View Javadoc
1   /*
2    * Licensed to the Apache Software Foundation (ASF) under one or more
3    * contributor license agreements.  See the NOTICE file distributed with
4    * this work for additional information regarding copyright ownership.
5    * The ASF licenses this file to You under the Apache License, Version 2.0
6    * (the "License"); you may not use this file except in compliance with
7    * the License.  You may obtain a copy of the License at
8    *
9    *  http://www.apache.org/licenses/LICENSE-2.0
10   *
11   * Unless required by applicable law or agreed to in writing, software
12   * distributed under the License is distributed on an "AS IS" BASIS,
13   * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
14   * See the License for the specific language governing permissions and
15   * limitations under the License.
16   */
17  
18  package org.apache.any23.extractor.html;
19  
20  import java.util.List;
21  
22  import org.apache.any23.extractor.ExtractionException;
23  import org.apache.any23.extractor.ExtractionResult;
24  import org.apache.any23.extractor.ExtractorDescription;
25  import org.apache.any23.extractor.TagSoupExtractionResult;
26  import org.apache.any23.extractor.html.HTMLDocument.TextField;
27  import org.apache.any23.vocab.Review;
28  import org.apache.any23.vocab.ReviewAggregate;
29  import org.apache.any23.vocab.VCard;
30  import org.eclipse.rdf4j.model.BNode;
31  import org.eclipse.rdf4j.model.Resource;
32  import org.eclipse.rdf4j.model.vocabulary.RDF;
33  import org.w3c.dom.Node;
34  
35  /**
36   * Extractor for the <a href="http://microformats.org/wiki/hreview-aggregate">hReview-aggregate</a> microformat.
37   * 
38   * @author Sebastien Richard
39   */
40  public class HReviewAggregateExtractor extends EntityBasedMicroformatExtractor {
41      private static final Review vREVIEW = Review.getInstance();
42      private static final ReviewAggregate vREVIEWAGG = ReviewAggregate.getInstance();
43      private static final VCard vVCARD = VCard.getInstance();
44  
45      @Override
46      public ExtractorDescription getDescription() {
47          return HReviewAggregateExtractorFactory.getDescriptionInstance();
48      }
49  
50      @Override
51      protected String getBaseClassName() {
52          return "hreview-aggregate";
53      }
54  
55      @Override
56      protected void resetExtractor() {
57          // Empty.
58      }
59  
60      @Override
61      protected boolean extractEntity(Node node, ExtractionResult out) throws ExtractionException {
62          BNode rev = getBlankNodeFor(node);
63          out.writeTriple(rev, RDF.TYPE, vREVIEWAGG.ReviewAggregate);
64          final HTMLDocumentTMLDocument.html#HTMLDocument">HTMLDocument fragment = new HTMLDocument(node);
65          addRating(fragment, rev);
66          addWorst(fragment, rev);
67          addBest(fragment, rev);
68          addAverage(fragment, rev);
69          addSummary(fragment, rev);
70          addType(fragment, rev);
71          addItem(fragment, rev);
72          addCount(fragment, rev);
73          addVotes(fragment, rev);
74  
75          final TagSoupExtractionResultorg/apache/any23/extractor/TagSoupExtractionResult.html#TagSoupExtractionResult">TagSoupExtractionResult tser = (TagSoupExtractionResult) out;
76          tser.addResourceRoot(DomUtils.getXPathListForNode(node), rev, this.getClass());
77  
78          return true;
79      }
80  
81      private void addType(HTMLDocument doc, Resource rev) {
82          TextField value = doc.getSingularTextField("type");
83          conditionallyAddStringProperty(value.source(), rev, vREVIEW.type, value.value());
84      }
85  
86      private void addItem(HTMLDocument root, BNode rev) throws ExtractionException {
87          List<Node> nodes = root.findAllByClassName("item");
88          for (Node node : nodes) {
89              Resource item = findDummy(new HTMLDocument(node));
90              addBNodeProperty(node, item, vREVIEW.hasReview, rev);
91          }
92      }
93  
94      private Resource findDummy(HTMLDocument item) throws ExtractionException {
95          Resource blank = getBlankNodeFor(item.getDocument());
96          TextField val = item.getSingularTextField("fn");
97          conditionallyAddStringProperty(val.source(), blank, vVCARD.fn, val.value());
98          final TextField url = item.getSingularUrlField("url");
99          conditionallyAddResourceProperty(blank, vVCARD.url, getHTMLDocument().resolveIRI(url.value()));
100         TextField pics[] = item.getPluralUrlField("photo");
101         for (TextField pic : pics) {
102             addIRIProperty(blank, vVCARD.photo, getHTMLDocument().resolveIRI(pic.value()));
103         }
104         return blank;
105     }
106 
107     private void addRating(HTMLDocument doc, Resource rev) {
108         HTMLDocument.TextField value = doc.getSingularTextField("rating");
109         conditionallyAddStringProperty(value.source(), rev, vREVIEW.rating, value.value());
110     }
111 
112     private void addWorst(HTMLDocument doc, Resource rev) {
113         HTMLDocument.TextField value = doc.getSingularTextField("worst");
114         conditionallyAddStringProperty(value.source(), rev, vREVIEWAGG.worst, value.value());
115     }
116 
117     private void addBest(HTMLDocument doc, Resource rev) {
118         HTMLDocument.TextField value = doc.getSingularTextField("best");
119         conditionallyAddStringProperty(value.source(), rev, vREVIEWAGG.best, value.value());
120     }
121 
122     private void addAverage(HTMLDocument doc, Resource rev) {
123         HTMLDocument.TextField value = doc.getSingularTextField("average");
124         conditionallyAddStringProperty(value.source(), rev, vREVIEWAGG.average, value.value());
125     }
126 
127     private void addCount(HTMLDocument doc, Resource rev) {
128         HTMLDocument.TextField value = doc.getSingularTextField("count");
129         conditionallyAddStringProperty(value.source(), rev, vREVIEWAGG.count, value.value());
130     }
131 
132     private void addVotes(HTMLDocument doc, Resource rev) {
133         HTMLDocument.TextField value = doc.getSingularTextField("votes");
134         conditionallyAddStringProperty(value.source(), rev, vREVIEWAGG.votes, value.value());
135     }
136 
137     private void addSummary(HTMLDocument doc, Resource rev) {
138         TextField value = doc.getSingularTextField("summary");
139         conditionallyAddStringProperty(value.source(), rev, vREVIEW.title, value.value());
140     }
141 }