DOC HOME SITE MAP MAN PAGES GNU INFO SEARCH PRINT BOOK
 

wcstod(3C)


wcstod, wcstof, wcstold -- convert wide string to floating point value

Synopsis

   #include <wchar.h>
   

double wcstod(const wchar_t *nptr, wchar_t **endptr);

float wcstof(const wchar_t *nptr, wchar_t **endptr);

long double wcstold(const wchar_t *nptr, wchar_t **endptr);

Description

wcstod returns, as a double-precision floating-point number, the wide character string pointed to by nptr. wcstof returns, as a single-precision floating-point number, the wide character string pointed to by nptr. wcstold returns, as a long double-precision floating-point number, the wide character string pointed to by nptr. Scanning occurs up to the first wide character that is unrecognized. The function recognizes an optional string that is composed of "white space" wide characters as defined by the iswspace function. The string is then followed by an optional sign then a sequence of digits optionally containing a decimal point character, followed by an exponential part (e or E) then another optional sign with an integer following it.

Also, instead of the regular decimal digit sequence, the string can be a hexadecimal floating value, an infinity, or a NaN. A hexadecimal floating value consists of 0x or OX followed by a sequence of hexadecimal digits optionally containing a decimal point character, followed by a binary exponent part p or P then an optional sign with an integer following it. The exponent part must be present if no decimal point character is present. An infinity is specified by the string inf or infinity case insensitive. A NaN is specified by nan case insensitive, followed by an optional sequence of zero or more alphanumeric or underscore _ characters between a pair of parenthesis. If the value of endptr is not null, a pointer to the wide character terminating the scan is returned in the location pointed to by endptr.

Return values

The function returns the value produced after the conversion process. If the function has not been performed then zero is returned and errno may be set to EINVAL.

If a correct value causes overflow, ±HUGE_VAL is returned, depending on the sign of the value, and errno is set to ERANGE.

If the value produced is correct but causes underflow, then zero will be returned with errno being set to ERANGE.

Errors

In the following conditions, these functions may fail and set errno to:

ERANGE
The value produced after the conversion process would cause either an overflow or underflow.

EINVAL
No conversion process could be carried out.

Usage

Zero and ±HUGE_VAL can be returned as a correct value after the conversion process. However, they can also be returned on error. To check for an error condition, zero should be assigned to errno followed by a call to one of these functions and then a check on errno. If the value of errno is non-zero it can be assumed that an error has occurred.

References

fscanf(3S), localeconv(3C), setlocale(3C), wchar(5), wcstol(3C)
© 2004 The SCO Group, Inc. All rights reserved.
UnixWare 7 Release 7.1.4 - 25 April 2004