[Summary view] [Print] [Text view]

   1  <?php
   2  /**
   3  * @version $Id: validation.php,v 1.2 2006/02/26 13:20:44 harryf Exp $
   4  * Tools for validing a UTF-8 string is well formed.
   5  * The Original Code is Mozilla Communicator client code.
   6  * The Initial Developer of the Original Code is
   7  * Netscape Communications Corporation.
   8  * Portions created by the Initial Developer are Copyright (C) 1998
   9  * the Initial Developer. All Rights Reserved.
  10  * Ported to PHP by Henri Sivonen (http://hsivonen.iki.fi)
  11  * Slight modifications to fit with phputf8 library by Harry Fuecks (hfuecks gmail com)
  12  * @see http://lxr.mozilla.org/seamonkey/source/intl/uconv/src/nsUTF8ToUnicode.cpp
  13  * @see http://lxr.mozilla.org/seamonkey/source/intl/uconv/src/nsUnicodeToUTF8.cpp
  14  * @see http://hsivonen.iki.fi/php-utf8/
  15  * @package utf8
  16  * @subpackage validation
  17  */
  18  
  19  //--------------------------------------------------------------------
  20  /**
  21  * Tests a string as to whether it's valid UTF-8 and supported by the
  22  * Unicode standard
  23  * Note: this function has been modified to simple return true or false
  24  * @author <hsivonen@iki.fi>
  25  * @param string UTF-8 encoded string
  26  * @return boolean true if valid
  27  * @see http://hsivonen.iki.fi/php-utf8/
  28  * @see utf8_compliant
  29  * @package utf8
  30  * @subpackage validation
  31  */
  32  function utf8_is_valid($str) {
  33      
  34      $mState = 0;     // cached expected number of octets after the current octet
  35                       // until the beginning of the next UTF8 character sequence
  36      $mUcs4  = 0;     // cached Unicode character
  37      $mBytes = 1;     // cached expected number of octets in the current sequence
  38      
  39      $len = strlen($str);
  40      
  41      for($i = 0; $i < $len; $i++) {
  42          
  43          $in = ord($str{$i});
  44          
  45          if ( $mState == 0) {
  46              
  47              // When mState is zero we expect either a US-ASCII character or a
  48              // multi-octet sequence.
  49              if (0 == (0x80 & ($in))) {
  50                  // US-ASCII, pass straight through.
  51                  $mBytes = 1;
  52                  
  53              } else if (0xC0 == (0xE0 & ($in))) {
  54                  // First octet of 2 octet sequence
  55                  $mUcs4 = ($in);
  56                  $mUcs4 = ($mUcs4 & 0x1F) << 6;
  57                  $mState = 1;
  58                  $mBytes = 2;
  59                  
  60              } else if (0xE0 == (0xF0 & ($in))) {
  61                  // First octet of 3 octet sequence
  62                  $mUcs4 = ($in);
  63                  $mUcs4 = ($mUcs4 & 0x0F) << 12;
  64                  $mState = 2;
  65                  $mBytes = 3;
  66                  
  67              } else if (0xF0 == (0xF8 & ($in))) {
  68                  // First octet of 4 octet sequence
  69                  $mUcs4 = ($in);
  70                  $mUcs4 = ($mUcs4 & 0x07) << 18;
  71                  $mState = 3;
  72                  $mBytes = 4;
  73                  
  74              } else if (0xF8 == (0xFC & ($in))) {
  75                  /* First octet of 5 octet sequence.
  76                  *
  77                  * This is illegal because the encoded codepoint must be either
  78                  * (a) not the shortest form or
  79                  * (b) outside the Unicode range of 0-0x10FFFF.
  80                  * Rather than trying to resynchronize, we will carry on until the end
  81                  * of the sequence and let the later error handling code catch it.
  82                  */
  83                  $mUcs4 = ($in);
  84                  $mUcs4 = ($mUcs4 & 0x03) << 24;
  85                  $mState = 4;
  86                  $mBytes = 5;
  87                  
  88              } else if (0xFC == (0xFE & ($in))) {
  89                  // First octet of 6 octet sequence, see comments for 5 octet sequence.
  90                  $mUcs4 = ($in);
  91                  $mUcs4 = ($mUcs4 & 1) << 30;
  92                  $mState = 5;
  93                  $mBytes = 6;
  94                  
  95              } else {
  96                  /* Current octet is neither in the US-ASCII range nor a legal first
  97                   * octet of a multi-octet sequence.
  98                   */
  99                  return FALSE;
 100                  
 101              }
 102          
 103          } else {
 104              
 105              // When mState is non-zero, we expect a continuation of the multi-octet
 106              // sequence
 107              if (0x80 == (0xC0 & ($in))) {
 108                  
 109                  // Legal continuation.
 110                  $shift = ($mState - 1) * 6;
 111                  $tmp = $in;
 112                  $tmp = ($tmp & 0x0000003F) << $shift;
 113                  $mUcs4 |= $tmp;
 114              
 115                  /**
 116                  * End of the multi-octet sequence. mUcs4 now contains the final
 117                  * Unicode codepoint to be output
 118                  */
 119                  if (0 == --$mState) {
 120                      
 121                      /*
 122                      * Check for illegal sequences and codepoints.
 123                      */
 124                      // From Unicode 3.1, non-shortest form is illegal
 125                      if (((2 == $mBytes) && ($mUcs4 < 0x0080)) ||
 126                          ((3 == $mBytes) && ($mUcs4 < 0x0800)) ||
 127                          ((4 == $mBytes) && ($mUcs4 < 0x10000)) ||
 128                          (4 < $mBytes) ||
 129                          // From Unicode 3.2, surrogate characters are illegal
 130                          (($mUcs4 & 0xFFFFF800) == 0xD800) ||
 131                          // Codepoints outside the Unicode range are illegal
 132                          ($mUcs4 > 0x10FFFF)) {
 133                          
 134                          return FALSE;
 135                          
 136                      }
 137                      
 138                      //initialize UTF8 cache
 139                      $mState = 0;
 140                      $mUcs4  = 0;
 141                      $mBytes = 1;
 142                  }
 143              
 144              } else {
 145                  /**
 146                  *((0xC0 & (*in) != 0x80) && (mState != 0))
 147                  * Incomplete multi-octet sequence.
 148                  */
 149                  
 150                  return FALSE;
 151              }
 152          }
 153      }
 154      return TRUE;
 155  }
 156  
 157  //--------------------------------------------------------------------
 158  /**
 159  * Tests whether a string complies as UTF-8. This will be much
 160  * faster than utf8_is_valid but will pass five and six octet
 161  * UTF-8 sequences, which are not supported by Unicode and
 162  * so cannot be displayed correctly in a browser. In other words
 163  * it is not as strict as utf8_is_valid but it's faster. If you use
 164  * is to validate user input, you place yourself at the risk that
 165  * attackers will be able to inject 5 and 6 byte sequences (which
 166  * may or may not be a significant risk, depending on what you are
 167  * are doing)
 168  * @see utf8_is_valid
 169  * @see http://www.php.net/manual/en/reference.pcre.pattern.modifiers.php#54805
 170  * @param string UTF-8 string to check
 171  * @return boolean TRUE if string is valid UTF-8
 172  * @package utf8
 173  * @subpackage validation
 174  */
 175  function utf8_compliant($str) {
 176      if ( strlen($str) == 0 ) {
 177          return TRUE;
 178      }
 179      // If even just the first character can be matched, when the /u
 180      // modifier is used, then it's valid UTF-8. If the UTF-8 is somehow
 181      // invalid, nothing at all will match, even if the string contains
 182      // some valid sequences
 183      return (preg_match('/^.{1}/us',$str,$ar) == 1);
 184  }
 185