TestRussianStem


1   package org.apache.lucene.analysis.ru;
2   
3   /**
4    * Copyright 2004 The Apache Software Foundation
5    *
6    * Licensed under the Apache License, Version 2.0 (the "License");
7    * you may not use this file except in compliance with the License.
8    * You may obtain a copy of the License at
9    *
10   *     http://www.apache.org/licenses/LICENSE-2.0
11   *
12   * Unless required by applicable law or agreed to in writing, software
13   * distributed under the License is distributed on an "AS IS" BASIS,
14   * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
15   * See the License for the specific language governing permissions and
16   * limitations under the License.
17   */
18  
19  import junit.framework.TestCase;
20  
21  import java.io.BufferedReader  ;
22  import java.io.File  ;
23  import java.io.InputStreamReader  ;
24  import java.io.FileInputStream  ;
25  import java.util.ArrayList  ;
26  
27  public class TestRussianStem extends TestCase
28  {
29      private ArrayList   words = new ArrayList  ();
30      private ArrayList   stems = new ArrayList  ();
31  
32      public TestRussianStem(String   name)
33      {
34          super(name);
35      }
36  
37      /**
38       * @see TestCase#setUp()
39       */
40      protected void setUp() throws Exception  
41      {
42          super.setUp();
43          //System.out.println(new java.util.Date());
44          String   str;
45          
46          File   dataDir = new File  (System.getProperty("dataDir"));
47  
48          // open and read words into an array list
49          BufferedReader   inWords =
50              new BufferedReader  (
51                  new InputStreamReader  (
52                      new FileInputStream  (new File  (dataDir, "/org/apache/lucene/analysis/ru/wordsUnicode.txt")),
53                      "Unicode"));
54          while ((str = inWords.readLine()) != null)
55          {
56              words.add(str);
57          }
58          inWords.close();
59  
60          // open and read stems into an array list
61          BufferedReader   inStems =
62              new BufferedReader  (
63                  new InputStreamReader  (
64                      new FileInputStream  (new File  (dataDir, "/org/apache/lucene/analysis/ru/stemsUnicode.txt")),
65                      "Unicode"));
66          while ((str = inStems.readLine()) != null)
67          {
68              stems.add(str);
69          }
70          inStems.close();
71      }
72  
73      /**
74       * @see TestCase#tearDown()
75       */
76      protected void tearDown() throws Exception  
77      {
78          super.tearDown();
79      }
80  
81      public void testStem()
82      {
83          for (int i = 0; i < words.size(); i++)
84          {
85              //if ( (i % 100) == 0 ) System.err.println(i);
86              String   realStem =
87                  RussianStemmer.stem(
88                      (String  ) words.get(i),
89                      RussianCharsets.UnicodeRussian);
90              assertEquals("unicode", stems.get(i), realStem);
91          }
92      }
93  
94      private String   printChars(String   output)
95      {
96          StringBuffer   s = new StringBuffer  ();
97          for (int i = 0; i < output.length(); i++)
98              {
99              s.append(output.charAt(i));
100         }
101         return s.toString();
102     }
103 }
104
A to Z: JavaDoc & Examples Daily Java News & Articles Open Source Projects Open Source Codes Free Computer Books Remove Frame
Popular Tags