toutf8.c

   1 /* toutf8.c     convert strings from system locale into UTF-8
   2  * Copyright (C) 2002  Simon Josefsson
   3  *
   4  * This file is part of libstringprep.
   5  *
   6  * Libstringprep is free software; you can redistribute it and/or
   7  * modify it under the terms of the GNU Lesser General Public
   8  * License as published by the Free Software Foundation; either
   9  * version 2.1 of the License, or (at your option) any later version.
  10  *
  11  * Libstringprep is distributed in the hope that it will be useful,
  12  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  13  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  14  * Lesser General Public License for more details.
  15  *
  16  * You should have received a copy of the GNU Lesser General Public
  17  * License along with libstringprep; if not, write to the Free Software
  18  * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
  19  *
  20  */
  21
  22 #ifdef HAVE_CONFIG_H
  23 #include "config.h"
  24 #endif
  25
  26 #ifdef HAVE_ICONV
  27
  28 #include <iconv.h>
  29
  30 #ifdef STDC_HEADERS
  31 #include <stdlib.h>
  32 #endif
  33 #if HAVE_STRING_H
  34 # if !STDC_HEADERS && HAVE_MEMORY_H
  35 #  include <memory.h>
  36 # endif
  37 # include <string.h>
  38 #endif
  39 #if HAVE_STRINGS_H
  40 # include <strings.h>
  41 #endif
  42
  43 #ifdef HAVE_ERRNO_H
  44 #include <errno.h>
  45 #endif
  46 extern int errno;
  47
  48 #if LOCALE_WORKS
  49 #include <langinfo.h>
  50 #include <locale.h>
  51 #endif
  52
  53 static const char *
  54 stringprep_locale_charset_slow ()
  55 {
  56   const char *charset = getenv ("CHARSET");
  57   char *p;
  58
  59   if (charset && *charset)
  60     return charset;
  61
  62 #if LOCALE_WORKS
  63   p = setlocale (LC_CTYPE, NULL);
  64   setlocale (LC_CTYPE, "");
  65
  66   charset = nl_langinfo (CODESET);
  67
  68   setlocale (LC_CTYPE, p);
  69 #endif
  70
  71   if (charset && *charset)
  72     return charset;
  73
  74   return "ASCII";
  75 }
  76
  77 static const char *stringprep_locale_charset_cache = NULL;
  78
  79 const char *
  80 stringprep_locale_charset ()
  81 {
  82   if (!stringprep_locale_charset_cache)
  83     stringprep_locale_charset_cache = stringprep_locale_charset_slow ();
  84
  85   return stringprep_locale_charset_cache;
  86 }
  87
  88 char *
  89 stringprep_convert (const char *str,
  90                     const char *to_codeset, const char *from_codeset)
  91 {
  92   iconv_t cd;
  93   char *dest;
  94   char *outp;
  95   char *p;
  96   size_t inbytes_remaining;
  97   size_t outbytes_remaining;
  98   size_t err;
  99   size_t outbuf_size;
 100   int have_error = 0;
 101   int len;
 102
 103   if (strcmp (to_codeset, from_codeset) == 0)
 104     return strdup (str);
 105
 106   cd = iconv_open (to_codeset, from_codeset);
 107
 108   if (cd == (iconv_t) - 1)
 109     return NULL;
 110
 111   len = strlen (str);
 112
 113   p = (char *) str;
 114   inbytes_remaining = len;
 115   outbuf_size = len + 1;        /* + 1 for nul in case len == 1 */
 116
 117   outbytes_remaining = outbuf_size - 1; /* -1 for nul */
 118   outp = dest = malloc (outbuf_size);
 119
 120 again:
 121
 122   err = iconv (cd, &p, &inbytes_remaining, &outp, &outbytes_remaining);
 123
 124   if (err == (size_t) - 1)
 125     {
 126       switch (errno)
 127         {
 128         case EINVAL:
 129           /* Incomplete text, do not report an error */
 130           break;
 131
 132         case E2BIG:
 133           {
 134             size_t used = outp - dest;
 135
 136             outbuf_size *= 2;
 137             dest = realloc (dest, outbuf_size);
 138
 139             outp = dest + used;
 140             outbytes_remaining = outbuf_size - used - 1;        /* -1 for nul */
 141
 142             goto again;
 143           }
 144           break;
 145
 146         case EILSEQ:
 147           have_error = 1;
 148           break;
 149
 150         default:
 151           have_error = 1;
 152           break;
 153         }
 154     }
 155
 156   *outp = '\0';
 157
 158   if ((p - str) != len)
 159     have_error = 1;
 160
 161   iconv_close (cd);
 162
 163   if (have_error)
 164     {
 165       free (dest);
 166       dest = NULL;
 167     }
 168
 169   return dest;
 170 }
 171
 172 #else
 173
 174 const char *
 175 stringprep_locale_charset ()
 176 {
 177   return NULL;
 178 }
 179
 180 char *
 181 stringprep_convert (const char *str,
 182                     const char *to_codeset, const char *from_codeset)
 183 {
 184   return NULL;
 185 }
 186
 187 #endif
 188
 189 char *
 190 stringprep_locale_to_utf8 (const char *str)
 191 {
 192   return stringprep_convert (str, "UTF-8", stringprep_locale_charset ());
 193 }