| Classes in this File | Line Coverage | Branch Coverage | Complexity | ||||
| SimpleLuceneAnalyzer |
|
| 1.0;1 |
| 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 or later | |
| 5 | * as published by the Free Software Foundation. This program is distributed | |
| 6 | * in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even | |
| 7 | * the 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 | * © CrossWire Bible Society, 2007 - 2016 | |
| 18 | * | |
| 19 | */ | |
| 20 | package org.crosswire.jsword.index.lucene.analysis; | |
| 21 | ||
| 22 | import java.io.Reader; | |
| 23 | ||
| 24 | import org.apache.lucene.analysis.ASCIIFoldingFilter; | |
| 25 | import org.apache.lucene.analysis.LowerCaseTokenizer; | |
| 26 | import org.apache.lucene.analysis.TokenStream; | |
| 27 | ||
| 28 | /** | |
| 29 | * Simple Analyzer providing same function as | |
| 30 | * org.apache.lucene.analysis.SimpleAnalyzer This is intended to be the default | |
| 31 | * analyzer for natural language fields. Additionally performs: Normalize | |
| 32 | * Diacritics (Changes Accented characters to their unaccented equivalent) for | |
| 33 | * ISO 8859-1 languages | |
| 34 | * | |
| 35 | * Note: Next Lucene release (beyond 2.2.0) will have a major performance | |
| 36 | * enhancement using method - public TokenStream reusableTokenStream(String | |
| 37 | * fieldName, Reader reader) We should use that. Ref: | |
| 38 | * https://issues.apache.org/jira/browse/LUCENE-969 | |
| 39 | * | |
| 40 | * @see gnu.lgpl.License The GNU Lesser General Public License for details. | |
| 41 | * @author Sijo Cherian | |
| 42 | */ | |
| 43 | public class SimpleLuceneAnalyzer extends AbstractBookAnalyzer { | |
| 44 | ||
| 45 | 0 | public SimpleLuceneAnalyzer() { |
| 46 | 0 | doStemming = false; |
| 47 | 0 | } |
| 48 | ||
| 49 | @Override | |
| 50 | public TokenStream tokenStream(String fieldName, Reader reader) { | |
| 51 | 0 | TokenStream result = new LowerCaseTokenizer(reader); |
| 52 | 0 | result = new ASCIIFoldingFilter(result); |
| 53 | 0 | return result; |
| 54 | } | |
| 55 | } |