntf("Enter number: .);

ntL'1l ::: getche () ;
printf("\n") ;

num - '0';
TI<L'"":l :::

if(nun>=Q && num<10) printf("%s", digits[num]);

return 0;
,

It? .:U;J MA S TER Y SKILL SCHECK

1 An array is a list of like-type variables.
2 The statement will not generate an error message because C
provides no bounds checking on array operations, but it is
,,,,Tfong because it causes count to be overrun.
3 :;;:include <stdio. h>

inr. main(void)

int statsl201. i. J;
int mode, count, oldcount. oldmode;

printf("Enter 20 numbers: \ n");

for(i=O; i<20; i++) scanf,"%d", &stats[i]);

oldcount ::: 0;
1* find the mode * /
 _533
MASlFRY SKlUS alECK

for(i=O; i<20; i++) {

mode = stats[i];
count 1: =
j* count the occurrences of this value */
for(j=i+l; j<20; j++)
if(mode==stats[j]) cou~t++;

/* if count is greater than old count, use new mode */

if (count>oldcount) {
ol~ode = mode:
oldcount = count;
)
)
printf (-The mode is %d\n", oldmode);

return 0;
)

4. int items[) = (1, 2, 3, 4, 5, 6, 7, B, 9, 10);

5. #include <stdio.h>
#include <string.h~

int main(voJ.d)
(
char str[aO];

do (
printf("Enter a string: M);
gets(str) ;
} while(strcmp("quit-, str»;

return 0;
)

6. ;* Computerized dictionary program. *j

#include <stdio.h>
#include <string.h>

int mal.n(void)
{
ehar diet[) [2) [40) = {
534 TtACIf YOURSW

" c

~house·, -a place of dwelling-,

"car" ·a vehicle" •
I

• computer" , "a thinking machine~.

·program·, .~ sequence of instructions·,
•• ••
) ;
char word [80 J ;
int l i

printf("Enter word: ");

gets (word) ;

'* =
i
look up the word */
OJ
/* search,while null string not yet encountered */
while(strcrnp(dict[iJ [0), "" )) (
H(!strcrnp(word. dict[i) [OJ)) {
printf ("meaning: %5·, dict [i] [1 J ) ;
break;
)
i++j
)
H(!strcrnp(dict[i}[OJ. "")
printf(-Not in dictionary\n");

return 0;
)

UMULA TlVE SKILLS CHECK

1. #include <stdio.h>
#include <string.h>

int main(void)
(
char str[80];
int i;

printf{"Enter a string: ·)i

gets (str);

/* pad the string if necessary *1

for(i=strlen(str); i<79; i++)
strcat(str, ",");
 ANSWERS
CUMIMTJ1IE SKlL1.S CHECK

printf(str} ;

return 0;
}

2. / * A simple coding program. */

linclude <stdio.h>
"include <string.h:>' ·

int main (void)

(
Ghar str[90];
.
l.n ....
~ .
1,
.
J;

printf("Enter message: .):

gets (str) ;

/ * code it '*/
1=0 : j = strlen(str) - 1:
whil"e (i<=j) {
if ( i < j) printf("'c%c·, str[i], str[j]) .
else printf ("%c·, str[i}};
i++; j--:
)

return 0;
}

3. linclude <stdio.h>
*include <string.h>

int main(void)
(
char str[80];
int spaces, periods. commas;
int i:

printf("Enter a string: "):

gets (str) ;

spaces = 0;
commas = 0;
periods = 0;
1lACII YOUIIIIEI.F

for(i=Oi i<strlen(str); iTT)

switch(str[i]) (
case ',': periods++;
break:
case ',': commas++;
break:
case' '0 spaces++:
)

printf("spaces: %d\n", spaces); I.

printf("commas: %d\n", commas),
printf(·p~riods: 'dR, periods};

return 0;
)
•
4. The getchar( ) function returns a character, not a string.
Hence, it cannot be used as shown. You must use geta( ) to
read a string from the keyboard.
S. /. A simple game of Hangman */
•
'include <stdio.h>
#include <string.h>

int main (void)

(
char word{) = ·concatenation";
char temp{] = "-------------";
char Chi
int i, count;

count = 0: j* count number of guesses */

do (
printf("'s\n", temp);
printf(·F~ter your guess: .):
ch = getchar () ;
printf("'n"):

/* see if letter matches any in word *j

for(i=O; i<strlen(word); iTT)
if(ch==wordlij) temp[i) = ch;
count++;
) while (strcmp(temp, word)),
 ANSWERS
REVIEW SKJUS CHECK
.,
537

printf("'s\n", temp};
printf ("You guessed the word and used %d guesses". count);
•
return 0;
}

APTER 6

EVIEW SKILLS CHECK

1 .• include <stdio.h>

int main(void)
{
int num[lO}. i, even, odd:

printf("Enter 10 integers: ");

for(i=O; 1<10; i++) seanf("td", &num[i]);

even = 0; odd = 0;
for(i=O; i< 10; i++) {
if{num[i]%2} odd = odd + num[i];
else even = even + num(i];
}

printf{·Sum of even numbers: %d\n", even);

printf("Sum of odd numbers: %dn, odd);

return 0;
}

2. #include <stdio.h>
_include <string.h>

int ma.in(void)
{
char pw[BO};
int i;

for(i=O; 1<3; i++) {

printf ("Password: .. );
538 lEACH YOURSELf
c

gets (pw) ;
if(!strcrnp(~Tristan·. pw» break;
)

if(i==3) printf("Access Denied");

else printf("Log-on Successful");

return 0;
)

3. The array, name, is not big enough to hold the string being
assigned to it.
4. A null string is a string that contains only the null character.
S. The strcpy( ) function copies the contents of one string into
another. The strcmp( ) function compares two strings and
returns less than zero if the first string is less than the second,
zero if the strings match, or gteater than zero if the first string is
greater than the second.
6. /* A Simple computerized telephone book. */

#include <stdio.h>
#include <string.h>

char phone 1 I 1211401 = (

"Fred", "555-1010",
"Barney", "555-1234" I

"Ralph", "555-2347".
"Tom". "555-8396",
""
) ;

int main(void)
(
char name[BO);
int i;

printf("Name? ");
gets (name) ;

for(i=O; phonelillOIIOl; i++)

if ( ! strcmp (name, phone [i] [0] ) )
printf ("number: %5". phone(i] [lJ);
 -539
£3 EXERCISES
"
return 0;
)

ERCISES

1. A pointer is a variable that contains the address of another

variable.
2. The pointer operators are the' and the &. The' operator
returns the value of the object pointed to by the pointer it
precedes. The & operator returns the address of the variable it
precedes.
3. The base type of a pointer is important because all pointer
arithmetic is done relative to it.
4. #include <stdio.h>

int main(void)
(
int i. "'P;

p = &i;

for(i=Oi i<lO; i++) printf("%d" "'p);

return 0;
)

XERCISES

1. You cannot multiply a pointer.

2. No, you can only add or subtract integer values.
3. 108

EXERCISES

1. No, you cannot change the value of a pointer that is generated

by using an array name without an index.
2. 8
540 TEACH YOURSElF

'" C

3. #include <stdio.h>
•
int main(void)
(
char str(BO). ·Pi

printf{"Enter a string: ") i

gets (str);

p = str;

/* While not at the end of the string and no

space has been enc0untered, increment p to
point to next character .
*/
while(*p && .p!=' ') p++;

printf (p) ;

return 0;
)

XERCISE
I. #include <stdio.h>

int main(void)
(
char ·one = .. one" ;
char "'two = .. two" ;
char ·three = "three" :

printf ("%5 %s %s\n" , one, two, three) ;

printf ("%5' %s %5 \n" • one, three, two) ;
printf ("%5 %s %5\n" , two, one, three) ;
printf ("%s h %s\n" , two, three, one) ;
printf ("%5 %s %5\n" , three, one, two) ;
printf ("%s %s %5 \n" , three, two, one) ;

return O·,
)
 · ANSWERS &41
66 EXERCISE
"
ERCISE

1. jinclude <stdio.h>
#include <string.h>

int main(void)
{
char *p[3] = {
"yes". "no",
"maybe - rephrase the question"
}
char str[80];

printf(MEnter your question: \n");

gets(str) ;

printf(p[strlen(str) % 3]):

return 0;
}

1. #include <stdio.h>

int main(void)
(
int i. 'p, **mp;

p = &i;
rnp = &p;
**mp = 10;
printf("%p %p %P", &i, p, mp);

return 0;
}
TEACH YOURSELF
C

ERCISES

1. #include <stdio.h>
#include' <string.h>

void mystrcat(char *to, char *from);

int main(void)
{
char str[80};

strcpy(str, -first part-);

mystrcat (str, ~ second part -);
printf (str);
•
return 0;
}

void mystrcat(char *to, char * from)

(
/* find the end of to */
while (*to) to++;

/* concatenate the string */

while(*from) *to++ = *from++;

/* add the null terminator */

*to = '\0';
}

2. #include <stdio . h>

void f (int *p);

int main(void)
{
int i;

f I&i I;

p r intf (-%d", i);

return 0;
}
 -15113
CUMIJU< 77Vf SJ<1f.LS CHECK ~

void flint 'p)

(
'p = -1;
)

S TERY SKIL L SCHECK

1. double .p;

2. *include <stdio.h>

int main (void)

(
int i, *p;

p = &i;
*p = 100;
printf("%d", il;

return 0 j
)

3. No. The pointer p has never been initialized to point to a valid

piece of memory that can hold a string.
4. Pointers and arrays are basically two ways oflooking at the
same thing. They are virtually interchangeable.
5. str[2]

*(str+2)

*(p+2)

6. 108

CUMULATIVE SKILLS CHECK

1. Pointers are often more convenient than array indexing and
may be faster in some cases.
c

2. iinclude <stdio.h>

int main (·void)

{
char strl80], *p;
int spaces:

printf (-Enter a string: M ) i

gets (str) ;

spaces = 0;
p = str;
while (.p) (
if(*p==' , ) spaces++;
p++ ;
}

printf (-Number of spaces: .%d- • spac.as) ;

return 0;
}

3. * ( (int *) count + ( 44 • 10 ) + 8) = 99;

APTER 7

VIEW SKILLS CHECK

1. The fragment assigns to i the value 19 indirectly using a pointer.

2. An array name with no index generates a pointer to the start of
the array.
3. Yes, the fragment is correct. It works because the compiler
creates a string table entry for the string 'this is a string' and
assigns p a pointer to the start of it.
4. iinclude <stdio. h>

int main (void)

{
 .atIM

double d. 'p;

p = &d:
'p = 100.99;

return 0;
}

S. tinclude <stdio. h>

int mystrlen(char *p) ';

int main (void)

(
char str(80);

printf (-Enter a string: .

) ;
gets (str) ;

printf(·Length is 'd-, myatrlen(atr»;

return 0;
}

int mystrlen(char .p)

(
int i;

i = 0:
while{'p) {
i++:
p++;
}
return i:
}

6. The fragment is correct. It displays c.

".
WICH YOURSELF
C

XERCISES

1. #include <stdio.h>

double avg();

int main (void)

{
printf("U', avg(),

return 0;
)

double avg ()
{
int i;
double sum, nUID;

sum =
0.0;
for(i=O; i<10: i++) {
printf(-Enter next number: .);
scanf("%lf', &num);
sum = sum + num;
)
return sum I 10.0:
)

2. #include <stdio.h>

double avg(void);

int main (void)

{
printf ("U", avg();

return 0;
}

double avg(void)
{
int i;
double sum, num;

sum = 0.0;
 7.2
ANSWERS
EXERCISES
.,
547

for(i=Oi i<10; i++) {

printf(-Enter next number: .,;
scanf("'lf", &nurn);
sum = sum + num;
)
return sum I 10.0;
)

3. The program is correct. However, the program would be better

if a fuJI function prototype were used when declaring myfunc( ).
4. double • Purge (void) ;

1. 'include <stdio.h>

int fact{int i};

int main (void)

{
printf (" 5 factorial is . .'d" , fact (5) ) ;

return 0;
)

int factCint i}
(
i£(i==l) return 1;
else return i ,.. fact(i-l):
}

2. The function will call itself repeatedly, until it crashes the

program, because there is no condition that prevents a recursive
caJl from occurring.
3. 'include <stdio.h>

void display(char *p};

int main (void)

{

return 0;
548 TEACH YOIJASW'

• C

void display(char *p)

{
H("pl {
printf (-'e·, *p);
display(p+l);
I
I

ERCISES
1. No. The function myfuN:( ) is being called with a pointer to
the first parameter instead of the parameter itself.
2. tinclude <stdio.h>

void prompt{char *msg. char ·str);

int main(void)
(
char strI80];

prompt(-Enter a string: • str);

printf(-Your string is: %s·, str);

return 0;
)

void prompt (char *msg, char *p)

{
printf (msg) ;
gets (p) ;
I

3. In call by value, the value of an argument is passed to a

function. In can by reference, the address of an argument is
passed to a function.

1. 'include <stdio.h>
'include <string.h>
 ANI\1J_
7.4 EXERaSES

tinclude <stdlib.h>

int,main(int argc, char *argv[l)

(
int i;

if(argc!=3) {
printf("You must specify two drgurnents . OJ;
exit(l) ;
}

i = strcmp(arqv[ll,
argv[2]);
if(i < 01 printf("'s > %5", argvl2}, argvlll I;
else if(i > 0) printf(-'s > %5·, argv[l], argv[2);
else printf(-They are the same-);

return 0:
}

2. 'include <stdio.h>
'include <string.h>
'include <itdlib.h>

int main(int argc, ehar *argv[])

(
if(argc!=3) {
printf("You must specify two numbers.-):
exit(1);
}

printf("'f", atof(argvllli • atof(argvI2lll;

return 0;
}

3. 'include <stdio.h>
'include <string.h>
'include <stdlib.h>

int main(int argc, char ·argv[)

(
if (Arge! =41 (

printf(·You must specify the ope ration .);

printf(·followed by two numbers.");
11110 TEACH YOURSElf
c

exit(l);
)

if(!strcmp('add', argv[l]))
printf("f', atof(argv[2]) + atof(argv[3]));
else if(!strcmp('subtract', argv[l]))
printf("f', atof(argv[2]) - atof(argv[3]));
else if(!strcmp('multiply', ArgYll]))
printf("f', atof(argv[2]) • atof(argv[3]));
if(!strcmp('divide', argv[l]))
prirtf("f', atof(argv[2]) I atof(argv[3]));

return 0;
J

ERCISE

1. _include <stdio.h>

int main (void)

{
double feet;

printf (-Enter feet: .);

scanf(-'lf-, &feet):
printf('Meters: 'f', f_to~(feet));

return 0;
J

/ * use old-style declaration. */

double f_to~(f)
double f;
{
return f I 3.28;
J

STERY SKILLS CHECK

1. A function that does not have parameters specifies void in the

parameter list of its prototype.
~
- -- - - ...

-
M.<STERY SKJUS CHECK
..,
•

2. A function prototype tells the compiler these three things: the

return type of the function, the type of its parameters, and the
number of its parameters. It is useful because it allows the
compiler to find errors if the function is called incorrectly.
3. Command-line arguments are passed to a C program through
the argc and argv parameters to maine ).
4. 'include <stdio. h>

void alpha{char ch);

int main (void)

(
alpha ( 'A');

return 0;

void alpha(char ch)

(-
printf(-'c·, ch);
if(ch < 'Z') alphalch+l);
}

S. 'include <stdio.h>
linclude <stdlib.h>

int mainCint argc, char *argv[])

(
.char *p;

ifCargc!=2) {
printfC·You need to specify a string·);
exit(l);
}

p " argv(l);

whileC*p) {
printf{-'c·, (*p)+l);
p++;
}
 return 0;
)

6. The prototype is shown here.

double myfunc(int x, int y, char ch);

7. Using the old-style function declaration, the function from

Exercise 6 looks like this.
double myfunc(x, y, chI
int x. y:
char ch i
(

8. The exit( ) function causes immediate program termination. It

also returns a value to the operating system.
9. The atoi( ) function converts its string argument into its
equivalent integer form . The string must represent (in string
form) a valid integer.

MULA TIVE SKILLS CHECK

1. linclude <stdio. h>
'include <strinq . h>
'include <stdlib . h>

int ~inCint argc. char *argv{]}

{
if{arqc!=2) {
printf(-Specify a password-):
exit(l) ;
)
if(!strcmp(arqv[lJ. "password"))
printf(-Access Permitted-):
else printfC-Access Denied-):

return 0;
)
 ANSWERS 553
CUMUlATIVE SlQLLS CHECK

2. *include <stdio.'h>
• include <ctype.h>

void string_up (char .p) ;

int main(void)
(
char str [) = "this is a test" ;

string_up (str) ;
printf (str);

return 0;
)

void string_up (char 'p)

(
while(*p) (
'p = toupper (.p) ;
p++;
)
)

3 . • include <stdio.h>

void avg(double *d. int nurn);

int main(void)
(
double nums[) = (1.0. 2.0. 3.0. ~.O. 5.0.
6.0. 7.0. ' 8.0. 9.0. 10.0);

avg(nums. 10);

return 0;
)

void avg(double ·d. int nurn)

(
double sum;
int temp;

temp = nurn-1;

for (sum=O; temp>=O; temp--)

 .,
554 nACIt YOURS£l.F
c .

sum = sum + d [temp] ;

.
printf(·Average is %f·, sum I (double) nurn);
)

4. A pointer contains the address of another variable. When a

pointer is passed to a function, the function may alter the
contents of tl)e object pointed to by the pointer. This is the
equivalent of call by reference.

HAPTER 8

EVIEW SKILLS CHECK

1. To allow the compiler to verify that a function is being called

correctly, you must include its prototype.
2. Function prototypes enable the compiler to provide stronger
type checking between the arguments used to call a function
and the parameters of the function. Also, it lets the compiler
confirm that the function is called with the proper number of
arguments.
3', #include <stdio.h>
#include <math.h>

double hypot(double 51. double 52);

int main(void)
(
printf("%f-, hypot(12.2, 1!f.21);

return 0;
)

double hypot(double 51. double 52)

{
double h;

h = 51*s1 + 52*52;
return sqrt (h) ;
)
 555
B. 1 EXERCISES
...

4. When a function·does not return a value, its return type should

be specified as void.
5. 'include <stdio.h>

int rstr!en(char *p);

int main (void)

(
printf(·%d-, rstrlen(Mhello there"»;

return 0;
)

int rstr!en(char *p)

{
iflOp} (
p++;
return l+rstrlen(p);
)
else return 0;
)

6. #include <stdio.h>

int main(int argc, char *argv[ll

(
printf(-There were %d arguments.\n", argc);
printf(-The last one is %5,-, argv[argc-ll);

return 0;
)

7. func (a. ch, d)

int a;
char Chi
double d;
{

ERCISES
1. #include <stdio.h>

#define MAX 100

$5& TEACH youRSEJi
.,. c

#Qefine COUNTBY 3

int main (void)

(
int i;

for{i=O; i<MAX; i++)

if(l (i%COUNTBYI) printf("'d" il;

return 0;
)

2. No, the fragment is wrong because a macro cannot be defined in

terms of another before the second macro is defined. Stated
differently, MIN is not defined when MAX is being defined.
3. As the macro is used, the fragment is wrong. The string needs to
be within double quotes.
4. Yes.

XERCISES

1. #include <stdio.h>

int main(void)
{
int i;

do {
i = getchar ( ) ;
if(i~=EOF) (
printf(-Error on input.-);
break;
)
if(putchar('. ')~~EOF) (
printfC-Error on output.-);
break;
)
) while ( (char) i 1~ '\n');

return 0;
)
 -
as EXEIICISES
.,
557

2. The putcbar( ) function outputs a character. It cannot output a

string.

1. 'include <conio.h>
'include <stdio.h>

int main (void)

(
char Chi

ch ~ getch I ) ;
printfl"'d", chI;

ret.urn OJ
)

2. 'include <stdio.h>
'include <conio.h>

int main (void)

(
do (
printf I"'c", '.');
) whilel!kbhitl));

return 0;
)

RCISES
•
2. No. The program is incorrect because gets{ ) must be called
with a pointer to an actual array.

RCISES

1. 'include <stdio.h>

int main(void)
(
558
"
unsigned long i;

for(i=2i i<=100; i++)

printf(-'-lOlu %-lOlu %-lOlu\n·, i, i*i. 1*i*1);

return 0;
}

2. printf("Clearance price: 40% off as marked-);

3. printf("%.2f", 1023.03};

ERCISES
,
1. .include <stdio.h>

int main (void)

(
char first{21J. middlel2lJ. lastl2l];

printf(-Enter your entire name: ")i

scanf("%20s%20s%20s-, first, middle, last);
printf("'s %5 %s·, first, middle, last);

return 0;
}

2. #include <stdio.h>

int main(void)
{
char nwnI80];

• printf ( • Enter a floating point number: .);

scanf("%[0-9.]', nwn);
printf (num) ;

return 0;
}

3. No, a character can only have a maximum field length of].

4. #include <stdio.h>

int main(void)
 MSt:1US
M4S7FRY SKlUS alECK
.,
5511

(
char ,trIBD);
double d;
int L num;

printf(-Enter a string, a double, and an integer: ");

scanf(~%s%lf%d%n·, str, &d, &i, &num);

printf (·Number of characters read: %d", nurn);

return 0;
)

5. #include <stdio.h>

int main{void)
(
unsigned u;

printf("Enter hexadecimal number: ");

scanfC"%x·, &u);
printf(-Decimal equivalent: tUM, u);

return 0;
)

ASTERY SKILLS CHECK

1. All these functions input a character from the keyboard. The

getchar( ) function is often implemented using line-buffered
110 which makes its use in interactive environments
undesirable. The getche( ) is an interactive equivalent to
getchar( ). The getch( ) function is the same as getche( )
except that it does not echo the character typed.
2. The %e specifier outputs a number in scientific notation using a
lowercase 'e'. The %E specifier outputs a number in scientific
notation using an 'E'.
3. A scanset is a set of characters that scanf( ) matches with
input. As long as the characters being read are part of the
scanset, scanf( ) continues to input them into the array pointed
to by the scanse!'s corresponding argument.
c

4 .• include <stdio.h>

int main (void)

(
char name[SO), date[SO), phone[SO);

printf(-Enter first name, birthdate .);

printf(-and phone number:\n-);
scanf("%s%S.tSs", name, date, phone);
printf(-'s %5 %s·, name, date. phone);

return 0;
}

5. The pats( ) function is much smaller and faster than printf( ).

But, it can only output strings.
6. 'include <stdio.h>

#define COUNT 100

int mainlvoid)
(
int i;

for(i=O: i<COUNT;i++)
printf("td ", i);

return 0;
)

7. EOF is a macro that stands for end-of-file. It is defined in

STOIO.H.

UMULATIVE SKILLS CHECK

1. linclude <stdio.h>

int main (void)

{
char name (9) [B01;
double b_avg[91;
int 1, h, 1 ;
 CUMIJU'oTTVE SKILLS CHECK
.,
181

double high. low. te~avg;

for(i=O; i<9; i++) {

printf("Entar name 'd: ., i+1);
scanf("'s", name(iJ),
printf(-Enter batting average: .):
scanf("'lf", &b_avg(iJ),
printf (" \n") ,
)

high = 0.0,
low = 1000.0,
team_avg = D.O.
for(i=O: i<9: i++' (
if(b_avg(iJ>high) (
h = i;
high = b_avg(iJ,
)
if (b_avg[i]<low) {
1 = i;
low = b3v9(iJ,
)
team_avg = team_avg+b_avg[i);
)
printf("The high is %s %f\n", name(hJ, b_avg(hJ I ,.
printf("The low is %s %f\n", name(lJ, b_avg(lJI'
printf(·T~e team average is %f", team_avg/9.0l;

return 0:
)

2. Note: There are many ways you could have written this
program. This one is simply representative.
/* An electronic card catalog. */
#include <stdio.h>
#include <stdlib.h>
#include <string.h>

Ide fine MAX 100

int menu (void) ;

void display (int i);
void author_search (voidl ;
void title_search(void);
562 TEACH YOURSElF
c

void enter(void):

char names [MAX) (80); /' author names '/

char titles [MAX) (80); /' titles '/
char pubs [MAX) (80); /' publisher '/

lnt top: 0: /* last location used *j

int main(void)
{
lnt choice;

do (
choice = menu();
switch (choice) (
case 1: enter(); /* enter books */
break;
case 2: author_search(); /* search by author */
break:
case 3: title_search(); ;* search by title */
break:
)
} whileCchoice!=4);

return 0;
}

/* Return a menu selection. */

menu (void)
(
char str(80);
lnt i;

printf ("Card Catalog: \n");

printfC" 1. Enter\n");
printf(" 2. Search by Author\n"};
printf(" 3. Search by Title\n"};
printf(" 4. Quit\n"};

do (
printfC"Choose your selection: ");
gets (str);
i = atai (str);
printfC"'n") ;
} while(i<l II i>4};
 ANSWERS B83
CUMULATIVE SKILLS CHECK

return i;
)

j* Enter books into database. */

void enter(void)
(
int i;

for(i=top; i<MAX; i++) {

printf(KEnter author name (ENTER to quit): .}:
getslnarnes[i]);
if(!*names{i]) break;
printf(8Enter title: ");
getsltitles[i]) ;
printf{-Enter publisher: .);
getslpubs[i]);
}

,' }
top = i;

/* Search by author. *j
void author_search (void)
(
char name[80);
int i, found;

printf' -Name: .. );
gets (name) ;

found = 0;
for(i=O; i<top; i++)
if (! strcmp{name. names (i) » (
display 1i) ;
found = 1;
printf I" \n") ;
}

if(!found) printf("Not Found\n");

}

j* Search by title.·;
void title_search (void)
{
char title[BO];
TEACH YOURSW

int i. found;

printf("Title: ");
gets (title);

found = 0;
for(i=O; i<top; i++)
if( !strcmpCtitle, titles[i]» (
display (i) ;
found = 1;
printf("\n") ;
)
ifC!found) printfC-Not Found\n-);
)

/* Display catalog entry . • /

void display(int i)
(
printfC-%s\n-, titles[iJ);
printfC-by %s\n-, names[i]);
printf("Published by %s\n", pubs[i»;
)

APTER 9

EVIEW SKILL SCHECK

1. The getchar( ) function is defined by the ANSI standard and is

used to input characters from the keyboard. However, in most
implementations, it uses line-buffered I/O, which makes it
impractical for interactive use. The getche( ) function is not
defined by the ANSI standard, but it is quite common and is
essentially an interactive version of gctchar( ).
2. When scanf( ) is reading a string, it stops when it encounters
the first whitespace character.
3. #include <stdio.h>

int isprime(int i);

int main(void)
{
 -- ...•
int i, count;

count 0; =
for(i=2; i<1001; itt)
'if(isprime(i)) {
printf ("%lOd". i);
count++;
if (count==4) (
printfC-\n-);
count = 0;
)
)

return 0;
)

int isprime(int i)
(
int j;

for(j=2; j<=(i/~); j+~)

H(I(i%j)) return 0;
retuJ"n 1 j
}

4. 'include <$tdio.h>

int main (void)

(
dO\lble d;
char Chi
char str(80J;

printt(·Ent~r a double. a character, and a string\n-);

scsnf("\lf%c\20s". kd. &ch. str),
printf,("\f 'e %5". 11. ' ch. str);

, 0:
ret\lrn '

5 . • inol\lde' .s~dio.h>

int main(voi<:l)
{
char str[80};
188 1EACII YOURSn1

• C

printf(MEnter leading digits followed by a string\n.):

scanf(",O[O-9J%s", str);
printfC-%s·, str)i

return 0:
)

XERCISES

1. #include' <stdio.h>
#include <stdlih.h>

int rnain(int argc, char *argv[])

(
FILE 'fp;
char chi

/* see if filename is specified */

if(argc!=2) (
printf(RFil~ name missing.\nR):
exit (1);
)

if((fp = fopen(argv[l]. "r"))==NULL) (

printfC·Cannot open file.\n.);
exit(1) ;
)

while((ch=fgetc(fp)) != EOF) putchar(ch) I

fclose(fp) ;

return 0;
)

2. 'include <stdio.h>
#include <stdlib.h>
#include <ctype.h>

int count[26]:

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

{
 9.2 -
EXERCISES
887
•

FILE *fp;
char ch:
int i:

j- see if file name is specified

if(argc!=2) (
*'
printf(·File name missing.\n");
exit(l);
}

if I Ifp = fopenlargv(l). "r") )==NULL) (

printf("Cannot open file.\n");
exit(1):
}

whilellch=fgetclfp)) !=EOF) (
ch = toupper(ch);
if(ch>='A' && ch<='Z') count[ch-'A')++;
}

for(i=O; i<26; i++)

printf("%c occurred %d times\n", i+'A', count[i.

fcloselfp) ;

return 0;
)

3. / * Copy a file. */
#include <stdio.h>
.include <stdlib.h>
#include <string.h>

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

(
FILE *from. *to;
char ch, watch:

/* see if correct number of command line arguments */

if (argc<3) (
printf{·Usage: copy <source> <destination>\n");
exit(1) ;
}

/* open source file *j

.
 if(( from = fopen(argv[lJ, ·r·»==NULL)(
printf(-Cannot open source file.\n-);
exJt(l) ;
)

1* open destination file */

if((to = fopen(argv[2J. ·w·»==NULL) (
print£(-Cannot open destination file.\n-):
exit(l) ;
J

if(argc==4 && !strcmp(argv[3J. ·watch·» watch = 1;

else watch = 0;

/* copy the file */

while((ch=fgetc(frorn» !=EOF) {
fputc (ch. to);
if (watch) putchar(ch);
J
fclose (from) ;
fc!ose(to) ;

return 0;
) - .
ERCISES
1. #include <stdio.h>
'include <stdlib.h>

int main(int argc, char *argv{])

(
FILE 'fp;

unsigned count;

/* see if file name is specified */

if(argc!=2) (
printf(-File name missing.\n-);
exit(l) ;
)

if (( fp = fopen (argv[lJ, ·rb·» ==NULL) (

printf(-Cannot open file.\n-);
 ANS\T.i!IAI
'.3 EXERCISES
...•
exit(l);
}

count = 0:
while (! feof (ip» (
fgetc (fp) ;
if (ferror(fp» {
printf (-File error. \n· ) :
exit(l) ;
J
count++ ;
}

printf(-File has %u bytes-, count-I);

fclose(fp) ;

return 0;
}

2. /. Exchange two files. */

#include <stdio.h>
#include <stdlib.h>
#include <string.h>

int main(int argc, char ·argv(])

{
FILE *fl, *f2, *temp;
char chi

/* see if correct number of command line arguments */

if (argc! =3) (
printf(-Usage: exchange<fl> <f2>\n-);
exit(l);
}

/* open first file */

if( (f1 = fopen(argv[l). Orb") J==NULL) (
printf(-Cannot open first file.\n-);
exit(l);
}

/. open second file ./

if«f2 = fopen(argv[2J."rb"»==NULL) (
printf(-Cannot open second file.\n-);
exit(l) ;
570
y

/* open temporary file */

if((ternp = fopen("temp.tmp", "wb"»==NULL) (
printfC-Cannot open t~porary file.\n-);
exit (1);
}

/* copy f1 to temp */
while (! feof (fl» (
ch = fgetc(fl);
if(!feof(fl» fputc(ch, temp);
}

fclose (f1) ;
/* open first file for output *j
if((fl = fopen(argv[1}, "wb"»==NULL) (
pr:ntfC-Cannot open first file.\n-);
exit(!);
}

/* copy £2 to £1 */
while(!feof(f2» (
ch = fge·tc (f2) ;
if(!feof(f2» fputc(ch, £1);
}
fclose(f2);
fclose(temp) ;

1* open second file for output */

if!(f2 = fopen(argv[2}, "wb")}==NULL) (
printfC-Cannot open ¥econd file.\n-);
exit(l) ;
)
/* open temp file for input */
if«temp = fopen("ternp.tmp" , "rb"»==NULL) (
printfC-Cannot open temporary file,\n*);
exit(1);
}

1* copy temp to £2 *1
while(!feof(ternp» {
ch = fgetc(temp);
if(!feof,ternp» fputc(ch, f2);
}
 571
...

fclose(fl) ;
fclose (f2) ;
fcloseCtemp) ;

return 0;
)

1 . / * A simple computerized teleph.one book. *I

.include <stdio.h>
#include <string.h>
'include <stdlib.h>

char names [100) [40J;

char numbers [lOOJ [40J;

int loc=Di

int menu (void) ;

void enter(void):
void load (void) ;
void save (void) ;
void find(void) j

int main (voiq,)

{
int choice:

do (
choice = menu();
switch (choice) {
case 1: enter();
break:
case 2: find();
break;
case 3: save ( ) ;
break:
case 4: load();
J
} while(choice!=5):

return 0;
 1IAaIYOURSB.F
C

j -Get menu choice. */

int menu(void)
{
int i:
char str[80};

printf ("1. Enter names and numbers\n");

printf("2. Find numbers\n");
printf("). Save directory to disk\n");
printf("4. Load directory from disk\n");
printf("S. Quit\n"};

do (
• printfC"Enter your choice: .);
gets(str);
i = atai (str);
printfC"'n") ;
} while(i<l II i>S};
return i;
}

void enter(void)
{

for(;lec<100; lac++"} {
H(loc<lOO) (
printf("Enter name and phone number:\n");
gets(names[loc]):
if(!*names[loc) break;
gets(numbers[loc);
}
}
}

void find(void)
(
char name[80j;
int i:

printf(·Ent~r name: .);

gets (name) ;

for(i=O; ie100; i++)

 9.4
ANSWERS
EXERCISES
.,
1173

if(!strcmp(name, names[i])}
printf(-%s %5\n-, names[i), numbers[i]):
)

void load (vo"id)

(
FILE *fp;

if«fp = fopen("phone" , "r"»==~LL) (

printf(-Cannot open file.\n~);
exit(l) ;
)

loc = 0;
while ( ! feaf (fp» (
fscanf(fp, "%S%5", names [loc] , numbers[loc]);
lac++;
)
fclose (fp) ;
)

void save (void)

{
FILE *fp;
int i;

H«fp = fopen("phon.", "w"))==NULL) {

printf("Cannot open file.\n");
exit(l) ;
)

for(i=O; i<lae; i++) (

fprintf(fp, -%5 %5· names(i]. numbers{i]);
)
fclose{fp) ;
)

2. 'include <stdio.h>
#include <stdlib.h>
'include <ctype.h>

int mainlint ar4c. char *argv ()

{ >,
FILE 'fp;
char Chi
5'14- _YOURSELf .
-y
c

char str[BO];
int count;

1* see if correct number of common line arguments * 1

if(argc!=2) (
printf(·Usage: display <file>\n"):
exit (1);
}

/* open the file *1

if((fp = fopen(argv[l]. "r"))==NULL) - (
printfC-Cannot open the file.\n·);
exite!) ;
}

count = 0;
while ( ! feof (fp)) (
fgets(str. 79. fp);
printfC-%s", str);
count++;

if (count==23) {
printf ("More? (y\n) .);
gets (str);
if{toupper(*str)=='N') break;
count = 0:
}
}

}
..
fclose(fp) ;

return 0;

3. /* Copy a file. '* /

iinclude <stdio.h>
'include <stdlib.h>
'include <string.h>

int main(int argc, char .argy[])

{
FILE -from . • to;
char st;r[128];

/* see if correct number of command line arguments *'

 .--.. &715
~= ...
9.4 CAUI~

if (argc<3) (
printf(nUsage: copy <source> <destination>\n-);
exit(l);
)

/* open source file -;

if«from = fopen(argv[l). 'r'»==NULL) (
printf(.Cannot open source f11e.\n");
exit (1);
)

,_ open destination file

if«to = fopen(arg,,'hL 'w'»==NULL) (
*'
printf{~Cannot open destination file.\n~);
exit(1) ;
)

/* copy the file */

while ( ! feof (from» {
fgets (str, 127, from);
if(ferror(from» (
print·f i. ~Error on input. \n");
break;
I
if feof (from»
(! fputs lstr, to);
if (ferror (to» (
printf("Error on output. \n-);
break:
I
I

if (fclose(from)==EOF) (
printf(~Error closing sourc~ fi1e.\n·);
exit(!) ;
I

if(fcl.ose(to)==EOF) (
printf{"Error closing destination fi1e.\n");
exit(l) ;
I

return 0;
I
 ERCISES

1. #include <stdio. h>

#include <stdlib.h>

int main (void)

(
FILE *fp1, *fp2;
double d;
int i:

if«fp1 ~ fopen("values", "wb"))~~NULL) {

printf (·Cannot open file. \n·) ;
exit (1) ;
)

H«fp2 ~ fopen("count", "wb"))~~NULL) (

printf(-Cannot open file.\n-):
exit (1) ;
)

d ~ 1.0;
for(i=O; d!=O.O && i<32766; i++) {
printf("Enter a number (0 to quit): .);
scanf("%lf", &d);
fwrite(&d, sizeof d, 1 , fpl);
)

fwrite(&i, sizeof i, 1, fp2};

fclose(fp1);
fclose (fp2) ;

return 0;
)

2. #include <stdio.h>
#include <stdlib.h>

int main (void)

(
FILE *fpl. *fp2;
double d;
int i;
 a77
"

i£«£p1 , £open("values", "rb"))==NULL) (

print£("Cannot open £ile.\n");
exit(l);
)

i£«£p2 = £open("count". "rb"))==NULL) (

printf(-Cannot open file.\n-);
exit(1) ;
}

fread(&i. sizeof i. 1, fp2); /* get count */

fore; i>O: i--) (

£read(&d, sizeof d, 1, £pl);
print£("'£\n", d);
}

idose (£pl) ;
£c1ose(fp2) ;

return OJ
)

ERCISES

1. 'include <stdio.h>
.include <stdlib.h>

int mainlint argc. char *argv[])

(
FILE -fp;
char Chi
long 1;

if(argc!=2) (
printf(-You must specify the file.\n·);
exit(l);
}

i f «fp = fopen(argv[l]. Orb") )== NULL) (

printf(-Cannot open file.\n-);
exit(1) ;
}
578 T£ACH YOURSW
c

fseek(fp. O. SEEK_END); /. find' end of file ./

1 = ftell (fp);

/* go back to the start of the file */

fseek(fp. O. SEEK_SET);
for ( ; 1>=0; 1 = 1 - 2L) (
ch = fgetc (fp) ;
put char (chi ;
fseek(fp. lL. SEEK_CUR);
}

fclose (fp) ;

return 0;
}

2 .• include <stdio.h>
linclude <stdlib.h>

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

(
FILE *fp;
unsigned char ch, val;

if(argc!=3) {
prin-tf (·Usage: find <filename> <value>-);
exit(l) ;
}

if«fp = fopen(argv[l). "rb"))==NULL) (

printfC-Cannot open file.\n-);
exit (1);
}

val = atoi(argv[2}};

while ( ! feof (fp)) (

eh = fgete (fp);
if(eh == val)
printf("Found value at 'ld\n". ftell(fp));
}

felose(fp) ;
 -878
9.B EXERCISE
.,

return 0:
1

1. #include <stdio.h>
.include <stdlib.h>
linclude <ctype.h>

int main(void)
(
char fname[BO];

printfC-Enter name of file to erase: .);

gets (fname) j

printfC-Are you sure? (YIN) .);

if(toupper(getchar(ll=='Y'}
if (remove (fname) )
printfC-'nFile not found or write' protected.\n-);

return 0;
1

ERCISE

1. /* Copy using- r_edirection.

Execute like this:

C>NAME < in > out

*'
linclude <stdio.h>

int main(void)
{
char Chi

while ( ! feof (stdin)) (

scanf(·'c·, &ch);
if(!feof(stdin}} printf("'c", ch};
1
 return 0;
)

ASTERY SKILLS CHECK

1. 'include <stdio.h>
'include <stdlib.h>
'include <ctype.h>
.
int main(int argc, char -argv[])
(
FILE tip;
char str(BO);

I- see if file name is specified -I

if(argc!=2) (
printf (-File name missing. \n·) ;
exit (1);
)

if «fp = fopen(argv(l], ·r·)) ==NULL) (

printf(-Cannot open file.\n-);
exit(l) i
)

while (! feof (fp» (

fgets (str, 79, fp);
if(!feof(fp» printf(·\S·, str);
printf (- ... More? (yin) .);
if(toupper(getchar(»=='N') break;
printf ( • \n·) ;
)

fclose (fp) ;

return 0;
)

2. I- Copy a file and convert to uppercase. -/

'include <stdio.h>
'include <stdlib.h>
'include <ctype.. h>
 -&8,
MAS1l'RY SJ(JllS CHECK

int main(int argc, char *argv[l)

{
FILE ·from, -to;
char Chi

/* see if correct number of command line ar~ents· */

if (arge! =3) {
printf (·Usage.: copy <source> <destination>\n"');
exit(l) ;
}

/ * open source file * /

if«from = fopen(argv[lL "r"»=.NULL) (
printf("Cannot open source file.\n"):
Qxit (1);
}

/. ope~ destinat10n file * /

if«to = fopen(argv[21. ·w")==NULLI {
printf (·Cannot open destination file. \n") ;
exit(l);
}

/* copy the file */

while (! feoE (from) {
ch = fgetc(from);
if(!feof(from») fputc(toupper(chl. to);
}
fclose(from) ;
fclose(to) ;

return 0;
}

3. The fprintf( ) and fseanf( ) functions operate exactly like

printf( ) and seanf( ), except that they work with files.
4. 'include <stdio.h>
'include <stdlib.h>

i~t main (void)

(
FILE *fp;
int i, nwn;
 582 TfACH VOURSW

" c

if«fp = fopen("rand", "wb"))==NULL) (

printf{-Cannot open file,\n-);
exit(l);
}

for(i=O: i<100; i++) (

nurn = rand ( );
fwrite(&num. sizeof nurn, 1, £p);
}

fc10se (fp) ;

return 0;
}

5. 'include <stdio.h>
jinclude <stdlib.h>

int main(void)
(
FILE '*fp;
int i. nurn;

H«fp = fopen("rand"\ "rb"))==NULL) (

printf("Cannot open file.\n");
exit (1);
, )

for(i=O; i<100; i++) (

fread(&num, sizeof nurn, 1, £p);
printf("'d\n·, nurn);
}

fc 1ose(fp) ;

return 0;
}

6. 'include <stdio.h>
'include <stdlib.h>

int main{void)
{
,
FILE *fp;
long i;
 -583
CUMUlATIVE SKIU.S CHECK
"
int num:

if«fp = fopen('rand', 'rb'»==NULL) (

printf(-Cannot open file.\n-);
exit(l);
)

printf('Which number 10-99)? ');

scanf(-'ld-, &i);
fseek (fp, i * sizeof (int) I SEEK_SET);
fread(&num, sizeof num, 1. fp);
printf 1' \d\n', nurn);

fclose 1fp) ;

return 0;
)

7. The 'console' I/O functions are simply special cases of the

.
general file system. ...

MULATltlE SKILLS CHECK

1 . / * An electronic card catalog. */

'include <stdio.h>
'incl~de <string.h>
'include <stdlib.h>

.define MAX 100

int menu (void) ;

void display(int il;
void author_search(void):
void title_search(void):
void enter(void);
void save (voidl ;
void load(void):

char names [MAX] [80]; /* author names * /

char titles[MAXJ [BOJ; /, titles ,/
char pubs[MAXJ [BOJ; /, publisher '/

int top = 0: /* last location used */

584 lIACII YOIIRSnf
c

int main (void)

(
int choice;

load(): /* read in catalog */

do (
choice = menu():
switch (choice) {
case 1: enter(); /* enter books */
break;
case 2: author_search(); /* search by author */
break:
case 3: title_search(); 1* search by, title */
break;
case 4: save();
}
} while(choice!=5);

return 0 i
}

/* Return a menu selection. */

rnenu(void)
(
int i;
char str(80];

printf{-Card Catalog:\n-);
printf(- 1. Enter\n-):
printf (. 2. Search by author\n·);
printf(" 3. Search . by Title\n");
printf(- 4. Save catalog\n-);
printf(" 5. Quit\n");

do (
printf(·Choose your selection: .);
gets (str) ;
i =
atoi(str);
printf (" \n") ;
} while(i<l II i>5);

return i;
}
 ___•.
CUMUlATIVE SKlUS CHECK

'* Enter books into database. */

void enter(void)
(
int i:

for(i=top: i<MAX: i++) {

printf("Enter author name (ENTER to quit) .);
gets (names (il ) ;
if(!*names[i]) break:
printf{"Enter title: .);
gets(titles[i»;
printf("Enter publisher: .);
gets(pubs[i» ;
)
top = i;
)

'* Search by author. */

void author_search(void)
{
char name [80] i
int i. found;

printf ("Name: ");

gets (name) ;

found: 0;
for(i=O; i<top; i++)
if{ ! strcmp {name, names (i) » {
display (i) ;
found: 1;
printf ( "\n" ) ;
)

if(!found) printf("Not Found\n");

)

j* Search by title. */
void title_search(void)
(
char tit1e[80);
int i, found:

print£(IOTitle: .):
gets(tit1e) ;
...
., TfACH YOURS£lF
C

found = 0;
for(i=O: i<top; i+.)
if(!strcmp(title, titlesli)) (
display (i) ;
found = 1;
printf (" In") ;
)
if(lfound) printf{~Not Found\n-);
)

/* Display catalog entry. */

void display{int i)
(
printfC-'s\n", titles[i]);
printf(-by %s\n-, names[i});
printf("Published by 'sIn", pubsli);
}

/* Load the catalog file. */

void load(void)
(
FILE jofp;

ifllfp = fopenl"cataloq", "r"»==NULL) (

printfC-Catalog file not on disk.\n-);
return;
}

fread(&top, sizeof top, 1, £Pl; I* -read count */

fread(names, sizeof names, 1, £pl;
fread(titles, ~izeof titles, 1, fp);
fread (pubs , sizeof pubs, 1. fp);

fclose I fp) ;
}

,
1* save the catalog file. */
void save(void)
(
FILE 'fp;

if(lfp = fopen("catalog", "W"»==NULL) (

printf(-Cannot op~n catalog file.\n-);
exit (1);
 1187
Ct/MIJLA 1M SKILlS CHECK
"
)

fwrite(&top, sizeof top, 1, fpl:

fwrite(names. sizeof names, 1, fp);

fwrite(titles, sizeof titles. 1, fp);

fwrite(pubs. sizeof pubs, 1, fpl:

fclose(fp) ;
)

2. /* Copy a file and remove tabs. */

iinclude <stdio.h>
'include <stdlib.h>
'include <string.h>

int main(int argc, char *argv{])

(
FILE ·from. *to;
char chi
int tab. count;

/* see if correct number of command line arguments *1

if(argc!=3) (
printfC-Usage: copy <source> <destination>\n-);
exit(l) ;
)

/* open source file *1

if«from = fopen(argv[l). "r")I==NULL) {
print;f ("Cq,nnot · open source file. Xn") ;
exit(l);
)

1* open destination file *1

if( (to = fopen(argv[2J. "W")I==NULL) (
printfC-Cannot open destination file.\n-);
exit(l); ,
)
1* copy the file *1
count = 0;
while(!feof(from» (
ch = fgetc (from) ;
if (ch== • \ t .) {
.for (tab = count; tab<8: tab++)
1188 lIACII YOURSW

• c

fputc (. '. to);

count = 0;
)
else {
i f (! feof (from» fputc (ch, to);
count++:
if(count==8 II ch=='\n') count = 0;

)
fclose (from) :
fclose (to) ;

return 0;
)

HAPTER 10

EVIEW SKILLS CHECK

1. 1* Copy a file. *1
'include <stdio.h>
#include <stdlib.h>

int main(int argc. char *argv[)

(
FILE *from. *tOj
char chi

1* see if correct number of command line arguments t/

if(argc!=3) (
printf(MUsage: copy <source> ~destination>\n-';
exit(!);
)

j* open source file +1

if«from = fopen(argv(lJ, "rb"»==NULL) C
printf (-Cannot. open source file. \n -) :
exit (1);
)

" It open destination file */

if«to = fopen(argv(2], "wb"»==NULL) (
printf(-Cannot open destination file.\n-);
 exit(l);
)

copy the file

j'* '*,
while(!feof(from)) (
ch = fgetc (from) ;
if(ferror(from)) (
printf(-Error on input.\n-);
break;
)
if(!feof(from)) fputc(ch, to),
tf ( ferror (to)) (
printf(-Error on output.\n-);
break:
)
)

if(fclose(froM)=~EOF) {
printf (-Error closing source file. \n-);
exit(l);
)

if(fclose(to)·=gQF) (
printf(-Error closin~ destination file.\n-):
exit(l) ;
)

retuI11 0;
)

2. linclude <sedio.h>
'inc1ud$ <stdlib.h>

int main (void)

(
FILl!: 'fp;

open file '*/

j'*
H«fp = fopen("myfile", "w"))~=NULL) (
printf("Cannot open file.\n");
exit(l),
)

fprintf(fp. -'s %.2£ \X 'c·, -this is a string-,

1230.23, OxlFFF, 'A');
uo
'f'

fclose(fp) ;

return 0;
)

3. #include <stdio.h>
'include <stdlib.h>

int main(vOid)
{
FILE -fPi
int countC20J, i;

/* open file */
if «fp = fopen ("TEMP", "wb"» ==NULL) {
printf (·Cannot open file. \n.) ;
exit(1);
)

for(i=O; i<20; i++) count[ij = i+l:

fwrite(count, sizeof count, 1, fp);

fclose (fp) ;

return 0;
)

4. 'include <stdio.h>
.include <stdlib.h>

int main (void)

(
FILE *fp;
int count (20) , i;

1* open file */
if({fp = fopen{"TEMP", "rb"»==NULL) (
printf("Cannot open file.\n");
exit (1);
)

fread(count, sizeof count, 1, fp):

 ~AN!!:IIl~:~JB~IS~ ~1
'0.1 EXERCISES

for(i=O; i<20: i++) printf(-'d· count[i]);

fclose(fp) ;

return 0;
)

5. stdin, stdout, and stderr are three streams that are opened
automatically when your C program begins executing. By
default they refer to the console, but in operating systems that
support 110 redirection, they can be redirected to other devices.
6. The printf( ) and scanf( ) functions are part of the C file
system. They are simply special case functions that
automatically use stdin and stdout.

XERCISES

1. j* A simple computerized telephone book. */

*include <stdio.h>
iinclude <string.h>
.include <stdlib.h>

*define MAX 100

struct phone_type {
char name (40J ;
int areacode;
char number (9 J ;
) phone (MAXJ ;

int loc=O;

int menu (void) ;

void enter(void);
void load (void) ;
void save (void) ;
void find(voidl;

int main (void)

(
int choice;
-~
c

do {
choice = menu ( ) :
switch (choice) (
case 1: enter();
break:
case 2: find () :
break;
case 3: save () :
break;
case 4: load () :
)
) while(choice!=S):

return 0:
)

/* Get menu choice . */

menu (void)
{
int i;
char .tr(80J:

printf (-1 . Enter names and numbers\n·);.

printf("2. Find numbers\n"):
printf(-3. Save directory to disk\n-);
printf(-'. Load directory from disk\n-);
printf("S. Quit\n"):
,
,
do (
printf("Bnter your choice: "):
geta(atr):
i • atoi (str):
print! ( "\rI" ) :
J while (i<l II i>5):
return i;
J

void enter(void)
{ ,
char temp (80J :

tor (: loc<lOO: locH) (

if(loc<lOO) {
printf (-Enter name: .):
gets(phone[locJ . name):
 ,a'

if(!*phone(loc).name) break;
printf(-Enter area code: .);
gets (temp);
phone {loc] .areacode = atoi(temp);
printf!"Enter number: .);
gets (phone [loc] . number) ;
)
)
)

void find{void}
(
char name [801 ;
int i;

printf("Enter name: .);

gets (name) ;
if(!*name) return;

for(i=O; i<100; iT+,

if (!strcmp(name. phone[i] .name»
printf("%s !%d) %s\n", phone[i] .name,
phone[i] .areacode. phone(iJ.n~~er);
)

void load {void)

(
FILE 'fp;

if«fp = fopen("phone", "r"ll==NULL) (

printf("Cannot open file.\n");
exit(l);
)

loe = 0;
while (! feof (fp» (
fseanf(fp, "'s%d's", phone[loe) . name,
&phone{loc1.areacode. phone[loc] .number);
loc++;
)
felose (fp) :
)

void save (void)

{
...• --
c

FILE "fp;
int i;

i f ( (fp = topen ("phone", "W")) -=NULL) (

printf("Cannot open file. In");
eXit(l);
)

for(iIl:O; i<loc; i++) (

fprintf(fp, "'s 'd '8" phone [i) . name ,
Phone(iJ.areacode, phone[i).number);
)
fclose (fp) ;
)

2. The variable i is a member of structure a_type. Therefore, it

cannot be used by itself. Instead, it must be accessed using 8 and
the dot operator, as shown here.
s.i ~ 10;

ERCISES

1. No. Since p is a pointer to a structure, you must use the arrow

operator, not the dot operator, to access a member.
2 .• include <stdio.h>
.include <time.h>

int main (void)

(

struct tm ·systime. .gmt;

time_t ti

t =time (NULL) ;
systime = localtime(&t);

printf(-Time is %.2d:%.2d:%.2d\n·, systime->tm_hour,

systime->t~n, systime->tm_sec);
gmt = gmtiw.e(&t);
printf("Coordinated Universal Time is %.2d:~.2d:'.2d\n.,
gmt->tm_hour.
gmt->tm_min, gmt->tm_sec);
 MIMRI
la. EIIEIICtSES
...•
printf('Date: ',2d/',2d/',2d', systime->~on+1,
systime->tm_mday, systime->tmLYear);

return 0;
)

ERCISES

1. '* A simple computerized telephone book. *'

'include <stdio.h>
'include <string.h>
'include <stdlib.h>

'define MAX 100

struct address {
II char street(40];
char city[40J;
char state[3];
char zip[12J;
) ;

struct phone_type {
char name [40J ;
int areacode;
char number [9] ;
struct address addr;
) phone [MAX] ;

int 10c=0;

int menu (void) ;

void enter(void);
void load(void);
void save (void) ;
void find(void);

int main (void)

(
int choice;

do (
choice = menu( );
 1tACH YOURSELF
C

switch (choice) {
case 1: enter ( );
break;
case 2: find( );
• break;
case 3: save ( ):
break;
case 4: load( ) ~
)
} while(choice!=5);

return 0;
)

Get menu choice. */

/*
rnenu(void)
(
lnt i;
char str[80];

printf(-1. Enter names and nurnbers\n-);

printf(~2. Find numbers\n-):
printfC-3. Save directory to disk\n-);
printfC-4. Load directory from disk\n.);
printf('S. Quit\n');

do (
printfC-Enter your choice: .);
gets (str);
i = atoi (str) ;
printf (. \n· ) ;
) while(i<l II i>S);
return i;
•
)

void enter(void)
{
char tempI80];

for (; loc<lOO; loc++) {

i f (10c<100) (
.
printfC-Enter name: .);
gets (phone {loc] .name):
if(!*phone(loc) . name) break;
printfC-Enter area code: .);
 -EXERaSES
H7
'If

gets Ctemp) ;
phone{loc] .areacode = atoi(temp);
printfC-Enter number: .);
gets Cphone [locJ .number);

j* input address info -j

printf(-Enter street address: .);
_ gets(phone[loc].addr.street);
print! (-Enter city: .);
getsCphone[locJ .addr.city);
printf(-Enter State: .);
gets (phone {loc] .addr.state):
printf ( . Enter zip code: .);
gets (phone(loc] .addr.zip):
)
)
)

void find(void)
{
char name[80};
int i:

printfC-Enter name: W);

gets (name) ;
if(!*name) return;

for l i=O; i<100; i++)

if(!strcmp(name, phone(i] .name» {
printfC"%s C%q) %s\n", phone[iJ .name,
phone[i] .areacode, phone{i] .number):
printf(-'s\n%s %5 %s\n", phone[i] .addr.street,
phone[i] .addr.city, phone[i] .addr.state,
phone [iJ . addr. zip) ;
)
)

void load(voidl
(
FILE *fPi

ifCCfp = fopenC "phone" , "rb"))==NULL) (

printf(-Cannot open file.\n-);
exit(l) ;
)
 loc = 0;
whi1e(!feof(fp)) (
fread(&phone[1ocl. sizeof phone[1ocl. 1. fp l;
loc++;
)
fc1ose(fp) ;
)

void saveCvoid)
(
FILE *fp;
int i;

if«(fp = fopen("phone". "wb"))==NULL) {

printfC-Cannot open file.\n·};
exit(l);
)

for(i=O; i<loc; i++) (

fwrite(&phone[il. sizeof phone[il. 1. fp);
)
fclose(fp) ;
)
\

XERCISES

1. 'include <stdio.h>

int mainCvoid}
(

int a, 3 ;
int b, 3;
int c, 2;
) bvar;

bvar.a =: -1;
bvar.b = 3;
bvar.c = 1;
printfC·'d %d %d-, bvar.a, bvar.b, bvar.c};

return 0;
)
 '0.5 EXEIfCIS£S
...
...

ERCISES
1. tinclude <stdio. h>
'include <stdlib.h>

union u_type (
'double d;
unsigned char c [8) ;
) ;

double ureadCFILE 'fp);

void uwrite(double num, FILE -fp):

int main (void)

C
FILE "fp;
double d;

i f CCfp = fopenC "myfile", "wb+")) ==NULL) C

printf(-Cannot open file.\n-}:
exitCl);
)

uwriteClOO.23, fp);
d = ureadCfp);
printfC"'lf", d);

return 0;
)

void uwrite {double num, FILE -fp}

(-
int i;
union u_type var;

var.d = num;
forli=O; i<8; i++) fputc(var.c{i], fp);
)

double ureadCFILE "fp)

{
int i;
union u_type var;
 ~---

800 1EACM YOURSElF

c

rewind (fp) ;
for(i=O; i<8; i++) var.c[iJ = fgetc(fp);

return var.d:
)

2. *include <stdio.h>

int rnain(void)
{
union t_type {
long 1:
lnt i:
} uvar;
,
uvar.l = OL; /* clear 1 */
uvar.i = 100;

printf(~%ld", uvar.l ) ;

return 0:
j

STERY SKILLS CHECK

1. A structure is a named group of related variables. A union
. defines a memory location snared by two or more vanables of
different types.
2. struct s_type {
char chi
float d ;
int i;
char str[80};
double balance;
} s_var:

3. Because p is a pointer to a structure, you must use the aITOW

operator to reference an element, not the dot operator.
4. #include <stdio.h>
#include <std l ib.h>

struct s_type {
char name [40 J ;
 '!.!<l"""
_ ~S. . .
.,.,.n £!!!
MASIERY SKJUS CI/ECIC ~

char phone [14] ;

int hours: ,
double walle;
} emp[lO];

int main (void)

(
FILE *fp;
int i;
char temp[eO];

if((fp = fopen('emp', 'wb'l)==NULL) {

printf (·cannot open EMP file. \n-);
exit(l);
)

for(i=Oi i<10; i+.) (

print£(-Enter name: .);
gets (emp[i] . name) ;
printf(-Enter telephone number: .);
gets (emp I i] . pl)one) ;
printf(-Enter hours worked: .);
gets (te.'!Ip);
emp[i] .hours = atoi(temp);
printf(MEnter hourly wage: .);
gets (temp) ;
ernp [i 1 . wage = atof (temp) ;
}

fwrite(emp, sizeof emp, 1, fp);

fclose (fp) ;

return 0;
}

S. 'include <stdio. h>

tinclude <stdlib.h>

struct s_type {
char name[40l;
char phone[14l;
int hours,
double wage;
) emp[l.O];
802
y

int rnain(void)
(
FILE ·fp;
int ii

if«fp = fopen("emp·, "rb"))==NULL) {

printf ("Cannot open EMP file. In") ;
exit(l) ;
J

fread(emp. sizeof amp, 1, fp);

for(i=O; 1<10; i+.} (
printf(",s 'sIn", emp[iJ . name , emp[i).phone);
printf("'d 'flnln", emp[iJ . hours , emp[iJ .wage);
}

fclose (fp) ,

return 0;
J

6. A bit"field is a structure member that specifies its length in bits

7. 'include <stdio.h>

int main (void)

{
union u_type (
short int ii
unsigned char c{2] ;
) uvar:

uvar.i :: 99;

printf("High order byte, 'uln", uvar.c[l));

printf("Low order byte, 'uln", uvar.c[OJ );

return 0;
}
 NJ.aa 80S
CUMUVoTM SK1U.S a£CK ~

MULATIVE SKILLS CHECK

1. 'include <stdio.h>

struct B_type {
int i;
char chi
double d;
} var1, var2;

int main (void)

(
varl.i = 100;
var2.i = 99;
var1.ch =- 'a';
var2.ch='b';
var1.d = 1. 0;
var2.d = 2.0;

printf(-varl: %d %c %f\n-, var1.i. varl.ch, varl.d):

printf(-var2: %d %c %f\n-, var2.i. var2.ch. var2.d):

struct_swap(&varl, &var2);

printf« -After swap: \n-):

printf (·varl: %d %c %f\n -. var1. i, varl. ch, varl. d) ;
printf(-var2: %d %c %f-, var2.i. var2.ch. var2.d};

return 0;
}

void struct_swap(struct s_type *i. struct s_type *j)

(
struct s..:,type temp;

temp = *i:
*i = *j;
*j =- temp;
)

2. j* Copy a file. *j
'include <stdio.h>
 jlinallJl;le <stdlib.b>
,
.-- - - .... 1

int main(int argc, char '*argvIl:)

{
FILE "'from. ·to;
union u_type { , •
int i;
char chi
} uvar;

/* see if correct number of command line arguments *j

if(argc!=31 {
printf("Usage: copy <source> <qestination>\n");
exit(ll;
)

'* open source file

if«from = fopen(argv(l}. "rb" "==NULL) {
*/

printf("Cannot open source f~le.\n·);

exit(l);
)

, • ••
'" /* open des.tl.llt\tlOll flle *1
.' if «to = fop;'n(argv[2}. "..b" "==NULL) {
printf ("Cannot opell d.e,stin.ation {ile. \n") ;
exit (11;
) ., ,
. ,
,* copy the file
t
"
*' -•
,
Eo'
" for ( ; i ) (
uvar. i = £getc (from) ;
if (uvar.i==EOF) break;
fputc (uvar. ch, to);
• ~ j:;') <> Co

fclose (from) ;
fclose(tol; ,

return 0; •
)
••
',.. "1 -
3. You cannot use a structure as an argument to ·acanf( ).
However, you can use a structure element as an argument, as
shown here.
.
c

scanf("'d·, &var.a):
. ,
 \ ... -- --~

4. 'include <string.h» "

'include <&tdio.h>

s~ruct B_type (
• ,
}
char str(80);
var; 1 • •
U
, ,•
11 • , L

void f(struct "_type i J ; f • ,

(
int main (..void)
(
strcpy(var.str, "this is original string");
f '( var} ;
.-
printf("'s". var.strl: - 4 ~

.
return 0;
}

void f(struct s_type i)

(
.. f ) "I
strcpy(i.str, "new string") ;
printf(·'s\n·, i.str)·
} 6 '

i £
, ,,
APTER 17.
"

VIEW SKILLS CHECK

"
1. Unclude <stdio : h>
" ,
-,
,
1
, ,. S.e ruct. nWlLt.ype j
i~t ~i
int sqr;
.~

,
'1: ,
•, int cu.be:
o 1

, , I } ·.nums ( 1'01. " !) , t I

int main (void)

•
(
int i;
!
foX"(i=l; i<1,1[ i++) { J , 1 ...' 3
nums[i-l].i = i;
8M _taURlIlJ

• C

nu..(i-l).~ - i·i,
nuaa(i-l).CUbe - i·i·i,
}

for(i-O, i<10: i++) (

printf("\d ", nu..(i).i):
printf("td ", nuaa(i).~):
printf("td\n", nu..li) . CUbe) :
}

return 0,
}

2. 'include <stdio.b>
union i_to_c (
char c(2):
ahort int i:
} ic:

int main (void)

I
printf(-Enter an integer .,:
scanf(-'bd-, ~ic . i) :
printf("Cbaracter representation of each ~e: tc tc",
ic . cIO), ic . cll)) :

return 0;
)

3. The fragment displays 8, the size of the largest element of

the union.
4. To access a structure member when actually uaing a etructure
variable, you must use the dot operator. The arrow operator iI
used when accesaing a member using a pointer to a structure.
5. A bit-field iI a structure element whole IIize II specified In bits.

ERCISES
1. The best variables to make into resIatm" types are k and m,
because they are accessed most frequently.
 "NO;I;;.S~.;.; ~
1102 ~

2. linclude <stdio.h>

void .~it(int value),

, int .... in(void)

(
sWlLit(lO) ,
sum_it (20) ,
sWILit(30) ,
sWILit(40);

return OJ
}

void s~it{int value)

(
static int sum=O;

sum = sum + value;

printf(-CUrrent value: 'd\n·, sum);
}

4. You cannot obtain the address of a register variable.

XERCISES
1 . linclude <stdio.h>

const double version # 6 . 01;

int main (void)

(
printf(-Version '.2f-, version):

return 0;
}

2. 'include <stdio.h>

char *mystrcpy(char *to, const char *from);

int main(void)
{
'11111111 _I_RIB'
... C

chlr 'p, str(80],

printfC-'. 'S·, P. str};

return 0:
)

char *mystrcpy(char *to. const char *from)

{
char -temp ;

temp == to;

while (*from) *to++ == -from++;

*to = '\0' ; /* null termdnator */

return temp:
)

ERCISES
,
2. e~um money {penny, nickel. quarter, half_dollar. dollar}:

3, No, you cannot output an enumeration constant as a sttingas is

attempted in the printf( ) statement,

1 . • include <stdio. h>

typedef unsigned long UL;

int rnain(void)
{
UL count;

count = 3~2323;

printfC - %lu·, coun t ) ;

 ",5 EXEROSES

return 0;
}

2. The typedef statement is out of order. The correct form of

typedefis

typedef oIdname newname;

RCISES

1. 'includ~ <stdio.h>
'include <stdlib.h>

int mainCint argc, char ·argv[1)

{
FILE * in. ·out;
unsigned char chi

if(argc!=3) (
printf(-Usage: code <in> <out>\p~);
exit (1);
) .'
if «in = fopen (argv [1] , "rb")) ==NULL) (
printf{-Cannot open input file.\n-);
exit(l);
}

if«out = fopen(argv[2], "wb"})==NULL) (

printf(·Cannot open output file.\n-);
exit(l);
}

while ( ! feof (in)) (

ch = fgetc (in) ;
if(!feof(in)) fputc(-ch, out) ;
} .-.. '

fclose(in) ;
fclose{out) ;

return 0;
}
810
"

2. 'include <stdio. h>

'include <stdlib.h>

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

!
FILE *in, "out:
unsigned char Chi

if(argc!=4) !
printf(~Usage: code <in> <out> <key>\n-):
eXit(l);
}

if(! in = fopen(argv[l]. "rb"))==NULL) !

printfC-Cannot open input file.\nW);
exit!l);
}

if«out. fopen (argv[2]. "wb"))==NULLI !

printf(-Cannot open output file.\n-);
exit(l);
}

while! ! feof (in)) !

ch = fgetc (in) ;
ch = *argv(3) ~ chi
if(!feof(in ) ) fputc(ch. out);
}

fclose (in) ;
fclose (ou t) ;

return 0;
}

3. a. 0000 0001
b. lUI lUI
c. lUI 1101
4. char chi
 NISWIIIS 811
• 11.6 EXERCISES

/ - To zero high order bit. AND with 127, which

in binary is 0111 1111. This causes the h i gh-
order bit to b~ zeroed and all other bits left
untouched.
*/
ch = ch & 127;

ERCISES

1. linclude <stdio.h>

int main(void)
(
int i, j, k;

printf("Enter a number: -r;

scanf("%d", &i) ;

j=i«l ;
k. = i » I,
printf("%d doubled, %d\n", i, i);
printf("'d halved, 'dO, i, k);

return 0;
}

2. 'include <stdio.h>
-
void rotatelunsigned char ·e);

int main(void)
(
unsigned char ch;
int i;

ch = 1;
for(i=O: i<16; i++) (
rotate (&ch) ;
printf("'u\n", ch):
}
612 lEACH YOURSELf

" C

return 0:
)

void rotate(unsigned char ·c)

(
union {
unsigned char ch[21:
unsigned U;
) rot;

rot.u = OJ /* clear 16 bits */

rot.ch[O] = ·c;

'*rot.ushift rot.u
=
integer left
« 1;
*/

/* See if a bit got shifted into e(l].

If so, OR it back onto the other end. */
i f Crot. ch[lJ) rot. ch (0) = rot. ch[ 0) I 1;

'c = rot .ch[O);

)

XERCISES

1. Itinclude <stdio.h>

int rnain(void)
(
int i. j, answer;

printfC-Enter two integers: .);

scanfC"'d'd", &i, &j);

answer = j ? i/j: 0;
printfC"'d", answer);

return 0;
)

2. count = a>b ? 100 0;

 -013
MASTfJlY SKILLS CHECK "

2. x &= Yi

3. #include <stdio.h>

int main(void)
(
int i;

forli~17; i<=1000; i+=17)

printf {"%d\n", il;

return 0;
}

1 #include <stdio. h>

int main(void)
(
int i, j, k;

for{i=Q. j=-50. k=i+j; i<100: i++, j++, k=i+j )

printf I"k = %d\n" , kl;

return 0;
}

2. 3

ASTERY SKILLS CHECK

1. The register specifier causes the C compiler to provide the

fastest access possible for the variable it precedes.
2. The const-specifier tells the C compiler that no statement in the
program·may modifY a variable declared aSCOJl8t. Also, a conat
pointer parameter may not be used to modifY the object pointed
to by the pointer. The volatile specifier tells the compiler thai
814
...

any variable it precedes may have its value changed in ways not
explicitly specified by the program.
3. ftinclude <stdio.h>

int main (void)

(
register int i. sum:

sum = 0;
for(i=l; i<101; i++)
swn = sum i;

printf ("%d", sum) ;

return 0;
}

4. Yes, the statement io valid. It creates another name for the ~'pe
long double.
s. 'include <stdio.h>
'include <coniOth>

int main (void)

(
char chI, ch2:
char mask, i;

printfC-Enter two characters: .);

chI = getche () ;
ch2 = getche ( , ;
printf (" \n", ;

mask = 1;
forCi=O; i<8; i++) {
i f « ....sk " chI I "" (....sk " ch2 I I
printf("bits td the same\n", il;
mask «= 1;
)

return 0;
}

6. The < < and> > are the left and right shift operators,
respectively.
'~ -- -- ,

A.W\tIERS . , .
C/JMIJlA 77IIE SIOUS CHECK '"

7 . c >= 10;

8. count = done? 0 : 100;

9. An enumeration is a list of named integer constants. Here is one

that enumerates the planets.
enum planets {Mercury. Venus. Earth , Mars, Jupiter.
Saturn, Neptune. Uranus, Pluto} :

MULATIVE SKILLS CHECK

1. 'include <stdio.h>

void show_binary(unsigned u}:

int main (void)

(
unsigned char ch, t1, t2;

ch = 100;
show_binary (ch) ;

t1 = chi
t2 = ch i

tl «= 4:
t2 »= 4;

ch = t1 I t2;

show_binary(ch);

return 0;
}

void show_binar-y(unsigned u)
{
unsigned n;

for {n=128; n>O: n=n / 2}

if(u & n l printf("l '1;
else printf("O ");
818 1IIICIIlOURSEU
c

2 . #include <stdio.h>
'include <8tdlib.h>

int main(int argc, char +argv£)

{
FILE -in;
unsigned char ch;

i f (argc! =2) (
printf(~Usage: code <in>\n-);
exit(l);
)

if«in = fopen(argv[lJ. "rb"»==NULL) (

printf(-Cannot open input file.\n·);
exit(l);
)

while ( ! feof (in» (

ch = fgetc(in);
if! !feof(in» putchar(-ch) ;
}

fclose (in) ;

return 0:
)

3. Yes, any type .o fvariable can be specified using register.

However, on some types, it may have no effect.
4. / * A simple computerized telephone book. */

'include <stdio.h>
'include <string.h>
'include <stdlib.h>

'define MAX 100

atruct address {
char street[40];
char city[40];
 -
CUMUlA7l1if SJ(JUS CHECK
817
"

char state [3) ;

char zip[12],
) ,
struct phone_type {
char name [40] ,
int areacode;
char number [ 9] ,
struct address1addri
) phone [MAX] ,

int lac =0;

int menu (void) ;

void enter(void);
void load (void) ;
·void save (void) ;
void find(void);

int main (void)

{
register int choice;

do (
choice ;:: menu () ;
switch(choice) (
case 1: enter{);
break;
case 2: find ( ) ;
break;
case 3: savel);
break;
case 4: load () ;
)
• } while(choice!=5);

return 0;
)

j* Get menu choice. */

menu (void)
{
register int i;
char str[eD],
618 TEACHYOURSELf
c

printf(-l. Enter names and number s \n-);

printf("2. Find numbers\n-);
printf("3. Save directory to di s k\n.):
printf(·4. Load directory from di s k \ n.);
printf("S. Quit\n" ) ;

do (
printf(-Enter your choice: .);
gets (str);
i = atoi (str);
printf (" \n") ;
) while(i<l II i>5);
return i;
)

void enter(void)
(
char temp [SO);

for (; loc<100: loc++) {

i f (loc<100) (
printf (" Enter name: .);
gets(phonelloc] .name);
if(!*phoneIloc] . name) break;
printf {PEnter a=ea code: .);
gets (temp) ;
phone; loc 1 . areacode = atoi (tem:;-.) ;
printi("Enter number: .):
gets (phone[lcc] .number);

/ * input address ~nfc ' /

printf(-Enter street adciress: .);
gets {phonelloc ] . addr.str~ et ) ;
print! ( "Enter city: .);
gets (phone[locJ . ad~.city);
printfC'Enter State: .);
gets (phone[loc] .addr.state);
printf{-Enter zip code: - ) ;
gets{phone[locJ . addr.zip);
)
)
)

void find (void)

[
 ANSWERS 619
~
CUMULATNE SKILLS CHECK

char name(80):
register lnt i;

printf(·Enter name: .);

gets (name) ;
if(!*name) return;

.for(i=O; i<100; i++)

if(!strcmp(name, phone[i}.name» {
print£("%s (%d) %s\n", phone[i] . name ,
phoneli] .areacode. phoneli} .number);
printf(-%s\n%s %s %5\n", phone{i].addr.street,
phone[i] .addr.city. phone{i] .addr.state,
phone(i}.addr.zip) ;
)
)

'void load(void)
(
FILE ·fpi

i£«fp = fopen("phone", "rb"»==NULL) (

printf("Cannot open fi1e.\n"):
exit(1):
)

loe = 0;
while ('! feo£ (fp» {
fread(&phone[loc], sizeof phone [loc], 1, £p};
loc++:
)
fclose'(fp) ;
)

void save(voidl
(
FILE *£p;
register int i;

if«£p = fopen("phone", "wb"»==NULL) (

printf("Cannot open £i1e.\n");
exit (.!) ;
)

for(i=O; i<loc; i++) {

620 TEACH YOURSW'

• C

fwritel&phone!iJ. sizeof phone!iJ. 1. fp);

J
fclose I fp) ;
J

APTER 72

EVIEW SKILLS CHECK

1. Modifying a variable with register causes the compiler to store

the. variable in such a way that access to it is as fast as poSSIble.
For integer and character types. this typically means. storing it
in a register of the CPU.
2. Because i is declared as const the function cannot modify any
object pointed to by it.
3. a. 11000100
b.11111111
c. 0011 1011
4. #include <stdio.h>

int main(void)
(
int i;

printf("Enter a number: .);

scanf ("'d", &i);

printf(·Doubled: %d\n·, i « 1):

printf(-Halved: %d\n-, i » 1);

return 0;
)

5. a = b = c = 1;
max = a<b ? 100 0;

i *= 2:
 821
y
12.2 EXERCISES

6. The extern modifier is principally used to inform the compiler'

about global variables defined in a different file. Placing extern
in front of a variable's declaration tells the compiler that the
variable is defined elsewhere, but allows the current file to
refer to it.

ERCISES

I. #define RANGE(i, min, max) ((i)«min)) II ((i»(max)) ? 1 0

2. #include <stdio.h>

#define ABS(i) (i)<O ? -(i) i

int main(void)
(
printf('%d 'd', ABS(-l), ABS(l));

return 0;

1 •. 2 EXERCISES
1. #include <stdio.h>

#define INT 0
#define FLOAT 1
_define PWR_TYPE INT

int main(void)
(
int e:
#if PWR_TYPE==FLOAT
double base. result;
#elif PWR_TYPE==INT
int base, result;
lendif

.if PWR_TYPE==FLOAT
printf("Enter floating point base: ");
scanf('%lf', &base);
822 TEACH YOURSElf
c

#elif PWR_TYPE==INT
printf{·Enter integer base: .):
scanf ("'d", &base);
iendif
printf("Enter integer exponent (greater than 0): .);
scanf ("'d", &e);

result = 1;
fore; e; e--)
result = result • base:

#if PWR_TYPE==FLOAT
printf("Result: %f-, result);
#elif PWR_TYPE==INT
printf(-Result: %d-, result);
tendif

return 0;
}

2. No. You cannot use an expression like IMIKE with #ifdef. Here
are two possible solutions.
*i fndef MIKE

#endif

/* or */

#if !defined MIKE

tend!f

ERCISES

2. The program displays one two.

 12.6
-
EXERCISES

XERCISES

2 . • include <stdio.h>
iinclude <stdlib.h>

int comp(const void *i. const void *j);

int main(void)
(
int sort[lOO), if key;
int ·Pi

for(i~O; i<100; i++)

sort{i] = rand();

qsort(sort, lOa. sizeof(int), comp);

for(i=O; i<100; i++)

printf("'d\n", sort[i);

printf(IOEnter number to find: ~);

scanf("'d", &key);
p = bsearch(&key, sort, 100. sizeof(int), comp);
if(p) printf("Nurnber is in array.\n");
else printf("Number not found.\n");

return 0;
)

int comp(const void *i. const void *j)

(
return .(int·)i - *(int*)j;
)

3. iinclude cstdio.h>

int swri(int a, int b) ;

int subtract (int a, int b) ;
int mul(int a, int b) ;
int div(int a, int b) ;
int modulus (int a, int b) ;

/* initialize the pointer array */

int (·p[S» (int x, int y) = (
 sum, subtract, mul, diy, modulus
} ;

int main (void)

(
int result;
int i, j, op;

printfC-Enter two numbers: H);

scanfC-'d%d-, &i, &j);
printf("O: add, 1: subtract, 2: multiply, 3: divide, "I;
printf("4: modulus\n"l;
do (
printf{"Enter number of operation: H);
scanfC-'d", &op);
) while(op<O II op>41;

result = ("p[op}) (i. j I;

printf("'d", resultl;

return 0;
) ---
int sum(int a, int b)
{
return a ...b;
}

int subtract(int a, int b)

{
return a-hi
}

int mul(int a, int b}

(
return a*b;
)

int div(int a, int b)

{
if(bl return alb;
else return 0;
}
 ANSAERS 825
'2.7 EXERCISES
...

int modulus(int a, int b)

(
if(b) return a~b;

else return 0 i
)

ERCISES

2. 'include <stdio. h>

'include <stdlib.h>

int main(void)
(
int *p, i;

p = malloc(lO*sizeof(int)}:
H(!p) (
printf("Allocation Error");
exit(l);
}

for(i=O; i<10; i++) p[iJ = i+1;

for(i=O; i<10; i++) printf("%d" *(p+i));

free II') ;

l-eturn 0;
}

3. The statement
'p : malloc (10);

should be
I' : malloc(10);

Also, the value returned by malloc( ) is not verified as a

valid pointer.
 •
626 TEACH YOURSElF
T C

MASTERY SKILLS CHECK

1. When you specif'y the file name within angle brackets, the
compiler searches for the file in an implementation-defined
manner. When you enclose the file name within double quotes
the compiler first tries some other implementation-defined
manner to find the file. If that fails, it restarts the search as if
you had enclosed the file name wIthin angle brackets.

2. #ifdef DEBUG
if(! (j%2» (
printf (-j = %d\n" , j );
J = 0;
)
#endif

3. _if DEBUG==l
if(!(j%2)) {
pri':1t:fC-j = %d\n" . j 1;
j = 0
}
#endif

4. To undefinc a macro name use #undef

5 __ FILE__ io a predefined macro that contains the name c f the
source file currently being compiled.
6. The # operator makes the argument it precedes mto a quoted
string. The - operator concatenates two arguments.
7. #include <stdio.h>
#include ~stdlib.h>
#include <string.h>

int comp(const void *i, const void -J)i

int main(voidj
(

char str{) = -this is a test of qsort·i

qsort(str, strlen(str). 1, comp);

printf (str);
 ANSWERS 627
CUMULA TlVE SKJUS CHECK
'"
return 0;
}

int cornp(const void *i , canst v oid *j)

{
return *(char*)i - *(char*)j;
)

8. #include <stdio.h>
#include <stdlib.n>

int main(void)
(
double *p;

p =
malloc(sizeof(doublel) i
H(!p} (
printf ("Allocation Error");
exit(l) ;
}

*p == 99.01;
printf("%f", *p);
free (p) :

return 0;
}

I
CUMULA TlVE SKILLS CHECK

1 /* An electronic card catalog . */

# i nclude <Std10.h>
#include <string.h>
#include <stdlib.h>

#define MAX 100

int menu (void) ;

void display(int i);
void author_search (void) ;
void title_ search{voi d);
void
void save (void) ;
void load (void) ;
628 TfACH YOURSRF
c

struct catalog {
char name [80] ; /' author name ' /
char title[80]; / ' title ' /
char pub[80]; / ' publisher ' /
unsigned date; / ' date of publication ' /
unsigned char ed; '*
edition rl
} ·cat [MAX]; /* notice that this declares a pointer array ,· ,

int top = 0: '* last location used */

int main(void)
(
int choice;

load(); / * read in catalog */

do (
choice = menu();
switch (choJ.ce) (
case 1: enter(); / * enter books */
break;
case 2: author_search(): /* search by author
break;
case 3: title_search(); /* search by titl~ ,
break:
case 4: save ( ) ;
}
} while (choice! =5) ;

return 0:
}

/ * Return a menu selection. */

int menu(void)
{
int i;
char str[aO];

printf ('Card Catalog: \n');

printf(- 1. Enter\n-);
printf(- 2. Search by Author\n-);
printfC- 3. Search by Title\n-);
printf(- 4. Save catalog\n-);
printf(' 5. Quit\n');
 -
CUMIlV< T1VE SIOLLS CHECJ<
629
...

do (
printf(-Choose your selection: ·'i
gets (str) ;
i = atai (str);
printf(·\n·) ;
}_waile{i<l II i>S);

return i;
}

,I 1t Enter books into database. */

void enter(void)
{
l.nt i;
char temp [80 J ;

for(i=top; i<MAX: i++l (

/ * allocate memory for book info * /
catli] = malloc{sizeof(struct catalog»;
if{!cat[iJ) (
printf("Out of memory.\n-); ·
return;
}

printf(MEnter author name (ENTER to quit): .};

gets (cat[il->name) ;
if(!*cat[i]->name) break;
printf ("Enter ,title: .);
gets(cat[i}->title);
printf{"Enter publisher: .);
gets{cat[iJ->pub);
printf(·Enter copyright date: HI;
gets (temp) ;
cat[i]->date = (unsigned) atoi(temp);
printf ("Enter edition: .. );
gets (temp) ;
cat[i)->ed = (unsigned char) atoi(temp);
}
top = i;
}

/* Sea~cn by author. */
void author_search(void)
. {
630 TlACH YOURSElf
c

char name [80 J ;

int i, found:

printf ("Name: .. ) i
gets (name) ;
found = 0:
forC i=O; i<top; i++)
if(!strcmp(name, cat(i]->name» {
display (i I ;
found = 1;
printf (. \nn);
}

if ( ! found) printf ("Not Found\n");

}

/ * Search by title . • /
void title_search(void)
(
char tHle[80);
int i, found;

printf(~Title: ");
gets (title);

found = 0;
for(i=O; i<top; i++)
if(!strcmp(title, cat[iJ->title» (
display (i) ;
found = 1;
printfC"\n");
}

if(!found) printfC "Not Found\n");

}

/* Display catalog entry. */

void display(int i)
{

printfC"%s\n", catCi]->title);
printfC"by %s\n", cat[i)->name);
printf("published by %s\n", cat [i}->publ ;
printfC"Copyright: %u. %u edition\n", cat[i]->date.
cat(i]->ed);
}
 ANSWERS 631
CUMULAT1Vf SKlLLS CHECK

/* Load the catalog file. */

void load (void)
(
FILE ·fpi
int i;

if{(fp = fopen("catalog", "rb"))==NULL) (

, printf("Catalog file not on disk.\n~);
return:
)

if (fread(&top, sizeor-top, 1, fp) 1) {

!= /* read count if
printf ("Error readihg' count. \n·) ;
exit (1);
)

for(i=O; i<top; i++) (

cat[i] :::: malloc(sizeof(struct catalog»;
if(!cat[iJ) (
printf(nOut of rnemory.\n");
top:::: i-1;
break;
)
~f(fread(cat[i), sizeof(struct catalog), 1. fp}!= 1) {
print f (" Error reading catalog data. \n") ;
exit (1);
)
)

fclose(fp) ;
)

;* Save the catalog file. */

void save(void)
(
FILE *fp;
in t. i;

if(fp = fopen("catalog", "wb"))==NULL) (

printf("Cannot open catalog file.\n");
exit(l);
)
 if (fwrite(&top, sizeof top, 1, fp) !~ 1) ( 1* write count *1
printf("Error writing count.\n");
exit(1);
J

for(i=O; i<top; i++)

if (fwrite(cat[iJ, sizeof(struct ' catalog), 1, fp)!= 1) (
printf("Error writing catalog data.\n");
exit(l);
J

fclose(fp) ;
)

2. linclude <stdio.h>

Idefine CODE_IT (ch) -ch

int rnain(void)
(
int ch;
printfC-Enter a character: .);
ch = getchar();
printf("'c coded is 'co , ch, CODE_IT(ch»;

return 0;
)
Index
•

ft, 15, 140, 167, 168, 191,214,302,342,358 Access modifiers, 349-352

ftft, 62, 63, 65·66 acos( ) function, 425
< >,378-379 AddreSSing, 32-bit VS. I6-bit, 475
<,42,62,63 Allocation, dynamic, 402-407, 440-444
<-,62,63 AND bitwise operator, 358-359, 359-360, 361
«, 363·364, 369 AND logical operator, 62
>,41,62,63 ANSI C Standard
>-,62,63 and function arguments, 32
».363-364 keywords, 35
',17,167·168,189 library functions, 3
\n,59,60 API (Application Program Interface), 474-475
"'. 358 API ENTRY caning convention, 476
" 100,462 aTgc parameter, 216-217, 219
,(comma), 4,12,13,34,154,155,370-371 Argument(s)
{ }, 3, 6, 7, 26, 29, 46, 84, 154, 155 command-line, 215-220
,(decimal),12 definition of, 4, 33
,(dot operator), 302, 314, 315, 330 function, 4, 32-34
.. , (ellipsis), 200 Argument list, 34
-, 12, 367·369 variabIe·length, 200
.... 42, 62, 63 argv parameter, 216-217, 267-268
!, 62, 63 Arithmetic
!-, 62, 63 expressions, 17-20
'> , 314·315,330 operators, 17
- - (decrement operator). 54-56, 172, 173, 182 pointer, 172-175
(), 4, 18, 19, 173, 175, 281, 380 Array(s)
%,13,17,18 bounds checking and, 140-141,295
, (period), 243 definition of, 138, 139
++ (increment operator), 54-56, 173, 175 function pointer, 397-400
',393·58 indexing, 139
",393·395 initializing, 154-158
?,365·367 multidimensional, 151-154
•" 4 one-dimensional, 139-144
", 12 of pointers, 186·187
; (semicolon), 2, 11, 12, 154 with pointers, accessing, 176-178, 179-181
Ie .1 , 20 string as character, 145-150,
I 1,21 of strings, 159·16~
[j, 139, 151, 216, 248 of structures, 303, 305-310
~, 358, 359, 360-361 unsized, 155-156
I, j58 Arrow operator, 314·315, 330
11,62,63,65-66 ASCII character
codes, values for, 72-73, 88
set, 60, 383
a asctime( ) function, 435
asin( ) function, 425·426
abon( ) function, 444-445
Assembly code, C as replacement for, lOa, 129 633
abs( ) function, 445
•
634 TEACH YOURSELF

" C

Assignmem(s) ASCII. See ASCII character

and arrays, 141, 142 constants, 12, 120
multiple-variable, 367·368 input, interactive, 233, 235-237
shorthand, 368, 369 input, line-buffered, 69-74, 233, 234
statement, 12 output with printf(), 12-13, 234
type conversion in, 129-131 output with putchar(), 233-235
atan() function. 426 reading and \\'riting in file J10, 262-268
atan2() function, 427 Class, Window, 477
atof() function, 217-218, 445-446 clock( ) function, 343-344, 436
at01() function, 150, 217-218, 446·447 CLOCKS_PER_SEC macro, 435. 436
atol() function, 217-218, 447 clock_t type, 344, 435
ATOM data type, 485 Code block, 46·48, 52·53, 230
auto storage class specifier, 113,339,458-459 Comma operator, 370-37l
Comments, 20-22
Compilation, conditional, 381-388
Compiler(s)
command line, 7
Background color of window, creating. 484-485
compiling C programs with C++, 8
Bina1J1 stream, 259
error messages and, 8-9
Bit·fields. 3p·32B
header files, 4
Bit-shift operators, 363-364
preprocessor directives, 4-5
Bitmaps, ~72
Concatenation, 147
Bitwise 'operators, 358-362
Conditional statements, 41
Block of j;ode,' 46-48, 52-53
CONIO.H header file, 72, 236
BOOl data type, 478
const access modlfier, 349-351, 459
Borland C++ compiler, 8, 9
bTf'ak statement, 459 Constants, 12, 119-122
in loop, 89-92 backslash-character, 58-61
character, 12, 120
in SWitch statement, 95, 98-99, 466
Brush, 484-485
floating-point, 119, 120-}21
bsearch( ) function, 402, 448-449 integer, 119, 121
Bubble sort, 143-144 numeric, 120-12]
BYTE data type, 478 octal and hexadecimal, 121
string, 121
continue statement, 92-94, 460
cos( ) fun..::tion, 428
cosh( ) function, 428-429
.C extension, 8 cprintf() function, 236, 237, 23B
C: The Complete Reference (Schildt), 258 CreateWindow() API function, 485-487
Call by reference, 211-212 cscanf( ), 72, 236
Can by value, 211·212 ctime( ) function, 436-437
Callback function, 477 CTYPE.H header file, 179,412
CALLBACK calling convention, 477 Current location (pOSition)
calloc() function, 440-441 definition of, 259
Case sensitivity and C, 3, 12, 35 detennining, 286
case statement, 95-96, 98, 99, 459, 466 to start of file, positioning, 290, 291-292
Cast, type, 1320-133 Cursor, mouse, 484
ceil( ) function, 427 CW_USEDEFAULT macro, 486
char data type, 10,459
signed and unsigned modifiers with, 108. 109
promotion to int, 126
variable in place of int, using, III
Character(s) Data type(s)
arrays, strings as, 145 basic, in C, 10-11
modifiers, C, 107·111
 INDEX 635
...

table of all C, 109

Windows, 478
__ DATE__ macro, 392·393 fabs( ) function, 429
Oebugging False and true in C, 41
#error directive and, 389, 390-391 fdosee ) function, 262, 294
example programs for, 384-387 feof() function, 269·270, 279, 281
#}ine directive for, 389, 390-391 ferror() function, 269, 270, 279, 281
Declaration VS. definition, 201 fflush() function, 291, 292
default statement, 95, 98, 460, 466 fgetc( ) function, 262-267
"define directive, 229-232, 377·378 fge1S( ) function, 274·276, 295·296
defined compile time operator, 383, 387-388 File(s)
DeHnition files, 490 access modes for, 260-261
DefWindowProc( ) API function, 489 closing, 262
Desktop model in Windows, 471·472 current location in. See Current location
Dialog boxes, 473 definition of, 259-260
difftime( ) function, 437 erasing, 290, 291
Directives, preprocessor, 4-5, 229-232, 388-391 errors in, checking for, 269-274
DispatchMessage( ) API function, 489 executable, 8
do loop, 84·86, 460 extension "'hen naming, 8
Domain error, 425 flushing disk buffer of, 262, 291, 292
Dot operator, 302, 314, 315, 330 header, 4
double data type, 10-11,461 linking, 339
DWORD data type, 478 object, 8
Dynamic allocation, 402-407, 440·444 opening, 260-261
random access to, 285·289
reading and writing any type of data in, 279-285
reading and writing text, 274-277
reading and writing bytes from or to, 262·268
#eHf directive, 381, 382, 387 renaming, 290
Heise directive, 381,382,383 source, 8
else statement, 44-45, 461 streams and, 259-260
and code blocks, 46, 48 FILE data type, 260
with nested if statements, 75·78 __ FILE __ macro, 392-393
target statements and, 48, 51, 52 float data type, 10-11, 462
#endif directive, 381, 382 Floating-point values, 10, 12
enum type speCifier, 461 floor( ) function, 429-430
Enumerations, 352·355 fopen() function, 260-261, 294
£OF macro, 233, 234, 239, 247, 262, 264, 266-267, for loop, 49·53, 462
269·270 infinite, 81
#error directive, 388-389, 389·390 nested, 87-88
.Errors variations, 79-81
and function prototypes, 197, 201·202 Format speCifiers
syntax, 8 for printf(), 13, 110, 241·243
warning messages and, 8·9 forscanf(), IS, 16, 11 0, 246·253
exit( ) function, 220, 449-450 Forward declaration/reference, 198-199
EXIT] AlLURE macro, 449 fprintf( ) function, 275, 276
EXIT_SUCCESS macro, 449 data conversion in, 278-279
exp( ) function, 429 . printi~g output to screen with, 293
Expressions fputc( ) function, 262·263
arithmetic, 17-20 fpulS() function, 274, 275-276
definition of, 17 fread( ) function, 279·285, 405
type conversions in, 126-128 free() function, 403, 404·407, 441-442
extern storage class specifier, 339-341, 347, 461 · fscanf( ) function, 275, 276 .
data conversion in, 278-279
638 1lACII YDURSW

• C

fseek( ) function. 285-288 Hoare, C.A.R., 452

ftell( ) function, 286-288 HUGE_VAL macro, 425
Function(s) HWND data type, 478
arguments, 4, 32·34, 200-201 HWND_DESKTOP macro, 486
callback, 477
calling, 4, 24
creating, 23-26 J
declaration liS. definition, 201
Icons, 472, 483
definition of, 2
IDI_APPlICAnON macro, 484
formal parameters of, 32-34
IDI_WINLOGO macro, 484
forward declaration/reference of, J9~]99
#if directive, 381·382, 386-387
general form of, 3, 197
library. See Library functions if statement, 41~3, 463
parameterized, 33-34 code blocks with, 4f>.48
else statement with, 44-45, 76-78
passing arguments to, 211-214
nested, 75--78
pointers, 395-401
relational operators in, 41-42
prototypes, 24, 26, 196-206
if·else-ifladder, 76-77, 382
returning pointers from. 204-205
#ifde.f directive, 381, 382-383, 384-386
returning values with, 27-31
structures passed to, 304, 313, 315 #ifndef directive, 381, 383
structures returned by , 304, 312 #include directive, 4-5, 378-379, 380
window, 476-477, 478, 489 Indirection, 168, 170
[write( ) function, 279-285, 405-407 multiple, 188-190
In-line code vs. function cans, 378
int data type, 10,473
as default function return value, 29
Integer(')
GOl (Graphics Device Interface). 474 size in If>.bit vs. 32·bit environments, 10, 107,
getc( ) function, 262-263 108,475
getch( ) function, 235--237 values for signed and unsigned, 108, 109
getchar() function, 71-74, 203-204, 233, 234 variables, 10
getche() function, 72-74, 233, 235-236, 294 Imegral promotions, automatic. 126
~etche( ) function, 236 Inte~e, corn~ndHbased, 14~150
GetMessage( ) API fundion, 488-489 110
gets() function, 145-146, 176, 190, 238-240, 295 console, 229-253
scanf() vs., 150, 248 file, 258-296
GetStockObject( ) API function, 484-485 redirection, 293-294
gmtime() function, 318, 435, 438 Sa also Streams
goto statement, 100-101, 462-463 isalnum() function, 413
Graphical User Interface (GUI), 471 isalpha() function, 413-414
Graphics Device Interface, 474 iscntr1() function, 414
isdigit() function, 415
i,graph() function, 415-416
islower() function, 416
.H extension, 4 isprint( ) function, 416-417
Handle, 478 ispunct() function, 417-418
HANDLE data type, 478 isspace() function, 418
Header fi1es, 4, 5, 203 i,upper() function, 418-419
Heap (memory region), 403, 440 i,xdigit() fpnction, 419
Hexadecimal
constants, specifying, 121
number system, 50, 121 J.
HGDlOBJ data type, 485
 637
INO£)(
,.

maine) function, 3, 6, 24
K command-line arguments and, 215-220
kbhil( ) function, 236, 237-238 and prototypes, 205-Z06
Keyboard maUce{) function, 403-405, 441, 442-443
inputting numbers from. 15-16 MATH.H header file, 27, 203, 425
reading characters from, 69-74, 233-240 Memory, dynamic allocation of, 402-407, 440
reading strings from, 145-146 Menus, 472-473
Keywords, e, 35-36, 458-468 Message(s)
for basic data types, table of, 10 loop, 477-478, 487-489
and Windows, 473, 477
Microsoft Visual C++, 8, 9, 72
L Modulus operator, 17, 18
Mouse and Windows, 472
Label, 100, 462 MSG structure, 478, 488, 489
labs( ) function, 450 Multitasking and Windows, 474
Library functions, 3-4, 412
dynamic allocation, 440-444
mathematics, 424-434
miscellaneous, 444-455
and prototypes in header files, 203 Naming conventions (Windows), 490
string and character, 412·424 NOT logical operator, 62
time and date, 434-440 NULL Illflcro, 260
#line directive, 389, 390-391, 393 Null
__ LlNE__ macro, 392-393 pointers, 169·170
Loadeursor( ) API function, 484 string, 150
Loadlcon( ) API function, 483-484 terminator, 145
localtime() function, 316-317, 435, 438-439
log( ) function, 430
logIO() function, 430-431
LONG data type, 478
Octal
long type modifier, 107-11 0, 464
constants, specifying, 121
longjmp{ ) function, 450-452, 454
number system, 60, 121
Loops l's complement operator, 358, 359, 360-361
exiting, 89·92
Operator(s)
forcing next iteration of, 92·94
arithmetic, 17, 18
infinite, 81
arrow, 314-315, 330
message, 477, 478
assignment, 12, 367·369
nested,87
bit-shift, 363-364
LPARAM data type, 488
bitwise, 35ft...362
LPeSTR data type, 478
comma, 370--371
LPSTR data type, 478
decrement, 54·56
, LPVOID data type, 486
dot, 302, 314, 315, 330
LRESULT data type, 477
increment, 54-56
logical,61-66
modulus, 17, 18
M precedence of, 372
Macro(s) relational, 41-12, 61·64
built-in (e), 391-393 ternary, 365-367
function-like, 377-378, 379-380, 393-394 unary, 17
substitution, 229-232 OR bitwise operator, 358, 361·362
OR logical operator, 62
63B TEACH YOURSW

" C

Parameters, 32·34, 211 Range error, 425

declaration, classic vs. modern, 220·223 rand( ) function, 244, 453-454, 455
formal, as local variables, 114 RAND_MAX maCro, 453
to main(), 216-217 realloce) function, 443-444
pointers as, 191-192, 211 , 212-214 Recursion, 207-210
Parity bit, 362 register storage class specifier, 339, 341-342,
POINT structure, 488 343-346,418
Pointer(s) RegisterClassEx( ) API function, 485
accessing array with, 176-178,179-181 rename( ) function, 290
arithmetic, 172-175, 179 remove( ) function , 290, 291
arrays of, 186-187 return statement, 28-30,418
base type of, 167, 168-169, 171 rewind() function , 290, 291-292
decrem~nting , 181-182
function , 395-401
incrementing, 173, ] 75, 181
indexing, 178-179
null , 169-170 scanf() format specifiers, 16, 110, 246-253
operators. 63-168 table of, 247
as parameters, 191-192, 211. 212-214 scanf() function , 15-16, 72, 246-253
to poimers, 188-190 and arrays , 140
returned from functions , 204-205 and gets() , 150
to string constants, 183-185 pointers and, 191
to structures, 314-317 and strings, 121,247-248,249-250,250--251
void (generic), 279 Scanset, 248, 250-251
PostQuitMessage() API function, 489 Scientific notation, 1 ]9-120,242
pow() function, 431 Scope rules, 112
f1pragma directive, 389, 391 SEEK_CUR macro, 286
Preprocessor, 4-5, 229, 388, 393. See also SEEK_END macro, 286
Directives, preprocessor SEEK_SET macro, 286
printf() format specifiers, 13, 110, 241-243 setjmp{) function, 450-45], 454-455
table of, 242 SETJMP.H header file , 450, 451
printf() function , 4, 12-14, 241-246 short type modifier, 107-111, 418
backslash-character constants for, 58-61 ShowWindow( ) API function, 487
perfonning disk file 110 with, 294 Sign flag , 108
and pointers, 174 signed type modifier, 107-111, 418
strings and, 121, 146 sin() function , 431-432
using putchar( ) instead of, 234 sinh( ) function, 432
using puts( ) instead of, 239 size_t type, 279-280, 440, 448, 452
Programs sizeofoperator, 281-282, 305, 330, 419
components of, 2-7 Sorting with arrays. 143-144, 4D0-401, 452-453
creating and compiling, 7-9 sqrt() function, 27-28, 132-133, 203, 433
Prototypes, 24 , 26, 196-206 srand( ) function , 455
putc( ) function, 262-263 Statement(s)
putchar( ) function, 233-235 assignment, 12
puts() function, 191,238,239, 240 conditional, 41
definition of, 2
nulI,81
selection, 41
static storage class specifier, 339, 342-343, 346-347,
qsort( ) function, 400-401, 452-453 419
Quicksort, 208, 400, 452 __SfOC__ macro, 392
stderr (standard eTTor) stream, 293, 294
 INDEX 639
"
stdin (standard input) stream, 293, 294, 295-296 time() function, 316. 317, 436, 438, 439-440
STDlO.H header file, 5, 145, 233, 238, 260, 279 286 TIME.H header file, 316, 344, 434
STDLI8.H header file, 150, 244, 401, 403, 440, 452, __TIME__ macro, 392-393
453 time_t type, 316, 434, 435
stdout (standard output) stream, 293-294 tm structure, 316, 434-435, 438
Storage class specifiers, 339-347 tolower() function, 179-181,424
stTcat() function, 147,420 tcuppeT() function, ]79-]81,424
strehr( ) function, 420-421 TranslateMessage( ) API function, 489
strcmp() function, 147,421-422 True and false in C, 41
strcpy() function, 146-147, ISO, 191, 192,422 Two's complement approach , 108-109
Streams, 259-260 Type casts, 132·]33
standard, 293-296 Type conversions
Stnng(s) in assignments, 129-131
arrays of, 159-162 in expressions, 126-128
as ch .. _cacter arrays, 14,5..150,412 Type modifiers, 107-111
command-based interface and, 149-150 Type promotions and prototypes. automatic, 200
concatenating, 147 typedef statement, 356-358, 467
definition of, 4, 145
nuH, 150
printf() and, 121, 146
scanf() and, 121,247-248, 250-251 UINT data type, 488
table, 159-162, 183, 186-187 Unary operators, 17
String constants lIundef directive, 389, 390
definitiQnof, 121, 183 , union keyword, 467
using pointers to, 183-185 Unions, 329-333
STRlNG,H header file, 146, 412 UNIX, 258
strlen() function, 148, 191,351,422 unsigned type modifier, 107-111, 467
strstr( ) function, 422-423 - UpdateWindow() API function, 487
strtok( ) function, 423-424
struct keyw·ord, 301, 419
S~ructurc(s), 300-324
arrays of, 303, 305-310 Values
definition of, 300 assigning, to variables, 12
general fonn of, 301 returning, from functions, 27-30
members, acce.!>sing, 302, 301t-30S, 314-315 Variables
nested,318-324 assigning values to. 12
passed to functions, 304, 313 automatic, 339
pOintf'TS to, 314-317 declaring, 10-12, 13-14, ]]2-114
ret11rned by functions, 304, 312 initializing, 123-125
size of, determining, 305 using register for fast access to, 341-342, 343-346
variables, 301, 302-303 Variables, global, ll, 112, 114-118
switch statement, 94-99, 420 extern and. 339-34!, 347
nested,96 initializing, 123
static, 343
Variables, local, II, 112-114, 115-119
I auto, JJ 3, 339
tan( ) function, 433 initializing, 123, li4-125
tanh( ) function, 433-434 static, 342-343. 346-347
Ternary operator. 365-367 void, 10, 23, 467
Text stream, 259 function prototypes and, 200, 201
Time pointers, 279
broken-down, 316-317, 435 used to denote no return value. 20]
calendar, 316-317, 434 volatile access modifier, 349, 351-352. 468
llACH YOURSElF

naming cpnventions for functions and variables,

490, 491
while loop, 8?-84, 468 programming philosophy, 47().471
WlN32,474-475 W-NDOWS.H header file, 478
WINAPI calling convention, 476 "/inMain(), 476, 477, 479, 482
Window WM_DESTROY message macro, 489
components of, 475-476 WM_QUIT message macro, 489
creating, 485·487 WNDCLASSEX structure, 478, 482-483
displaying, 487 WORD data type, 478
style, macros for, 486 WPARAM data type, 488
Window class, 477 WS_OVERLAPPEDWINDOW macro, 486
Window function, 476-477, 478, 489'
Windows
application basics, 476-478
appJication skeleton, 478-489
data types, common, 478 XOR logical operation, 64--66
message-based interaction with programs, XOR bitwise operator, 358, 359
473-474
mouse and, 472
B.... I-"t:U;n~ pru~r-dmmi n g a uthor Herb Schildt has
lauJ!hl million ... to program in toda~'s most popu lar ., Disconr the power of functions
lan~ua2es. ."\0,", in this th ird edition of th e best. .I C se console and file J/O

"illin~ Ie (h" rlf ( . hi' pro, en plan for .. \\"ork with structures and unioll'i
sdccess bas bnn updated, e'paDded. and ., Explore advanced data tJpt's and opertors
rDhanct'd. There trul~ is no betler \\a~ to learn C. ./ Gain insight into C's preprocessor
" Learn how C can be used to create Windows
"riUm "jth Herb's uncompromising clarit~ and
programs
ant'Dtion 10 dt'lail. I ch \ ur: elf ( hegins" ith
the fUDdamentah. cm er... all the e"'tolials. and Best of all. "ou'lIlearn \\-ilh Schildt's pronn
concluds _hh a 100,," at \ome or(." ~ mo.,t " masterJ"method: the high)) ef1'ectin J"et \'enatiJe
ad~ ... nced reatures. \Iong the \\a~ 3rt' plent~ of a pproach that has helped millions become skilltd
prncticul t:\amp1rs. self-e~alualion c;.ldll checks. programmers. This approach lets JOu "ork at your
and exertisn.. an.'l~\ers to all e\.en'i'll'S are at the 0 "0 pace. in )'our own "3). E\'en if)ou han' found

bad or the book. so _'OU can ea... il~ ~our C confusing in the past. Schildt's clear. paced
prog.-.ss. presentation will make eHn adnnced topics easy
to understand .
Beca use C forms the basis for C++. aOer
Insidl' ~oar \\UJ
completing 1.. 14 1 OU f ( . lOU will ha\'e
., I.earn tbe !l.truclure of a C program the nect'ssar), foundation to ad~'ance to C++ .
'" l ndrrsland each of C\ program conlrol " ·hen it comes to teahing C. Herb schildt has it
statnnent
dOYtn to a science. And in this rnised and updated
." l: umiDt" ebta t~ pt'\. ,' ariable~. and e\.pressions best\ell er. he re~eals the formula that will make
." E_'-plore arra~ '0 and Mrin,::\
lO U a C programmer- the quickest. easiest. and
~ Lam about pointer, most elfecli\"{' wa~ possible!