functions/decode/utf_8.php

   1 <?php
   2 /**
   3  * functions/decode/utf-8.php - utf-8 decoding functions
   4  *
   5  * Copyright (c) 2003-2004 The SquirrelMail Project Team
   6  * Licensed under the GNU GPL. For full terms see the file COPYING.
   7  *
   8  * This file contains utf-8 decoding function that is needed to read
   9  * utf-8 encoded mails in non-utf-8 locale.
  10  *
  11  * Every decoded character consists of n bytes. First byte is octal
  12  * 300-375, other bytes - always octals 200-277.
  13  *<pre>
  14  * Ranges (first byte):
  15  *                oct     dec    hex
  16  * Two byte   - 300-337 192-223 C0-DF
  17  * Three byte - 340-357 224-239 E0-EF
  18  * Four byte  - 360-367 240-247 F0-F7
  19  * Five byte  - 370-373 248-251 F8-FB
  20  * Six byte   - 374-375 252-253 FC-FD
  21  *
  22  * \a\b characters are decoded to html code calculated with formula:
  23  *  octdec(a-300)*64 + octdec(b-200)
  24  *
  25  * \a\b\c characters are decoded to html code calculated with formula:
  26  *  octdec(a-340)*64^2 + octdec(b-200)*64 + octdec(c-200)
  27  *
  28  * \a\b\c\d characters are decoded to html code calculated with formula:
  29  *  octdec(a-360)*64^3 + octdec(b-200)*64^2 +
  30  *  + octdec(c-200)*64 + octdec(d-200)
  31  *
  32  * \a\b\c\d\e characters are decoded to html code calculated with formula:
  33  *  octdec(a-370)*64^4 + octdec(b-200)*64^3 +
  34  *  + octdec(c-200)*64^2 + octdec(d-200)*64 + octdec(e-200)
  35  *
  36  * \a\b\c\d\e\f characters are decoded to html code calculated with formula:
  37  *  octdec(a-374)*64^5 + octdec(b-200)*64^4 + octdec(c-200)*64^3 +
  38  *  + octdec(d-200)*64^2 + octdec(e-200)*64 + octdec(f-200)
  39  *</pre>
  40  * @version $Id$
  41  * @package squirrelmail
  42  * @subpackage decode
  43  */
  44
  45 /**
  46  * Decode utf-8 strings
  47  * @param string $string Encoded string
  48  * @return string Decoded string
  49  */
  50 function charset_decode_utf_8 ($string) {
  51     global $squirrelmail_language;
  52
  53     // Japanese translation uses mbstring function to read utf-8
  54     if ($squirrelmail_language == 'ja_JP')
  55         return $string;
  56
  57     // don't do decoding when there are no 8bit symbols
  58     if (! sq_is8bit($string,'utf-8'))
  59         return $string;
  60
  61     // decode six byte unicode characters
  62     /* (i think currently there is no such symbol)
  63     $string = preg_replace("/([\374-\375])([\200-\277])([\200-\277])([\200-\277])([\200-\277])([\200-\277])/e",
  64     "'&#'.((ord('\\1')-252)*1073741824+(ord('\\2')-200)*16777216+(ord('\\3')-200)*262144+(ord('\\4')-128)*4096+(ord('\\5')-128)*64+(ord('\\6')-128)).';'",
  65     $string);
  66     */
  67
  68     // decode five byte unicode characters
  69     /* (i think currently there is no such symbol)
  70     $string = preg_replace("/([\370-\373])([\200-\277])([\200-\277])([\200-\277])([\200-\277])/e",
  71     "'&#'.((ord('\\1')-248)*16777216+(ord('\\2')-200)*262144+(ord('\\3')-128)*4096+(ord('\\4')-128)*64+(ord('\\5')-128)).';'",
  72     $string);
  73     */
  74
  75     // decode four byte unicode characters
  76     $string = preg_replace("/([\360-\367])([\200-\277])([\200-\277])([\200-\277])/e",
  77     "'&#'.((ord('\\1')-240)*262144+(ord('\\2')-128)*4096+(ord('\\3')-128)*64+(ord('\\4')-128)).';'",
  78     $string);
  79
  80     // decode three byte unicode characters
  81     $string = preg_replace("/([\340-\357])([\200-\277])([\200-\277])/e",
  82     "'&#'.((ord('\\1')-224)*4096+(ord('\\2')-128)*64+(ord('\\3')-128)).';'",
  83     $string);
  84
  85     // decode two byte unicode characters
  86     $string = preg_replace("/([\300-\337])([\200-\277])/e",
  87     "'&#'.((ord('\\1')-192)*64+(ord('\\2')-128)).';'",
  88     $string);
  89
  90     // remove broken unicode
  91     $string = preg_replace("/[\200-\237]|\240|[\241-\377]/",'?',$string);
  92
  93     return $string;
  94 }
  95 ?>