Data types in C language

Jun 7, 2018

1 - The predefined types

C is a typed language. This means in particular that any variable, constant or function is of a specific type. The type of an object defines how it is represented in memory.
The computer's memory is composed of a continuous sequence of bytes. Each byte of the memory ischaracterized by its address, which is an integer. Two contiguous bytes in memory have addresses that differ of a unit. When a variable is defined, it is assigned an address. This variable will correspond to a zone memory whose length (the number of bytes) is fixed by the type.

The memory size corresponding to the different types depends on the compilers; however, the ANSI standard specifies a number of constraints.
The basic types in C are characters, integers and floats (real numbers). They are designated by
the following keywords:

char
int
float
double
short
long
unsigned

The character type char

The keyword char designates an object of type character. A char can contain any element of the character set of the machine used. Most of the time, a char object is coded on a byte; it is the most elementary object in C. The character set used generally corresponds to ASCII coding (on 7 bits). The most machines now use the ISO-8859 (8-bit) character set, whose first 128 characters are ASCII characters. The last 128 characters (encoded on 8 bits) are used for characters specific to different languages.

One of the peculiarities of the char type in C is that it can be likened to an integer: any object of type char can be used in an expression that uses integer objects. For example, if c is of type char, the expression c + 1 is valid. It designates the next character in the ASCII code. Thus, the following program prints the character 'B'.

#include < stdio.h>

int main() { 
char c = 'E';
printf("%c", c + 1); 
return 0;
}

Integers types

The keyword for the integer type is int. An object of type int is represented by a "natural" word of the machine used, 32 bits for a DEC alpha or an Intel PC.
The int type may be preceded by a precision attribute (short or long) and / or a representation attribute (unsigned). An object of type short int has at least the size of a float and at most the size of an int. In general, a short int is coded on 16 bits. An object of type long int has at least the size of an int (64 bits on a DEC alpha, 32 bits on an Intel PC).