1 /* -*- buffer-read-only: t -*- vi: set ro: */
2 /* DO NOT EDIT! GENERATED AUTOMATICALLY! */
3 /* Determine a canonical name for the current locale's character encoding.
5 Copyright (C) 2000-2006, 2008 Free Software Foundation, Inc.
7 This program is free software; you can redistribute it and/or modify
8 it under the terms of the GNU General Public License as published by
9 the Free Software Foundation; either version 3, or (at your option)
12 This program is distributed in the hope that it will be useful,
13 but WITHOUT ANY WARRANTY; without even the implied warranty of
14 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
15 GNU General Public License for more details.
17 You should have received a copy of the GNU General Public License along
18 with this program; if not, write to the Free Software Foundation,
19 Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */
21 /* Written by Bruno Haible <bruno@clisp.org>. */
26 #include "localcharset.h"
33 #if defined _WIN32 || defined __WIN32__
38 /* Assume EMX program runs on OS/2, even if compiled under DOS. */
44 #if !defined WIN32_NATIVE
45 # if HAVE_LANGINFO_CODESET
46 # include <langinfo.h>
48 # if 0 /* see comment below */
53 # define WIN32_LEAN_AND_MEAN
56 #elif defined WIN32_NATIVE
57 # define WIN32_LEAN_AND_MEAN
65 #if ENABLE_RELOCATABLE
66 # include "relocatable.h"
68 # define relocate(pathname) (pathname)
73 # include "configmake.h"
76 #if defined _WIN32 || defined __WIN32__ || defined __CYGWIN__ || defined __EMX__ || defined __DJGPP__
77 /* Win32, Cygwin, 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 (void)
114 cp = charset_aliases;
117 #if !(defined VMS || defined WIN32_NATIVE || defined __CYGWIN__)
120 const char *base = "charset.alias";
123 /* Make it possible to override the charset.alias location. This is
124 necessary for running the testsuite before "make install". */
125 dir = getenv ("CHARSETALIASDIR");
126 if (dir == NULL || dir[0] == '\0')
127 dir = relocate (LIBDIR);
129 /* Concatenate dir and base into freshly allocated file_name. */
131 size_t dir_len = strlen (dir);
132 size_t base_len = strlen (base);
133 int add_slash = (dir_len > 0 && !ISSLASH (dir[dir_len - 1]));
134 file_name = (char *) malloc (dir_len + add_slash + base_len + 1);
135 if (file_name != NULL)
137 memcpy (file_name, dir, dir_len);
139 file_name[dir_len] = DIRECTORY_SEPARATOR;
140 memcpy (file_name + dir_len + add_slash, base, base_len + 1);
144 if (file_name == NULL || (fp = fopen (file_name, "r")) == NULL)
145 /* Out of memory or file not found, treat it as empty. */
149 /* Parse the file's contents. */
150 char *res_ptr = NULL;
164 if (c == '\n' || c == ' ' || c == '\t')
168 /* Skip comment, to end of line. */
171 while (!(c == EOF || c == '\n'));
177 if (fscanf (fp, "%50s %50s", buf1, buf2) < 2)
181 old_res_ptr = res_ptr;
184 res_size = l1 + 1 + l2 + 1;
185 res_ptr = (char *) malloc (res_size + 1);
189 res_size += l1 + 1 + l2 + 1;
190 res_ptr = (char *) realloc (res_ptr, res_size + 1);
196 if (old_res_ptr != NULL)
200 strcpy (res_ptr + res_size - (l2 + 1) - (l1 + 1), buf1);
201 strcpy (res_ptr + res_size - (l2 + 1), buf2);
208 *(res_ptr + res_size) = '\0';
213 if (file_name != NULL)
219 /* To avoid the troubles of an extra file charset.alias_vms in the
220 sources of many GNU packages, simply inline the aliases here. */
221 /* The list of encodings is taken from the OpenVMS 7.3-1 documentation
222 "Compaq C Run-Time Library Reference Manual for OpenVMS systems"
223 section 10.7 "Handling Different Character Sets". */
224 cp = "ISO8859-1" "\0" "ISO-8859-1" "\0"
225 "ISO8859-2" "\0" "ISO-8859-2" "\0"
226 "ISO8859-5" "\0" "ISO-8859-5" "\0"
227 "ISO8859-7" "\0" "ISO-8859-7" "\0"
228 "ISO8859-8" "\0" "ISO-8859-8" "\0"
229 "ISO8859-9" "\0" "ISO-8859-9" "\0"
231 "eucJP" "\0" "EUC-JP" "\0"
232 "SJIS" "\0" "SHIFT_JIS" "\0"
233 "DECKANJI" "\0" "DEC-KANJI" "\0"
234 "SDECKANJI" "\0" "EUC-JP" "\0"
236 "eucTW" "\0" "EUC-TW" "\0"
237 "DECHANYU" "\0" "DEC-HANYU" "\0"
238 "DECHANZI" "\0" "GB2312" "\0"
240 "DECKOREAN" "\0" "EUC-KR" "\0";
243 # if defined WIN32_NATIVE || defined __CYGWIN__
244 /* To avoid the troubles of installing a separate file in the same
245 directory as the DLL and of retrieving the DLL's directory at
246 runtime, simply inline the aliases here. */
248 cp = "CP936" "\0" "GBK" "\0"
249 "CP1361" "\0" "JOHAB" "\0"
250 "CP20127" "\0" "ASCII" "\0"
251 "CP20866" "\0" "KOI8-R" "\0"
252 "CP20936" "\0" "GB2312" "\0"
253 "CP21866" "\0" "KOI8-RU" "\0"
254 "CP28591" "\0" "ISO-8859-1" "\0"
255 "CP28592" "\0" "ISO-8859-2" "\0"
256 "CP28593" "\0" "ISO-8859-3" "\0"
257 "CP28594" "\0" "ISO-8859-4" "\0"
258 "CP28595" "\0" "ISO-8859-5" "\0"
259 "CP28596" "\0" "ISO-8859-6" "\0"
260 "CP28597" "\0" "ISO-8859-7" "\0"
261 "CP28598" "\0" "ISO-8859-8" "\0"
262 "CP28599" "\0" "ISO-8859-9" "\0"
263 "CP28605" "\0" "ISO-8859-15" "\0"
264 "CP38598" "\0" "ISO-8859-8" "\0"
265 "CP51932" "\0" "EUC-JP" "\0"
266 "CP51936" "\0" "GB2312" "\0"
267 "CP51949" "\0" "EUC-KR" "\0"
268 "CP51950" "\0" "EUC-TW" "\0"
269 "CP54936" "\0" "GB18030" "\0"
270 "CP65001" "\0" "UTF-8" "\0";
274 charset_aliases = cp;
280 /* Determine the current locale's character encoding, and canonicalize it
281 into one of the canonical names listed in config.charset.
282 The result must not be freed; it is statically allocated.
283 If the canonical name cannot be determined, the result is a non-canonical
290 locale_charset (void)
295 #if !(defined WIN32_NATIVE || defined OS2)
297 # if HAVE_LANGINFO_CODESET
299 /* Most systems support nl_langinfo (CODESET) nowadays. */
300 codeset = nl_langinfo (CODESET);
303 /* Cygwin 2006 does not have locales. nl_langinfo (CODESET) always
304 returns "US-ASCII". As long as this is not fixed, return the suffix
305 of the locale name from the environment variables (if present) or
306 the codepage as a number. */
307 if (codeset != NULL && strcmp (codeset, "US-ASCII") == 0)
310 static char buf[2 + 10 + 1];
312 locale = getenv ("LC_ALL");
313 if (locale == NULL || locale[0] == '\0')
315 locale = getenv ("LC_CTYPE");
316 if (locale == NULL || locale[0] == '\0')
317 locale = getenv ("LANG");
319 if (locale != NULL && locale[0] != '\0')
321 /* If the locale name contains an encoding after the dot, return
323 const char *dot = strchr (locale, '.');
327 const char *modifier;
330 /* Look for the possible @... trailer and remove it, if any. */
331 modifier = strchr (dot, '@');
332 if (modifier == NULL)
334 if (modifier - dot < sizeof (buf))
336 memcpy (buf, dot, modifier - dot);
337 buf [modifier - dot] = '\0';
343 /* Woe32 has a function returning the locale's codepage as a number. */
344 sprintf (buf, "CP%u", GetACP ());
351 /* On old systems which lack it, use setlocale or getenv. */
352 const char *locale = NULL;
354 /* But most old systems don't have a complete set of locales. Some
355 (like SunOS 4 or DJGPP) have only the C locale. Therefore we don't
356 use setlocale here; it would return "C" when it doesn't support the
357 locale name the user has set. */
359 locale = setlocale (LC_CTYPE, NULL);
361 if (locale == NULL || locale[0] == '\0')
363 locale = getenv ("LC_ALL");
364 if (locale == NULL || locale[0] == '\0')
366 locale = getenv ("LC_CTYPE");
367 if (locale == NULL || locale[0] == '\0')
368 locale = getenv ("LANG");
372 /* On some old systems, one used to set locale = "iso8859_1". On others,
373 you set it to "language_COUNTRY.charset". In any case, we resolve it
374 through the charset.alias file. */
379 #elif defined WIN32_NATIVE
381 static char buf[2 + 10 + 1];
383 /* Woe32 has a function returning the locale's codepage as a number. */
384 sprintf (buf, "CP%u", GetACP ());
390 static char buf[2 + 10 + 1];
394 /* Allow user to override the codeset, as set in the operating system,
395 with standard language environment variables. */
396 locale = getenv ("LC_ALL");
397 if (locale == NULL || locale[0] == '\0')
399 locale = getenv ("LC_CTYPE");
400 if (locale == NULL || locale[0] == '\0')
401 locale = getenv ("LANG");
403 if (locale != NULL && locale[0] != '\0')
405 /* If the locale name contains an encoding after the dot, return it. */
406 const char *dot = strchr (locale, '.');
410 const char *modifier;
413 /* Look for the possible @... trailer and remove it, if any. */
414 modifier = strchr (dot, '@');
415 if (modifier == NULL)
417 if (modifier - dot < sizeof (buf))
419 memcpy (buf, dot, modifier - dot);
420 buf [modifier - dot] = '\0';
425 /* Resolve through the charset.alias file. */
430 /* OS/2 has a function returning the locale's codepage as a number. */
431 if (DosQueryCp (sizeof (cp), cp, &cplen))
435 sprintf (buf, "CP%u", cp[0]);
443 /* The canonical name cannot be determined. */
447 for (aliases = get_charset_aliases ();
449 aliases += strlen (aliases) + 1, aliases += strlen (aliases) + 1)
450 if (strcmp (codeset, aliases) == 0
451 || (aliases[0] == '*' && aliases[1] == '\0'))
453 codeset = aliases + strlen (aliases) + 1;
457 /* Don't return an empty string. GNU libc and GNU libiconv interpret
458 the empty string as denoting "the locale's character encoding",
459 thus GNU libiconv would call this function a second time. */
460 if (codeset[0] == '\0')