gnu/dist/grep/intl/localcharset.c

   1 /*      $NetBSD$        */
   2
   3 /* Determine a canonical name for the current locale's character encoding.
   4
   5    Copyright (C) 2000-2002 Free Software Foundation, Inc.
   6
   7    This program is free software; you can redistribute it and/or modify it
   8    under the terms of the GNU Library General Public License as published
   9    by the Free Software Foundation; either version 2, or (at your option)
  10    any later version.
  11
  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 GNU
  15    Library General Public License for more details.
  16
  17    You should have received a copy of the GNU Library General Public
  18    License along with this program; if not, write to the Free Software
  19    Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307,
  20    USA.  */
  21
  22 /* Written by Bruno Haible <haible@clisp.cons.org>.  */
  23
  24 #ifdef HAVE_CONFIG_H
  25 # include <config.h>
  26 #endif
  27
  28 #if HAVE_STDDEF_H
  29 # include <stddef.h>
  30 #endif
  31
  32 #include <stdio.h>
  33 #if HAVE_STRING_H
  34 # include <string.h>
  35 #else
  36 # include <strings.h>
  37 #endif
  38 #if HAVE_STDLIB_H
  39 # include <stdlib.h>
  40 #endif
  41
  42 #if defined _WIN32 || defined __WIN32__
  43 # undef WIN32   /* avoid warning on mingw32 */
  44 # define WIN32
  45 #endif
  46
  47 #if defined __EMX__
  48 /* Assume EMX program runs on OS/2, even if compiled under DOS.  */
  49 # define OS2
  50 #endif
  51
  52 #if !defined WIN32
  53 # if HAVE_LANGINFO_CODESET
  54 #  include <langinfo.h>
  55 # else
  56 #  if HAVE_SETLOCALE
  57 #   include <locale.h>
  58 #  endif
  59 # endif
  60 #elif defined WIN32
  61 # define WIN32_LEAN_AND_MEAN
  62 # include <windows.h>
  63 #endif
  64 #if defined OS2
  65 # define INCL_DOS
  66 # include <os2.h>
  67 #endif
  68
  69 #if defined _WIN32 || defined __WIN32__ || defined __EMX__ || defined __DJGPP__
  70   /* Win32, OS/2, DOS */
  71 # define ISSLASH(C) ((C) == '/' || (C) == '\\')
  72 #endif
  73
  74 #ifndef DIRECTORY_SEPARATOR
  75 # define DIRECTORY_SEPARATOR '/'
  76 #endif
  77
  78 #ifndef ISSLASH
  79 # define ISSLASH(C) ((C) == DIRECTORY_SEPARATOR)
  80 #endif
  81
  82 #ifdef HAVE_GETC_UNLOCKED
  83 # undef getc
  84 # define getc getc_unlocked
  85 #endif
  86
  87 /* The following static variable is declared 'volatile' to avoid a
  88    possible multithread problem in the function get_charset_aliases. If we
  89    are running in a threaded environment, and if two threads initialize
  90    'charset_aliases' simultaneously, both will produce the same value,
  91    and everything will be ok if the two assignments to 'charset_aliases'
  92    are atomic. But I don't know what will happen if the two assignments mix.  */
  93 #if __STDC__ != 1
  94 # define volatile /* empty */
  95 #endif
  96 /* Pointer to the contents of the charset.alias file, if it has already been
  97    read, else NULL.  Its format is:
  98    ALIAS_1 '\0' CANONICAL_1 '\0' ... ALIAS_n '\0' CANONICAL_n '\0' '\0'  */
  99 static const char * volatile charset_aliases;
 100
 101 /* Return a pointer to the contents of the charset.alias file.  */
 102 static const char *
 103 get_charset_aliases ()
 104 {
 105   const char *cp;
 106
 107   cp = charset_aliases;
 108   if (cp == NULL)
 109     {
 110 #if !defined WIN32
 111       FILE *fp;
 112       const char *dir = LIBDIR;
 113       const char *base = "charset.alias";
 114       char *file_name;
 115
 116       /* Concatenate dir and base into freshly allocated file_name.  */
 117       {
 118         size_t dir_len = strlen (dir);
 119         size_t base_len = strlen (base);
 120         int add_slash = (dir_len > 0 && !ISSLASH (dir[dir_len - 1]));
 121         file_name = (char *) malloc (dir_len + add_slash + base_len + 1);
 122         if (file_name != NULL)
 123           {
 124             memcpy (file_name, dir, dir_len);
 125             if (add_slash)
 126               file_name[dir_len] = DIRECTORY_SEPARATOR;
 127             memcpy (file_name + dir_len + add_slash, base, base_len + 1);
 128           }
 129       }
 130
 131       if (file_name == NULL || (fp = fopen (file_name, "r")) == NULL)
 132         /* Out of memory or file not found, treat it as empty.  */
 133         cp = "";
 134       else
 135         {
 136           /* Parse the file's contents.  */
 137           int c;
 138           char buf1[50+1];
 139           char buf2[50+1];
 140           char *res_ptr = NULL;
 141           size_t res_size = 0;
 142           size_t l1, l2;
 143
 144           for (;;)
 145             {
 146               c = getc (fp);
 147               if (c == EOF)
 148                 break;
 149               if (c == '\n' || c == ' ' || c == '\t')
 150                 continue;
 151               if (c == '#')
 152                 {
 153                   /* Skip comment, to end of line.  */
 154                   do
 155                     c = getc (fp);
 156                   while (!(c == EOF || c == '\n'));
 157                   if (c == EOF)
 158                     break;
 159                   continue;
 160                 }
 161               ungetc (c, fp);
 162               if (fscanf (fp, "%50s %50s", buf1, buf2) < 2)
 163                 break;
 164               l1 = strlen (buf1);
 165               l2 = strlen (buf2);
 166               if (res_size == 0)
 167                 {
 168                   res_size = l1 + 1 + l2 + 1;
 169                   res_ptr = (char *) malloc (res_size + 1);
 170                 }
 171               else
 172                 {
 173                   res_size += l1 + 1 + l2 + 1;
 174                   res_ptr = (char *) realloc (res_ptr, res_size + 1);
 175                 }
 176               if (res_ptr == NULL)
 177                 {
 178                   /* Out of memory. */
 179                   res_size = 0;
 180                   break;
 181                 }
 182               strcpy (res_ptr + res_size - (l2 + 1) - (l1 + 1), buf1);
 183               strcpy (res_ptr + res_size - (l2 + 1), buf2);
 184             }
 185           fclose (fp);
 186           if (res_size == 0)
 187             cp = "";
 188           else
 189             {
 190               *(res_ptr + res_size) = '\0';
 191               cp = res_ptr;
 192             }
 193         }
 194
 195       if (file_name != NULL)
 196         free (file_name);
 197
 198 #else
 199
 200       /* To avoid the troubles of installing a separate file in the same
 201          directory as the DLL and of retrieving the DLL's directory at
 202          runtime, simply inline the aliases here.  */
 203
 204 # if defined WIN32
 205       cp = "CP936" "\0" "GBK" "\0"
 206            "CP1361" "\0" "JOHAB" "\0";
 207 # endif
 208 #endif
 209
 210       charset_aliases = cp;
 211     }
 212
 213   return cp;
 214 }
 215
 216 /* Determine the current locale's character encoding, and canonicalize it
 217    into one of the canonical names listed in config.charset.
 218    The result must not be freed; it is statically allocated.
 219    If the canonical name cannot be determined, the result is a non-canonical
 220    name.  */
 221
 222 #ifdef STATIC
 223 STATIC
 224 #endif
 225 const char *
 226 locale_charset ()
 227 {
 228   const char *codeset;
 229   const char *aliases;
 230
 231 #if !(defined WIN32 || defined OS2)
 232
 233 # if HAVE_LANGINFO_CODESET
 234
 235   /* Most systems support nl_langinfo (CODESET) nowadays.  */
 236   codeset = nl_langinfo (CODESET);
 237
 238 # else
 239
 240   /* On old systems which lack it, use setlocale or getenv.  */
 241   const char *locale = NULL;
 242
 243   /* But most old systems don't have a complete set of locales.  Some
 244      (like SunOS 4 or DJGPP) have only the C locale.  Therefore we don't
 245      use setlocale here; it would return "C" when it doesn't support the
 246      locale name the user has set.  */
 247 #  if HAVE_SETLOCALE && 0
 248   locale = setlocale (LC_CTYPE, NULL);
 249 #  endif
 250   if (locale == NULL || locale[0] == '\0')
 251     {
 252       locale = getenv ("LC_ALL");
 253       if (locale == NULL || locale[0] == '\0')
 254         {
 255           locale = getenv ("LC_CTYPE");
 256           if (locale == NULL || locale[0] == '\0')
 257             locale = getenv ("LANG");
 258         }
 259     }
 260
 261   /* On some old systems, one used to set locale = "iso8859_1". On others,
 262      you set it to "language_COUNTRY.charset". In any case, we resolve it
 263      through the charset.alias file.  */
 264   codeset = locale;
 265
 266 # endif
 267
 268 #elif defined WIN32
 269
 270   static char buf[2 + 10 + 1];
 271
 272   /* Win32 has a function returning the locale's codepage as a number.  */
 273   sprintf (buf, "CP%u", GetACP ());
 274   codeset = buf;
 275
 276 #elif defined OS2
 277
 278   const char *locale;
 279   static char buf[2 + 10 + 1];
 280   ULONG cp[3];
 281   ULONG cplen;
 282
 283   /* Allow user to override the codeset, as set in the operating system,
 284      with standard language environment variables.  */
 285   locale = getenv ("LC_ALL");
 286   if (locale == NULL || locale[0] == '\0')
 287     {
 288       locale = getenv ("LC_CTYPE");
 289       if (locale == NULL || locale[0] == '\0')
 290         locale = getenv ("LANG");
 291     }
 292   if (locale != NULL && locale[0] != '\0')
 293     {
 294       /* If the locale name contains an encoding after the dot, return it.  */
 295       const char *dot = strchr (locale, '.');
 296
 297       if (dot != NULL)
 298         {
 299           const char *modifier;
 300
 301           dot++;
 302           /* Look for the possible @... trailer and remove it, if any.  */
 303           modifier = strchr (dot, '@');
 304           if (modifier == NULL)
 305             return dot;
 306           if (modifier - dot < sizeof (buf))
 307             {
 308               memcpy (buf, dot, modifier - dot);
 309               buf [modifier - dot] = '\0';
 310               return buf;
 311             }
 312         }
 313
 314       /* Resolve through the charset.alias file.  */
 315       codeset = locale;
 316     }
 317   else
 318     {
 319       /* OS/2 has a function returning the locale's codepage as a number.  */
 320       if (DosQueryCp (sizeof (cp), cp, &cplen))
 321         codeset = "";
 322       else
 323         {
 324           sprintf (buf, "CP%u", cp[0]);
 325           codeset = buf;
 326         }
 327     }
 328
 329 #endif
 330
 331   if (codeset == NULL)
 332     /* The canonical name cannot be determined.  */
 333     codeset = "";
 334
 335   /* Resolve alias. */
 336   for (aliases = get_charset_aliases ();
 337        *aliases != '\0';
 338        aliases += strlen (aliases) + 1, aliases += strlen (aliases) + 1)
 339     if (strcmp (codeset, aliases) == 0
 340         || (aliases[0] == '*' && aliases[1] == '\0'))
 341       {
 342         codeset = aliases + strlen (aliases) + 1;
 343         break;
 344       }
 345
 346   return codeset;
 347 }