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