K&R demonstrates an example of initializing a pointer array as follows:
char *month_name(int n)
{
static char *name[] = {
"Illegal month",
"January", "February", "March",
"April", "May", "June",
"July", "August", "September",
"October", "November", "December"
};
return (n < 1 || n > 12) ? name[0] : name[n];
}
K&R states that "this is an ideal application for an internal static array". But why is this so? Why not just make this an automatic pointer array?
If the array was an automatic variable, each function call would have to initialize thirteen pointers, and then return one of them to the calling code.
With the array as a static variable, it is initialized once (before the main() is called, I believe; at any rate, before the function is first used), and thereafter remains unchanged. This is likely to be more efficient, and probably measurably more efficient (though it might not matter unless you're using the month_name() function very intensively).
It would probably be better if there were some const qualifiers lurking in the code; the caller of the function should certainly not modify the result string:
const char *month_name(int n)
{
static const char * const name[] =
{
"Illegal month",
"January", "February", "March",
"April", "May", "June",
"July", "August", "September",
"October", "November", "December"
};
return (n < 1 || n > 12) ? name[0] : name[n];
}
