atof - convert string to double-precision number

Syntax	#include <stdlib.h> double atof(const char *str);
Description	atof() converts an EBCDIC string pointed to by str into a floating-point number of type double. The string to be converted may be formatted as follows: [{tab|'BLANK'}...][+|-][digit...][.][digit...][{E|e}[+|-]digit...] or [{tab|'BLANK'}...][+|-]0{X|x}[ hexdigit...][.][hexdigit...][{P|p}[+|-]digit ...] All control characters for white space are legal for tab (see definition under isspace()). The atof(str) function differs from strtod(str,(char**)NULL) only in error handling.
Return val.	Floating-point number of type double
		for strings formatted as described above and representing a numeric value that is within the permissible floating point range.
	Extension 0 HUGE_VAL	for strings which do not correspond to the syntax described above. for strings whose numeric value lies outside the permissible floating-point range. errno is set to indicate the error.
Errors	atof() will fail if: ERANGE      The return value causes an overflow or underflow. (End)
Notes	atof() is completely contained in strtod(). However, the function continues to be offered because it is used in many existing applications. The decimal character in the string to be converted is affected by the locale (category LC_NUMERIC). The decimal point is the default. atof() also recognizes strings that begin with digits but then end with any character. atof() cuts off the numeric part, converts it according to the above description, and ignores the rest.
See also	atoi(), atol(), strtod(), strtol(), strtoul(), stdlib.h.