2011-04-27 23:42:16 +02:00
|
|
|
/* $NetBSD: _wcstod.h,v 1.1 2006/04/15 12:17:23 tnozaki Exp $ */
|
|
|
|
|
|
|
|
/*-
|
|
|
|
* Copyright (c) 2002 Tim J. Robbins
|
|
|
|
* All rights reserved.
|
|
|
|
*
|
|
|
|
* Redistribution and use in source and binary forms, with or without
|
|
|
|
* modification, are permitted provided that the following conditions
|
|
|
|
* are met:
|
|
|
|
* 1. Redistributions of source code must retain the above copyright
|
|
|
|
* notice, this list of conditions and the following disclaimer.
|
|
|
|
* 2. Redistributions in binary form must reproduce the above copyright
|
|
|
|
* notice, this list of conditions and the following disclaimer in the
|
|
|
|
* documentation and/or other materials provided with the distribution.
|
|
|
|
*
|
|
|
|
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
|
|
|
|
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
|
|
|
|
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
|
|
|
|
* ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
|
|
|
|
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
|
|
|
|
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
|
|
|
|
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
|
|
|
|
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
|
|
|
|
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
|
|
|
|
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
|
|
|
|
* SUCH DAMAGE.
|
|
|
|
*
|
|
|
|
* Original version ID:
|
|
|
|
* FreeBSD: /repoman/r/ncvs/src/lib/libc/locale/wcstod.c,v 1.4 2004/04/07 09:47:56 tjr Exp
|
|
|
|
* NetBSD: wcstod.c,v 1.8 2006/04/13 01:25:13 tnozaki Exp
|
|
|
|
*/
|
|
|
|
|
|
|
|
/*
|
|
|
|
* function template for wcstof, wcstod, wcstold.
|
|
|
|
*
|
|
|
|
* parameters:
|
2013-03-11 19:00:30 +01:00
|
|
|
* _FUNCNAME : function name
|
|
|
|
* _RETURN_TYPE : return type
|
|
|
|
* _STRTOD_FUNC : real conversion function
|
2011-04-27 23:42:16 +02:00
|
|
|
*/
|
|
|
|
#ifndef __WCSTOD_H_
|
|
|
|
#define __WCSTOD_H_
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Convert a string to a double-precision number.
|
|
|
|
*
|
|
|
|
* This is the wide-character counterpart of strto{f,d,ld}(). So that
|
|
|
|
* we do not have to duplicate the code of strto{f,d,ld}() here,
|
|
|
|
* we convert the supplied wide character string to multibyte and
|
|
|
|
* call strto{f,d,ld}() on the result.
|
|
|
|
* This assumes that the multibyte encoding is compatible with ASCII
|
|
|
|
* for at least the digits, radix character and letters.
|
|
|
|
*/
|
|
|
|
_RETURN_TYPE
|
|
|
|
_FUNCNAME(const wchar_t * __restrict nptr, wchar_t ** __restrict endptr)
|
|
|
|
{
|
2013-03-11 19:00:30 +01:00
|
|
|
const wchar_t *src, *start;
|
|
|
|
_RETURN_TYPE val;
|
|
|
|
char *buf, *end;
|
|
|
|
size_t bufsiz, len;
|
2011-04-27 23:42:16 +02:00
|
|
|
|
2013-03-11 19:00:30 +01:00
|
|
|
_DIAGASSERT(nptr != NULL);
|
|
|
|
/* endptr may be null */
|
2011-04-27 23:42:16 +02:00
|
|
|
|
2013-03-11 19:00:30 +01:00
|
|
|
src = nptr;
|
|
|
|
while (iswspace((wint_t)*src) != 0)
|
|
|
|
++src;
|
|
|
|
if (*src == L'\0')
|
|
|
|
goto no_convert;
|
2011-04-27 23:42:16 +02:00
|
|
|
|
2013-03-11 19:00:30 +01:00
|
|
|
/*
|
|
|
|
* Convert the supplied numeric wide char. string to multibyte.
|
|
|
|
*
|
|
|
|
* We could attempt to find the end of the numeric portion of the
|
|
|
|
* wide char. string to avoid converting unneeded characters but
|
|
|
|
* choose not to bother; optimising the uncommon case where
|
|
|
|
* the input string contains a lot of text after the number
|
|
|
|
* duplicates a lot of strto{f,d,ld}()'s functionality and
|
|
|
|
* slows down the most common cases.
|
|
|
|
*/
|
|
|
|
start = src;
|
|
|
|
len = wcstombs(NULL, src, 0);
|
|
|
|
if (len == (size_t)-1)
|
|
|
|
/* errno = EILSEQ */
|
|
|
|
goto no_convert;
|
2011-04-27 23:42:16 +02:00
|
|
|
|
2013-03-11 19:00:30 +01:00
|
|
|
_DIAGASSERT(len > 0);
|
2011-04-27 23:42:16 +02:00
|
|
|
|
2013-03-11 19:00:30 +01:00
|
|
|
bufsiz = len;
|
|
|
|
buf = (void *)malloc(bufsiz + 1);
|
|
|
|
if (buf == NULL)
|
|
|
|
/* errno = ENOMEM */
|
|
|
|
goto no_convert;
|
2011-04-27 23:42:16 +02:00
|
|
|
|
2013-03-11 19:00:30 +01:00
|
|
|
len = wcstombs(buf, src, bufsiz + 1);
|
2011-04-27 23:42:16 +02:00
|
|
|
|
2013-03-11 19:00:30 +01:00
|
|
|
_DIAGASSERT(len == bufsiz);
|
|
|
|
_DIAGASSERT(buf[len] == '\0');
|
2011-04-27 23:42:16 +02:00
|
|
|
|
2013-03-11 19:00:30 +01:00
|
|
|
/* Let strto{f,d,ld}() do most of the work for us. */
|
|
|
|
val = _STRTOD_FUNC(buf, &end);
|
|
|
|
if (buf == end) {
|
|
|
|
free(buf);
|
|
|
|
goto no_convert;
|
|
|
|
}
|
2011-04-27 23:42:16 +02:00
|
|
|
|
2013-03-11 19:00:30 +01:00
|
|
|
/*
|
|
|
|
* We only know where the number ended in the _multibyte_
|
|
|
|
* representation of the string. If the caller wants to know
|
|
|
|
* where it ended, count multibyte characters to find the
|
|
|
|
* corresponding position in the wide char string.
|
|
|
|
*/
|
|
|
|
if (endptr != NULL)
|
|
|
|
/* XXX Assume each wide char is one byte. */
|
|
|
|
*endptr = __UNCONST(start + (size_t)(end - buf));
|
2011-04-27 23:42:16 +02:00
|
|
|
|
2013-03-11 19:00:30 +01:00
|
|
|
free(buf);
|
2011-04-27 23:42:16 +02:00
|
|
|
|
2013-03-11 19:00:30 +01:00
|
|
|
return val;
|
2011-04-27 23:42:16 +02:00
|
|
|
|
|
|
|
no_convert:
|
2013-03-11 19:00:30 +01:00
|
|
|
if (endptr != NULL)
|
|
|
|
*endptr = __UNCONST(nptr);
|
|
|
|
return 0.0;
|
2011-04-27 23:42:16 +02:00
|
|
|
}
|
|
|
|
#endif /*__WCSTOD_H_*/
|