functions/decode/utf-8.php

   1 <?php
   2 /*
   3  * decode/utf-8.php
   4  * $Id$
   5  *
   6  * Copyright (c) 1999-2003 The SquirrelMail Project Team
   7  * Licensed under the GNU GPL. For full terms see the file COPYING.
   8  *
   9  * This file contains utf-8 decoding function that is needed to read
  10  * utf-8 encoded mails in non-utf-8 locale.
  11  *
  12  * Every decoded character consists of n bytes. First byte is octal
  13  * 300-375, other bytes - always octals 200-277.
  14  *
  15  * \a\b characters are decoded to html code octdec(a-300)*64 + octdec(b-200)
  16  * \a\b\c characters are decoded to html code octdec(a-340)*64*64 + octdec(b-200)*64 + octdec(c-200)
  17  *
  18  * decoding cycle is unfinished. please test and report problems to tokul@users.sourceforge.net
  19  *
  20  */
  21 function charset_decode_utf8 ($string) {
  22     global $default_charset, $languages, $sm_notAlias;
  23
  24     if (strtolower($default_charset) == 'utf-8')
  25         return $string;
  26     if (strtolower($languages[$sm_notAlias]['CHARSET']) == 'utf-8')
  27         return $string;
  28
  29     /* Only do the slow convert if there are 8-bit characters */
  30     if (! ereg("[\200-\377]", $string))
  31         return $string;
  32
  33     // decode three byte unicode characters
  34     $string = preg_replace("/([\340-\357])([\200-\277])([\200-\277])/e",
  35     "'&#'.((ord('\\1')-224)*4096+(ord('\\2')-128)*64+(ord('\\3')-128)).';'",
  36     $string);
  37
  38     // decode two byte unicode characters
  39     $string = preg_replace("/([\300-\337])([\200-\277])/e",
  40     "'&#'.((ord('\\1')-192)*64+(ord('\\2')-128)).';'",
  41     $string);
  42
  43     return $string;
  44 }
  45
  46 ?>