get('Core', 'AcceptFullDocuments')) {
$is_full = $this->extractBody($string, true);
}
$doc = new DOMDocument();
$doc->encoding = 'UTF-8'; // technically does nothing, but whatever
// replace and escape the CDATA sections, since parsing under HTML
// mode won't get 'em.
$string = $this->escapeCDATA($string);
if (!$is_full) {
// preprocess string, essential for UTF-8
$string =
''.
''.
''.$string.'';
}
@$doc->loadHTML($string); // mute all errors, handle it transparently
return $this->tokenizeDOM(
$doc->childNodes->item(1)-> // html
getElementsByTagName('body')->item(0) // body
);
}
/**
* Recursive function that tokenizes a node, putting it into an accumulator.
*
* @param $node DOMNode to be tokenized.
* @param $tokens Array-list of already tokenized tokens.
* @param $collect Says whether or start and close are collected, set to
* false at first recursion because it's the implicit DIV
* tag you're dealing with.
* @returns Tokens of node appended to previously passed tokens.
*/
protected function tokenizeDOM($node, $tokens = array(), $collect = false) {
// recursive goodness!
// intercept non element nodes
if ( !($node instanceof DOMElement) ) {
if ($node instanceof DOMComment) {
$tokens[] = new HTMLPurifier_Token_Comment($node->data);
} elseif ($node instanceof DOMText ||
$node instanceof DOMCharacterData) {
$tokens[] = new HTMLPurifier_Token_Text($node->data);
}
// quite possibly, the object wasn't handled, that's fine
return $tokens;
}
// We still have to make sure that the element actually IS empty
if (!$node->hasChildNodes()) {
if ($collect) {
$tokens[] = new HTMLPurifier_Token_Empty(
$node->tagName,
$this->transformAttrToAssoc($node->attributes)
);
}
} else {
if ($collect) { // don't wrap on first iteration
$tokens[] = new HTMLPurifier_Token_Start(
$tag_name = $node->tagName, // somehow, it get's dropped
$this->transformAttrToAssoc($node->attributes)
);
}
foreach ($node->childNodes as $node) {
// remember, it's an accumulator. Otherwise, we'd have
// to use array_merge
$tokens = $this->tokenizeDOM($node, $tokens, true);
}
if ($collect) {
$tokens[] = new HTMLPurifier_Token_End($tag_name);
}
}
return $tokens;
}
/**
* Converts a DOMNamedNodeMap of DOMAttr objects into an assoc array.
*
* @param $attribute_list DOMNamedNodeMap of DOMAttr objects.
* @returns Associative array of attributes.
*/
protected function transformAttrToAssoc($attribute_list) {
$attribute_array = array();
// undocumented behavior
foreach ($attribute_list as $key => $attr) {
$attribute_array[$key] = $attr->value;
}
return $attribute_array;
}
}
?>