[project @ Make yadis META tag search ignore comments]

tailor · tailor · commit b8d3eff26808 · 2007-04-26T18:19:11.000Z
diff --git a/Auth/Yadis/ParseHTML.php b/Auth/Yadis/ParseHTML.php
@@ -27,6 +27,12 @@ class Auth_Yadis_ParseHTML {
      */
     var $_re_flags = "si";
 
+    /**
+     * @access private
+     */
+    var $_removed_re =
+           "<!--.*?-->|<!\[CDATA\[.*?\]\]>|<script\b(?!:)[^>]*>.*?<\/script>";
+
     /**
      * @access private
      */
@@ -43,6 +49,10 @@ function Auth_Yadis_ParseHTML()
                                     $this->_attr_find,
                                     $this->_re_flags);
 
+        $this->_removed_re = sprintf("/%s/%s",
+                                     $this->_removed_re,
+                                     $this->_re_flags);
+
         $this->_entity_replacements = array(
                                             'amp' => '&',
                                             'lt' => '<',
@@ -146,6 +156,10 @@ function tagPattern($tag_names, $close, $self_close)
      */
     function getMetaTags($html_string)
     {
+        $html_string = preg_replace($this->_removed_re,
+                                    "",
+                                    $html_string);
+
         $key_tags = array($this->tagPattern('html', false, false),
                           $this->tagPattern('head', false, false),
                           $this->tagPattern('head', true, false),
diff --git a/Tests/Auth/Yadis/data/test1-parsehtml.txt b/Tests/Auth/Yadis/data/test1-parsehtml.txt
@@ -96,6 +96,14 @@ None
 <html><head><body><meta http-equiv="X-XRDS-Location" content="found">
 
 None
+<!-- <meta> is inside comment -->
+<html>
+  <head>
+    <!--<meta http-equiv="X-XRDS-Location" content="found">-->
+  </head>
+</html>
+
+None
 <!-- <meta> is inside of <body> -->
 <html>
   <head>
