lucene-java-3.4.0/lucene/contrib/xml-query-parser/src/java/org/apache/lucene/xmlparser/builders/LikeThisQueryBuilder.java

   1 /*
   2  * Created on 25-Jan-2006
   3  */
   4 package org.apache.lucene.xmlparser.builders;
   5
   6 import java.io.IOException;
   7 import java.io.StringReader;
   8 import java.util.HashSet;
   9 import java.util.Set;
  10
  11 import org.apache.lucene.analysis.Analyzer;
  12 import org.apache.lucene.analysis.TokenStream;
  13 import org.apache.lucene.analysis.tokenattributes.CharTermAttribute;
  14 import org.apache.lucene.search.similar.MoreLikeThisQuery;
  15 import org.apache.lucene.search.Query;
  16 import org.apache.lucene.xmlparser.DOMUtils;
  17 import org.apache.lucene.xmlparser.ParserException;
  18 import org.apache.lucene.xmlparser.QueryBuilder;
  19 import org.w3c.dom.Element;
  20 /**
  21  * Licensed to the Apache Software Foundation (ASF) under one or more
  22  * contributor license agreements.  See the NOTICE file distributed with
  23  * this work for additional information regarding copyright ownership.
  24  * The ASF licenses this file to You under the Apache License, Version 2.0
  25  * (the "License"); you may not use this file except in compliance with
  26  * the License.  You may obtain a copy of the License at
  27  *
  28  *     http://www.apache.org/licenses/LICENSE-2.0
  29  *
  30  * Unless required by applicable law or agreed to in writing, software
  31  * distributed under the License is distributed on an "AS IS" BASIS,
  32  * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
  33  * See the License for the specific language governing permissions and
  34  * limitations under the License.
  35  */
  36
  37 /**
  38  *
  39  */
  40 public class LikeThisQueryBuilder implements QueryBuilder {
  41
  42         private Analyzer analyzer;
  43         String defaultFieldNames [];
  44         int defaultMaxQueryTerms=20;
  45         int defaultMinTermFrequency=1;
  46         float defaultPercentTermsToMatch=30; //default is a 3rd of selected terms must match
  47
  48         public LikeThisQueryBuilder(Analyzer analyzer,String [] defaultFieldNames)
  49         {
  50                 this.analyzer=analyzer;
  51                 this.defaultFieldNames=defaultFieldNames;
  52         }
  53
  54         /* (non-Javadoc)
  55          * @see org.apache.lucene.xmlparser.QueryObjectBuilder#process(org.w3c.dom.Element)
  56          */
  57         public Query getQuery(Element e) throws ParserException {
  58                 String fieldsList=e.getAttribute("fieldNames"); //a comma-delimited list of fields
  59                 String fields[]=defaultFieldNames;
  60                 if((fieldsList!=null)&&(fieldsList.trim().length()>0))
  61                 {
  62                         fields=fieldsList.trim().split(",");
  63                         //trim the fieldnames
  64                         for (int i = 0; i < fields.length; i++) {
  65                                 fields[i]=fields[i].trim();
  66                         }
  67                 }
  68
  69                 //Parse any "stopWords" attribute
  70                 //TODO MoreLikeThis needs to ideally have per-field stopWords lists - until then
  71                 //I use all analyzers/fields to generate multi-field compatible stop list
  72                 String stopWords=e.getAttribute("stopWords");
  73                 Set<String> stopWordsSet=null;
  74                 if((stopWords!=null)&&(fields!=null))
  75                 {
  76                     stopWordsSet=new HashSet<String>();
  77                     for (int i = 0; i < fields.length; i++)
  78             {
  79                 try
  80                 {
  81                   TokenStream ts = analyzer.reusableTokenStream(fields[i],new StringReader(stopWords));
  82                   CharTermAttribute termAtt = ts.addAttribute(CharTermAttribute.class);
  83                   ts.reset();
  84                         while(ts.incrementToken()) {
  85                             stopWordsSet.add(termAtt.toString());
  86                         }
  87                         ts.end();
  88                         ts.close();
  89                 }
  90                 catch(IOException ioe)
  91                 {
  92                     throw new ParserException("IoException parsing stop words list in "
  93                             +getClass().getName()+":"+ioe.getLocalizedMessage());
  94                 }
  95             }
  96                 }
  97
  98
  99                 MoreLikeThisQuery mlt=new MoreLikeThisQuery(DOMUtils.getText(e),fields,analyzer, fields[0]);
 100                 mlt.setMaxQueryTerms(DOMUtils.getAttribute(e,"maxQueryTerms",defaultMaxQueryTerms));
 101                 mlt.setMinTermFrequency(DOMUtils.getAttribute(e,"minTermFrequency",defaultMinTermFrequency));
 102                 mlt.setPercentTermsToMatch(DOMUtils.getAttribute(e,"percentTermsToMatch",defaultPercentTermsToMatch)/100);
 103                 mlt.setStopWords(stopWordsSet);
 104                 int minDocFreq=DOMUtils.getAttribute(e,"minDocFreq",-1);
 105                 if(minDocFreq>=0)
 106                 {
 107                         mlt.setMinDocFreq(minDocFreq);
 108                 }
 109
 110                 mlt.setBoost(DOMUtils.getAttribute(e,"boost",1.0f));
 111
 112                 return mlt;
 113         }
 114
 115
 116
 117 }