September 18

C has a rather small set of data types:

• int represents integers, with the ``natural'' precision for the platform you're on (typically, but not necessarily 32 bits these days). More or less precision can be specified using the modifiers short or long. You may also specify unsigned, whereas signed is the default. A literal such as 4321 is an int, if you append an L (e.g. 4321L it is a long int. The suffix U indicates unsigned. A leading 0 indicates an octal (base 8) integer, a leading 0x or 0X indicates a hexadecimal (base 16) integer.
• char is a single byte, representing a single character in the local character set. These may be signed or unsigned. A character literal is delimited by single quotes (e.g. 'x'). Certain characters can be represented by an escape sequence, e.g. \n for newline, \t for tab. chars are actually small integers, and can participate in integer arithmetic.
• double represents a number with a fractional part with double precision (how much precision is, again, machine-dependent). More precision may be specified with long, less by using type float. Include a decimal point (e.g. 123.4) or an exponent (e.g. 1e-2) to indicate double, add the suffix F to indicate float

How much precision? This is specified in the header <float.h> for the particular machine you are on.

Variables are declared before use. The declaration statement may include initialization:

int i, num, k= 3;
char a= 'T', *cp;
int *ip= &k;

num= k+3;
cp= &a;

A declaration may have the prefix const, which which case warnings (may) occur if there is an attempt to change its value after it is initialized.