L-electron-Rare/moodle
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

MDL-14898 upgraded html purifier to 3.1.0 in HEAD

Browse Source
This commit is contained in:
skodak
2008-06-16 16:39:09 +00:00
parent b553527e99
commit eb203ee412
290 changed files with 4780 additions and 3432 deletions
Download Patch File Download Diff File Expand all files Collapse all files
lib/htmlpurifier/HTMLPurifier/Lexer/DOMLex.php
+9 -11
View File
@@ -1,8 +1,5 @@
<?php
require_once 'HTMLPurifier/Lexer.php';
require_once 'HTMLPurifier/TokenFactory.php';
/**
* Parser that uses PHP 5's DOM extension (part of the core).
*
@@ -34,11 +31,11 @@ class HTMLPurifier_Lexer_DOMLex extends HTMLPurifier_Lexer
public function __construct() {
// setup the factory
parent::HTMLPurifier_Lexer();
parent::__construct();
$this->factory = new HTMLPurifier_TokenFactory();
}
public function tokenizeHTML($html, $config, &$context) {
public function tokenizeHTML($html, $config, $context) {
$html = $this->normalize($html, $config, $context);
@@ -47,9 +44,9 @@ class HTMLPurifier_Lexer_DOMLex extends HTMLPurifier_Lexer
if ($config->get('Core', 'AggressivelyFixLt')) {
$char = '[^a-z!\/]';
$comment = "/<!--(.*?)(-->|\z)/is";
$html = preg_replace_callback($comment, array('HTMLPurifier_Lexer_DOMLex', 'callbackArmorCommentEntities'), $html);
$html = preg_replace_callback($comment, array($this, 'callbackArmorCommentEntities'), $html);
$html = preg_replace("/<($char)/i", '&lt;\\1', $html);
$html = preg_replace_callback($comment, array('HTMLPurifier_Lexer_DOMLex', 'callbackUndoCommentSubst'), $html); // fix comments
$html = preg_replace_callback($comment, array($this, 'callbackUndoCommentSubst'), $html); // fix comments
}
// preprocess html, essential for UTF-8
@@ -94,7 +91,7 @@ class HTMLPurifier_Lexer_DOMLex extends HTMLPurifier_Lexer
$last = end($tokens);
$data = $node->data;
// (note $node->tagname is already normalized)
if ($last instanceof HTMLPurifier_Token_Start && $last->name == 'script') {
if ($last instanceof HTMLPurifier_Token_Start && ($last->name == 'script' || $last->name == 'style')) {
$new_data = trim($data);
if (substr($new_data, 0, 4) === '<!--') {
$data = substr($new_data, 4);
@@ -175,7 +172,7 @@ class HTMLPurifier_Lexer_DOMLex extends HTMLPurifier_Lexer
* Callback function for undoing escaping of stray angled brackets
* in comments
*/
function callbackUndoCommentSubst($matches) {
public function callbackUndoCommentSubst($matches) {
return '<!--' . strtr($matches[1], array('&amp;'=>'&','&lt;'=>'<')) . $matches[2];
}
@@ -183,14 +180,14 @@ class HTMLPurifier_Lexer_DOMLex extends HTMLPurifier_Lexer
* Callback function that entity-izes ampersands in comments so that
* callbackUndoCommentSubst doesn't clobber them
*/
function callbackArmorCommentEntities($matches) {
public function callbackArmorCommentEntities($matches) {
return '<!--' . str_replace('&', '&amp;', $matches[1]) . $matches[2];
}
/**
* Wraps an HTML fragment in the necessary HTML
*/
function wrapHTML($html, $config, &$context) {
protected function wrapHTML($html, $config, $context) {
$def = $config->getDefinition('HTML');
$ret = '';
@@ -203,6 +200,7 @@ class HTMLPurifier_Lexer_DOMLex extends HTMLPurifier_Lexer
$ret .= '<html><head>';
$ret .= '<meta http-equiv="Content-Type" content="text/html; charset=utf-8" />';
// No protection if $html contains a stray </div>!
$ret .= '</head><body><div>'.$html.'</div></body></html>';
return $ret;
}