libidn/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., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
  19  *
  20  */
  21
  22 #if 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 LOCALE_WORKS 1
  44 #endif
  45
  46 #if LOCALE_WORKS
  47 # include <langinfo.h>
  48 # include <locale.h>
  49 #endif
  50
  51 #ifdef _LIBC
  52 # define stringprep_locale_charset() nl_langinfo (CODESET)
  53 #else
  54 /**
  55  * stringprep_locale_charset - return charset used in current locale
  56  *
  57  * Find out current locale charset.  The function respect the CHARSET
  58  * environment variable, but typically uses nl_langinfo(CODESET) when
  59  * it is supported.  It fall back on "ASCII" if CHARSET isn't set and
  60  * nl_langinfo isn't supported or return anything.
  61  *
  62  * Note that this function return the application's locale's preferred
  63  * charset (or thread's locale's preffered charset, if your system
  64  * support thread-specific locales).  It does not return what the
  65  * system may be using.  Thus, if you receive data from external
  66  * sources you cannot in general use this function to guess what
  67  * charset it is encoded in.  Use stringprep_convert from the external
  68  * representation into the charset returned by this function, to have
  69  * data in the locale encoding.
  70  *
  71  * Return value: Return the character set used by the current locale.
  72  *   It will never return NULL, but use "ASCII" as a fallback.
  73  **/
  74 const char *
  75 stringprep_locale_charset (void)
  76 {
  77   const char *charset = getenv ("CHARSET");     /* flawfinder: ignore */
  78
  79   if (charset && *charset)
  80     return charset;
  81
  82 # ifdef LOCALE_WORKS
  83   charset = nl_langinfo (CODESET);
  84
  85   if (charset && *charset)
  86     return charset;
  87 # endif
  88
  89   return "ASCII";
  90 }
  91 #endif
  92
  93 /**
  94  * stringprep_convert - encode string using new character set
  95  * @str: input zero-terminated string.
  96  * @to_codeset: name of destination character set.
  97  * @from_codeset: name of origin character set, as used by @str.
  98  *
  99  * Convert the string from one character set to another using the
 100  * system's iconv() function.
 101  *
 102  * Return value: Returns newly allocated zero-terminated string which
 103  *   is @str transcoded into to_codeset.
 104  **/
 105 char *
 106 stringprep_convert (const char *str,
 107                     const char *to_codeset, const char *from_codeset)
 108 {
 109 #if HAVE_ICONV
 110   return iconv_string (str, from_codeset, to_codeset);
 111 #else
 112   char *p;
 113   fprintf (stderr, "libidn: warning: libiconv not installed, cannot "
 114            "convert data to UTF-8\n");
 115   p = malloc (strlen (str) + 1);
 116   if (!p)
 117     return NULL;
 118   return strcpy (p, str);
 119 #endif
 120 }
 121
 122 /**
 123  * stringprep_locale_to_utf8 - convert locale encoded string to UTF-8
 124  * @str: input zero terminated string.
 125  *
 126  * Convert string encoded in the locale's character set into UTF-8 by
 127  * using stringprep_convert().
 128  *
 129  * Return value: Returns newly allocated zero-terminated string which
 130  *   is @str transcoded into UTF-8.
 131  **/
 132 char *
 133 stringprep_locale_to_utf8 (const char *str)
 134 {
 135   return stringprep_convert (str, "UTF-8", stringprep_locale_charset ());
 136 }
 137
 138 /**
 139  * stringprep_utf8_to_locale - encode UTF-8 string to locale encoding
 140  * @str: input zero terminated string.
 141  *
 142  * Convert string encoded in UTF-8 into the locale's character set by
 143  * using stringprep_convert().
 144  *
 145  * Return value: Returns newly allocated zero-terminated string which
 146  *   is @str transcoded into the locale's character set.
 147  **/
 148 char *
 149 stringprep_utf8_to_locale (const char *str)
 150 {
 151   return stringprep_convert (str, stringprep_locale_charset (), "UTF-8");
 152 }