iconv-detect.c

   1 /* -*- Mode: C; tab-width: 8; indent-tabs-mode: t; c-basic-offset: 8 -*- */
   2 /*
   3  * Copyright (C) 1999-2008 Novell, Inc. (www.novell.com)
   4  *
   5  * This library is free software: you can redistribute it and/or modify it
   6  * under the terms of the GNU Lesser General Public License as published by
   7  * the Free Software Foundation.
   8  *
   9  * This library is distributed in the hope that it will be useful, but
  10  * WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
  11  * or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License
  12  * for more details.
  13  *
  14  * You should have received a copy of the GNU Lesser General Public License
  15  * along with this library. If not, see <http://www.gnu.org/licenses/>.
  16  *
  17  * Authors: Jeffrey Stedfast <fejj@ximian.com>
  18  */
  19
  20 #include <stdio.h>
  21 #include <stdlib.h>
  22 #include <string.h>
  23 #include <iconv.h>
  24
  25 enum {
  26         ISO_UNSUPPORTED          = 0,
  27
  28         /* iso-8859-1 */
  29         ISO_DASH_D_DASH_D_LOWER  = (1 << 0),
  30         ISO_DASH_D_DASH_D        = (1 << 1),
  31         ISO_D_DASH_D             = (1 << 2),
  32         ISO_D_D                  = (1 << 3),
  33         ISO_UNDER_D_DASH_D       = (1 << 4),
  34         NO_ISO_D_DASH_D          = (1 << 5),
  35
  36         /* iso-10646-1 */
  37         /*ISO_DASH_D_DASH_D_LOWER  = (1 << 0),*/
  38         /*ISO_DASH_D_DASH_D        = (1 << 1),*/
  39         /*ISO_D_DASH_D             = (1 << 2),*/
  40         ISO_DASH_D_LOWER           = (1 << 3),
  41         ISO_DASH_D                 = (1 << 4),
  42         ISO_D                      = (1 << 5),
  43         UCS4                       = (1 << 6),
  44
  45         /* iso-2022-jp */
  46         ISO_DASH_D_DASH_S_LOWER  = (1 << 0),
  47         ISO_DASH_D_DASH_S        = (1 << 1),
  48         ISO_D_DASH_S             = (1 << 2),
  49 };
  50
  51
  52 typedef struct {
  53         char *charset;
  54         char *format;
  55         int id;
  56 } CharInfo;
  57
  58
  59 static CharInfo iso8859_tests[] = {
  60         { "iso-8859-1",  "iso-%d-%d", ISO_DASH_D_DASH_D_LOWER },
  61         { "ISO-8859-1",  "ISO-%d-%d", ISO_DASH_D_DASH_D },
  62         { "ISO8859-1",   "ISO%d-%d",  ISO_D_DASH_D },
  63         { "ISO88591",    "ISO%d%d",   ISO_D_D },
  64         { "ISO_8859-1",  "ISO_%d-%d", ISO_UNDER_D_DASH_D },
  65         { "8859-1",      "%d-%d",     NO_ISO_D_DASH_D },
  66 };
  67
  68 static int num_iso8859_tests = sizeof (iso8859_tests) / sizeof (CharInfo);
  69
  70 static CharInfo iso2022_tests[] = {
  71         { "iso-2022-jp", "iso-%d-%s", ISO_DASH_D_DASH_S_LOWER },
  72         { "ISO-2022-JP", "ISO-%d-%s", ISO_DASH_D_DASH_S },
  73         { "ISO2022-JP",  "ISO%d-%s",  ISO_D_DASH_S },
  74 };
  75
  76 static int num_iso2022_tests = sizeof (iso2022_tests) / sizeof (CharInfo);
  77
  78 static CharInfo iso10646_tests[] = {
  79         { "iso-10646-1", "iso-%d-%d",  ISO_DASH_D_DASH_D_LOWER },
  80         { "ISO-10646-1", "ISO-%d-%d",  ISO_DASH_D_DASH_D },
  81         { "ISO10646-1",  "ISO%d-%d",   ISO_D_DASH_D },
  82         { "iso-10646",   "iso-%d",     ISO_DASH_D_LOWER },
  83         { "ISO-10646",   "ISO-%d",     ISO_DASH_D },
  84         { "ISO10646",    "ISO%d",      ISO_D },
  85         { "UCS-4BE",     "UCS-4BE",    UCS4 },
  86 };
  87
  88 static int num_iso10646_tests = sizeof (iso10646_tests) / sizeof (CharInfo);
  89
  90 static int
  91 test_iconv (void)
  92 {
  93         char *jp = "\x1B\x24\x42\x46\x7C\x4B\x5C\x38\x6C";
  94         char *utf8 = "\xE6\x97\xA5\xE6\x9C\xAC\xE8\xAA\x9E";
  95         char *transbuf = malloc (10), *trans = transbuf;
  96         iconv_t cd;
  97         size_t jp_len = strlen (jp), utf8_len = 10;
  98         size_t utf8_real_len = strlen (utf8);
  99
 100         cd = iconv_open ("UTF-8", "ISO-2022-JP");
 101         if (cd == (iconv_t) -1)
 102                 return 0;
 103
 104         if (iconv (cd, &jp, &jp_len, &trans, &utf8_len) == -1 || jp_len != 0) {
 105                 iconv_close (cd);
 106                 return 0;
 107         }
 108         if (memcmp (utf8, transbuf, utf8_real_len) != 0) {
 109                 iconv_close (cd);
 110                 return 0;
 111         }
 112
 113         iconv_close (cd);
 114
 115         return 1;
 116 }
 117
 118 int
 119 main (int argc,
 120       char **argv)
 121 {
 122         unsigned int iso8859, iso2022, iso10646;
 123         CharInfo *info;
 124         iconv_t cd;
 125         FILE *fp;
 126         int i;
 127
 128         if (!test_iconv ())
 129                 return 1;
 130
 131         fp = fopen ("iconv-detect.h", "w");
 132         if (fp == NULL)
 133                 return 255;
 134
 135         fprintf (fp, "/* This is an auto-generated header, DO NOT EDIT! */\n\n");
 136
 137         iso8859 = ISO_UNSUPPORTED;
 138         info = iso8859_tests;
 139         /*printf ("#define DEFAULT_ISO_FORMAT(iso,codepage)\t");*/
 140         for (i = 0; i < num_iso8859_tests; i++) {
 141                 cd = iconv_open (info[i].charset, "UTF-8");
 142                 if (cd != (iconv_t) -1) {
 143                         iconv_close (cd);
 144                         /*printf ("(\"%s\", (iso), (codepage))\n", info[i].format);*/
 145                         fprintf (stderr, "System prefers %s\n", info[i].charset);
 146                         iso8859 = info[i].id;
 147                         break;
 148                 }
 149         }
 150
 151         if (iso8859 == ISO_UNSUPPORTED) {
 152                 fprintf (stderr, "System doesn't support any ISO-8859-1 formats\n");
 153                 fprintf (fp, "#define ICONV_ISO_D_FORMAT \"%s\"\n", info[0].format);
 154         } else {
 155                 fprintf (fp, "#define ICONV_ISO_D_FORMAT \"%s\"\n", info[i].format);
 156         }
 157
 158         iso2022 = ISO_UNSUPPORTED;
 159         info = iso2022_tests;
 160         /*printf ("#define ISO_2022_FORMAT(iso,codepage)\t");*/
 161         for (i = 0; i < num_iso2022_tests; i++) {
 162                 cd = iconv_open (info[i].charset, "UTF-8");
 163                 if (cd != (iconv_t) -1) {
 164                         iconv_close (cd);
 165                         /*printf ("(\"%s\", (iso), (codepage))\n", info[i].format);*/
 166                         fprintf (stderr, "System prefers %s\n", info[i].charset);
 167                         iso2022 = info[i].id;
 168                         break;
 169                 }
 170         }
 171
 172         if (iso2022 == ISO_UNSUPPORTED) {
 173                 fprintf (stderr, "System doesn't support any ISO-2022 formats\n");
 174                 fprintf (fp, "#define ICONV_ISO_S_FORMAT \"%s\"\n", info[0].format);
 175         } else {
 176                 fprintf (fp, "#define ICONV_ISO_S_FORMAT \"%s\"\n", info[i].format);
 177         }
 178
 179         iso10646 = ISO_UNSUPPORTED;
 180         info = iso10646_tests;
 181         /*printf ("#define ISO_10646_FORMAT(iso,codepage)\t");*/
 182         for (i = 0; i < num_iso10646_tests; i++) {
 183                 cd = iconv_open (info[i].charset, "UTF-8");
 184                 if (cd != (iconv_t) -1) {
 185                         iconv_close (cd);
 186                         /*if (info[i].id < ISO_DASH_D_LOWER)
 187                                 printf ("(\"%s\", (iso), (codepage))\n", info[i].format);
 188                         else
 189                         printf ("(\"%s\", (iso))\n", info[i].format);*/
 190                         fprintf (stderr, "System prefers %s\n", info[i].charset);
 191                         iso10646 = info[i].id;
 192                         break;
 193                 }
 194         }
 195
 196         /* we don't need a printf format for iso-10646 because there is only 1 */
 197         if (iso10646 == ISO_UNSUPPORTED) {
 198                 fprintf (stderr, "System doesn't support any ISO-10646-1 formats\n");
 199                 fprintf (fp, "#define ICONV_10646 \"%s\"\n", info[0].charset);
 200         } else {
 201                 fprintf (fp, "#define ICONV_10646 \"%s\"\n", info[i].charset);
 202         }
 203
 204         fclose (fp);
 205
 206         return 0;
 207 }