Add support for Schema.org/Recipe

java/org/chromium/distiller/SchemaOrgParserAccessor.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 org.chromium.distiller.proto.DomDistillerProtos.TimingInfo;
 
 import com.google.gwt.dom.client.Element;
 
 import java.util.ArrayList;
 import java.util.Collections;
 import java.util.Comparator;
 import java.util.List;
 
 /**
  * This class instantiates SchemaOrgParser and implements MarkupParser.Accessor interface to provide
  * access to properties that SchemaOrgParser has parsed.
  */
 public class SchemaOrgParserAccessor implements MarkupParser.Accessor {
 
+    private static List<SchemaOrgParser.Type> supportedTypes;
     private SchemaOrgParser mParser;
     private final Element mRoot;
     private final TimingInfo mTimingInfo;
+    private static final String ENGLISH_LANGUAGE = "en";
 
     /**
      * The object that instantiates SchemaOrgParser and implements its MarkupParser.Accessor
      * interface.
      */
     public SchemaOrgParserAccessor(Element root) {
         this(root, (TimingInfo) null);
     }
 
     public SchemaOrgParserAccessor(Element root, TimingInfo timingInfo) {
         mRoot = root;
         mTimingInfo = timingInfo;
     }
 
+    static {
+        supportedTypes = new ArrayList<>();
+        supportedTypes.add(SchemaOrgParser.Type.RECIPE);
+    }
+
     private void init() {
         if (mParser == null) {
             mParser = new SchemaOrgParser(mRoot, mTimingInfo);
         }
     }
 
     @Override
     public String getTitle() {
         init();
         String title = "";
@@ skipping 128 matching lines @@
         List<SchemaOrgParser.ArticleItem> articles = mParser.getArticleItems();
         return articles.isEmpty() ? null : articles.get(0).getArticle();
     }
 
     @Override
     public boolean optOut() {
         return false;
     }
 
     /**
+     * Get schema.org parsed content.
+     *
+     * Content is generated when a main entity is identified or
+     * has only one entity item parsed and it is supported
+     *
+     * @return HTML output from the supported item.
+     */
+    @Override
+    public String getStructuredData() {
+        init();

[wychen, 2016/05/31 21:56:34]

Only init() when the page is in English.

[dalmirsilva, 2016/07/06 17:53:34]

Done.

+        String output = "";
+        if (DomUtil.getLanguage(mRoot).contains(ENGLISH_LANGUAGE)) {

[wychen, 2016/05/31 21:56:34]

Strictly speaking, the lang code should "start with" en, not just containing it.

[dalmirsilva, 2016/07/06 17:53:34]

Done.

+            SchemaOrgParser.ThingItem entity = mParser.getMainEntity();
+            List<SchemaOrgParser.ThingItem> topLevelEntities =
+                    mParser.getTopLevelEntities();
+            if (entity == null && topLevelEntities.size() == 1) {
+                entity = topLevelEntities.get(0);
+            }
+            if (entity != null && supportedTypes.contains(entity.getType())) {
+                output = entity.generateOutput();
+            }
+        }
+        return output;
+    }
+
+    /**
      * Sort a list of {@link SchemaOrgParser.ArticleItem}s by their area
      * in descending order.
      * @param articles List of {@link SchemaOrgParser.ArticleItem}s to
      * be sorted.
      * @return The sorted list.
      */
     private List<SchemaOrgParser.ArticleItem> sortArticlesByArea(
             List<SchemaOrgParser.ArticleItem> articles) {
         List<SchemaOrgParser.ArticleItem> sortedArticles =
                 new ArrayList<>(articles);
         Collections.sort(sortedArticles,
                 new Comparator<SchemaOrgParser.ArticleItem>() {
                     @Override
                     public int compare(SchemaOrgParser.ArticleItem i1,
                                        SchemaOrgParser.ArticleItem i2) {
                         int area1 = DomUtil.getArea(i1.getElement());
                         int area2 = DomUtil.getArea(i2.getElement());
                         if (area1 > area2) {
                             return -1;
                         } else if (area1 < area2) {
                             return 1;
                         }
                         return 0;
                     }
                 });
         return sortedArticles;
     }
 }