Fix for keeping lists structure

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.LogUtil;
 import org.chromium.distiller.TableClassifier;
@@ skipping 82 matching lines @@
             for (EmbedExtractor extractor : extractors) {
                 WebElement embed = extractor.extract(e);
                 if (embed != null) {
                     builder.embed(embed);
                     return false;
                 }
             }
         }
 
         switch (e.getTagName()) {
+            case "OL":
+            case "UL":
+            case "LI":
+                builder.tag(new WebTag(e.getTagName(), WebTag.TagType.START));

[wychen, 2015/08/11 00:12:15]

We could move this out of the switch and use canBeNested() instead.

[Marcelo Correa, 2015/08/11 00:21:01]

Sure, But this wouldn't change the pattern how the elements are filtered?
I mean, excluding the embed extractor and hidden elements, everything else is
filtered by the switch.

+                break;
             case "BR":
                 builder.lineBreak(e);
                 return false;
             // Skip data tables, keep track of them to be extracted by RelevantElementsFinder
             // later.
             case "TABLE":
                 TableClassifier.Type type = TableClassifier.table(TableElement.as(e));
                 logTableInfo(e, type);
                 if (type == TableClassifier.Type.DATA) {
                     builder.dataTable(e);
@@ skipping 22 matching lines @@
             case "LINK":
             case "NOSCRIPT":
                 return false;
         }
         builder.startElement(e);
         return true;
     }
 
     @Override
     public void exit(Node n) {
+        if (n.getNodeType() == Node.ELEMENT_NODE) {
+            Element e = Element.as(n);
+            switch (e.getTagName()) {
+                case "OL":
+                case "UL":
+                case "LI":
+                    builder.tag(new WebTag(e.getTagName(), WebTag.TagType.END));

[wychen, 2015/08/11 00:12:15]

And here.

+                    break;
+            }
+        }
         builder.endElement();
     }
 
     private static void logVisibilityInfo(Element e, boolean visible) {
         if (!LogUtil.isLoggable(LogUtil.DEBUG_LEVEL_VISIBILITY_INFO)) return;
         Style style = DomUtil.getComputedStyle(e);
         LogUtil.logToConsole((visible ? "KEEP " : "SKIP ") + e.getTagName() +
                 ": id=" + e.getId() +
                 ", dsp=" + style.getDisplay() +
                 ", vis=" + style.getVisibility() +
                 ", opaq=" + style.getOpacity());
     }
 
     private static void logTableInfo(Element e, TableClassifier.Type type) {
         if (!LogUtil.isLoggable(LogUtil.DEBUG_LEVEL_VISIBILITY_INFO)) return;
         Element parent = e.getParentElement();
         LogUtil.logToConsole("TABLE: " + type +
                 ", id=" + e.getId() +
                 ", class=" + e.getClassName() +
                 ", parent=[" + parent.getTagName() +
                 ", id=" + parent.getId() +
                 ", class=" + parent.getClassName() +
                 "]");
     }
 }