1 /* Determine a canonical name for the current locale's character encoding.
3 Copyright (C) 2000-2004 Free Software Foundation, Inc.
5 This program is free software; you can redistribute it and/or modify
6 it under the terms of the GNU General Public License as published by
7 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
13 GNU General Public License for more details.
15 You should have received a copy of the GNU General Public License along
16 with this program; if not, write to the Free Software Foundation,
17 Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */
19 /* Written by Bruno Haible <bruno@clisp.org>. */
26 #include "localcharset.h"
42 #if defined _WIN32 || defined __WIN32__
43 # undef WIN32 /* avoid warning on mingw32 */
48 /* Assume EMX program runs on OS/2, even if compiled under DOS. */
53 # if HAVE_LANGINFO_CODESET
54 # include <langinfo.h>
61 # define WIN32_LEAN_AND_MEAN
69 #if ENABLE_RELOCATABLE
70 # include "relocatable.h"
72 # define relocate(pathname) (pathname)
75 #if defined _WIN32 || defined __WIN32__ || defined __CYGWIN__ || defined __EMX__ || defined __DJGPP__
76 /* Win32, Cygwin, OS/2, DOS */
77 # define ISSLASH(C) ((C) == '/' || (C) == '\\')
80 #ifndef DIRECTORY_SEPARATOR
81 # define DIRECTORY_SEPARATOR '/'
85 # define ISSLASH(C) ((C) == DIRECTORY_SEPARATOR)
88 #if HAVE_DECL_GETC_UNLOCKED
90 # define getc getc_unlocked
93 /* The following static variable is declared 'volatile' to avoid a
94 possible multithread problem in the function get_charset_aliases. If we
95 are running in a threaded environment, and if two threads initialize
96 'charset_aliases' simultaneously, both will produce the same value,
97 and everything will be ok if the two assignments to 'charset_aliases'
98 are atomic. But I don't know what will happen if the two assignments mix. */
100 # define volatile /* empty */
102 /* Pointer to the contents of the charset.alias file, if it has already been
103 read, else NULL. Its format is:
104 ALIAS_1 '\0' CANONICAL_1 '\0' ... ALIAS_n '\0' CANONICAL_n '\0' '\0' */
105 static const char * volatile charset_aliases;
107 /* Return a pointer to the contents of the charset.alias file. */
109 get_charset_aliases ()
113 cp = charset_aliases;
116 #if !(defined VMS || defined WIN32)
119 const char *base = "charset.alias";
122 /* Make it possible to override the charset.alias location. This is
123 necessary for running the testsuite before "make install". */
124 dir = getenv ("CHARSETALIASDIR");
125 if (dir == NULL || dir[0] == '\0')
126 dir = relocate (LIBDIR);
128 /* Concatenate dir and base into freshly allocated file_name. */
130 size_t dir_len = strlen (dir);
131 size_t base_len = strlen (base);
132 int add_slash = (dir_len > 0 && !ISSLASH (dir[dir_len - 1]));
133 file_name = (char *) malloc (dir_len + add_slash + base_len + 1);
134 if (file_name != NULL)
136 memcpy (file_name, dir, dir_len);
138 file_name[dir_len] = DIRECTORY_SEPARATOR;
139 memcpy (file_name + dir_len + add_slash, base, base_len + 1);
143 if (file_name == NULL || (fp = fopen (file_name, "r")) == NULL)
144 /* Out of memory or file not found, treat it as empty. */
148 /* Parse the file's contents. */
149 char *res_ptr = NULL;
163 if (c == '\n' || c == ' ' || c == '\t')
167 /* Skip comment, to end of line. */
170 while (!(c == EOF || c == '\n'));
176 if (fscanf (fp, "%50s %50s", buf1, buf2) < 2)
180 old_res_ptr = res_ptr;
183 res_size = l1 + 1 + l2 + 1;
184 res_ptr = (char *) malloc (res_size + 1);
188 res_size += l1 + 1 + l2 + 1;
189 res_ptr = (char *) realloc (res_ptr, res_size + 1);
195 if (old_res_ptr != NULL)
199 strcpy (res_ptr + res_size - (l2 + 1) - (l1 + 1), buf1);
200 strcpy (res_ptr + res_size - (l2 + 1), buf2);
207 *(res_ptr + res_size) = '\0';
212 if (file_name != NULL)
218 /* To avoid the troubles of an extra file charset.alias_vms in the
219 sources of many GNU packages, simply inline the aliases here. */
220 /* The list of encodings is taken from the OpenVMS 7.3-1 documentation
221 "Compaq C Run-Time Library Reference Manual for OpenVMS systems"
222 section 10.7 "Handling Different Character Sets". */
223 cp = "ISO8859-1" "\0" "ISO-8859-1" "\0"
224 "ISO8859-2" "\0" "ISO-8859-2" "\0"
225 "ISO8859-5" "\0" "ISO-8859-5" "\0"
226 "ISO8859-7" "\0" "ISO-8859-7" "\0"
227 "ISO8859-8" "\0" "ISO-8859-8" "\0"
228 "ISO8859-9" "\0" "ISO-8859-9" "\0"
230 "eucJP" "\0" "EUC-JP" "\0"
231 "SJIS" "\0" "SHIFT_JIS" "\0"
232 "DECKANJI" "\0" "DEC-KANJI" "\0"
233 "SDECKANJI" "\0" "EUC-JP" "\0"
235 "eucTW" "\0" "EUC-TW" "\0"
236 "DECHANYU" "\0" "DEC-HANYU" "\0"
237 "DECHANZI" "\0" "GB2312" "\0"
239 "DECKOREAN" "\0" "EUC-KR" "\0";
243 /* To avoid the troubles of installing a separate file in the same
244 directory as the DLL and of retrieving the DLL's directory at
245 runtime, simply inline the aliases here. */
247 cp = "CP936" "\0" "GBK" "\0"
248 "CP1361" "\0" "JOHAB" "\0"
249 "CP20127" "\0" "ASCII" "\0"
250 "CP20866" "\0" "KOI8-R" "\0"
251 "CP21866" "\0" "KOI8-RU" "\0"
252 "CP28591" "\0" "ISO-8859-1" "\0"
253 "CP28592" "\0" "ISO-8859-2" "\0"
254 "CP28593" "\0" "ISO-8859-3" "\0"
255 "CP28594" "\0" "ISO-8859-4" "\0"
256 "CP28595" "\0" "ISO-8859-5" "\0"
257 "CP28596" "\0" "ISO-8859-6" "\0"
258 "CP28597" "\0" "ISO-8859-7" "\0"
259 "CP28598" "\0" "ISO-8859-8" "\0"
260 "CP28599" "\0" "ISO-8859-9" "\0"
261 "CP28605" "\0" "ISO-8859-15" "\0";
265 charset_aliases = cp;
271 /* Determine the current locale's character encoding, and canonicalize it
272 into one of the canonical names listed in config.charset.
273 The result must not be freed; it is statically allocated.
274 If the canonical name cannot be determined, the result is a non-canonical
286 #if !(defined WIN32 || defined OS2)
288 # if HAVE_LANGINFO_CODESET
290 /* Most systems support nl_langinfo (CODESET) nowadays. */
291 codeset = nl_langinfo (CODESET);
295 /* On old systems which lack it, use setlocale or getenv. */
296 const char *locale = NULL;
298 /* But most old systems don't have a complete set of locales. Some
299 (like SunOS 4 or DJGPP) have only the C locale. Therefore we don't
300 use setlocale here; it would return "C" when it doesn't support the
301 locale name the user has set. */
302 # if HAVE_SETLOCALE && 0
303 locale = setlocale (LC_CTYPE, NULL);
305 if (locale == NULL || locale[0] == '\0')
307 locale = getenv ("LC_ALL");
308 if (locale == NULL || locale[0] == '\0')
310 locale = getenv ("LC_CTYPE");
311 if (locale == NULL || locale[0] == '\0')
312 locale = getenv ("LANG");
316 /* On some old systems, one used to set locale = "iso8859_1". On others,
317 you set it to "language_COUNTRY.charset". In any case, we resolve it
318 through the charset.alias file. */
325 static char buf[2 + 10 + 1];
327 /* Woe32 has a function returning the locale's codepage as a number. */
328 sprintf (buf, "CP%u", GetACP ());
334 static char buf[2 + 10 + 1];
338 /* Allow user to override the codeset, as set in the operating system,
339 with standard language environment variables. */
340 locale = getenv ("LC_ALL");
341 if (locale == NULL || locale[0] == '\0')
343 locale = getenv ("LC_CTYPE");
344 if (locale == NULL || locale[0] == '\0')
345 locale = getenv ("LANG");
347 if (locale != NULL && locale[0] != '\0')
349 /* If the locale name contains an encoding after the dot, return it. */
350 const char *dot = strchr (locale, '.');
354 const char *modifier;
357 /* Look for the possible @... trailer and remove it, if any. */
358 modifier = strchr (dot, '@');
359 if (modifier == NULL)
361 if (modifier - dot < sizeof (buf))
363 memcpy (buf, dot, modifier - dot);
364 buf [modifier - dot] = '\0';
369 /* Resolve through the charset.alias file. */
374 /* OS/2 has a function returning the locale's codepage as a number. */
375 if (DosQueryCp (sizeof (cp), cp, &cplen))
379 sprintf (buf, "CP%u", cp[0]);
387 /* The canonical name cannot be determined. */
391 for (aliases = get_charset_aliases ();
393 aliases += strlen (aliases) + 1, aliases += strlen (aliases) + 1)
394 if (strcmp (codeset, aliases) == 0
395 || (aliases[0] == '*' && aliases[1] == '\0'))
397 codeset = aliases + strlen (aliases) + 1;
401 /* Don't return an empty string. GNU libc and GNU libiconv interpret
402 the empty string as denoting "the locale's character encoding",
403 thus GNU libiconv would call this function a second time. */
404 if (codeset[0] == '\0')