1
/* Determine a canonical name for the current locale's character encoding.
3
Copyright (C) 2000-2003 Free Software Foundation, Inc.
5
This program is free software; you can redistribute it and/or modify it
6
under the terms of the GNU Library General Public License as published
7
by the Free Software Foundation; either version 2, or (at your option)
10
This program is distributed in the hope that it will be useful,
11
but WITHOUT ANY WARRANTY; without even the implied warranty of
12
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
13
Library General Public License for more details.
15
You should have received a copy of the GNU Library General Public
16
License along with this program; if not, write to the Free Software
17
Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307,
20
/* Written by Bruno Haible <bruno@clisp.org>. */
27
#include "localcharset.h"
43
#if defined _WIN32 || defined __WIN32__
44
# undef WIN32 /* avoid warning on mingw32 */
49
/* Assume EMX program runs on OS/2, even if compiled under DOS. */
54
# if HAVE_LANGINFO_CODESET
55
# include <langinfo.h>
62
# define WIN32_LEAN_AND_MEAN
70
#if ENABLE_RELOCATABLE
71
# include "relocatable.h"
73
# define relocate(pathname) (pathname)
76
#if defined _WIN32 || defined __WIN32__ || defined __EMX__ || defined __DJGPP__
77
/* Win32, OS/2, DOS */
78
# define ISSLASH(C) ((C) == '/' || (C) == '\\')
81
#ifndef DIRECTORY_SEPARATOR
82
# define DIRECTORY_SEPARATOR '/'
86
# define ISSLASH(C) ((C) == DIRECTORY_SEPARATOR)
89
#if HAVE_DECL_GETC_UNLOCKED
91
# define getc getc_unlocked
94
/* The following static variable is declared 'volatile' to avoid a
95
possible multithread problem in the function get_charset_aliases. If we
96
are running in a threaded environment, and if two threads initialize
97
'charset_aliases' simultaneously, both will produce the same value,
98
and everything will be ok if the two assignments to 'charset_aliases'
99
are atomic. But I don't know what will happen if the two assignments mix. */
101
# define volatile /* empty */
103
/* Pointer to the contents of the charset.alias file, if it has already been
104
read, else NULL. Its format is:
105
ALIAS_1 '\0' CANONICAL_1 '\0' ... ALIAS_n '\0' CANONICAL_n '\0' '\0' */
106
static const char * volatile charset_aliases;
108
/* Return a pointer to the contents of the charset.alias file. */
110
get_charset_aliases ()
114
cp = charset_aliases;
117
#if !(defined VMS || defined WIN32)
119
const char *dir = relocate (LIBDIR);
120
const char *base = "charset.alias";
123
/* Concatenate dir and base into freshly allocated file_name. */
125
size_t dir_len = strlen (dir);
126
size_t base_len = strlen (base);
127
int add_slash = (dir_len > 0 && !ISSLASH (dir[dir_len - 1]));
128
file_name = (char *) malloc (dir_len + add_slash + base_len + 1);
129
if (file_name != NULL)
131
memcpy (file_name, dir, dir_len);
133
file_name[dir_len] = DIRECTORY_SEPARATOR;
134
memcpy (file_name + dir_len + add_slash, base, base_len + 1);
138
if (file_name == NULL || (fp = fopen (file_name, "r")) == NULL)
139
/* Out of memory or file not found, treat it as empty. */
143
/* Parse the file's contents. */
147
char *res_ptr = NULL;
156
if (c == '\n' || c == ' ' || c == '\t')
160
/* Skip comment, to end of line. */
163
while (!(c == EOF || c == '\n'));
169
if (fscanf (fp, "%50s %50s", buf1, buf2) < 2)
175
res_size = l1 + 1 + l2 + 1;
176
res_ptr = (char *) malloc (res_size + 1);
180
res_size += l1 + 1 + l2 + 1;
181
res_ptr = (char *) realloc (res_ptr, res_size + 1);
189
strcpy (res_ptr + res_size - (l2 + 1) - (l1 + 1), buf1);
190
strcpy (res_ptr + res_size - (l2 + 1), buf2);
197
*(res_ptr + res_size) = '\0';
202
if (file_name != NULL)
208
/* To avoid the troubles of an extra file charset.alias_vms in the
209
sources of many GNU packages, simply inline the aliases here. */
210
/* The list of encodings is taken from the OpenVMS 7.3-1 documentation
211
"Compaq C Run-Time Library Reference Manual for OpenVMS systems"
212
section 10.7 "Handling Different Character Sets". */
213
cp = "ISO8859-1" "\0" "ISO-8859-1" "\0"
214
"ISO8859-2" "\0" "ISO-8859-2" "\0"
215
"ISO8859-5" "\0" "ISO-8859-5" "\0"
216
"ISO8859-7" "\0" "ISO-8859-7" "\0"
217
"ISO8859-8" "\0" "ISO-8859-8" "\0"
218
"ISO8859-9" "\0" "ISO-8859-9" "\0"
220
"eucJP" "\0" "EUC-JP" "\0"
221
"SJIS" "\0" "SHIFT_JIS" "\0"
222
"DECKANJI" "\0" "DEC-KANJI" "\0"
223
"SDECKANJI" "\0" "EUC-JP" "\0"
225
"eucTW" "\0" "EUC-TW" "\0"
226
"DECHANYU" "\0" "DEC-HANYU" "\0"
227
"DECHANZI" "\0" "GB2312" "\0"
229
"DECKOREAN" "\0" "EUC-KR" "\0";
233
/* To avoid the troubles of installing a separate file in the same
234
directory as the DLL and of retrieving the DLL's directory at
235
runtime, simply inline the aliases here. */
237
cp = "CP936" "\0" "GBK" "\0"
238
"CP1361" "\0" "JOHAB" "\0"
239
"CP20127" "\0" "ASCII" "\0"
240
"CP20866" "\0" "KOI8-R" "\0"
241
"CP21866" "\0" "KOI8-RU" "\0"
242
"CP28591" "\0" "ISO-8859-1" "\0"
243
"CP28592" "\0" "ISO-8859-2" "\0"
244
"CP28593" "\0" "ISO-8859-3" "\0"
245
"CP28594" "\0" "ISO-8859-4" "\0"
246
"CP28595" "\0" "ISO-8859-5" "\0"
247
"CP28596" "\0" "ISO-8859-6" "\0"
248
"CP28597" "\0" "ISO-8859-7" "\0"
249
"CP28598" "\0" "ISO-8859-8" "\0"
250
"CP28599" "\0" "ISO-8859-9" "\0"
251
"CP28605" "\0" "ISO-8859-15" "\0";
255
charset_aliases = cp;
261
/* Determine the current locale's character encoding, and canonicalize it
262
into one of the canonical names listed in config.charset.
263
The result must not be freed; it is statically allocated.
264
If the canonical name cannot be determined, the result is a non-canonical
276
#if !(defined WIN32 || defined OS2)
278
# if HAVE_LANGINFO_CODESET
280
/* Most systems support nl_langinfo (CODESET) nowadays. */
281
codeset = nl_langinfo (CODESET);
285
/* On old systems which lack it, use setlocale or getenv. */
286
const char *locale = NULL;
288
/* But most old systems don't have a complete set of locales. Some
289
(like SunOS 4 or DJGPP) have only the C locale. Therefore we don't
290
use setlocale here; it would return "C" when it doesn't support the
291
locale name the user has set. */
292
# if HAVE_SETLOCALE && 0
293
locale = setlocale (LC_CTYPE, NULL);
295
if (locale == NULL || locale[0] == '\0')
297
locale = getenv ("LC_ALL");
298
if (locale == NULL || locale[0] == '\0')
300
locale = getenv ("LC_CTYPE");
301
if (locale == NULL || locale[0] == '\0')
302
locale = getenv ("LANG");
306
/* On some old systems, one used to set locale = "iso8859_1". On others,
307
you set it to "language_COUNTRY.charset". In any case, we resolve it
308
through the charset.alias file. */
315
static char buf[2 + 10 + 1];
317
/* Woe32 has a function returning the locale's codepage as a number. */
318
sprintf (buf, "CP%u", GetACP ());
324
static char buf[2 + 10 + 1];
328
/* Allow user to override the codeset, as set in the operating system,
329
with standard language environment variables. */
330
locale = getenv ("LC_ALL");
331
if (locale == NULL || locale[0] == '\0')
333
locale = getenv ("LC_CTYPE");
334
if (locale == NULL || locale[0] == '\0')
335
locale = getenv ("LANG");
337
if (locale != NULL && locale[0] != '\0')
339
/* If the locale name contains an encoding after the dot, return it. */
340
const char *dot = strchr (locale, '.');
344
const char *modifier;
347
/* Look for the possible @... trailer and remove it, if any. */
348
modifier = strchr (dot, '@');
349
if (modifier == NULL)
351
if (modifier - dot < sizeof (buf))
353
memcpy (buf, dot, modifier - dot);
354
buf [modifier - dot] = '\0';
359
/* Resolve through the charset.alias file. */
364
/* OS/2 has a function returning the locale's codepage as a number. */
365
if (DosQueryCp (sizeof (cp), cp, &cplen))
369
sprintf (buf, "CP%u", cp[0]);
377
/* The canonical name cannot be determined. */
381
for (aliases = get_charset_aliases ();
383
aliases += strlen (aliases) + 1, aliases += strlen (aliases) + 1)
384
if (strcmp (codeset, aliases) == 0
385
|| (aliases[0] == '*' && aliases[1] == '\0'))
387
codeset = aliases + strlen (aliases) + 1;
391
/* Don't return an empty string. GNU libc and GNU libiconv interpret
392
the empty string as denoting "the locale's character encoding",
393
thus GNU libiconv would call this function a second time. */
394
if (codeset[0] == '\0')