intl/localcharset.c

   1 /* Determine a canonical name for the current locale's character encoding.
   2
   3    Copyright (C) 2000-2002 Free Software Foundation, Inc.
   4
   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)
   8    any later version.
   9
  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.
  14
  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,
  18    USA.  */
  19
  20 /* Written by Bruno Haible <haible@clisp.cons.org>.  */
  21
  22 #ifdef HAVE_CONFIG_H
  23 # include <config.h>
  24 #endif
  25
  26 #if HAVE_STDDEF_H
  27 # include <stddef.h>
  28 #endif
  29
  30 #include <stdio.h>
  31 #if HAVE_STRING_H
  32 # include <string.h>
  33 #else
  34 # include <strings.h>
  35 #endif
  36 #if HAVE_STDLIB_H
  37 # include <stdlib.h>
  38 #endif
  39
  40 #if defined _WIN32 || defined __WIN32__
  41 # undef WIN32   /* avoid warning on mingw32 */
  42 # define WIN32
  43 #endif
  44
  45 #if defined __EMX__
  46 /* Assume EMX program runs on OS/2, even if compiled under DOS.  */
  47 # define OS2
  48 #endif
  49
  50 #if !defined WIN32
  51 # if HAVE_LANGINFO_CODESET
  52 #  include <langinfo.h>
  53 # else
  54 #  if HAVE_SETLOCALE
  55 #   include <locale.h>
  56 #  endif
  57 # endif
  58 #elif defined WIN32
  59 # define WIN32_LEAN_AND_MEAN
  60 # include <windows.h>
  61 #endif
  62 #if defined OS2
  63 # define INCL_DOS
  64 # include <os2.h>
  65 #endif
  66
  67 #if defined _WIN32 || defined __WIN32__ || defined __EMX__ || defined __DJGPP__
  68   /* Win32, OS/2, DOS */
  69 # define ISSLASH(C) ((C) == '/' || (C) == '\\')
  70 #endif
  71
  72 #ifndef DIRECTORY_SEPARATOR
  73 # define DIRECTORY_SEPARATOR '/'
  74 #endif
  75
  76 #ifndef ISSLASH
  77 # define ISSLASH(C) ((C) == DIRECTORY_SEPARATOR)
  78 #endif
  79
  80 #ifdef HAVE_GETC_UNLOCKED
  81 # undef getc
  82 # define getc getc_unlocked
  83 #endif
  84
  85 /* The following static variable is declared 'volatile' to avoid a
  86    possible multithread problem in the function get_charset_aliases. If we
  87    are running in a threaded environment, and if two threads initialize
  88    'charset_aliases' simultaneously, both will produce the same value,
  89    and everything will be ok if the two assignments to 'charset_aliases'
  90    are atomic. But I don't know what will happen if the two assignments mix.  */
  91 #if __STDC__ != 1
  92 # define volatile /* empty */
  93 #endif
  94 /* Pointer to the contents of the charset.alias file, if it has already been
  95    read, else NULL.  Its format is:
  96    ALIAS_1 '\0' CANONICAL_1 '\0' ... ALIAS_n '\0' CANONICAL_n '\0' '\0'  */
  97 static const char * volatile charset_aliases;
  98
  99 /* Return a pointer to the contents of the charset.alias file.  */
 100 static const char *
 101 get_charset_aliases ()
 102 {
 103   const char *cp;
 104
 105   cp = charset_aliases;
 106   if (cp == NULL)
 107     {
 108 #if !defined WIN32
 109       FILE *fp;
 110       const char *dir = LIBDIR;
 111       const char *base = "charset.alias";
 112       char *file_name;
 113
 114       /* Concatenate dir and base into freshly allocated file_name.  */
 115       {
 116         size_t dir_len = strlen (dir);
 117         size_t base_len = strlen (base);
 118         int add_slash = (dir_len > 0 && !ISSLASH (dir[dir_len - 1]));
 119         file_name = (char *) malloc (dir_len + add_slash + base_len + 1);
 120         if (file_name != NULL)
 121           {
 122             memcpy (file_name, dir, dir_len);
 123             if (add_slash)
 124               file_name[dir_len] = DIRECTORY_SEPARATOR;
 125             memcpy (file_name + dir_len + add_slash, base, base_len + 1);
 126           }
 127       }
 128
 129       if (file_name == NULL || (fp = fopen (file_name, "r")) == NULL)
 130         /* Out of memory or file not found, treat it as empty.  */
 131         cp = "";
 132       else
 133         {
 134           /* Parse the file's contents.  */
 135           int c;
 136           char buf1[50+1];
 137           char buf2[50+1];
 138           char *res_ptr = NULL;
 139           size_t res_size = 0;
 140           size_t l1, l2;
 141
 142           for (;;)
 143             {
 144               c = getc (fp);
 145               if (c == EOF)
 146                 break;
 147               if (c == '\n' || c == ' ' || c == '\t')
 148                 continue;
 149               if (c == '#')
 150                 {
 151                   /* Skip comment, to end of line.  */
 152                   do
 153                     c = getc (fp);
 154                   while (!(c == EOF || c == '\n'));
 155                   if (c == EOF)
 156                     break;
 157                   continue;
 158                 }
 159               ungetc (c, fp);
 160               if (fscanf (fp, "%50s %50s", buf1, buf2) < 2)
 161                 break;
 162               l1 = strlen (buf1);
 163               l2 = strlen (buf2);
 164               if (res_size == 0)
 165                 {
 166                   res_size = l1 + 1 + l2 + 1;
 167                   res_ptr = (char *) malloc (res_size + 1);
 168                 }
 169               else
 170                 {
 171                   res_size += l1 + 1 + l2 + 1;
 172                   res_ptr = (char *) realloc (res_ptr, res_size + 1);
 173                 }
 174               if (res_ptr == NULL)
 175                 {
 176                   /* Out of memory. */
 177                   res_size = 0;
 178                   break;
 179                 }
 180               strcpy (res_ptr + res_size - (l2 + 1) - (l1 + 1), buf1);
 181               strcpy (res_ptr + res_size - (l2 + 1), buf2);
 182             }
 183           fclose (fp);
 184           if (res_size == 0)
 185             cp = "";
 186           else
 187             {
 188               *(res_ptr + res_size) = '\0';
 189               cp = res_ptr;
 190             }
 191         }
 192
 193       if (file_name != NULL)
 194         free (file_name);
 195
 196 #else
 197
 198       /* To avoid the troubles of installing a separate file in the same
 199          directory as the DLL and of retrieving the DLL's directory at
 200          runtime, simply inline the aliases here.  */
 201
 202 # if defined WIN32
 203       cp = "CP936" "\0" "GBK" "\0"
 204            "CP1361" "\0" "JOHAB" "\0";
 205 # endif
 206 #endif
 207
 208       charset_aliases = cp;
 209     }
 210
 211   return cp;
 212 }
 213
 214 /* Determine the current locale's character encoding, and canonicalize it
 215    into one of the canonical names listed in config.charset.
 216    The result must not be freed; it is statically allocated.
 217    If the canonical name cannot be determined, the result is a non-canonical
 218    name.  */
 219
 220 #ifdef STATIC
 221 STATIC
 222 #endif
 223 const char *
 224 locale_charset ()
 225 {
 226   const char *codeset;
 227   const char *aliases;
 228
 229 #if !(defined WIN32 || defined OS2)
 230
 231 # if HAVE_LANGINFO_CODESET
 232
 233   /* Most systems support nl_langinfo (CODESET) nowadays.  */
 234   codeset = nl_langinfo (CODESET);
 235
 236 # else
 237
 238   /* On old systems which lack it, use setlocale or getenv.  */
 239   const char *locale = NULL;
 240
 241   /* But most old systems don't have a complete set of locales.  Some
 242      (like SunOS 4 or DJGPP) have only the C locale.  Therefore we don't
 243      use setlocale here; it would return "C" when it doesn't support the
 244      locale name the user has set.  */
 245 #  if HAVE_SETLOCALE && 0
 246   locale = setlocale (LC_CTYPE, NULL);
 247 #  endif
 248   if (locale == NULL || locale[0] == '\0')
 249     {
 250       locale = getenv ("LC_ALL");
 251       if (locale == NULL || locale[0] == '\0')
 252         {
 253           locale = getenv ("LC_CTYPE");
 254           if (locale == NULL || locale[0] == '\0')
 255             locale = getenv ("LANG");
 256         }
 257     }
 258
 259   /* On some old systems, one used to set locale = "iso8859_1". On others,
 260      you set it to "language_COUNTRY.charset". In any case, we resolve it
 261      through the charset.alias file.  */
 262   codeset = locale;
 263
 264 # endif
 265
 266 #elif defined WIN32
 267
 268   static char buf[2 + 10 + 1];
 269
 270   /* Win32 has a function returning the locale's codepage as a number.  */
 271   sprintf (buf, "CP%u", GetACP ());
 272   codeset = buf;
 273
 274 #elif defined OS2
 275
 276   const char *locale;
 277   static char buf[2 + 10 + 1];
 278   ULONG cp[3];
 279   ULONG cplen;
 280
 281   /* Allow user to override the codeset, as set in the operating system,
 282      with standard language environment variables.  */
 283   locale = getenv ("LC_ALL");
 284   if (locale == NULL || locale[0] == '\0')
 285     {
 286       locale = getenv ("LC_CTYPE");
 287       if (locale == NULL || locale[0] == '\0')
 288         locale = getenv ("LANG");
 289     }
 290   if (locale != NULL && locale[0] != '\0')
 291     {
 292       /* If the locale name contains an encoding after the dot, return it.  */
 293       const char *dot = strchr (locale, '.');
 294
 295       if (dot != NULL)
 296         {
 297           const char *modifier;
 298
 299           dot++;
 300           /* Look for the possible @... trailer and remove it, if any.  */
 301           modifier = strchr (dot, '@');
 302           if (modifier == NULL)
 303             return dot;
 304           if (modifier - dot < sizeof (buf))
 305             {
 306               memcpy (buf, dot, modifier - dot);
 307               buf [modifier - dot] = '\0';
 308               return buf;
 309             }
 310         }
 311
 312       /* Resolve through the charset.alias file.  */
 313       codeset = locale;
 314     }
 315   else
 316     {
 317       /* OS/2 has a function returning the locale's codepage as a number.  */
 318       if (DosQueryCp (sizeof (cp), cp, &cplen))
 319         codeset = "";
 320       else
 321         {
 322           sprintf (buf, "CP%u", cp[0]);
 323           codeset = buf;
 324         }
 325     }
 326
 327 #endif
 328
 329   if (codeset == NULL)
 330     /* The canonical name cannot be determined.  */
 331     codeset = "";
 332
 333   /* Resolve alias. */
 334   for (aliases = get_charset_aliases ();
 335        *aliases != '\0';
 336        aliases += strlen (aliases) + 1, aliases += strlen (aliases) + 1)
 337     if (strcmp (codeset, aliases) == 0
 338         || (aliases[0] == '*' && aliases[1] == '\0'))
 339       {
 340         codeset = aliases + strlen (aliases) + 1;
 341         break;
 342       }
 343
 344   return codeset;
 345 }