Fix partially hidden article

java/org/chromium/distiller/webdocument/DomConverter.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.webdocument;
 
 import org.chromium.distiller.DomUtil;
 import org.chromium.distiller.DomWalker;
+import org.chromium.distiller.JavaScript;
 import org.chromium.distiller.LogUtil;
 import org.chromium.distiller.TableClassifier;
 
 import com.google.gwt.dom.client.Element;
 import com.google.gwt.dom.client.Node;
 import com.google.gwt.dom.client.Style;
 import com.google.gwt.dom.client.TableElement;
 import com.google.gwt.dom.client.Text;
 import org.chromium.distiller.extractors.embeds.EmbedExtractor;
 import org.chromium.distiller.extractors.embeds.ImageExtractor;
@@ skipping 11 matching lines @@
  * elements that should not be in the created document. Some of these skipped elements (hidden
  * elements and data tables) are available for retrieval after processing.
  */
 public class DomConverter implements DomWalker.Visitor {
     private final WebDocumentBuilderInterface builder;
     private final Set<Node> hiddenElements;
     private final List<EmbedExtractor> extractors;
     // For quick lookup of tags that could possibly be embeds.
     private final HashSet<String> embedTagNames;
 
+    private boolean isMobileFriendly;
+    private Element articleElement;
+
     public DomConverter(WebDocumentBuilderInterface builder) {
-        hiddenElements = new HashSet<Node>();
+        hiddenElements = new HashSet<>();
         this.builder = builder;
 
-        extractors = new ArrayList<EmbedExtractor>();
+        extractors = new ArrayList<>();
         extractors.add(new ImageExtractor());
         extractors.add(new TwitterExtractor());
         extractors.add(new VimeoExtractor());
         extractors.add(new YouTubeExtractor());
 
         embedTagNames = new HashSet<>();
         for (EmbedExtractor extractor : extractors) {
             embedTagNames.addAll(extractor.getRelevantTagNames());
         }
     }
 
+    public void setIsMobileFriendly(boolean mobileFriendly) {
+        isMobileFriendly = mobileFriendly;
+    }
+
+    public void setArticleElement(Element article) {
+        articleElement = article;
+    }
+
     public final Set<Node> getHiddenElements() {
         return hiddenElements;
     }
 
     @Override
     public void skip(Element e) {
         builder.skipElement(e);
     }
 
     @Override
     public boolean visit(Node n) {
         switch (n.getNodeType()) {
             case Node.TEXT_NODE:
                 builder.textNode(Text.as(n));
                 return false;
             case Node.ELEMENT_NODE:
                 return visitElement(Element.as(n));
             default:
                 return false;
         }
     }
 
     private boolean visitElement(Element e) {
         // Skip invisible or uninteresting elements.
         boolean visible = DomUtil.isVisible(e);
-        logVisibilityInfo(e, visible);
+        boolean keepAnyway = false;
         if (!visible) {
+            if (isMobileFriendly && articleElement != null && JavaScript.contains(articleElement, e)
+                    && DomUtil.hasClassName(e, "hidden")) {

[mdjones, 2016/08/24 22:45:51]

Do you think it is possible/worth while to isolate this code into an extractor?
You could have it hold a reference to hiddenElements and keep track of it that
way.

+                // Process more hidden elements in a marked article in mobile-friendly pages
+                // because some sites hide the lower part of the article.
+                // See crbug.com/599121
+                keepAnyway = true;
+            }
+        }
+        logVisibilityInfo(e, visible || keepAnyway);
+        if (!visible && !keepAnyway) {
             hiddenElements.add(e);
             return false;
         }
 
         // Node-type specific extractors check for elements they are interested in here. Everything
         // else will be filtered through the switch below.
 
         // Check for embedded elements that might be extracted.
         if (embedTagNames.contains(e.getTagName())) {
             // If the tag is marked as interesting, check the extractors.
@@ skipping 78 matching lines @@
         Element parent = e.getParentElement();
         LogUtil.logToConsole("TABLE: " + type +
                 ", id=" + e.getId() +
                 ", class=" + e.getClassName() +
                 ", parent=[" + parent.getTagName() +
                 ", id=" + parent.getId() +
                 ", class=" + parent.getClassName() +
                 "]");
     }
 }