lib/htmlpurifier/HTMLPurifier/EntityParser.php

   1 <?php
   2
   3 // if want to implement error collecting here, we'll need to use some sort
   4 // of global data (probably trigger_error) because it's impossible to pass
   5 // $config or $context to the callback functions.
   6
   7 /**
   8  * Handles referencing and derefencing character entities
   9  */
  10 class HTMLPurifier_EntityParser
  11 {
  12
  13     /**
  14      * Reference to entity lookup table.
  15      * @type HTMLPurifier_EntityLookup
  16      */
  17     protected $_entity_lookup;
  18
  19     /**
  20      * Callback regex string for parsing entities.
  21      * @type string
  22      */
  23     protected $_substituteEntitiesRegex =
  24         '/&(?:[#]x([a-fA-F0-9]+)|[#]0*(\d+)|([A-Za-z_:][A-Za-z0-9.\-_:]*));?/';
  25         //     1. hex             2. dec      3. string (XML style)
  26
  27     /**
  28      * Decimal to parsed string conversion table for special entities.
  29      * @type array
  30      */
  31     protected $_special_dec2str =
  32             array(
  33                     34 => '"',
  34                     38 => '&',
  35                     39 => "'",
  36                     60 => '<',
  37                     62 => '>'
  38             );
  39
  40     /**
  41      * Stripped entity names to decimal conversion table for special entities.
  42      * @type array
  43      */
  44     protected $_special_ent2dec =
  45             array(
  46                     'quot' => 34,
  47                     'amp'  => 38,
  48                     'lt'   => 60,
  49                     'gt'   => 62
  50             );
  51
  52     /**
  53      * Substitutes non-special entities with their parsed equivalents. Since
  54      * running this whenever you have parsed character is t3h 5uck, we run
  55      * it before everything else.
  56      *
  57      * @param string $string String to have non-special entities parsed.
  58      * @return string Parsed string.
  59      */
  60     public function substituteNonSpecialEntities($string)
  61     {
  62         // it will try to detect missing semicolons, but don't rely on it
  63         return preg_replace_callback(
  64             $this->_substituteEntitiesRegex,
  65             array($this, 'nonSpecialEntityCallback'),
  66             $string
  67         );
  68     }
  69
  70     /**
  71      * Callback function for substituteNonSpecialEntities() that does the work.
  72      *
  73      * @param array $matches  PCRE matches array, with 0 the entire match, and
  74      *                  either index 1, 2 or 3 set with a hex value, dec value,
  75      *                  or string (respectively).
  76      * @return string Replacement string.
  77      */
  78
  79     protected function nonSpecialEntityCallback($matches)
  80     {
  81         // replaces all but big five
  82         $entity = $matches[0];
  83         $is_num = (@$matches[0][1] === '#');
  84         if ($is_num) {
  85             $is_hex = (@$entity[2] === 'x');
  86             $code = $is_hex ? hexdec($matches[1]) : (int) $matches[2];
  87             // abort for special characters
  88             if (isset($this->_special_dec2str[$code])) {
  89                 return $entity;
  90             }
  91             return HTMLPurifier_Encoder::unichr($code);
  92         } else {
  93             if (isset($this->_special_ent2dec[$matches[3]])) {
  94                 return $entity;
  95             }
  96             if (!$this->_entity_lookup) {
  97                 $this->_entity_lookup = HTMLPurifier_EntityLookup::instance();
  98             }
  99             if (isset($this->_entity_lookup->table[$matches[3]])) {
 100                 return $this->_entity_lookup->table[$matches[3]];
 101             } else {
 102                 return $entity;
 103             }
 104         }
 105     }
 106
 107     /**
 108      * Substitutes only special entities with their parsed equivalents.
 109      *
 110      * @notice We try to avoid calling this function because otherwise, it
 111      * would have to be called a lot (for every parsed section).
 112      *
 113      * @param string $string String to have non-special entities parsed.
 114      * @return string Parsed string.
 115      */
 116     public function substituteSpecialEntities($string)
 117     {
 118         return preg_replace_callback(
 119             $this->_substituteEntitiesRegex,
 120             array($this, 'specialEntityCallback'),
 121             $string
 122         );
 123     }
 124
 125     /**
 126      * Callback function for substituteSpecialEntities() that does the work.
 127      *
 128      * This callback has same syntax as nonSpecialEntityCallback().
 129      *
 130      * @param array $matches  PCRE-style matches array, with 0 the entire match, and
 131      *                  either index 1, 2 or 3 set with a hex value, dec value,
 132      *                  or string (respectively).
 133      * @return string Replacement string.
 134      */
 135     protected function specialEntityCallback($matches)
 136     {
 137         $entity = $matches[0];
 138         $is_num = (@$matches[0][1] === '#');
 139         if ($is_num) {
 140             $is_hex = (@$entity[2] === 'x');
 141             $int = $is_hex ? hexdec($matches[1]) : (int) $matches[2];
 142             return isset($this->_special_dec2str[$int]) ?
 143                 $this->_special_dec2str[$int] :
 144                 $entity;
 145         } else {
 146             return isset($this->_special_ent2dec[$matches[3]]) ?
 147                 $this->_special_ent2dec[$matches[3]] :
 148                 $entity;
 149         }
 150     }
 151 }
 152
 153 // vim: et sw=4 sts=4