Package joshua.corpus.vocab

Examples of joshua.corpus.vocab.Vocabulary

85
86
87
88
89
90
91
92
93

  // TODO: fuse createVocabulary and createCorpusArray together to avoid allocating the trivial array.
  public static CorpusArray createCorpusArray(String inputFilename)
  throws IOException {
    Vocabulary vocabulary = new Vocabulary();
    int[] ws = Vocabulary.initializeVocabulary(inputFilename, vocabulary, true);
    return createCorpusArray(inputFilename, vocabulary, ws[0], ws[1]);
  }
View Full Code Here
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
  int numBuiltInTerminals = 4;
 
  @Test
  public void basicVocabTest() {
   
    Vocabulary vocab1 = new Vocabulary();
    Vocabulary vocab2 = new Vocabulary(new HashSet<String>());
   
    Assert.assertEquals(vocab1, vocab2);
   
    Assert.assertFalse(vocab1.intToString.isEmpty());
//    Assert.assertTrue(vocab1.intToString.get(0)==Vocabulary.UNKNOWN_WORD_STRING);
    Assert.assertFalse(vocab1.getWords().isEmpty());
    Assert.assertTrue(vocab1.getWord(0)==Vocabulary.UNKNOWN_WORD_STRING);
    Assert.assertEquals(vocab1.getWords(), vocab1.intToString.values());

    Assert.assertEquals(vocab1.size(), numBuiltInSymbols);
    Assert.assertEquals(vocab1.getWord(Vocabulary.UNKNOWN_WORD), Vocabulary.UNKNOWN_WORD_STRING);

    //Assert.assertEquals(vocab1.getID("sample"), Vocabulary.UNKNOWN_WORD);
    //Assert.assertEquals(vocab1.getID(null), Vocabulary.UNKNOWN_WORD);

    Assert.assertFalse(vocab1.terminalToInt.isEmpty());
    Assert.assertEquals(vocab1.terminalToInt.size(), this.numBuiltInTerminals);
//    Assert.assertFalse(vocab1.isFixed);
//   
//    vocab1.fixVocabulary();
//    Assert.assertTrue(vocab1.isFixed);
   
    Assert.assertEquals(vocab1.getID(SymbolTable.X_STRING), -1);
    Assert.assertEquals(vocab1.getID(SymbolTable.X1_STRING), -2);
    Assert.assertEquals(vocab1.getID(SymbolTable.X2_STRING), -3);
   
    Assert.assertEquals(vocab1.getWord(-1), SymbolTable.X_STRING);
    Assert.assertEquals(vocab1.getWord(-2), SymbolTable.X1_STRING);
    Assert.assertEquals(vocab1.getWord(-3), SymbolTable.X2_STRING);
   
   
   
    Assert.assertFalse(vocab2.intToString.isEmpty());
//    Assert.assertTrue(vocab2.intToString.get(0)==Vocabulary.UNKNOWN_WORD_STRING);
    Assert.assertFalse(vocab2.getWords().isEmpty());
//    Assert.assertTrue(vocab2.getWord(0)==Vocabulary.UNKNOWN_WORD_STRING);
    Assert.assertEquals(vocab2.getWords(), vocab2.intToString.values());

    Assert.assertEquals(vocab2.size(), numBuiltInSymbols);
    Assert.assertEquals(vocab2.getWord(Vocabulary.UNKNOWN_WORD), Vocabulary.UNKNOWN_WORD_STRING);

//    Assert.assertEquals(vocab2.getID("sample"), Vocabulary.UNKNOWN_WORD);
//    Assert.assertEquals(vocab2.getID(null), Vocabulary.UNKNOWN_WORD);
   
    Assert.assertFalse(vocab2.terminalToInt.isEmpty());
View Full Code Here
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
      sourcePrintStream.println(corpusString);
      sourcePrintStream.close();
      sourceFileName = sourceFile.getAbsolutePath();
    }
   
    Vocabulary vocab = new Vocabulary();
    Vocabulary.initializeVocabulary(sourceFileName, vocab, true);
   
    Assert.assertEquals(vocab.getWord(vocab.getID("it")), "it");
    Assert.assertEquals(vocab.getWord(vocab.getID("makes")), "makes");
    Assert.assertEquals(vocab.getWord(vocab.getID("him")), "him");
    Assert.assertEquals(vocab.getWord(vocab.getID("and")), "and");
    Assert.assertEquals(vocab.getWord(vocab.getID("mars")), "mars");
    Assert.assertEquals(vocab.getWord(vocab.getID(",")), ",");
    Assert.assertEquals(vocab.getWord(vocab.getID("sets")), "sets");
    Assert.assertEquals(vocab.getWord(vocab.getID("on")), "on");
    Assert.assertEquals(vocab.getWord(vocab.getID("takes")), "takes");
    Assert.assertEquals(vocab.getWord(vocab.getID("off")), "off");
   
//    Assert.assertEquals(vocab.getWord(vocab.getID("persuades")), Vocabulary.UNKNOWN_WORD_STRING);
//    Assert.assertEquals(vocab.getWord(vocab.getID("disheartens")), Vocabulary.UNKNOWN_WORD_STRING);
  }
View Full Code Here
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
    String filename = "data/tiny.en";
    int numSentences = 5;  // Should be 5 sentences in tiny.en
    int numWords = 89;     // Should be 89 words in tiny.en
    int numUniqWords = 60; // Should be 60 unique words in tiny.en
   
    Vocabulary vocab = new Vocabulary();
    Vocabulary vocab2 = new Vocabulary();
   
    Assert.assertTrue(vocab.equals(vocab2));
    Assert.assertTrue(vocab2.equals(vocab));
    Assert.assertEquals(vocab, vocab2);
   
    try {
      int[] result = Vocabulary.initializeVocabulary(filename, vocab, true);
      Assert.assertNotNull(result);
      Assert.assertEquals(result.length, 2);
      Assert.assertEquals(result[0], numWords);
      Assert.assertEquals(result[1], numSentences); 
     
//      Assert.assertTrue(vocab.isFixed);
      Assert.assertEquals(vocab.size(), numUniqWords+numBuiltInSymbols);
     
    } catch (IOException e) {
      Assert.fail("Could not load file " + filename);
    }
   
    Assert.assertFalse(vocab.equals(vocab2));
   
    try {
      int[] result = Vocabulary.initializeVocabulary(filename, vocab2, true);
      Assert.assertNotNull(result);
      Assert.assertEquals(result.length, 2);
      Assert.assertEquals(result[0], numWords);
      Assert.assertEquals(result[1], numSentences); 
     
//      Assert.assertTrue(vocab2.isFixed);
      Assert.assertEquals(vocab2.size(), numUniqWords+numBuiltInSymbols);
     
    } catch (IOException e) {
      Assert.fail("Could not load file " + filename);
    }
   
View Full Code Here
91
92
93
94
95
96
97
98
99
100
101
102
103
104
        sourcePrintStream.println(sentence);
      }
      sourcePrintStream.close();
      String corpusFileName = sourceFile.getAbsolutePath();
     
      Vocabulary symbolTable;
     
      logger.fine("Constructing vocabulary from file " + corpusFileName);
      symbolTable = new Vocabulary();
      int[] lengths = Vocabulary.initializeVocabulary(corpusFileName, symbolTable, true);

      logger.fine("Constructing corpus array from file " + corpusFileName);
      Corpus corpus = SuffixArrayFactory.createCorpusArray(corpusFileName, symbolTable, lengths[0], lengths[1]);
View Full Code Here
129
130
131
132
133
134
135
136
137
138
139
   
   
    try {
     
      // FIX: can't use createVocabulary(String) because we set numWords and numSentences
      Vocabulary vocab = new Vocabulary();
      Vocabulary.initializeVocabulary(filename, vocab, true);
      CorpusArray corpus = SuffixArrayFactory.createCorpusArray(filename, vocab, numWords, numSentences);
     
      corpus.write(filename+".corpus", filename+".vocab", "UTF-8");
     
View Full Code Here
TOP

Related Classes of joshua.corpus.vocab.Vocabulary

  • joshua.aligner.AlignCandidates
  • joshua.corpus.BasicPhrase
  • joshua.corpus.CorpusArray
  • joshua.corpus.CorpusArrayTest
  • joshua.corpus.lexprob.WriteLexProbs
  • joshua.corpus.suffix_array.AbstractHierarchicalPhrasesTest
  • joshua.corpus.suffix_array.BasicPhrase
  • joshua.corpus.suffix_array.Compile
  • joshua.corpus.suffix_array.ConvertCorpus
  • joshua.corpus.suffix_array.FrequentClassesTest
    • Copyright © 2018 www.massapicom. All rights reserved.
    All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.