001: package org.apache.lucene.analysis.snowball;
002:
003: /* ====================================================================
004: * The Apache Software License, Version 1.1
005: *
006: * Copyright (c) 2004 The Apache Software Foundation. All rights
007: * reserved.
008: *
009: * Redistribution and use in source and binary forms, with or without
010: * modification, are permitted provided that the following conditions
011: * are met:
012: *
013: * 1. Redistributions of source code must retain the above copyright
014: * notice, this list of conditions and the following disclaimer.
015: *
016: * 2. Redistributions in binary form must reproduce the above copyright
017: * notice, this list of conditions and the following disclaimer in
018: * the documentation and/or other materials provided with the
019: * distribution.
020: *
021: * 3. The end-user documentation included with the redistribution,
022: * if any, must include the following acknowledgment:
023: * "This product includes software developed by the
024: * Apache Software Foundation (http://www.apache.org/)."
025: * Alternately, this acknowledgment may appear in the software itself,
026: * if and wherever such third-party acknowledgments normally appear.
027: *
028: * 4. The names "Apache" and "Apache Software Foundation" and
029: * "Apache Lucene" must not be used to endorse or promote products
030: * derived from this software without prior written permission. For
031: * written permission, please contact apache@apache.org.
032: *
033: * 5. Products derived from this software may not be called "Apache",
034: * "Apache Lucene", nor may "Apache" appear in their name, without
035: * prior written permission of the Apache Software Foundation.
036: *
037: * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESSED OR IMPLIED
038: * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
039: * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
040: * DISCLAIMED. IN NO EVENT SHALL THE APACHE SOFTWARE FOUNDATION OR
041: * ITS CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
042: * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
043: * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF
044: * USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
045: * ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
046: * OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT
047: * OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
048: * SUCH DAMAGE.
049: * ====================================================================
050: *
051: * This software consists of voluntary contributions made by many
052: * individuals on behalf of the Apache Software Foundation. For more
053: * information on the Apache Software Foundation, please see
054: * <http://www.apache.org/>.
055: */
056:
057: import java.io.*;
058:
059: import junit.framework.*;
060:
061: import org.apache.lucene.analysis.*;
062:
063: public class TestSnowball extends TestCase {
064:
065: public void assertAnalyzesTo(Analyzer a, String input,
066: String[] output) throws Exception {
067: TokenStream ts = a
068: .tokenStream("dummy", new StringReader(input));
069: for (int i = 0; i < output.length; i++) {
070: Token t = ts.next();
071: assertNotNull(t);
072: assertEquals(output[i], t.termText());
073: }
074: assertNull(ts.next());
075: ts.close();
076: }
077:
078: public void testEnglish() throws Exception {
079: Analyzer a = new SnowballAnalyzer("English");
080: assertAnalyzesTo(a, "he abhorred accents", new String[] { "he",
081: "abhor", "accent" });
082: }
083:
084: public void testFilterTokens() throws Exception {
085: final Token tok = new Token("accents", 2, 7, "wrd");
086: tok.setPositionIncrement(3);
087:
088: SnowballFilter filter = new SnowballFilter(new TokenStream() {
089: public Token next() {
090: return tok;
091: }
092: }, "English");
093:
094: Token newtok = filter.next();
095:
096: assertEquals("accent", newtok.termText());
097: assertEquals(2, newtok.startOffset());
098: assertEquals(7, newtok.endOffset());
099: assertEquals("wrd", newtok.type());
100: assertEquals(3, newtok.getPositionIncrement());
101: }
102: }
|