lib/toutf8.c

   1 /* toutf8.c --- Convert strings from system locale into UTF-8.
   2  * Copyright (C) 2002, 2003, 2004, 2005  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., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA
  19  *
  20  */
  21
  22 #ifdef HAVE_CONFIG_H
  23 # include "config.h"
  24 #endif
  25
  26 /* Get prototypes. */
  27 #include "stringprep.h"
  28
  29 /* Get fprintf. */
  30 #include <stdio.h>
  31
  32 /* Get getenv. */
  33 #include <stdlib.h>
  34
  35 /* Get strlen. */
  36 #include <string.h>
  37
  38 /* Get iconv_string. */
  39 #include "iconvme.h"
  40
  41 #ifdef _LIBC
  42 # define HAVE_ICONV 1
  43 # define HAVE_LOCALE_H 1
  44 # define HAVE_LANGINFO_CODESET 1
  45 #endif
  46
  47 #if HAVE_LOCALE_H
  48 # include <locale.h>
  49 #endif
  50
  51 #if HAVE_LANGINFO_CODESET
  52 # include <langinfo.h>
  53 #endif
  54
  55 #ifdef _LIBC
  56 # define stringprep_locale_charset() nl_langinfo (CODESET)
  57 #else
  58 /**
  59  * stringprep_locale_charset - return charset used in current locale
  60  *
  61  * Find out current locale charset.  The function respect the CHARSET
  62  * environment variable, but typically uses nl_langinfo(CODESET) when
  63  * it is supported.  It fall back on "ASCII" if CHARSET isn't set and
  64  * nl_langinfo isn't supported or return anything.
  65  *
  66  * Note that this function return the application's locale's preferred
  67  * charset (or thread's locale's preffered charset, if your system
  68  * support thread-specific locales).  It does not return what the
  69  * system may be using.  Thus, if you receive data from external
  70  * sources you cannot in general use this function to guess what
  71  * charset it is encoded in.  Use stringprep_convert from the external
  72  * representation into the charset returned by this function, to have
  73  * data in the locale encoding.
  74  *
  75  * Return value: Return the character set used by the current locale.
  76  *   It will never return NULL, but use "ASCII" as a fallback.
  77  **/
  78 const char *
  79 stringprep_locale_charset (void)
  80 {
  81   const char *charset = getenv ("CHARSET");     /* flawfinder: ignore */
  82
  83   if (charset && *charset)
  84     return charset;
  85
  86 # ifdef HAVE_LANGINFO_CODESET
  87   charset = nl_langinfo (CODESET);
  88
  89   if (charset && *charset)
  90     return charset;
  91 # endif
  92
  93   return "ASCII";
  94 }
  95 #endif
  96
  97 /**
  98  * stringprep_convert - encode string using new character set
  99  * @str: input zero-terminated string.
 100  * @to_codeset: name of destination character set.
 101  * @from_codeset: name of origin character set, as used by @str.
 102  *
 103  * Convert the string from one character set to another using the
 104  * system's iconv() function.
 105  *
 106  * Return value: Returns newly allocated zero-terminated string which
 107  *   is @str transcoded into to_codeset.
 108  **/
 109 char *
 110 stringprep_convert (const char *str,
 111                     const char *to_codeset, const char *from_codeset)
 112 {
 113 #if HAVE_ICONV
 114   return iconv_string (str, from_codeset, to_codeset);
 115 #else
 116   char *p;
 117   fprintf (stderr, "libidn: warning: libiconv not installed, cannot "
 118            "convert data to UTF-8\n");
 119   p = malloc (strlen (str) + 1);
 120   if (!p)
 121     return NULL;
 122   return strcpy (p, str);
 123 #endif
 124 }
 125
 126 /**
 127  * stringprep_locale_to_utf8 - convert locale encoded string to UTF-8
 128  * @str: input zero terminated string.
 129  *
 130  * Convert string encoded in the locale's character set into UTF-8 by
 131  * using stringprep_convert().
 132  *
 133  * Return value: Returns newly allocated zero-terminated string which
 134  *   is @str transcoded into UTF-8.
 135  **/
 136 char *
 137 stringprep_locale_to_utf8 (const char *str)
 138 {
 139   return stringprep_convert (str, "UTF-8", stringprep_locale_charset ());
 140 }
 141
 142 /**
 143  * stringprep_utf8_to_locale - encode UTF-8 string to locale encoding
 144  * @str: input zero terminated string.
 145  *
 146  * Convert string encoded in UTF-8 into the locale's character set by
 147  * using stringprep_convert().
 148  *
 149  * Return value: Returns newly allocated zero-terminated string which
 150  *   is @str transcoded into the locale's character set.
 151  **/
 152 char *
 153 stringprep_utf8_to_locale (const char *str)
 154 {
 155   return stringprep_convert (str, stringprep_locale_charset (), "UTF-8");
 156 }