001/** 
002 * Copyright (c) 2007-2008, Regents of the University of Colorado 
003 * All rights reserved.
004 * 
005 * Redistribution and use in source and binary forms, with or without
006 * modification, are permitted provided that the following conditions are met:
007 * 
008 * Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer. 
009 * Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution. 
010 * Neither the name of the University of Colorado at Boulder nor the names of its contributors may be used to endorse or promote products derived from this software without specific prior written permission. 
011 * 
012 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
013 * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
014 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
015 * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
016 * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
017 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
018 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
019 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
020 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
021 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
022 * POSSIBILITY OF SUCH DAMAGE. 
023 */
024package org.cleartk.corpus.ace2005;
025
026import java.io.File;
027import java.io.IOException;
028import java.net.URI;
029import java.net.URISyntaxException;
030import java.util.ArrayList;
031import java.util.List;
032import java.util.regex.Matcher;
033import java.util.regex.Pattern;
034
035import org.apache.uima.UimaContext;
036import org.apache.uima.analysis_engine.AnalysisEngineProcessException;
037import org.apache.uima.cas.CAS;
038import org.apache.uima.cas.CASException;
039import org.apache.uima.jcas.JCas;
040import org.apache.uima.jcas.cas.FSArray;
041import org.apache.uima.resource.ResourceInitializationException;
042import org.cleartk.ne.type.Chunk;
043import org.cleartk.ne.type.NamedEntity;
044import org.cleartk.ne.type.NamedEntityMention;
045import org.jdom2.Document;
046import org.jdom2.Element;
047import org.jdom2.JDOMException;
048import org.jdom2.input.SAXBuilder;
049import org.apache.uima.fit.component.JCasAnnotator_ImplBase;
050import org.apache.uima.fit.descriptor.SofaCapability;
051import org.apache.uima.fit.util.FSCollectionFactory;
052
053/**
054 * <br>
055 * Copyright (c) 2007-2008, Regents of the University of Colorado <br>
056 * All rights reserved.
057 * 
058 * 
059 * This parser works with the ACE2005CollectionReader to pre-populate the JCas with ACE named
060 * entities. Currently, it only loads information from the entity and entity_mention tags and adds
061 * NamedEntity and NamedEntityMention objects to the JCas.
062 * 
063 * @author Philip Ogren
064 * 
065 */
066@SofaCapability(
067    inputSofas = { Ace2005Constants.ACE_2005_APF_URI_VIEW, CAS.NAME_DEFAULT_SOFA },
068    outputSofas = {})
069public class Ace2005GoldAnnotator extends JCasAnnotator_ImplBase {
070
071  Pattern ampPattern;
072
073  @Override
074  public void initialize(UimaContext aContext) throws ResourceInitializationException {
075    super.initialize(aContext);
076    ampPattern = Pattern.compile(Pattern.quote("&amp;"));
077  }
078
079  @Override
080  public void process(JCas jCas) throws AnalysisEngineProcessException {
081    try {
082      String apfUri = jCas.getView(Ace2005Constants.ACE_2005_APF_URI_VIEW).getSofaDataURI();
083      JCas initialView = jCas.getView(CAS.NAME_DEFAULT_SOFA);
084      String documentText = initialView.getDocumentText();
085      SAXBuilder builder = new SAXBuilder();
086      builder.setDTDHandler(null);
087      URI sofaDataURI = new URI(apfUri);
088      Document doc = builder.build(new File(sofaDataURI));
089
090      Element apfSource = doc.getRootElement();
091      Element apfDocument = apfSource.getChild("document");
092      for (Element apfEntity : apfDocument.getChildren("entity")) {
093        NamedEntity namedEntity = new NamedEntity(initialView);
094        namedEntity.setEntityType(apfEntity.getAttributeValue("TYPE"));
095        namedEntity.setEntitySubtype(apfEntity.getAttributeValue("SUBTYPE"));
096        namedEntity.setEntityClass(apfEntity.getAttributeValue("CLASS"));
097        namedEntity.setEntityId(apfEntity.getAttributeValue("ID"));
098        namedEntity.addToIndexes();
099
100        List<NamedEntityMention> mentions = new ArrayList<NamedEntityMention>();
101
102        for (Element entityMention : apfEntity.getChildren("entity_mention")) {
103          int start = Integer.parseInt(entityMention.getChild("extent").getChild("charseq").getAttributeValue(
104              "START"));
105          int end = Integer.parseInt(entityMention.getChild("extent").getChild("charseq").getAttributeValue(
106              "END"));
107          String givenText = entityMention.getChild("extent").getChild("charseq").getText();
108          String parsedText = documentText.substring(start, end + 1);
109          Matcher ampMatcher = ampPattern.matcher(parsedText);
110          parsedText = ampMatcher.replaceAll("&");
111
112          NamedEntityMention mention = new NamedEntityMention(initialView, start, end + 1);
113          mention.setMentionId(entityMention.getAttributeValue("ID"));
114          mention.setMentionType(entityMention.getAttributeValue("TYPE"));
115          mention.setMentionedEntity(namedEntity);
116
117          Chunk chunk = new Chunk(initialView, start, end + 1);
118          mention.setAnnotation(chunk);
119
120          int headStart = Integer.parseInt(entityMention.getChild("head").getChild("charseq").getAttributeValue(
121              "START"));
122          int headEnd = Integer.parseInt(entityMention.getChild("head").getChild("charseq").getAttributeValue(
123              "END"));
124          Chunk head = new Chunk(initialView, headStart, headEnd + 1);
125          mention.setHead(head);
126
127          mention.addToIndexes();
128          mentions.add(mention);
129
130          givenText = givenText.replaceAll("\\s+", " ");
131          parsedText = givenText.replaceAll("\\s+", " ");
132
133          // if(!givenText.equals(parsedText))
134          // {
135          //
136          // System.out.println("given text and parsed text differ.");
137          // System.out.println(givenText);
138          // System.out.println(parsedText);
139          // System.out.println(apfDocument.getAttributeValue("DOCID"));
140          // System.out.println(apfEntity.getAttributeValue("ID"));
141          // System.out.println(documentText);
142          // }
143        }
144        namedEntity.setMentions(new FSArray(jCas, mentions.size()));
145        FSCollectionFactory.fillArrayFS(namedEntity.getMentions(), mentions);
146      }
147
148    } catch (CASException ce) {
149      throw new AnalysisEngineProcessException(ce);
150    } catch (IOException ioe) {
151      throw new AnalysisEngineProcessException(ioe);
152    } catch (JDOMException je) {
153      throw new AnalysisEngineProcessException(je);
154    } catch (URISyntaxException use) {
155      throw new AnalysisEngineProcessException(use);
156    }
157  }
158
159}