C Language Tutorial
Table of Contents:
1. A First Program
2. Let's Compute
3. Loops
4. Symbolic Constants
5. Conditionals
6. Pointers
7. Arrays
8. Character Arrays
9. I/O Capabilities
10. Functions
11. Command-line Arguments
12. Graphical Interfaces: Dialog Boxes
--------------------------------------------------------------------------------
This section contains a brief introduction to the C language. It is intended as a tutorial on the language, and aims at getting a reader new to C started as quickly as possible. It is certainly not intended as a substitute for any of the numerous textbooks on C.
The best way to learn a new ``human'' language is to speak it right from the outset, listening and repeating, leaving the intricacies of the grammar for later. The same applies to computer languages--to learn C, we must start writing C programs as quickly as possible.
An excellent textbook on C by two well-known and widely respected authors is:
The C Programming Language -- ANSI C
Brian W. C. Kernighan & Dennis M. Ritchie
Prentice Hall, 1988
Dennis Ritchie designed and implemented the first C compiler on a PDP-11 (a prehistoric machine by today's standards, yet one which had enormous influence on modern scientific computation). The C language was based on two (now defunct) languages: BCPL, written by Martin Richards, and B, written by Ken Thompson in 1970 for the first UNIX system on a PDP-7. The original ``official'' C language was the ``K & R'' C, the nickname coming from the names of the two authors of the original ``The C Programming Language''. In 1988, the American National Standards Institute (ANSI) adopted a ``new and improved'' version of C, known today as ``ANSI C''. This is the version described in the current edition of ``The C Programming Language -- ANSI C''. The ANSI version contains many revisions to the syntax and the internal workings of the language, the major ones being improved calling syntax for procedures and standarization of most (but, unfortunately, not quite all!) system libraries.
1. A First Program
Let's be polite and start by saluting the world! Type the following program into your favorite editor:
--------------------------------------------------------------------------------
#include < stdio.h>
void main()
{
printf("nHello Worldn");
}
--------------------------------------------------------------------------------
Save the code in the file hello.c, then compile it by typing:
gcc hello.c
This creates an executable file a.out, which is then executed simply by typing its name. The result is that the characters `` Hello World'' are printed out, preceded by an empty line.
A C program contains functions and variables. The functions specify the tasks to be performed by the program. The ``main'' function establishes the overall logic of the code. It is normally kept short and calls different functions to perform the necessary sub-tasks. All C codes must have a ``main'' function.
Our hello.c code calls printf, an output function from the I/O (input/output) library (defined in the file stdio.h). The original C language did not have any built-in I/O statements whatsoever. Nor did it have much arithmetic functionality. The original language was really not intended for ''scientific'' or ''technical'' computation.. These functions are now performed by standard libraries, which are now part of ANSI C. The K & R textbook lists the content of these and other standard libraries in an appendix.
The printf line prints the message ``Hello World'' on ``stdout'' (the output stream corresponding to the X-terminal window in which you run the code); ``n'' prints a ``new line'' character, which brings the cursor onto the next line. By construction, printf never inserts this character on its own: the following program would produce the same result:
--------------------------------------------------------------------------------
#include < stdio.h>
void main()
{
printf("n");
printf("Hello World");
printf("n");
}
--------------------------------------------------------------------------------
Try leaving out the ``n'' lines and see what happens.
The first statement ``#include < stdio.h>'' includes a specification of the C I/O library. All variables in C must be explicitly defined before use: the ``.h'' files are by convention ``header files'' which contain definitions of variables and functions necessary for the functioning of a program, whether it be in a user-written section of code, or as part of the standard C libaries. The directive ``#include'' tells the C compiler to insert the contents of the specified file at that point in the code. The ``< ...>'' notation instructs the compiler to look for the file in certain ``standard'' system directories.
The void preceeding ``main'' indicates that main is of ``void'' type--that is, it has no type associated with it, meaning that it cannot return a result on execution.
The ``;'' denotes the end of a statement. Blocks of statements are put in braces {...}, as in the definition of functions. All C statements are defined in free format, i.e., with no specified layout or column assignment. Whitespace (tabs or spaces) is never significant, except inside quotes as part of a character string. The following program would produce exactly the same result as our earlier example:
#include < stdio.h>
void main(){printf("nHello Worldn");}
The reasons for arranging your programs in lines and indenting to show structure should be obvious!
--------------------------------------------------------------------------------
2. Let's Compute
The following program, sine.c, computes a table of the sine function for angles between 0 and 360 degrees.
--------------------------------------------------------------------------------
/************************/
/* Table of */
/* Sine Function */
/************************/
/* Michel Vallieres */
/* Written: Winter 1995 */
#include < stdio.h>
#include < math.h>
void main()
{
int angle_degree;
double angle_radian, pi, value;
/* Print a header */
printf ("nCompute a table of the sine functionnn");
/* obtain pi once for all */
/* or just use pi = M_PI, where
M_PI is defined in math.h */
pi = 4.0*atan(1.0);
printf ( " Value of PI = %f nn", pi );
printf ( " angle Sine n" );
angle_degree=0; /* initial angle value */
/* scan over angle */
while ( angle_degree <= 360 ) /* loop until angle_degree > 360 */
{
angle_radian = pi * angle_degree/180.0 ;
value = sin(angle_radian);
printf ( " %3d %f n ", angle_degree, value );
angle_degree = angle_degree + 10; /* increment the loop index */
}
}
--------------------------------------------------------------------------------
The code starts with a series of comments indicating its the purpose, as well as its author. It is considered good programming style to identify and document your work (although, sadly, most people only do this as an afterthought). Comments can be written anywhere in the code: any characters between /* and */ are ignored by the compiler and can be used to make the code easier to understand. The use of variable names that are meaningful within the context of the problem is also a good idea.
The #include statements now also include the header file for the standard mathematics library math.h. This statement is needed to define the calls to the trigonometric functions atan and sin. Note also that the compilation must include the mathematics library explicitly by typing
gcc sine.c -lm
Variable names are arbitrary (with some compiler-defined maximum length, typically 32 characters). C uses the following standard variable types:
int -> integer variable
short -> short integer
long -> long integer
float -> single precision real (floating point) variable
double -> double precision real (floating point) variable
char -> character variable (single byte)
The compilers checks for consistency in the types of all variables used in any code. This feature is intended to prevent mistakes, in particular in mistyping variable names. Calculations done in the math library routines are usually done in double precision arithmetic (64 bits on most workstations). The actual number of bytes used in the internal storage of these data types depends on the machine being used.
The printf function can be instructed to print integers, floats and strings properly. The general syntax is
printf( "format", variables );
where "format" specifies the converstion specification and variables is a list of quantities to print. Some useful formats are
%.nd integer (optional n = number of columns; if 0, pad with zeroes)
%m.nf float or double (optional m = number of columns,
n = number of decimal places)
%ns string (optional n = number of columns)
%c character
n t to introduce new line or tab
g ring the bell (``beep'') on the terminal
--------------------------------------------------------------------------------
3. Loops
Most real programs contain some construct that loops within the program, performing repetitive actions on a stream of data or a region of memory. There are several ways to loop in C. Two of the most common are the while loop:
while (expression)
{
...block of statements to execute...
}
and the for loop:
for (expression_1; expression_2; expression_3)
{
...block of statements to execute...
}
The while loop continues to loop until the conditional expression becomes false. The condition is tested upon entering the loop. Any logical construction (see below for a list) can be used in this context.
The for loop is a special case, and is equivalent to the following while loop:
expression_1;
while (expression_2)
{
...block of statements...
expression_3;
}
For instance, the following structure is often encountered:
i = initial_i;
while (i <= i_max)
{
...block of statements...
i = i + i_increment;
}
This structure may be rewritten in the easier syntax of the for loop as:
for (i = initial_i; i <= i_max; i = i + i_increment)
{
...block of statements...
}
Infinite loops are possible (e.g. for(;;)), but not too good for your computer budget! C permits you to write an infinite loop, and provides the break statement to ``breakout '' of the loop. For example, consider the following (admittedly not-so-clean) re-write of the previous loop:
angle_degree = 0;
for ( ; ; )
{
...block of statements...
angle_degree = angle_degree + 10;
if (angle_degree == 360) break;
}
The conditional if simply asks whether angle_degree is equal to 360 or not; if yes, the loop is stopped.
--------------------------------------------------------------------------------
4. Symbolic Constants
You can define constants of any type by using the #define compiler directive. Its syntax is simple--for instance
#define ANGLE_MIN 0
#define ANGLE_MAX 360
would define ANGLE_MIN and ANGLE_MAX to the values 0 and 360, respectively. C distinguishes between lowercase and uppercase letters in variable names. It is customary to use capital letters in defining global constants.
--------------------------------------------------------------------------------
5. Conditionals
Conditionals are used within the if and while constructs:
if (conditional_1)
{
...block of statements executed if conditional_1 is true...
}
else if (conditional_2)
{
...block of statements executed if conditional_2 is true...
}
else
{
...block of statements executed otherwise...
}
and any variant that derives from it, either by omitting branches or by including nested conditionals.
Conditionals are logical operations involving comparison of quantities (of the same type) using the conditional operators:
< smaller than
<= smaller than or equal to
== equal to
!= not equal to
>= greater than or equal to
> greater than
and the boolean operators
&& and
|| or
! not
Another conditional use is in the switch construct:
switch (expression)
{
case const_expression_1:
{
...block of statements...
break;
}
case const_expression_2:
{
...block of statements...
break;
}
default:
{
...block of statements..
}
}
The appropriate block of statements is executed according to the value of the expression, compared with the constant expressions in the case statement. The break statements insure that the statements in the cases following the chosen one will not be executed. If you would want to execute these statements, then you would leave out the break statements. This construct is particularly useful in handling input variables.
--------------------------------------------------------------------------------
6. Pointers
The C language allows the programmer to ``peek and poke'' directly into memory locations. This gives great flexibility and power to the language, but it also one of the great hurdles that the beginner must overcome in using the language.
All variables in a program reside in memory; the statements
float x;
x = 6.5;
request that the compiler reserve 4 bytes of memory (on a 32-bit computer) for the floating-point variable x, then put the ``value'' 6.5 in it.
Sometimes we want to know where a variable resides in memory. The address (location in memory) of any variable is obtained by placing the operator ``&'' before its name. Therefore &x is the address of x. C allows us to go one stage further and define a variable, called a pointer, that contains the address of (i.e. ``points to'') other variables. For example:
float x;
float* px;
x = 6.5;
px = &x;
defines px to be a pointer to objects of type float, and sets it equal to the address of x:
Pointer use for a variable
The content of the memory location referenced by a pointer is obtained using the ``*'' operator (this is called dereferencing the pointer). Thus, *px refers to the value of x.
C allows us to perform arithmetic operations using pointers, but beware that the ``unit'' in pointer arithmetic is the size (in bytes) of the object to which the pointer points. For example, if px is a pointer to a variable x of type float, then the expression px + 1 refers not to the next bit or byte in memory but to the location of the next float after x (4 bytes away on most workstations); if x were of type double, then px + 1 would refer to a location 8 bytes (the size of a double)away, and so on. Only if x is of type char will px + 1 actually refer to the next byte in memory.
Thus, in
char* pc;
float* px;
float x;
x = 6.5;
px = &x;
pc = (char*) px;
(the (char*) in the last line is a ``cast'', which converts one data type to another), px and pc both point to the same location in memory--the address of x--but px + 1 and pc + 1 point to different memory locations.
Consider the following simple code.
--------------------------------------------------------------------------------
void main()
{
float x, y; /* x and y are of float type */
float *fp, *fp2; /* fp and fp2 are pointers to float */
x = 6.5; /* x now contains the value 6.5 */
/* print contents and address of x */
printf("Value of x is %f, address of x %ldn", x, &x);
fp = &x; /* fp now points to location of x */
/* print the contents of fp */
printf("Value in memory location fp is %fn", *fp);
/* change content of memory location */
*fp = 9.2;
printf("New value of x is %f = %f n", *fp, x);
/* perform arithmetic */
*fp = *fp + 1.5;
printf("Final value of x is %f = %f n", *fp, x);
/* transfer values */
y = *fp;
fp2 = fp;
printf("Transfered value into y = %f and fp2 = %f n", y, *fp2);
}
--------------------------------------------------------------------------------
Run this code to see the results of these different operations. Note that, while the value of a pointer (if you print it out with printf) is typically a large integer, denoting some particular memory location in the computer, pointers are not integers--they are a completely different data type.
--------------------------------------------------------------------------------
7. Arrays
Arrays of any type can be formed in C. The syntax is simple:
type name[dim];
In C, arrays starts at position 0. The elements of the array occupy adjacent locations in memory. C treats the name of the array as if it were a pointer to the first element--this is important in understanding how to do arithmetic with arrays. Thus, if v is an array, *v is the same thing as v[0], *(v+1) is the same thing as v[1], and so on:
Pointer use for an array
Consider the following code, which illustrates the use of pointers:
--------------------------------------------------------------------------------
#define SIZE 3
void main()
{
float x[SIZE];
float *fp;
int i;
/* initialize the array x */
/* use a "cast" to force i */
/* into the equivalent float */
for (i = 0; i < SIZE; i++)
x[i] = 0.5*(float)i;
/* print x */
for (i = 0; i < SIZE; i++)
printf(" %d %f n", i, x[i]);
/* make fp point to array x */
fp = x;
/* print via pointer arithmetic */
/* members of x are adjacent to */
/* each other in memory */
/* *(fp+i) refers to content of */
/* memory location (fp+i) or x[i] */
for (i = 0; i < SIZE; i++)
printf(" %d %f n", i, *(fp+i));
}
--------------------------------------------------------------------------------
(The expression ``i++'' is C shorthand for ``i = i + 1''.) Since x[i] means the i-th element of the array x, and fp = x points to the start of the x array, then *(fp+i) is the content of the memory address i locations beyond fp, that is, x[i].
--------------------------------------------------------------------------------
8. Character Arrays
A string constant , such as
"I am a string"
is an array of characters. It is represented internally in C by the ASCII characters in the string, i.e., ``I'', blank, ``a'', ``m'',... for the above string, and terminated by the special null character ``
