2005-08-04 18:38:57 +02:00
|
|
|
/*****************************************************************************
|
2006-03-21 17:42:34 +01:00
|
|
|
* unicode.c: Unicode <-> locale functions
|
2005-08-04 18:38:57 +02:00
|
|
|
*****************************************************************************
|
2006-02-12 20:48:50 +01:00
|
|
|
* Copyright (C) 2005-2006 the VideoLAN team
|
2010-10-08 19:30:11 +02:00
|
|
|
* Copyright © 2005-2010 Rémi Denis-Courmont
|
2005-08-04 18:38:57 +02:00
|
|
|
*
|
|
|
|
* Authors: Rémi Denis-Courmont <rem # videolan.org>
|
|
|
|
*
|
|
|
|
* This program is free software; you can redistribute it and/or modify
|
|
|
|
* it under the terms of the GNU General Public License as published by
|
|
|
|
* the Free Software Foundation; either version 2 of the License, or
|
|
|
|
* (at your option) any later version.
|
|
|
|
*
|
|
|
|
* This program is distributed in the hope that it will be useful,
|
|
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
|
|
* GNU General Public License for more details.
|
|
|
|
*
|
|
|
|
* You should have received a copy of the GNU General Public License
|
|
|
|
* along with this program; if not, write to the Free Software
|
2006-01-13 00:10:04 +01:00
|
|
|
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston MA 02110-1301, USA.
|
2005-08-04 18:38:57 +02:00
|
|
|
*****************************************************************************/
|
|
|
|
|
|
|
|
/*****************************************************************************
|
|
|
|
* Preamble
|
|
|
|
*****************************************************************************/
|
2008-01-23 22:50:58 +01:00
|
|
|
#ifdef HAVE_CONFIG_H
|
|
|
|
# include "config.h"
|
|
|
|
#endif
|
|
|
|
|
2008-05-31 20:56:22 +02:00
|
|
|
#include <vlc_common.h>
|
2011-02-17 13:14:52 +01:00
|
|
|
|
|
|
|
#include "libvlc.h"
|
2006-11-26 15:20:34 +01:00
|
|
|
#include <vlc_charset.h>
|
2006-02-11 22:46:06 +01:00
|
|
|
|
2006-02-13 14:16:08 +01:00
|
|
|
#include <assert.h>
|
2006-02-12 12:39:00 +01:00
|
|
|
|
2006-02-11 16:13:56 +01:00
|
|
|
#include <stdio.h>
|
2006-02-21 12:53:49 +01:00
|
|
|
#include <stdarg.h>
|
2006-03-27 13:26:50 +02:00
|
|
|
#include <stdlib.h>
|
2006-02-11 22:46:06 +01:00
|
|
|
#include <sys/types.h>
|
2006-02-13 21:06:22 +01:00
|
|
|
#ifdef UNDER_CE
|
|
|
|
# include <tchar.h>
|
2010-10-28 04:58:57 +02:00
|
|
|
#elif defined(WIN32)
|
|
|
|
# include <io.h>
|
2006-02-13 21:06:22 +01:00
|
|
|
#endif
|
2010-01-25 23:05:53 +01:00
|
|
|
#include <errno.h>
|
2010-10-08 20:02:23 +02:00
|
|
|
#include <wctype.h>
|
2006-02-13 21:11:48 +01:00
|
|
|
|
2008-05-03 23:02:44 +02:00
|
|
|
/**
|
|
|
|
* Releases (if needed) a localized or uniformized string.
|
|
|
|
* @param str non-NULL return value from FromLocale() or ToLocale().
|
|
|
|
*/
|
2006-11-11 12:27:58 +01:00
|
|
|
void LocaleFree (const char *str)
|
|
|
|
{
|
2010-07-18 11:06:05 +02:00
|
|
|
#ifdef ASSUME_UTF8
|
|
|
|
(void) str;
|
2008-03-03 02:45:45 +01:00
|
|
|
#else
|
2010-07-18 11:06:05 +02:00
|
|
|
free ((char *)str);
|
2006-02-12 12:39:00 +01:00
|
|
|
#endif
|
2006-11-11 12:27:58 +01:00
|
|
|
}
|
|
|
|
|
2006-02-12 12:39:00 +01:00
|
|
|
|
2006-03-22 12:24:45 +01:00
|
|
|
/**
|
2008-05-03 23:02:44 +02:00
|
|
|
* Converts a string from the system locale character encoding to UTF-8.
|
2006-03-22 12:24:45 +01:00
|
|
|
*
|
2008-05-03 23:02:44 +02:00
|
|
|
* @param locale nul-terminated string to convert
|
2006-03-22 12:24:45 +01:00
|
|
|
*
|
|
|
|
* @return a nul-terminated UTF-8 string, or NULL in case of error.
|
|
|
|
* To avoid memory leak, you have to pass the result to LocaleFree()
|
|
|
|
* when it is no longer needed.
|
|
|
|
*/
|
2006-11-11 12:27:58 +01:00
|
|
|
char *FromLocale (const char *locale)
|
2006-02-12 12:39:00 +01:00
|
|
|
{
|
2010-07-18 11:06:05 +02:00
|
|
|
#ifdef ASSUME_UTF8
|
|
|
|
return (char *)locale;
|
|
|
|
#else
|
2010-10-30 12:00:23 +02:00
|
|
|
return locale ? FromCharset ("", locale, strlen(locale)) : NULL;
|
2010-07-18 11:06:05 +02:00
|
|
|
#endif
|
2006-02-12 12:39:00 +01:00
|
|
|
}
|
|
|
|
|
2008-05-03 23:02:44 +02:00
|
|
|
/**
|
|
|
|
* converts a string from the system locale character encoding to utf-8,
|
|
|
|
* the result is always allocated on the heap.
|
|
|
|
*
|
|
|
|
* @param locale nul-terminated string to convert
|
|
|
|
*
|
|
|
|
* @return a nul-terminated utf-8 string, or null in case of error.
|
|
|
|
* The result must be freed using free() - as with the strdup() function.
|
|
|
|
*/
|
2006-11-11 12:27:58 +01:00
|
|
|
char *FromLocaleDup (const char *locale)
|
2006-02-12 20:48:50 +01:00
|
|
|
{
|
2010-07-18 11:06:05 +02:00
|
|
|
#ifdef ASSUME_UTF8
|
|
|
|
return strdup (locale);
|
|
|
|
#else
|
2010-10-30 12:00:23 +02:00
|
|
|
return FromCharset ("", locale, strlen(locale));
|
2010-07-18 11:06:05 +02:00
|
|
|
#endif
|
2006-02-12 20:48:50 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
|
2006-03-22 12:24:45 +01:00
|
|
|
/**
|
2008-05-03 23:02:44 +02:00
|
|
|
* ToLocale: converts an UTF-8 string to local system encoding.
|
2006-03-22 12:24:45 +01:00
|
|
|
*
|
|
|
|
* @param utf8 nul-terminated string to be converted
|
|
|
|
*
|
|
|
|
* @return a nul-terminated string, or NULL in case of error.
|
|
|
|
* To avoid memory leak, you have to pass the result to LocaleFree()
|
|
|
|
* when it is no longer needed.
|
|
|
|
*/
|
2006-11-11 12:27:58 +01:00
|
|
|
char *ToLocale (const char *utf8)
|
2006-02-12 12:39:00 +01:00
|
|
|
{
|
2010-07-18 11:06:05 +02:00
|
|
|
#ifdef ASSUME_UTF8
|
|
|
|
return (char *)utf8;
|
|
|
|
#else
|
2010-10-30 12:00:23 +02:00
|
|
|
size_t outsize;
|
|
|
|
return utf8 ? ToCharset ("", utf8, &outsize) : NULL;
|
2010-07-18 11:06:05 +02:00
|
|
|
#endif
|
2006-02-12 12:39:00 +01:00
|
|
|
}
|
|
|
|
|
2006-02-21 12:53:49 +01:00
|
|
|
|
2008-05-03 23:02:44 +02:00
|
|
|
/**
|
|
|
|
* converts a string from UTF-8 to the system locale character encoding,
|
|
|
|
* the result is always allocated on the heap.
|
|
|
|
*
|
|
|
|
* @param utf8 nul-terminated string to convert
|
|
|
|
*
|
|
|
|
* @return a nul-terminated string, or null in case of error.
|
|
|
|
* The result must be freed using free() - as with the strdup() function.
|
|
|
|
*/
|
2008-04-30 19:04:05 +02:00
|
|
|
char *ToLocaleDup (const char *utf8)
|
2006-02-12 12:39:00 +01:00
|
|
|
{
|
2010-07-18 11:06:05 +02:00
|
|
|
#ifdef ASSUME_UTF8
|
|
|
|
return strdup (utf8);
|
|
|
|
#else
|
2010-10-30 12:00:23 +02:00
|
|
|
size_t outsize;
|
|
|
|
return ToCharset ("", utf8, &outsize);
|
2010-07-18 11:06:05 +02:00
|
|
|
#endif
|
2006-02-12 12:39:00 +01:00
|
|
|
}
|
|
|
|
|
2008-05-03 23:02:44 +02:00
|
|
|
/**
|
|
|
|
* Formats an UTF-8 string as vfprintf(), then print it, with
|
|
|
|
* appropriate conversion to local encoding.
|
|
|
|
*/
|
2006-08-31 16:55:20 +02:00
|
|
|
int utf8_vfprintf( FILE *stream, const char *fmt, va_list ap )
|
2006-02-21 12:53:49 +01:00
|
|
|
{
|
2010-08-21 11:24:56 +02:00
|
|
|
#ifdef ASSUME_UTF8
|
|
|
|
return vfprintf (stream, fmt, ap);
|
|
|
|
#else
|
2006-03-12 23:52:25 +01:00
|
|
|
char *str;
|
2010-07-22 19:34:39 +02:00
|
|
|
int res;
|
|
|
|
|
2010-08-21 11:24:56 +02:00
|
|
|
# if defined( WIN32 ) && !defined( UNDER_CE )
|
2010-07-22 19:34:39 +02:00
|
|
|
/* Writing to the console is a lot of fun on Microsoft Windows.
|
|
|
|
* If you use the standard I/O functions, you must use the OEM code page,
|
|
|
|
* which is different from the usual ANSI code page. Or maybe not, if the
|
|
|
|
* user called "chcp". Anyway, we prefer Unicode. */
|
|
|
|
int fd = _fileno (stream);
|
|
|
|
if (likely(fd != -1) && _isatty (fd))
|
|
|
|
{
|
|
|
|
res = vasprintf (&str, fmt, ap);
|
|
|
|
if (unlikely(res == -1))
|
|
|
|
return -1;
|
|
|
|
|
|
|
|
size_t wlen = 2 * (res + 1);
|
|
|
|
wchar_t *wide = malloc (wlen);
|
|
|
|
if (likely(wide != NULL))
|
|
|
|
{
|
|
|
|
wlen = MultiByteToWideChar (CP_UTF8, 0, str, res + 1, wide, wlen);
|
|
|
|
if (wlen > 0)
|
|
|
|
{
|
|
|
|
HANDLE h = (HANDLE)(intptr_t)_get_osfhandle (fd);
|
|
|
|
DWORD out;
|
|
|
|
|
|
|
|
WriteConsoleW (h, wide, wlen - 1, &out, NULL);
|
|
|
|
}
|
|
|
|
else
|
|
|
|
res = -1;
|
|
|
|
free (wide);
|
|
|
|
}
|
|
|
|
else
|
|
|
|
res = -1;
|
|
|
|
free (str);
|
|
|
|
return res;
|
|
|
|
}
|
2010-08-21 11:24:56 +02:00
|
|
|
# endif
|
2010-07-22 19:34:39 +02:00
|
|
|
|
2010-08-21 11:24:56 +02:00
|
|
|
res = vasprintf (&str, fmt, ap);
|
2010-07-22 19:34:39 +02:00
|
|
|
if (unlikely(res == -1))
|
2006-03-12 23:52:25 +01:00
|
|
|
return -1;
|
|
|
|
|
2010-08-21 11:24:56 +02:00
|
|
|
char *ansi = ToLocaleDup (str);
|
|
|
|
free (str);
|
|
|
|
|
2010-11-03 16:18:10 +01:00
|
|
|
if (ansi == NULL)
|
|
|
|
return -1;
|
2010-08-21 11:24:56 +02:00
|
|
|
fputs (ansi, stream);
|
|
|
|
free (ansi);
|
2006-03-12 23:52:25 +01:00
|
|
|
return res;
|
2010-08-21 11:24:56 +02:00
|
|
|
#endif
|
2006-02-21 12:53:49 +01:00
|
|
|
}
|
|
|
|
|
2008-05-03 23:02:44 +02:00
|
|
|
/**
|
|
|
|
* Formats an UTF-8 string as fprintf(), then print it, with
|
|
|
|
* appropriate conversion to local encoding.
|
|
|
|
*/
|
2006-02-21 12:53:49 +01:00
|
|
|
int utf8_fprintf( FILE *stream, const char *fmt, ... )
|
|
|
|
{
|
2006-03-12 23:52:25 +01:00
|
|
|
va_list ap;
|
|
|
|
int res;
|
2006-02-21 12:53:49 +01:00
|
|
|
|
2006-03-12 23:52:25 +01:00
|
|
|
va_start( ap, fmt );
|
|
|
|
res = utf8_vfprintf( stream, fmt, ap );
|
|
|
|
va_end( ap );
|
|
|
|
return res;
|
2006-02-21 12:53:49 +01:00
|
|
|
}
|
|
|
|
|
2006-03-22 13:26:11 +01:00
|
|
|
|
2010-10-08 19:30:11 +02:00
|
|
|
/**
|
|
|
|
* Converts the first character from a UTF-8 sequence into a code point.
|
|
|
|
*
|
|
|
|
* @param str an UTF-8 bytes sequence
|
|
|
|
* @return 0 if str points to an empty string, i.e. the first character is NUL;
|
|
|
|
* number of bytes that the first character occupies (from 1 to 4) otherwise;
|
|
|
|
* -1 if the byte sequence was not a valid UTF-8 sequence.
|
|
|
|
*/
|
2011-02-10 18:43:58 +01:00
|
|
|
size_t vlc_towc (const char *str, uint32_t *restrict pwc)
|
2005-08-04 18:38:57 +02:00
|
|
|
{
|
2006-12-15 15:16:08 +01:00
|
|
|
uint8_t *ptr = (uint8_t *)str;
|
2006-09-03 17:01:52 +02:00
|
|
|
assert (str != NULL);
|
|
|
|
|
2010-10-08 19:30:11 +02:00
|
|
|
uint8_t c = ptr[0];
|
2006-12-15 15:16:08 +01:00
|
|
|
|
2010-10-08 19:30:11 +02:00
|
|
|
if (unlikely(c == '\0'))
|
|
|
|
{
|
|
|
|
*pwc = 0;
|
|
|
|
return 0;
|
|
|
|
}
|
2006-12-15 15:16:08 +01:00
|
|
|
|
2010-10-08 19:30:11 +02:00
|
|
|
if (unlikely(c > 0xF4))
|
|
|
|
return -1;
|
2006-12-15 15:16:08 +01:00
|
|
|
|
2010-10-08 19:30:11 +02:00
|
|
|
int charlen = clz8 (c ^ 0xFF);
|
|
|
|
switch (charlen)
|
|
|
|
{
|
|
|
|
case 0: // 7-bit ASCII character -> OK
|
|
|
|
*pwc = c;
|
|
|
|
return 1;
|
2006-12-15 15:16:08 +01:00
|
|
|
|
2010-10-08 19:30:11 +02:00
|
|
|
case 1: // continuation byte -> error
|
|
|
|
return -1;
|
|
|
|
}
|
2006-12-15 15:16:08 +01:00
|
|
|
|
2010-10-08 19:30:11 +02:00
|
|
|
assert (charlen >= 2 && charlen <= 4);
|
2006-12-15 15:16:08 +01:00
|
|
|
|
2010-10-08 19:30:11 +02:00
|
|
|
uint32_t cp = c & ~((0xff >> (7 - charlen)) << (7 - charlen));
|
|
|
|
for (int i = 1; i < charlen; i++)
|
|
|
|
{
|
|
|
|
assert (cp < (1 << 26));
|
|
|
|
c = ptr[i];
|
2006-12-15 15:16:08 +01:00
|
|
|
|
2010-10-08 19:30:11 +02:00
|
|
|
if (unlikely((c >> 6) != 2)) // not a continuation byte
|
|
|
|
return -1;
|
2006-03-12 23:52:25 +01:00
|
|
|
|
2010-10-08 19:30:11 +02:00
|
|
|
cp = (cp << 6) | (ptr[i] & 0x3f);
|
2005-08-04 18:38:57 +02:00
|
|
|
}
|
|
|
|
|
2010-10-08 19:30:11 +02:00
|
|
|
switch (charlen)
|
|
|
|
{
|
|
|
|
case 4:
|
|
|
|
if (unlikely(cp > 0x10FFFF)) // beyond Unicode
|
|
|
|
return -1;
|
|
|
|
case 3:
|
|
|
|
if (unlikely(cp >= 0xD800 && cp < 0xC000)) // UTF-16 surrogate
|
|
|
|
return -1;
|
|
|
|
case 2:
|
|
|
|
if (unlikely(cp < 128)) // ASCII overlong
|
|
|
|
return -1;
|
|
|
|
if (unlikely(cp < (1u << (5 * charlen - 3)))) // overlong
|
|
|
|
return -1;
|
|
|
|
}
|
|
|
|
*pwc = cp;
|
|
|
|
return charlen;
|
2005-08-04 18:38:57 +02:00
|
|
|
}
|
2005-10-02 18:52:52 +02:00
|
|
|
|
2010-10-08 20:02:23 +02:00
|
|
|
/**
|
|
|
|
* Look for an UTF-8 string within another one in a case-insensitive fashion.
|
|
|
|
* Beware that this is quite slow. Contrary to strcasestr(), this function
|
|
|
|
* works regardless of the system character encoding, and handles multibyte
|
|
|
|
* code points correctly.
|
|
|
|
|
|
|
|
* @param haystack string to look into
|
|
|
|
* @param needle string to look for
|
|
|
|
* @return a pointer to the first occurence of the needle within the haystack,
|
|
|
|
* or NULL if no occurence were found.
|
|
|
|
*/
|
|
|
|
char *vlc_strcasestr (const char *haystack, const char *needle)
|
|
|
|
{
|
|
|
|
ssize_t s;
|
|
|
|
|
|
|
|
do
|
|
|
|
{
|
|
|
|
const char *h = haystack, *n = needle;
|
|
|
|
|
|
|
|
for (;;)
|
|
|
|
{
|
|
|
|
uint32_t cph, cpn;
|
|
|
|
|
|
|
|
s = vlc_towc (n, &cpn);
|
|
|
|
if (s == 0)
|
|
|
|
return (char *)haystack;
|
|
|
|
if (unlikely(s < 0))
|
|
|
|
return NULL;
|
|
|
|
n += s;
|
|
|
|
|
|
|
|
s = vlc_towc (h, &cph);
|
|
|
|
if (s <= 0 || towlower (cph) != towlower (cpn))
|
|
|
|
break;
|
|
|
|
h += s;
|
|
|
|
}
|
|
|
|
|
|
|
|
s = vlc_towc (haystack, &(uint32_t) { 0 });
|
|
|
|
haystack += s;
|
|
|
|
}
|
|
|
|
while (s != 0);
|
|
|
|
|
|
|
|
return NULL;
|
|
|
|
}
|
2010-10-08 19:30:11 +02:00
|
|
|
|
2006-03-22 13:26:11 +01:00
|
|
|
/**
|
2008-05-03 23:02:44 +02:00
|
|
|
* Replaces invalid/overlong UTF-8 sequences with question marks.
|
2006-03-22 13:26:11 +01:00
|
|
|
* Note that it is not possible to convert from Latin-1 to UTF-8 on the fly,
|
|
|
|
* so we don't try that, even though it would be less disruptive.
|
|
|
|
*
|
|
|
|
* @return str if it was valid UTF-8, NULL if not.
|
|
|
|
*/
|
|
|
|
char *EnsureUTF8( char *str )
|
|
|
|
{
|
2010-10-08 19:30:11 +02:00
|
|
|
char *ret = str;
|
|
|
|
size_t n;
|
|
|
|
uint32_t cp;
|
|
|
|
|
|
|
|
while ((n = vlc_towc (str, &cp)) != 0)
|
|
|
|
if (likely(n != (size_t)-1))
|
|
|
|
str += n;
|
|
|
|
else
|
|
|
|
{
|
|
|
|
*str++ = '?';
|
|
|
|
ret = NULL;
|
|
|
|
}
|
|
|
|
return ret;
|
2006-03-22 13:26:11 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
/**
|
2008-05-03 23:02:44 +02:00
|
|
|
* Checks whether a string is a valid UTF-8 byte sequence.
|
2006-03-22 13:26:11 +01:00
|
|
|
*
|
|
|
|
* @param str nul-terminated string to be checked
|
|
|
|
*
|
|
|
|
* @return str if it was valid UTF-8, NULL if not.
|
|
|
|
*/
|
|
|
|
const char *IsUTF8( const char *str )
|
|
|
|
{
|
2010-10-08 19:30:11 +02:00
|
|
|
size_t n;
|
|
|
|
uint32_t cp;
|
|
|
|
|
|
|
|
while ((n = vlc_towc (str, &cp)) != 0)
|
|
|
|
if (likely(n != (size_t)-1))
|
|
|
|
str += n;
|
|
|
|
else
|
|
|
|
return NULL;
|
|
|
|
return str;
|
2006-03-22 13:26:11 +01:00
|
|
|
}
|
2010-01-25 23:05:53 +01:00
|
|
|
|
|
|
|
/**
|
|
|
|
* Converts a string from the given character encoding to utf-8.
|
|
|
|
*
|
|
|
|
* @return a nul-terminated utf-8 string, or null in case of error.
|
|
|
|
* The result must be freed using free().
|
|
|
|
*/
|
2010-01-25 23:15:06 +01:00
|
|
|
char *FromCharset(const char *charset, const void *data, size_t data_size)
|
2010-01-25 23:05:53 +01:00
|
|
|
{
|
|
|
|
vlc_iconv_t handle = vlc_iconv_open ("UTF-8", charset);
|
|
|
|
if (handle == (vlc_iconv_t)(-1))
|
|
|
|
return NULL;
|
|
|
|
|
|
|
|
char *out = NULL;
|
|
|
|
for(unsigned mul = 4; mul < 8; mul++ )
|
|
|
|
{
|
|
|
|
size_t in_size = data_size;
|
|
|
|
const char *in = data;
|
|
|
|
size_t out_max = mul * data_size;
|
|
|
|
char *tmp = out = malloc (1 + out_max);
|
|
|
|
if (!out)
|
|
|
|
break;
|
|
|
|
|
|
|
|
if (vlc_iconv (handle, &in, &in_size, &tmp, &out_max) != (size_t)(-1)) {
|
|
|
|
*tmp = '\0';
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
free(out);
|
|
|
|
out = NULL;
|
|
|
|
|
|
|
|
if (errno != E2BIG)
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
vlc_iconv_close(handle);
|
|
|
|
return out;
|
|
|
|
}
|
|
|
|
|
2010-10-26 19:48:46 +02:00
|
|
|
/**
|
|
|
|
* Converts a nul-terminated UTF-8 string to a given character encoding.
|
|
|
|
* @param charset iconv name of the character set
|
|
|
|
* @param in nul-terminated UTF-8 string
|
|
|
|
* @param outsize pointer to hold the byte size of result
|
|
|
|
*
|
|
|
|
* @return A pointer to the result, which must be released using free().
|
|
|
|
* The UTF-8 nul terminator is included in the conversion if the target
|
|
|
|
* character encoding supports it. However it is not included in the returned
|
|
|
|
* byte size.
|
|
|
|
* In case of error, NULL is returned and the byte size is undefined.
|
|
|
|
*/
|
|
|
|
void *ToCharset(const char *charset, const char *in, size_t *outsize)
|
|
|
|
{
|
|
|
|
vlc_iconv_t hd = vlc_iconv_open (charset, "UTF-8");
|
|
|
|
if (hd == (vlc_iconv_t)(-1))
|
|
|
|
return NULL;
|
|
|
|
|
|
|
|
const size_t inlen = strlen (in);
|
|
|
|
void *res;
|
|
|
|
|
|
|
|
for (unsigned mul = 4; mul < 16; mul++)
|
|
|
|
{
|
|
|
|
size_t outlen = mul * (inlen + 1);
|
|
|
|
res = malloc (outlen);
|
|
|
|
if (unlikely(res == NULL))
|
|
|
|
break;
|
|
|
|
|
|
|
|
const char *inp = in;
|
|
|
|
char *outp = res;
|
2010-11-03 23:25:07 +01:00
|
|
|
size_t inb = inlen;
|
2010-11-03 23:40:41 +01:00
|
|
|
size_t outb = outlen - mul;
|
2010-10-26 19:48:46 +02:00
|
|
|
|
|
|
|
if (vlc_iconv (hd, &inp, &inb, &outp, &outb) != (size_t)(-1))
|
|
|
|
{
|
2010-11-03 23:49:43 +01:00
|
|
|
*outsize = outlen - mul - outb;
|
2010-11-03 23:40:41 +01:00
|
|
|
outb += mul;
|
2010-10-26 19:48:46 +02:00
|
|
|
inb = 1; /* append nul terminator if possible */
|
|
|
|
if (vlc_iconv (hd, &inp, &inb, &outp, &outb) != (size_t)(-1))
|
|
|
|
break;
|
|
|
|
if (errno == EILSEQ) /* cannot translate nul terminator!? */
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
free (res);
|
2010-11-03 23:25:07 +01:00
|
|
|
res = NULL;
|
2010-10-26 19:48:46 +02:00
|
|
|
if (errno != E2BIG) /* conversion failure */
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
vlc_iconv_close (hd);
|
|
|
|
return res;
|
|
|
|
}
|
|
|
|
|