Strip unwanted classNames from all nodes

java/org/chromium/distiller/DomUtil.java

 // Copyright 2014 The Chromium Authors. All rights reserved.
 // Use of this source code is governed by a BSD-style license that can be
 // found in the LICENSE file.
 
 package org.chromium.distiller;
 
 import com.google.gwt.core.client.JsArray;
 import com.google.gwt.core.client.JsArrayString;
 import com.google.gwt.dom.client.AnchorElement;
 import com.google.gwt.dom.client.Document;
@@ skipping 400 matching lines @@
         for (var key in elementsWithSrc) {
             if (elementsWithSrc[key].src) {
                 elementsWithSrc[key].src = elementsWithSrc[key].src;
             }
         }
     }-*/;
 
     /**
      * Strips some attribute from certain tags in the tree rooted at |rootNode|, including root.
      * @param tagNames The tag names to be processed. ["*"] means all.
+     * TODO(crbug.com/654108): We should convert to whitelisting for all the tags.
      */
     @SuppressWarnings("unused")
     public static void stripAttributeFromTags(Node rootNode, String attribute, String[] tagNames) {
         Element root = Element.as(rootNode);
         for (String tag: tagNames) {
             if (root.getTagName().equals(tag) || tag.equals("*")) {
                 root.removeAttribute(attribute);
             }
         }
 
@@ skipping 36 matching lines @@
     }
 
     /**
      * Strips all "target" attributes from anchor nodes in the tree rooted at |rootNode|
      */
     public static void stripTargetAttributes(Node rootNode) {
         stripAttributeFromTags(rootNode, "TARGET", new String[]{"A"});
     }
 
     /**
+     * Strips unwanted classNames from all nodes in the tree rooted at |root|.
+     * TODO(crbug.com/654109): "caption" is essential for styling, but all classNames should
+     *                         be removed eventually.
+     */
+    public static void stripUnwantedClassNames(Node root) {
+        if (root.getNodeType() == Node.ELEMENT_NODE) {
+            Element element = Element.as(root);
+            if (element.hasAttribute("class")) {
+                stripUnwantedClassName(element);
+            }
+        }
+        NodeList<Element> elems = DomUtil.querySelectorAll(root, "[class]");
+        for (int i = 0; i < elems.getLength(); i++) {
+            stripUnwantedClassName(elems.getItem(i));
+        }
+    }
+
+    private static void stripUnwantedClassName(Element elem) {
+        if (elem.getClassName().contains("caption")) {
+            elem.setClassName("caption");
+        } else {
+            elem.removeAttribute("class");
+        }
+    }
+
+    /**
      * Get a list of relevant nodes from a subtree.
      * @param root The root of the subtree.
      * @return A list of relevant nodes.
      */
     public static List<Node> getOutputNodes(Node root) {
         final List<Node> nodes = new ArrayList<>();
         new DomWalker(new DomWalker.Visitor() {
             @Override
             public boolean visit(Node n) {
                 switch (n.getNodeType()) {
@@ skipping 49 matching lines @@
     }-*/;
 
     public static native Document createHTMLDocument(Document doc) /*-{
         return doc.implementation.createHTMLDocument();
     }-*/;
 
     public static native Element getFirstElementChild(Document document) /*-{
         return document.firstElementChild;
     }-*/;
 }