[2.0.1] Improve special case handling for <script>

- DirectLex now honors comments with greater than or less than signs in them
- Comments are transformed into script elements, ending comments are scrapped
- Buggy generator code rewritten to be more error-proof
- AttrValidator checks if token has attributes before processing
- Remove invalid documentation from Scripting
- "Commenting" of script elements switched to the more advanced version

git-svn-id: http://htmlpurifier.org/svnroot/htmlpurifier/trunk@1189 48356398-32a2-884e-a903-53898d9a118a

library/HTMLPurifier/Lexer.php

@@ -259,7 +259,19 @@ class HTMLPurifier_Lexer
      */
     function escapeCDATA($string) {
         return preg_replace_callback(
-            '/<!\[CDATA\[(.+?)\]\]>/',
+            '/<!\[CDATA\[(.+?)\]\]>/s',
+            array('HTMLPurifier_Lexer', 'CDATACallback'),
+            $string
+        );
+    }
+    
+    /**
+     * Special CDATA case that is especiall convoluted for <script>
+     */
+    function escapeCommentedCDATA($string) {
+        // <!--//--><![CDATA[//><!--
+        return preg_replace_callback(
+            '#<!--//--><!\[CDATA\[//><!--(.+?)//--><!\]\]>#s',
             array('HTMLPurifier_Lexer', 'CDATACallback'),
             $string
         );
@@ -291,6 +303,11 @@ class HTMLPurifier_Lexer
             $html = $this->extractBody($html);
         }
         
+        if ($config->get('HTML', 'Trusted')) {
+            // escape convoluted CDATA
+            $html = $this->escapeCommentedCDATA($html);
+        }
+        
         // escape CDATA
         $html = $this->escapeCDATA($html);