1   /**
2    * Distribution License:
3    * JSword is free software; you can redistribute it and/or modify it under
4    * the terms of the GNU Lesser General Public License, version 2.1 as published by
5    * the Free Software Foundation. This program is distributed in the hope
6    * that it will be useful, but WITHOUT ANY WARRANTY; without even the
7    * implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
8    * See the GNU Lesser General Public License for more details.
9    *
10   * The License is available on the internet at:
11   *       http://www.gnu.org/copyleft/lgpl.html
12   * or by writing to:
13   *      Free Software Foundation, Inc.
14   *      59 Temple Place - Suite 330
15   *      Boston, MA 02111-1307, USA
16   *
17   * Copyright: 2009
18   *     The copyright to this program is held by it's authors.
19   *
20   * ID: $Id:  $
21   */
22  package org.crosswire.jsword.index.lucene.analysis;
23  
24  import java.io.IOException;
25  import java.io.Reader;
26  
27  import org.apache.lucene.analysis.LowerCaseFilter;
28  import org.apache.lucene.analysis.StopFilter;
29  import org.apache.lucene.analysis.TokenStream;
30  import org.apache.lucene.analysis.ar.ArabicAnalyzer;
31  import org.apache.lucene.analysis.ar.ArabicLetterTokenizer;
32  import org.apache.lucene.analysis.ar.ArabicNormalizationFilter;
33  import org.apache.lucene.analysis.ar.ArabicStemFilter;
34  import org.apache.lucene.util.Version;
35  
36  /**
37   * An Analyzer whose {@link TokenStream} is built from a
38   * {@link ArabicLetterTokenizer} filtered with {@link LowerCaseFilter},
39   * {@link ArabicNormalizationFilter}, {@link ArabicStemFilter} (optional) and
40   * Arabic {@link StopFilter} (optional).
41   * 
42   * @see gnu.lgpl.License for license details.<br>
43   *      The copyright to this program is held by it's authors.
44   * @author DM Smith [dmsmith555 at yahoo dot com]
45   */
46  public class ArabicLuceneAnalyzer extends AbstractBookAnalyzer {
47      public ArabicLuceneAnalyzer() throws IOException {
48          stopSet = ArabicAnalyzer.getDefaultStopSet();
49      }
50  
51      /* (non-Javadoc)
52       * @see org.apache.lucene.analysis.Analyzer#tokenStream(java.lang.String, java.io.Reader)
53       */
54      @Override
55      public final TokenStream tokenStream(String fieldName, Reader reader) {
56          TokenStream result = new ArabicLetterTokenizer(reader);
57          result = new LowerCaseFilter(result);
58          result = new ArabicNormalizationFilter(result);
59          if (doStopWords && stopSet != null) {
60              result = new StopFilter(false, result, stopSet);
61          }
62  
63          if (doStemming) {
64              result = new ArabicStemFilter(result);
65          }
66  
67          return result;
68      }
69  
70      /* (non-Javadoc)
71       * @see org.apache.lucene.analysis.Analyzer#reusableTokenStream(java.lang.String, java.io.Reader)
72       */
73      @Override
74      public TokenStream reusableTokenStream(String fieldName, Reader reader) throws IOException {
75          SavedStreams streams = (SavedStreams) getPreviousTokenStream();
76          if (streams == null) {
77              streams = new SavedStreams(new ArabicLetterTokenizer(reader));
78              streams.setResult(new LowerCaseFilter(streams.getResult()));
79              streams.setResult(new ArabicNormalizationFilter(streams.getResult()));
80  
81              if (doStopWords && stopSet != null) {
82                  streams.setResult(new StopFilter(StopFilter.getEnablePositionIncrementsVersionDefault(matchVersion), streams.getResult(), stopSet));
83              }
84  
85              if (doStemming) {
86                  streams.setResult(new ArabicStemFilter(streams.getResult()));
87              }
88  
89              setPreviousTokenStream(streams);
90          } else {
91              streams.getSource().reset(reader);
92          }
93          return streams.getResult();
94      }
95  
96      private final Version matchVersion = Version.LUCENE_29;
97  }
98