src/charsets.c

   1 /* Text conversion from one charset to another.
   2
   3    Copyright (C) 2001 Walery Studennikov <despair@sama.ru>
   4
   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 of the License, or
   8    (at your option) 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
  13    GNU General Public License for more details.
  14
  15    You should have received a copy of the GNU General Public License
  16    along with this program; if not, write to the Free Software
  17    Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
  18  */
  19
  20 #include <config.h>
  21
  22 #ifdef HAVE_CHARSET
  23 #include <stdlib.h>
  24 #include <stdio.h>
  25 #include <string.h>
  26 #include <iconv.h>
  27
  28 #include "global.h"
  29 #include "charsets.h"
  30
  31 int n_codepages = 0;
  32
  33 struct codepage_desc *codepages;
  34
  35 unsigned char conv_displ[256];
  36 unsigned char conv_input[256];
  37
  38 int
  39 load_codepages_list (void)
  40 {
  41     int result = -1;
  42     FILE *f;
  43     char *fname;
  44     char buf[256];
  45     extern char *mc_home;
  46     extern int display_codepage;
  47     char *default_codepage = NULL;
  48
  49     fname = concat_dir_and_file (mc_home, CHARSETS_INDEX);
  50     if (!(f = fopen (fname, "r"))) {
  51         fprintf (stderr, _("Warning: file %s not found\n"), fname);
  52         g_free (fname);
  53         return -1;
  54     }
  55     g_free (fname);
  56
  57     for (n_codepages = 0; fgets (buf, sizeof (buf), f);)
  58         if (buf[0] != '\n' && buf[0] != '\0' && buf[0] != '#')
  59             ++n_codepages;
  60     rewind (f);
  61
  62     codepages = g_new0 (struct codepage_desc, n_codepages + 1);
  63
  64     for (n_codepages = 0; fgets (buf, sizeof buf, f);) {
  65         /* split string into id and cpname */
  66         char *p = buf;
  67         int buflen = strlen (buf);
  68
  69         if (*p == '\n' || *p == '\0' || *p == '#')
  70             continue;
  71
  72         if (buflen > 0 && buf[buflen - 1] == '\n')
  73             buf[buflen - 1] = '\0';
  74         while (*p != '\t' && *p != ' ' && *p != '\0')
  75             ++p;
  76         if (*p == '\0')
  77             goto fail;
  78
  79         *p++ = 0;
  80
  81         while (*p == '\t' || *p == ' ')
  82             ++p;
  83         if (*p == '\0')
  84             goto fail;
  85
  86         if (strcmp (buf, "default") == 0) {
  87             default_codepage = g_strdup (p);
  88             continue;
  89         }
  90
  91         codepages[n_codepages].id = g_strdup (buf);
  92         codepages[n_codepages].name = g_strdup (p);
  93         ++n_codepages;
  94     }
  95
  96     if (default_codepage) {
  97         display_codepage = get_codepage_index (default_codepage);
  98         g_free (default_codepage);
  99     }
 100
 101     result = n_codepages;
 102   fail:
 103     fclose (f);
 104     return result;
 105 }
 106
 107 void
 108 free_codepages_list (void)
 109 {
 110     if (n_codepages > 0) {
 111         int i;
 112         for (i = 0; i < n_codepages; i++) {
 113             g_free (codepages[i].id);
 114             g_free (codepages[i].name);
 115         }
 116         n_codepages = 0;
 117         g_free (codepages);
 118         codepages = 0;
 119     }
 120 }
 121
 122 #define OTHER_8BIT "Other_8_bit"
 123
 124 char *
 125 get_codepage_id (int n)
 126 {
 127     return (n < 0) ? OTHER_8BIT : codepages[n].id;
 128 }
 129
 130 int
 131 get_codepage_index (const char *id)
 132 {
 133     int i;
 134     if (strcmp (id, OTHER_8BIT) == 0)
 135         return -1;
 136     for (i = 0; codepages[i].id; ++i)
 137         if (strcmp (id, codepages[i].id) == 0)
 138             return i;
 139     return -1;
 140 }
 141
 142 static char
 143 translate_character (iconv_t cd, char c)
 144 {
 145     char outbuf[4], *obuf;
 146     size_t ibuflen, obuflen, count;
 147
 148     ICONV_CONST char *ibuf = &c;
 149     obuf = outbuf;
 150     ibuflen = 1;
 151     obuflen = 4;
 152
 153     count = iconv (cd, &ibuf, &ibuflen, &obuf, &obuflen);
 154     if (count >= 0 && ibuflen == 0)
 155         return outbuf[0];
 156
 157     return UNKNCHAR;
 158 }
 159
 160 char errbuf[255];
 161
 162 /*
 163  * FIXME: This assumes that ASCII is always the first encoding
 164  * in mc.charsets
 165  */
 166 #define CP_ASCII 0
 167
 168 char *
 169 init_translation_table (int cpsource, int cpdisplay)
 170 {
 171     int i;
 172     iconv_t cd;
 173     char *cpsour, *cpdisp;
 174
 175     /* Fill inpit <-> display tables */
 176
 177     if (cpsource < 0 || cpdisplay < 0 || cpsource == cpdisplay) {
 178         for (i = 0; i <= 255; ++i) {
 179             conv_displ[i] = i;
 180             conv_input[i] = i;
 181         }
 182         return NULL;
 183     }
 184
 185     for (i = 0; i <= 127; ++i) {
 186         conv_displ[i] = i;
 187         conv_input[i] = i;
 188     }
 189
 190     cpsour = codepages[cpsource].id;
 191     cpdisp = codepages[cpdisplay].id;
 192
 193     /* display <- inpit table */
 194
 195     cd = iconv_open (cpdisp, cpsour);
 196     if (cd == (iconv_t) - 1) {
 197         g_snprintf (errbuf, sizeof (errbuf),
 198                     _("Cannot translate from %s to %s"), cpsour, cpdisp);
 199         return errbuf;
 200     }
 201
 202     for (i = 128; i <= 255; ++i)
 203         conv_displ[i] = translate_character (cd, i);
 204
 205     iconv_close (cd);
 206
 207     /* inpit <- display table */
 208
 209     cd = iconv_open (cpsour, cpdisp);
 210     if (cd == (iconv_t) - 1) {
 211         g_snprintf (errbuf, sizeof (errbuf),
 212                     _("Cannot translate from %s to %s"), cpdisp, cpsour);
 213         return errbuf;
 214     }
 215
 216     for (i = 128; i <= 255; ++i) {
 217         unsigned char ch;
 218         ch = translate_character (cd, i);
 219         conv_input[i] = (ch == UNKNCHAR) ? i : ch;
 220     }
 221
 222     iconv_close (cd);
 223
 224     return NULL;
 225 }
 226
 227 void
 228 convert_to_display (char *str)
 229 {
 230     if (!str)
 231         return;
 232
 233     while (*str) {
 234         *str = conv_displ[(unsigned char) *str];
 235         str++;
 236     }
 237 }
 238
 239 void
 240 convert_from_input (char *str)
 241 {
 242     if (!str)
 243         return;
 244
 245     while (*str) {
 246         *str = conv_input[(unsigned char) *str];
 247         str++;
 248     }
 249 }
 250 #endif                          /* HAVE_CHARSET */