toutf8.c

   1 /* toutf8.c     Convert strings from system locale into UTF-8.
   2  * Copyright (C) 2002, 2003  Simon Josefsson
   3  *
   4  * This file is part of GNU Libidn.
   5  *
   6  * GNU Libidn 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  * GNU Libidn 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 GNU Libidn; if not, write to the Free Software
  18  * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
  19  *
  20  */
  21
  22 #include "internal.h"
  23
  24 #ifdef HAVE_ICONV
  25
  26 #include <iconv.h>
  27
  28 #if LOCALE_WORKS
  29 #include <langinfo.h>
  30 #include <locale.h>
  31 #endif
  32
  33 static const char *
  34 stringprep_locale_charset_slow (void)
  35 {
  36   const char *charset = getenv ("CHARSET");     /* flawfinder: ignore */
  37
  38   if (charset && *charset)
  39     return charset;
  40
  41 #if LOCALE_WORKS
  42   {
  43     char *p;
  44
  45     p = setlocale (LC_CTYPE, NULL);
  46     setlocale (LC_CTYPE, "");
  47
  48     charset = nl_langinfo (CODESET);
  49
  50     setlocale (LC_CTYPE, p);
  51
  52     if (charset && *charset)
  53       return charset;
  54   }
  55 #endif
  56
  57   return "ASCII";
  58 }
  59
  60 static const char *stringprep_locale_charset_cache = NULL;
  61
  62 /**
  63  * stringprep_locale_charset:
  64  *
  65  * Return value: Return the character set used by the system locale.
  66  *   It will never return NULL, but use "ASCII" as a fallback.
  67  **/
  68 const char *
  69 stringprep_locale_charset (void)
  70 {
  71   if (!stringprep_locale_charset_cache)
  72     stringprep_locale_charset_cache = stringprep_locale_charset_slow ();
  73
  74   return stringprep_locale_charset_cache;
  75 }
  76
  77 /**
  78  * stringprep_convert:
  79  * @str: input zero-terminated string.
  80  * @to_codeset: name of destination character set.
  81  * @from_codeset: name of origin character set, as used by @str.
  82  *
  83  * Convert the string from one character set to another using the
  84  * system's iconv() function.
  85  *
  86  * Return value: Returns newly allocated zero-terminated string which
  87  *   is @str transcoded into to_codeset.
  88  **/
  89 char *
  90 stringprep_convert (const char *str,
  91                     const char *to_codeset, const char *from_codeset)
  92 {
  93   iconv_t cd;
  94   char *dest;
  95   char *outp;
  96   char *p, *startp;
  97   size_t inbytes_remaining;
  98   size_t outbytes_remaining;
  99   size_t err;
 100   size_t outbuf_size;
 101   int have_error = 0;
 102   int len;
 103
 104   if (strcmp (to_codeset, from_codeset) == 0)
 105     return strdup (str);
 106
 107   cd = iconv_open (to_codeset, from_codeset);
 108
 109   if (cd == (iconv_t) - 1)
 110     return NULL;
 111
 112   p = strdup(str);
 113   if (p == NULL)
 114     return NULL;
 115   len = strlen (p);
 116   startp = p;
 117   inbytes_remaining = len;
 118   outbuf_size = len + 1;        /* + 1 for nul in case len == 1 */
 119
 120   outbytes_remaining = outbuf_size - 1; /* -1 for nul */
 121   outp = dest = malloc (outbuf_size);
 122
 123 again:
 124
 125   err = iconv (cd, &p, &inbytes_remaining, &outp, &outbytes_remaining);
 126
 127   if (err == (size_t) - 1)
 128     {
 129       switch (errno)
 130         {
 131         case EINVAL:
 132           /* Incomplete text, do not report an error */
 133           break;
 134
 135         case E2BIG:
 136           {
 137             size_t used = outp - dest;
 138
 139             outbuf_size *= 2;
 140             dest = realloc (dest, outbuf_size);
 141
 142             outp = dest + used;
 143             outbytes_remaining = outbuf_size - used - 1;        /* -1 for nul */
 144
 145             goto again;
 146           }
 147           break;
 148
 149         case EILSEQ:
 150           have_error = 1;
 151           break;
 152
 153         default:
 154           have_error = 1;
 155           break;
 156         }
 157     }
 158
 159   *outp = '\0';
 160
 161   if ((p - startp) != len)
 162     have_error = 1;
 163
 164
 165   free(startp);
 166
 167   iconv_close (cd);
 168
 169   if (have_error)
 170     {
 171       free (dest);
 172       dest = NULL;
 173     }
 174
 175   return dest;
 176 }
 177
 178 #else
 179
 180 const char *
 181 stringprep_locale_charset ()
 182 {
 183   return "ASCII";
 184 }
 185
 186 char *
 187 stringprep_convert (const char *str,
 188                     const char *to_codeset, const char *from_codeset)
 189 {
 190   fprintf (stderr,
 191            "warning: cannot convert data to UTF-8, returning source\n");
 192   fprintf (stderr, "warning: this indicate a badly installed GNU Libidn\n");
 193   return strdup (str);
 194 }
 195
 196 #endif
 197
 198 /**
 199  * stringprep_locale_to_utf8:
 200  * @str: input zero terminated string.
 201  *
 202  * Convert string encoded in the locale's character set into UTF-8 by
 203  * using stringprep_convert().
 204  *
 205  * Return value: Returns newly allocated zero-terminated string which
 206  *   is @str transcoded into UTF-8.
 207  **/
 208 char *
 209 stringprep_locale_to_utf8 (const char *str)
 210 {
 211   return stringprep_convert (str, "UTF-8", stringprep_locale_charset ());
 212 }
 213
 214 /**
 215  * stringprep_utf8_to_locale:
 216  * @str: input zero terminated string.
 217  *
 218  * Convert string encoded in UTF-8 into the locale's character set by
 219  * using stringprep_convert().
 220  *
 221  * Return value: Returns newly allocated zero-terminated string which
 222  *   is @str transcoded into the locale's character set.
 223  **/
 224 char *
 225 stringprep_utf8_to_locale (const char *str)
 226 {
 227   return stringprep_convert (str, stringprep_locale_charset (), "UTF-8");
 228 }