1 /* Determine a canonical name for the current locale's character encoding.
3 Copyright (C) 2000-2001 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 <haible@clisp.cons.org>. */
40 #if defined _WIN32 || defined __WIN32__
41 # undef WIN32 /* avoid warning on mingw32 */
46 # if HAVE_LANGINFO_CODESET
47 # include <langinfo.h>
54 # define WIN32_LEAN_AND_MEAN
58 #ifndef DIRECTORY_SEPARATOR
59 # define DIRECTORY_SEPARATOR '/'
63 # define ISSLASH(C) ((C) == DIRECTORY_SEPARATOR)
66 /* The following static variable is declared 'volatile' to avoid a
67 possible multithread problem in the function get_charset_aliases. If we
68 are running in a threaded environment, and if two threads initialize
69 'charset_aliases' simultaneously, both will produce the same value,
70 and everything will be ok if the two assignments to 'charset_aliases'
71 are atomic. But I don't know what will happen if the two assignments mix. */
72 /* Pointer to the contents of the charset.alias file, if it has already been
73 read, else NULL. Its format is:
74 ALIAS_1 '\0' CANONICAL_1 '\0' ... ALIAS_n '\0' CANONICAL_n '\0' '\0' */
75 static char * volatile charset_aliases;
77 /* Return a pointer to the contents of the charset.alias file. */
79 get_charset_aliases ()
88 const char *dir = LIBDIR;
89 const char *base = "charset.alias";
92 /* Concatenate dir and base into freshly allocated file_name. */
94 size_t dir_len = strlen (dir);
95 size_t base_len = strlen (base);
96 int add_slash = (dir_len > 0 && !ISSLASH (dir[dir_len - 1]));
97 file_name = (char *) malloc (dir_len + add_slash + base_len + 1);
98 if (file_name != NULL)
100 memcpy (file_name, dir, dir_len);
102 file_name[dir_len] = DIRECTORY_SEPARATOR;
103 memcpy (file_name + dir_len + add_slash, base, base_len + 1);
107 if (file_name == NULL || (fp = fopen (file_name, "r")) == NULL)
108 /* Out of memory or file not found, treat it as empty. */
112 /* Parse the file's contents. */
116 char *res_ptr = NULL;
125 if (c == '\n' || c == ' ' || c == '\t')
129 /* Skip comment, to end of line. */
132 while (!(c == EOF || c == '\n'));
138 if (fscanf(fp, "%50s %50s", buf1, buf2) < 2)
144 res_size = l1 + 1 + l2 + 1;
145 res_ptr = malloc (res_size + 1);
149 res_size += l1 + 1 + l2 + 1;
150 res_ptr = realloc (res_ptr, res_size + 1);
158 strcpy (res_ptr + res_size - (l2 + 1) - (l1 + 1), buf1);
159 strcpy (res_ptr + res_size - (l2 + 1), buf2);
166 *(res_ptr + res_size) = '\0';
171 if (file_name != NULL)
176 /* To avoid the troubles of installing a separate file in the same
177 directory as the DLL and of retrieving the DLL's directory at
178 runtime, simply inline the aliases here. */
180 cp = "CP936" "\0" "GBK" "\0"
181 "CP1361" "\0" "JOHAB" "\0";
184 charset_aliases = cp;
190 /* Determine the current locale's character encoding, and canonicalize it
191 into one of the canonical names listed in config.charset.
192 The result must not be freed; it is statically allocated.
193 If the canonical name cannot be determined, the result is a non-canonical
207 # if HAVE_LANGINFO_CODESET
209 /* Most systems support nl_langinfo (CODESET) nowadays. */
210 codeset = nl_langinfo (CODESET);
214 /* On old systems which lack it, use setlocale or getenv. */
215 const char *locale = NULL;
217 /* But most old systems don't have a complete set of locales. Some
218 (like SunOS 4 or DJGPP) have only the C locale. Therefore we don't
219 use setlocale here; it would return "C" when it doesn't support the
220 locale name the user has set. */
221 # if HAVE_SETLOCALE && 0
222 locale = setlocale (LC_CTYPE, NULL);
224 if (locale == NULL || locale[0] == '\0')
226 locale = getenv ("LC_ALL");
227 if (locale == NULL || locale[0] == '\0')
229 locale = getenv ("LC_CTYPE");
230 if (locale == NULL || locale[0] == '\0')
231 locale = getenv ("LANG");
235 /* On some old systems, one used to set locale = "iso8859_1". On others,
236 you set it to "language_COUNTRY.charset". In any case, we resolve it
237 through the charset.alias file. */
244 static char buf[2 + 10 + 1];
246 /* Win32 has a function returning the locale's codepage as a number. */
247 sprintf (buf, "CP%u", GetACP ());
253 /* The canonical name cannot be determined. */
255 else if (codeset[0] != '\0')
258 for (aliases = get_charset_aliases ();
260 aliases += strlen (aliases) + 1, aliases += strlen (aliases) + 1)
261 if (!strcmp (codeset, aliases))
263 codeset = aliases + strlen (aliases) + 1;