1 /*****************************************************************************
2 * charset.c: Locale's character encoding stuff.
3 *****************************************************************************
4 * See also unicode.c for Unicode to locale conversion helpers.
6 * Copyright (C) 2003-2006 the VideoLAN team
9 * Authors: Derk-Jan Hartman <thedj at users.sf.net>
11 * RĂ©mi Denis-Courmont
13 * vlc_current_charset() an adaption of mp_locale_charset():
15 * Copyright (C) 2001-2003 The Mape Project
16 * Written by Karel Zak <zakkr@zf.jcu.cz>.
18 * This program is free software; you can redistribute it and/or modify
19 * it under the terms of the GNU General Public License as published by
20 * the Free Software Foundation; either version 2 of the License, or
21 * (at your option) any later version.
23 * This program is distributed in the hope that it will be useful,
24 * but WITHOUT ANY WARRANTY; without even the implied warranty of
25 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
26 * GNU General Public License for more details.
28 * You should have received a copy of the GNU General Public License
29 * along with this program; if not, write to the Free Software
30 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston MA 02110-1301, USA.
31 *****************************************************************************/
38 # if HAVE_LANGINFO_CODESET
39 # include <langinfo.h>
56 typedef struct VLCCharsetAlias
58 char *psz_alias, *psz_name;
62 * The libcharset load all from external text file, but it's strange and
63 * slow solution, we rather use array(s) compiled into source. In the
64 * "good" libc this is not needful -- for example in linux.
66 * Please, put to this funtion exotic aliases only. The libc 'iconv' knows
67 * a lot of basic aliases (check it first by iconv -l).
70 #if (defined OS2 || !HAVE_LANGINFO_CODESET) && !defined WIN32
71 static const char* vlc_encoding_from_language( const char *l )
73 /* check for language (and perhaps country) codes */
74 if (strstr(l, "zh_TW")) return "Big5";
75 if (strstr(l, "zh_HK")) return "Big5HKSCS"; /* no MIME charset */
76 if (strstr(l, "zh")) return "GB2312";
77 if (strstr(l, "th")) return "TIS-620";
78 if (strstr(l, "ja")) return "EUC-JP";
79 if (strstr(l, "ko")) return "EUC-KR";
80 if (strstr(l, "ru")) return "KOI8-R";
81 if (strstr(l, "uk")) return "KOI8-U";
82 if (strstr(l, "pl") || strstr(l, "hr") ||
83 strstr(l, "hu") || strstr(l, "cs") ||
84 strstr(l, "sk") || strstr(l, "sl")) return "ISO-8859-2";
85 if (strstr(l, "eo") || strstr(l, "mt")) return "ISO-8859-3";
86 if (strstr(l, "lt") || strstr(l, "la")) return "ISO-8859-4";
87 if (strstr(l, "bg") || strstr(l, "be") ||
88 strstr(l, "mk") || strstr(l, "uk")) return "ISO-8859-5";
89 if (strstr(l, "ar")) return "ISO-8859-6";
90 if (strstr(l, "el")) return "ISO-8859-7";
91 if (strstr(l, "he") || strstr(l, "iw")) return "ISO-8859-8";
92 if (strstr(l, "tr")) return "ISO-8859-9";
93 if (strstr(l, "th")) return "ISO-8859-11";
94 if (strstr(l, "lv")) return "ISO-8859-13";
95 if (strstr(l, "cy")) return "ISO-8859-14";
96 if (strstr(l, "et")) return "ISO-8859-15"; /* all latin1 could be iso15 as well */
97 if (strstr(l, "ro")) return "ISO-8859-2"; /* or ISO-8859-16 */
98 if (strstr(l, "am") || strstr(l, "vi")) return "UTF-8";
99 /* We don't know. This ain't working go to default. */
104 static const char* vlc_charset_aliases( const char *psz_name )
109 VLCCharsetAlias aliases[] =
112 { "CP1361", "JOHAB" },
113 { "CP20127", "ASCII" },
114 { "CP20866", "KOI8-R" },
115 { "CP21866", "KOI8-RU" },
116 { "CP28591", "ISO-8859-1" },
117 { "CP28592", "ISO-8859-2" },
118 { "CP28593", "ISO-8859-3" },
119 { "CP28594", "ISO-8859-4" },
120 { "CP28595", "ISO-8859-5" },
121 { "CP28596", "ISO-8859-6" },
122 { "CP28597", "ISO-8859-7" },
123 { "CP28598", "ISO-8859-8" },
124 { "CP28599", "ISO-8859-9" },
125 { "CP28605", "ISO-8859-15" },
129 VLCCharsetAlias aliases[] =
131 { "IBM-850", "CP850" },
132 { "IBM-856", "CP856" },
133 { "IBM-921", "ISO-8859-13" },
134 { "IBM-922", "CP922" },
135 { "IBM-932", "CP932" },
136 { "IBM-943", "CP943" },
137 { "IBM-1046", "CP1046" },
138 { "IBM-1124", "CP1124" },
139 { "IBM-1129", "CP1129" },
140 { "IBM-1252", "CP1252" },
141 { "IBM-EUCCN", "GB2312" },
142 { "IBM-EUCJP", "EUC-JP" },
143 { "IBM-EUCKR", "EUC-KR" },
144 { "IBM-EUCTW", "EUC-TW" },
148 VLCCharsetAlias aliases[] =
150 { "ROMAN8", "HP-ROMAN8" },
151 { "ARABIC8", "HP-ARABIC8" },
152 { "GREEK8", "HP-GREEK8" },
153 { "HEBREW8", "HP-HEBREW8" },
154 { "TURKISH8", "HP-TURKISH8" },
155 { "KANA8", "HP-KANA8" },
156 { "HP15CN", "GB2312" },
160 VLCCharsetAlias aliases[] =
162 { "EUCCN", "GB2312" },
166 VLCCharsetAlias aliases[] =
168 { "KSC5601", "CP949" },
169 { "SDECKANJI", "EUC-JP" },
170 { "TACTIS", "TIS-620" },
174 VLCCharsetAlias aliases[] =
177 { "CNS11643", "EUC-TW" },
178 { "5601", "EUC-KR" },
179 { "JOHAP92", "JOHAB" },
180 { "PCK", "SHIFT_JIS" },
181 { "2533", "TIS-620" },
185 VLCCharsetAlias aliases[] =
188 { "EUCCN", "GB2312" },
192 VLCCharsetAlias aliases[] = {{NULL, NULL}};
195 for (a = aliases; a->psz_alias; a++)
196 if (strcasecmp (a->psz_alias, psz_name) == 0)
199 /* we return original name beacuse iconv() probably will know
200 * something better about name if we don't know it :-) */
204 /* Returns charset from "language_COUNTRY.charset@modifier" string */
205 #if (defined OS2 || !HAVE_LANGINFO_CODESET) && !defined WIN32
206 static void vlc_encoding_from_locale( char *psz_locale, char *psz_charset )
208 char *psz_dot = strchr( psz_locale, '.' );
210 if( psz_dot != NULL )
212 const char *psz_modifier;
216 /* Look for the possible @... trailer and remove it, if any. */
217 psz_modifier = strchr( psz_dot, '@' );
219 if( psz_modifier == NULL )
221 strcpy( psz_charset, psz_dot );
224 if( 0 < ( psz_modifier - psz_dot )
225 && ( psz_modifier - psz_dot ) < 2 + 10 + 1 )
227 memcpy( psz_charset, psz_dot, psz_modifier - psz_dot );
228 psz_charset[ psz_modifier - psz_dot ] = '\0';
232 /* try language mapping */
233 strcpy( psz_charset, vlc_encoding_from_language( psz_locale ) );
237 vlc_bool_t vlc_current_charset( char **psz_charset )
239 const char *psz_codeset;
241 #if !(defined WIN32 || defined OS2 || defined __APPLE__)
243 # if HAVE_LANGINFO_CODESET
244 /* Most systems support nl_langinfo( CODESET ) nowadays. */
245 psz_codeset = nl_langinfo( CODESET );
246 if( !strcmp( psz_codeset, "ANSI_X3.4-1968" ) )
247 psz_codeset = "ASCII";
249 /* On old systems which lack it, use setlocale or getenv. */
250 const char *psz_locale = NULL;
251 char buf[2 + 10 + 1];
253 /* But most old systems don't have a complete set of locales. Some
254 * (like SunOS 4 or DJGPP) have only the C locale. Therefore we don't
255 * use setlocale here; it would return "C" when it doesn't support the
256 * locale name the user has set. Darwin's setlocale is broken. */
257 # if HAVE_SETLOCALE && !__APPLE__
258 psz_locale = setlocale( LC_ALL, NULL );
260 if( psz_locale == NULL || psz_locale[0] == '\0' )
262 psz_locale = getenv( "LC_ALL" );
263 if( psz_locale == NULL || psz_locale[0] == '\0' )
265 psz_locale = getenv( "LC_CTYPE" );
266 if( psz_locale == NULL || psz_locale[0] == '\0')
267 psz_locale = getenv( "LANG" );
271 /* On some old systems, one used to set locale = "iso8859_1". On others,
272 * you set it to "language_COUNTRY.charset". Darwin only has LANG :( */
273 vlc_encoding_from_locale( (char *)psz_locale, buf );
275 # endif /* HAVE_LANGINFO_CODESET */
277 #elif defined __APPLE__
279 /* Darwin is always using UTF-8 internally. */
280 psz_codeset = "UTF-8";
284 char buf[2 + 10 + 1];
286 /* Woe32 has a function returning the locale's codepage as a number. */
287 sprintf( buf, "CP%u", GetACP() );
292 const char *psz_locale;
293 char buf[2 + 10 + 1];
297 /* Allow user to override the codeset, as set in the operating system,
298 * with standard language environment variables. */
299 psz_locale = getenv( "LC_ALL" );
300 if( psz_locale == NULL || psz_locale[0] == '\0' )
302 psz+locale = getenv( "LC_CTYPE" );
303 if( psz_locale == NULL || locale[0] == '\0' )
304 locale = getenv( "LANG" );
306 if( psz_locale != NULL && psz_locale[0] != '\0' )
307 vlc_encoding_from_locale( psz_locale, buf );
311 /* OS/2 has a function returning the locale's codepage as a number. */
312 if( DosQueryCp( sizeof( cp ), cp, &cplen ) )
316 sprintf( buf, "CP%u", cp[0] );
321 if( psz_codeset == NULL )
322 /* The canonical name cannot be determined. */
325 psz_codeset = vlc_charset_aliases( psz_codeset );
327 /* Don't return an empty string. GNU libc and GNU libiconv interpret
328 * the empty string as denoting "the locale's character encoding",
329 * thus GNU libiconv would call this function a second time. */
330 if( psz_codeset[0] == '\0' )
332 /* Last possibility is 'CHARSET' enviroment variable */
333 if( !( psz_codeset = getenv( "CHARSET" ) ) )
334 psz_codeset = "ISO-8859-1";
338 *psz_charset = strdup(psz_codeset);
340 if( !strcasecmp(psz_codeset, "UTF8") || !strcasecmp(psz_codeset, "UTF-8") )
346 char *__vlc_fix_readdir_charset( vlc_object_t *p_this, const char *psz_string )
349 if ( p_this->p_libvlc->iconv_macosx != (vlc_iconv_t)-1 )
351 const char *psz_in = psz_string;
352 size_t i_in = strlen(psz_in);
353 size_t i_out = i_in * 2;
354 char *psz_utf8 = malloc(i_out + 1);
355 char *psz_out = psz_utf8;
357 vlc_mutex_lock( &p_this->p_libvlc->iconv_lock );
358 size_t i_ret = vlc_iconv( p_this->p_libvlc->iconv_macosx,
359 &psz_in, &i_in, &psz_out, &i_out );
360 vlc_mutex_unlock( &p_this->p_libvlc->iconv_lock );
361 if( i_ret == (size_t)-1 || i_in )
364 "failed to convert \"%s\" from HFS+ charset (%s)",
365 psz_string, strerror(errno) );
367 return strdup( psz_string );
375 return strdup( psz_string );
379 * There are two decimal separators in the computer world-wide locales:
380 * dot (which is the american default), and comma (which is used in France,
381 * the country with the most VLC developers, among others).
383 * i18n_strtod() has the same prototype as ANSI C strtod() but it accepts
384 * either decimal separator when deserializing the string to a float number,
385 * independant of the local computer setting.
387 double i18n_strtod( const char *str, char **end )
394 d = strtod( str, end );
397 if(( e == ',' ) || ( e == '.' ))
399 char dup[strlen( str ) + 1];
405 dup[*end - str] = ( e == ',' ) ? '.' : ',';
406 d = strtod( dup, end );
412 * i18n_atof() has the same prototype as ANSI C atof() but it accepts
413 * either decimal separator when deserializing the string to a float number,
414 * independant of the local computer setting.
416 double i18n_atof( const char *str )
418 return i18n_strtod( str, NULL );
423 * us_strtod() has the same prototype as ANSI C strtod() but it expects
424 * a dot as decimal separator regardless of the system locale.
426 double us_strtod( const char *str, char **end )
428 char dup[strlen( str ) + 1], *ptr;
432 ptr = strchr( dup, ',' );
436 d = strtod( dup, &ptr );
438 *end = (char *)&str[ptr - dup];
444 * us_atof() has the same prototype as ANSI C atof() but it expects a dot
445 * as decimal separator, regardless of the system locale.
447 double us_atof( const char *str )
449 return us_strtod( str, NULL );