lib/htmlpurifier/HTMLPurifier/Lexer/DOMLex.php

   1 <?php
   2
   3 require_once 'HTMLPurifier/Lexer.php';
   4 require_once 'HTMLPurifier/TokenFactory.php';
   5
   6 /**
   7  * Parser that uses PHP 5's DOM extension (part of the core).
   8  *
   9  * In PHP 5, the DOM XML extension was revamped into DOM and added to the core.
  10  * It gives us a forgiving HTML parser, which we use to transform the HTML
  11  * into a DOM, and then into the tokens.  It is blazingly fast (for large
  12  * documents, it performs twenty times faster than
  13  * HTMLPurifier_Lexer_DirectLex,and is the default choice for PHP 5.
  14  *
  15  * @note Any empty elements will have empty tokens associated with them, even if
  16  * this is prohibited by the spec. This is cannot be fixed until the spec
  17  * comes into play.
  18  *
  19  * @note PHP's DOM extension does not actually parse any entities, we use
  20  *       our own function to do that.
  21  *
  22  * @warning DOM tends to drop whitespace, which may wreak havoc on indenting.
  23  *          If this is a huge problem, due to the fact that HTML is hand
  24  *          edited and you are unable to get a parser cache that caches the
  25  *          the output of HTML Purifier while keeping the original HTML lying
  26  *          around, you may want to run Tidy on the resulting output or use
  27  *          HTMLPurifier_DirectLex
  28  */
  29
  30 class HTMLPurifier_Lexer_DOMLex extends HTMLPurifier_Lexer
  31 {
  32
  33     private $factory;
  34
  35     public function __construct() {
  36         // setup the factory
  37         parent::HTMLPurifier_Lexer();
  38         $this->factory = new HTMLPurifier_TokenFactory();
  39     }
  40
  41     public function tokenizeHTML($string, $config, &$context) {
  42
  43         $string = $this->normalize($string, $config, $context);
  44
  45         // preprocess string, essential for UTF-8
  46         $string =
  47             '<!DOCTYPE html '.
  48                 'PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"'.
  49                 '"http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">'.
  50             '<html><head>'.
  51             '<meta http-equiv="Content-Type" content="text/html;'.
  52                 ' charset=utf-8" />'.
  53             '</head><body><div>'.$string.'</div></body></html>';
  54
  55         $doc = new DOMDocument();
  56         $doc->encoding = 'UTF-8'; // technically does nothing, but whatever
  57
  58         // DOM will toss errors if the HTML its parsing has really big
  59         // problems, so we're going to mute them. This can cause problems
  60         // if a custom error handler that doesn't implement error_reporting
  61         // is set, as noted by a Drupal plugin of HTML Purifier. Consider
  62         // making our own error reporter to temporarily load in
  63         @$doc->loadHTML($string);
  64
  65         $tokens = array();
  66         $this->tokenizeDOM(
  67             $doc->getElementsByTagName('html')->item(0)-> // html
  68                   getElementsByTagName('body')->item(0)-> // body
  69                   getElementsByTagName('div')->item(0) // div
  70             , $tokens);
  71         return $tokens;
  72     }
  73
  74     /**
  75      * Recursive function that tokenizes a node, putting it into an accumulator.
  76      *
  77      * @param $node     DOMNode to be tokenized.
  78      * @param $tokens   Array-list of already tokenized tokens.
  79      * @param $collect  Says whether or start and close are collected, set to
  80      *                  false at first recursion because it's the implicit DIV
  81      *                  tag you're dealing with.
  82      * @returns Tokens of node appended to previously passed tokens.
  83      */
  84     protected function tokenizeDOM($node, &$tokens, $collect = false) {
  85         // recursive goodness!
  86
  87         // intercept non element nodes. WE MUST catch all of them,
  88         // but we're not getting the character reference nodes because
  89         // those should have been preprocessed
  90         if ($node->nodeType === XML_TEXT_NODE ||
  91                   $node->nodeType === XML_CDATA_SECTION_NODE) {
  92             $tokens[] = $this->factory->createText($node->data);
  93             return;
  94         } elseif ($node->nodeType === XML_COMMENT_NODE) {
  95             $tokens[] = $this->factory->createComment($node->data);
  96             return;
  97         } elseif (
  98             // not-well tested: there may be other nodes we have to grab
  99             $node->nodeType !== XML_ELEMENT_NODE
 100         ) {
 101             return;
 102         }
 103
 104         $attr = $node->hasAttributes() ?
 105             $this->transformAttrToAssoc($node->attributes) :
 106             array();
 107
 108         // We still have to make sure that the element actually IS empty
 109         if (!$node->childNodes->length) {
 110             if ($collect) {
 111                 $tokens[] = $this->factory->createEmpty($node->tagName, $attr);
 112             }
 113         } else {
 114             if ($collect) { // don't wrap on first iteration
 115                 $tokens[] = $this->factory->createStart(
 116                     $tag_name = $node->tagName, // somehow, it get's dropped
 117                     $attr
 118                 );
 119             }
 120             foreach ($node->childNodes as $node) {
 121                 // remember, it's an accumulator. Otherwise, we'd have
 122                 // to use array_merge
 123                 $this->tokenizeDOM($node, $tokens, true);
 124             }
 125             if ($collect) {
 126                 $tokens[] = $this->factory->createEnd($tag_name);
 127             }
 128         }
 129
 130     }
 131
 132     /**
 133      * Converts a DOMNamedNodeMap of DOMAttr objects into an assoc array.
 134      *
 135      * @param $attribute_list DOMNamedNodeMap of DOMAttr objects.
 136      * @returns Associative array of attributes.
 137      */
 138     protected function transformAttrToAssoc($node_map) {
 139         // NamedNodeMap is documented very well, so we're using undocumented
 140         // features, namely, the fact that it implements Iterator and
 141         // has a ->length attribute
 142         if ($node_map->length === 0) return array();
 143         $array = array();
 144         foreach ($node_map as $attr) {
 145             $array[$attr->name] = $attr->value;
 146         }
 147         return $array;
 148     }
 149
 150 }
 151
 152 ?>