Package joshua.corpus.suffix_array

Examples of joshua.corpus.suffix_array.SuffixArray

186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
      alignmentFileName = alignmentFile.getAbsolutePath();
    }
   
    CorpusArray sourceCorpusArray =
      SuffixArrayFactory.createCorpusArray(sourceFileName);
    SuffixArray sourceSuffixArray =
      SuffixArrayFactory.createSuffixArray(sourceCorpusArray, SuffixArray.DEFAULT_CACHE_CAPACITY);
   
    CorpusArray targetCorpusArray =
      SuffixArrayFactory.createCorpusArray(targetFileName);
    SuffixArray targetSuffixArray =
      SuffixArrayFactory.createSuffixArray(targetCorpusArray, SuffixArray.DEFAULT_CACHE_CAPACITY);

    Alignments alignmentArray = SuffixArrayFactory.createAlignments(alignmentFileName, sourceSuffixArray, targetSuffixArray);

    return new SampledLexProbs(Integer.MAX_VALUE, sourceSuffixArray, targetSuffixArray, alignmentArray, Cache.DEFAULT_CAPACITY, false);
View Full Code Here
86
87
88
89
90
91
92
93
94
95
96
97
98
99
      }
    }
    Alignments alignments = new AlignmentArray(alignedTargetIndices, alignedSourceIndices, 1);
   
    CorpusArray targetCorpus = new CorpusArray(sentenceF, sentenceStartPositions, vocab);
    SuffixArray targetSuffixes = new SuffixArray(targetCorpus);

    CorpusArray sourceCorpus = new CorpusArray(sentence, sentenceStartPositions, vocab);
    SuffixArray sourceSuffixes = new SuffixArray(sourceCorpus);
    ParallelCorpusGrammarFactory parallelCorpus = new ParallelCorpusGrammarFactory(sourceSuffixes, targetSuffixes, alignments, null, Integer.MAX_VALUE, maxPhraseSpan, maxPhraseLength, maxNonterminals, 2, Float.MIN_VALUE, JoshuaConfiguration.phrase_owner, JoshuaConfiguration.default_non_terminal, JoshuaConfiguration.oovFeatureCost);
   
//    tree = new PrefixTree(vocab, maxPhraseSpan, maxPhraseLength, maxNonterminals);
    tree = new PrefixTree(parallelCorpus);
    Assert.assertNotNull(tree);
View Full Code Here
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
      for(int i = 0; i < corpusSentence.size(); i++) {
        corpus[i] = corpusSentence.getWordID(i);
      }
     
      CorpusArray corpusArray = new CorpusArray(corpus, sentenceStartPositions, sourceVocab);
      suffixArray = new SuffixArray(corpusArray);
     


      int[] targetSentenceStartPositions = {0};
     
      BasicPhrase targetCorpusSentence = new BasicPhrase(targetCorpusString, targetVocab);
      Assert.assertEquals(targetCorpusSentence.size(), 18);
     
      int[] targetCorpus = new int[targetCorpusSentence.size()];
      for(int i = 0; i < targetCorpusSentence.size(); i++) {
        targetCorpus[i] = targetCorpusSentence.getWordID(i);
      }
     

     

     
      CorpusArray targetCorpusArray = new CorpusArray(targetCorpus, targetSentenceStartPositions, targetVocab);
      targetSuffixArray = new SuffixArray(targetCorpusArray);

     
      int[] lowestAlignedTargetIndex = new int[corpusSentence.size()];
      int[] highestAlignedTargetIndex = new int[corpusSentence.size()];
      int[] lowestAlignedSourceIndex = new int[targetCorpusSentence.size()];
View Full Code Here
62
63
64
65
66
67
68
69
70
71
    }
   
    CorpusArray corpusArray = new CorpusArray(corpus, sentences, vocab);
   
    if (binaryFileName==null || binaryFileName.trim().length()==0)
      suffixArray = new SuffixArray(corpusArray);
    else
      suffixArray = new MemoryMappedSuffixArray(binaryFileName, corpusArray, MemoryMappedSuffixArray.DEFAULT_CACHE_CAPACITY);
   
  }
View Full Code Here
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
    }
   
    this.sourceCorpusArray =
      SuffixArrayFactory.createCorpusArray(sourceFileName);
    this.sourceVocab = (Vocabulary) sourceCorpusArray.getVocabulary();
    SuffixArray sourceSuffixArray =
      SuffixArrayFactory.createSuffixArray(sourceCorpusArray, SuffixArray.DEFAULT_CACHE_CAPACITY);

    this.targetCorpusArray =
      SuffixArrayFactory.createCorpusArray(targetFileName);
    this.targetVocab = (Vocabulary) targetCorpusArray.getVocabulary();
    SuffixArray targetSuffixArray =
      SuffixArrayFactory.createSuffixArray(targetCorpusArray, SuffixArray.DEFAULT_CACHE_CAPACITY);

    this.alignmentArray =
      SuffixArrayFactory.createAlignments(alignmentFileName, sourceSuffixArray, targetSuffixArray);
View Full Code Here
105
106
107
108
109
110
111
112
113
114
115
116
117
   
    targetCorpusArray =
      SuffixArrayFactory.createCorpusArray(targetFileName);

    if (alignmentsType.equals("AlignmentArray")) {
      SuffixArray targetSuffixArray =
        SuffixArrayFactory.createSuffixArray(targetCorpusArray, SuffixArray.DEFAULT_CACHE_CAPACITY);
      SuffixArray sourceSuffixArray =
        SuffixArrayFactory.createSuffixArray(sourceCorpusArray, SuffixArray.DEFAULT_CACHE_CAPACITY);
      alignments =
        SuffixArrayFactory.createAlignments(alignmentFileName, sourceSuffixArray, targetSuffixArray);
    } else if (alignmentsType.equals("AlignmentGrids")) {
      alignments = new AlignmentGrids(new Scanner(new File(alignmentFileName)), sourceCorpusArray, targetCorpusArray, 3);     
View Full Code Here
TOP

Related Classes of joshua.corpus.suffix_array.SuffixArray

  • joshua.corpus.alignment.AlignmentsTest
  • joshua.corpus.lexprob.BetterLexProbsTest
  • joshua.corpus.lexprob.SampledLexProbs
  • joshua.corpus.suffix_array.SuffixArrayTest
  • joshua.prefix_tree.PrefixTreeAdvancedTest
  • joshua.prefix_tree.PrefixTreeTest
  • java.io.FileOutputStream
    • 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.