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., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  18  */
  19
  20 #include <config.h>
  21
  22 #ifdef HAVE_CHARSET
  23
  24 #include <stdio.h>
  25 #include <stdlib.h>
  26 #include <string.h>
  27
  28 #include <iconv.h>
  29
  30 #include "global.h"
  31 #include "charsets.h"
  32
  33 int n_codepages = 0;
  34
  35 struct codepage_desc *codepages;
  36
  37 unsigned char conv_displ[256];
  38 unsigned char conv_input[256];
  39
  40 int
  41 load_codepages_list (void)
  42 {
  43     int result = -1;
  44     FILE *f;
  45     char *fname;
  46     char buf[256];
  47     extern char *mc_home;
  48     extern int display_codepage;
  49     char *default_codepage = NULL;
  50
  51     fname = concat_dir_and_file (mc_home, CHARSETS_INDEX);
  52     if (!(f = fopen (fname, "r"))) {
  53         fprintf (stderr, _("Warning: file %s not found\n"), fname);
  54         g_free (fname);
  55         return -1;
  56     }
  57     g_free (fname);
  58
  59     for (n_codepages = 0; fgets (buf, sizeof (buf), f);)
  60         if (buf[0] != '\n' && buf[0] != '\0' && buf[0] != '#')
  61             ++n_codepages;
  62     rewind (f);
  63
  64     codepages = g_new0 (struct codepage_desc, n_codepages + 1);
  65
  66     for (n_codepages = 0; fgets (buf, sizeof buf, f);) {
  67         /* split string into id and cpname */
  68         char *p = buf;
  69         int buflen = strlen (buf);
  70
  71         if (*p == '\n' || *p == '\0' || *p == '#')
  72             continue;
  73
  74         if (buflen > 0 && buf[buflen - 1] == '\n')
  75             buf[buflen - 1] = '\0';
  76         while (*p != '\t' && *p != ' ' && *p != '\0')
  77             ++p;
  78         if (*p == '\0')
  79             goto fail;
  80
  81         *p++ = 0;
  82
  83         while (*p == '\t' || *p == ' ')
  84             ++p;
  85         if (*p == '\0')
  86             goto fail;
  87
  88         if (strcmp (buf, "default") == 0) {
  89             default_codepage = g_strdup (p);
  90             continue;
  91         }
  92
  93         codepages[n_codepages].id = g_strdup (buf);
  94         codepages[n_codepages].name = g_strdup (p);
  95         ++n_codepages;
  96     }
  97
  98     if (default_codepage) {
  99         display_codepage = get_codepage_index (default_codepage);
 100         g_free (default_codepage);
 101     }
 102
 103     result = n_codepages;
 104   fail:
 105     fclose (f);
 106     return result;
 107 }
 108
 109 void
 110 free_codepages_list (void)
 111 {
 112     if (n_codepages > 0) {
 113         int i;
 114         for (i = 0; i < n_codepages; i++) {
 115             g_free (codepages[i].id);
 116             g_free (codepages[i].name);
 117         }
 118         n_codepages = 0;
 119         g_free (codepages);
 120         codepages = 0;
 121     }
 122 }
 123
 124 #define OTHER_8BIT "Other_8_bit"
 125
 126 const char *
 127 get_codepage_id (int n)
 128 {
 129     return (n < 0) ? OTHER_8BIT : codepages[n].id;
 130 }
 131
 132 int
 133 get_codepage_index (const char *id)
 134 {
 135     int i;
 136     if (strcmp (id, OTHER_8BIT) == 0)
 137         return -1;
 138     for (i = 0; codepages[i].id; ++i)
 139         if (strcmp (id, codepages[i].id) == 0)
 140             return i;
 141     return -1;
 142 }
 143
 144 static char
 145 translate_character (iconv_t cd, char c)
 146 {
 147     char outbuf[4], *obuf;
 148     size_t ibuflen, obuflen, count;
 149
 150     ICONV_CONST char *ibuf = &c;
 151     obuf = outbuf;
 152     ibuflen = 1;
 153     obuflen = 4;
 154
 155     count = iconv (cd, &ibuf, &ibuflen, &obuf, &obuflen);
 156     if (count != ((size_t) -1) && ibuflen == 0)
 157         return outbuf[0];
 158
 159     return UNKNCHAR;
 160 }
 161
 162 char errbuf[255];
 163
 164 /*
 165  * FIXME: This assumes that ASCII is always the first encoding
 166  * in mc.charsets
 167  */
 168 #define CP_ASCII 0
 169
 170 const char *
 171 init_translation_table (int cpsource, int cpdisplay)
 172 {
 173     int i;
 174     iconv_t cd;
 175     const char *cpsour, *cpdisp;
 176
 177     /* Fill inpit <-> display tables */
 178
 179     if (cpsource < 0 || cpdisplay < 0 || cpsource == cpdisplay) {
 180         for (i = 0; i <= 255; ++i) {
 181             conv_displ[i] = i;
 182             conv_input[i] = i;
 183         }
 184         return NULL;
 185     }
 186
 187     for (i = 0; i <= 127; ++i) {
 188         conv_displ[i] = i;
 189         conv_input[i] = i;
 190     }
 191
 192     cpsour = codepages[cpsource].id;
 193     cpdisp = codepages[cpdisplay].id;
 194
 195     /* display <- inpit table */
 196
 197     cd = iconv_open (cpdisp, cpsour);
 198     if (cd == (iconv_t) - 1) {
 199         g_snprintf (errbuf, sizeof (errbuf),
 200                     _("Cannot translate from %s to %s"), cpsour, cpdisp);
 201         return errbuf;
 202     }
 203
 204     for (i = 128; i <= 255; ++i)
 205         conv_displ[i] = translate_character (cd, i);
 206
 207     iconv_close (cd);
 208
 209     /* inpit <- display table */
 210
 211     cd = iconv_open (cpsour, cpdisp);
 212     if (cd == (iconv_t) - 1) {
 213         g_snprintf (errbuf, sizeof (errbuf),
 214                     _("Cannot translate from %s to %s"), cpdisp, cpsour);
 215         return errbuf;
 216     }
 217
 218     for (i = 128; i <= 255; ++i) {
 219         unsigned char ch;
 220         ch = translate_character (cd, i);
 221         conv_input[i] = (ch == UNKNCHAR) ? i : ch;
 222     }
 223
 224     iconv_close (cd);
 225
 226     return NULL;
 227 }
 228
 229 void
 230 convert_to_display (char *str)
 231 {
 232     if (!str)
 233         return;
 234
 235     while (*str) {
 236         *str = conv_displ[(unsigned char) *str];
 237         str++;
 238     }
 239 }
 240
 241 void
 242 convert_from_input (char *str)
 243 {
 244     if (!str)
 245         return;
 246
 247     while (*str) {
 248         *str = conv_input[(unsigned char) *str];
 249         str++;
 250     }
 251 }
 252 #endif                          /* HAVE_CHARSET */