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 #include "strutil.h"            /* utf-8 functions */
  33 #include "main.h"
  34
  35 int n_codepages = 0;
  36
  37 struct codepage_desc *codepages;
  38
  39 unsigned char conv_displ[256];
  40 unsigned char conv_input[256];
  41
  42 int
  43 load_codepages_list (void)
  44 {
  45     int result = -1;
  46     FILE *f;
  47     char *fname;
  48     char buf[256];
  49     extern char *mc_home;
  50     extern int display_codepage;
  51     char *default_codepage = NULL;
  52
  53     fname = concat_dir_and_file (mc_home, CHARSETS_INDEX);
  54     if (!(f = fopen (fname, "r"))) {
  55         fprintf (stderr, _("Warning: file %s not found\n"), fname);
  56         g_free (fname);
  57         return -1;
  58     }
  59     g_free (fname);
  60
  61     for (n_codepages = 0; fgets (buf, sizeof (buf), f);)
  62         if (buf[0] != '\n' && buf[0] != '\0' && buf[0] != '#')
  63             ++n_codepages;
  64     rewind (f);
  65
  66     codepages = g_new0 (struct codepage_desc, n_codepages + 1);
  67
  68     for (n_codepages = 0; fgets (buf, sizeof buf, f);) {
  69         /* split string into id and cpname */
  70         char *p = buf;
  71         int buflen = strlen (buf);
  72
  73         if (*p == '\n' || *p == '\0' || *p == '#')
  74             continue;
  75
  76         if (buflen > 0 && buf[buflen - 1] == '\n')
  77             buf[buflen - 1] = '\0';
  78         while (*p != '\t' && *p != ' ' && *p != '\0')
  79             ++p;
  80         if (*p == '\0')
  81             goto fail;
  82
  83         *p++ = 0;
  84
  85         while (*p == '\t' || *p == ' ')
  86             ++p;
  87         if (*p == '\0')
  88             goto fail;
  89
  90         if (strcmp (buf, "default") == 0) {
  91             default_codepage = g_strdup (p);
  92             continue;
  93         }
  94
  95         codepages[n_codepages].id = g_strdup (buf);
  96         codepages[n_codepages].name = g_strdup (p);
  97         ++n_codepages;
  98     }
  99
 100     if (default_codepage) {
 101         display_codepage = get_codepage_index (default_codepage);
 102         g_free (default_codepage);
 103     }
 104
 105     result = n_codepages;
 106   fail:
 107     fclose (f);
 108     return result;
 109 }
 110
 111 void
 112 free_codepages_list (void)
 113 {
 114     if (n_codepages > 0) {
 115         int i;
 116         for (i = 0; i < n_codepages; i++) {
 117             g_free (codepages[i].id);
 118             g_free (codepages[i].name);
 119         }
 120         n_codepages = 0;
 121         g_free (codepages);
 122         codepages = 0;
 123     }
 124 }
 125
 126 #define OTHER_8BIT "Other_8_bit"
 127
 128 const char *
 129 get_codepage_id (int n)
 130 {
 131     return (n < 0) ? OTHER_8BIT : codepages[n].id;
 132 }
 133
 134 int
 135 get_codepage_index (const char *id)
 136 {
 137     int i;
 138     if (strcmp (id, OTHER_8BIT) == 0)
 139         return -1;
 140     for (i = 0; codepages[i].id; ++i)
 141         if (strcmp (id, codepages[i].id) == 0)
 142             return i;
 143     return -1;
 144 }
 145
 146 static char
 147 translate_character (iconv_t cd, char c)
 148 {
 149     char outbuf[4], *obuf;
 150     size_t ibuflen, obuflen, count;
 151
 152     ICONV_CONST char *ibuf = &c;
 153     obuf = outbuf;
 154     ibuflen = 1;
 155     obuflen = 4;
 156
 157     count = iconv (cd, &ibuf, &ibuflen, &obuf, &obuflen);
 158     if (count != ((size_t) -1) && ibuflen == 0)
 159         return outbuf[0];
 160
 161     return UNKNCHAR;
 162 }
 163
 164 char errbuf[255];
 165
 166 /*
 167  * FIXME: This assumes that ASCII is always the first encoding
 168  * in mc.charsets
 169  */
 170 #define CP_ASCII 0
 171
 172 const char *
 173 init_translation_table (int cpsource, int cpdisplay)
 174 {
 175     int i;
 176     iconv_t cd;
 177     const char *cpsour, *cpdisp;
 178
 179     /* Fill inpit <-> display tables */
 180
 181     if (cpsource < 0 || cpdisplay < 0 || cpsource == cpdisplay) {
 182         for (i = 0; i <= 255; ++i) {
 183             conv_displ[i] = i;
 184             conv_input[i] = i;
 185         }
 186         return NULL;
 187     }
 188
 189     for (i = 0; i <= 127; ++i) {
 190         conv_displ[i] = i;
 191         conv_input[i] = i;
 192     }
 193
 194     cpsour = codepages[cpsource].id;
 195     cpdisp = codepages[cpdisplay].id;
 196
 197     /* display <- inpit table */
 198
 199     cd = iconv_open (cpdisp, cpsour);
 200     if (cd == (iconv_t) - 1) {
 201         g_snprintf (errbuf, sizeof (errbuf),
 202                     _("Cannot translate from %s to %s"), cpsour, cpdisp);
 203         return errbuf;
 204     }
 205
 206     for (i = 128; i <= 255; ++i)
 207         conv_displ[i] = translate_character (cd, i);
 208
 209     iconv_close (cd);
 210
 211     /* inpit <- display table */
 212
 213     cd = iconv_open (cpsour, cpdisp);
 214     if (cd == (iconv_t) - 1) {
 215         g_snprintf (errbuf, sizeof (errbuf),
 216                     _("Cannot translate from %s to %s"), cpdisp, cpsour);
 217         return errbuf;
 218     }
 219
 220     for (i = 128; i <= 255; ++i) {
 221         unsigned char ch;
 222         ch = translate_character (cd, i);
 223         conv_input[i] = (ch == UNKNCHAR) ? i : ch;
 224     }
 225
 226     iconv_close (cd);
 227
 228     return NULL;
 229 }
 230
 231 void
 232 convert_to_display (char *str)
 233 {
 234     if (!str)
 235         return;
 236
 237     while (*str) {
 238         *str = conv_displ[(unsigned char) *str];
 239         str++;
 240     }
 241 }
 242
 243 void
 244 convert_from_input (char *str)
 245 {
 246     if (!str)
 247         return;
 248
 249     while (*str) {
 250         *str = conv_input[(unsigned char) *str];
 251         str++;
 252     }
 253 }
 254
 255 unsigned char
 256 convert_from_utf_to_current (const char *str)
 257 {
 258     if (!str)
 259         return '.';
 260
 261     unsigned char ch;
 262     char *cp_from = NULL;
 263     GIConv conv;
 264     GString *translated_data;
 265
 266     translated_data = g_string_new ("");
 267     cp_from =  get_codepage_id ( source_codepage );
 268     conv = str_crt_conv_from (cp_from);
 269
 270     if (conv != INVALID_CONV) {
 271         if (str_convert (conv, (char *) str, translated_data) != ESTR_FAILURE) {
 272             ch = translated_data->str[0];
 273         } else {
 274             ch = '.';
 275         }
 276         str_close_conv (conv);
 277     }
 278     g_string_free (translated_data, TRUE);
 279     return ch;
 280
 281 }
 282
 283 unsigned char
 284 convert_from_utf_to_current_c (const int input_char)
 285 {
 286     unsigned char str[6 + 1];
 287     unsigned char buf_ch[6 + 1];
 288     unsigned char ch = '.';
 289
 290     char *cp_from = NULL;
 291     GIConv conv;
 292     GString *translated_data;
 293     int res = 0;
 294
 295     res = g_unichar_to_utf8 (input_char, str);
 296     if ( res == 0 ) {
 297         return ch;
 298     }
 299     str[6] = '\0';
 300
 301     cp_from =  get_codepage_id ( source_codepage );
 302     conv = str_crt_conv_from (cp_from);
 303
 304     if (conv != INVALID_CONV) {
 305         switch (str_translate_char (conv, str, res, buf_ch, sizeof(buf_ch))) {
 306         case 0:
 307             ch = buf_ch[0];
 308             break;
 309         case 1:
 310             ch = '.';
 311             break;
 312         case 2:
 313             ch = '.';
 314         }
 315         str_close_conv (conv);
 316     }
 317     return ch;
 318
 319 }
 320
 321 #endif                          /* HAVE_CHARSET */