Package opennlp.tools.util

Examples of opennlp.tools.util.Span

31
32
33
34
35
36
37
38
39
40
41
  @Test
  public void testRetrievingContent() {
   
    SentenceSample sample = new SentenceSample("1. 2.",
        new Span(0, 2), new Span(3, 5));
   
    assertEquals("1. 2.", sample.getDocument());
    assertEquals(new Span(0, 2), sample.getSentences()[0]);
    assertEquals(new Span(3, 5), sample.getSentences()[1]);
  }
View Full Code Here
47
48
49
50
51
52
53
    assertFalse(createPredSample().equals(createGoldSample()));
    assertFalse(createPredSample().equals(new Object()));
  }
 
  public static SentenceSample createGoldSample() {
    return new SentenceSample("1. 2.", new Span(0, 2), new Span(3, 5));
  }
View Full Code Here
51
52
53
54
55
56
57
  public static SentenceSample createGoldSample() {
    return new SentenceSample("1. 2.", new Span(0, 2), new Span(3, 5));
  }

  public static SentenceSample createPredSample() {
    return new SentenceSample("1. 2.", new Span(0, 1), new Span(2, 5));
  }
View Full Code Here
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
    Span[] preds = chunker.chunkAsSpans(toks1, tags1);
    System.out.println(Arrays.toString(preds));

    assertEquals(10, preds.length);
    assertEquals(new Span(0, 1, "NP"), preds[0]);
    assertEquals(new Span(1, 2, "VP"), preds[1]);
    assertEquals(new Span(2, 4, "NP"), preds[2]);
    assertEquals(new Span(4, 5, "VP"), preds[3]);
    assertEquals(new Span(5, 6, "SBAR"), preds[4]);
    assertEquals(new Span(6, 7, "NP"), preds[5]);
    assertEquals(new Span(7, 9, "VP"), preds[6]);
    assertEquals(new Span(9, 13, "NP"), preds[7]);
    assertEquals(new Span(13, 14, "PP"), preds[8]);
    assertEquals(new Span(14, 16, "NP"), preds[9]);

  }
View Full Code Here
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
    //gather up potential tokens
    int end = d.length();
    for (int i = 0; i < end; i++) {
      if (StringUtil.isWhitespace(d.charAt(i))) {
        if (inTok) {
          tokens.add(new Span(tokStart, i));
          inTok = false;
          tokStart = -1;
        }
      }
      else {
        if (!inTok) {
          tokStart = i;
          inTok = true;
        }
      }
    }
   
    if (inTok) {
      tokens.add(new Span(tokStart, end));
    }
   
    return tokens.toArray(new Span[tokens.size()]);
  }
View Full Code Here
83
84
85
86
87
88
89
90
91
92
93
        sentence.append(' ');
      }
     
      int beginIndex = sentence.length();
      sentence.append(tokens[i]);
      mergedTokenSpans.add(new Span(beginIndex, sentence.length()));
    }
   
    text = sentence.toString();
    tokenSpans = Collections.unmodifiableList(mergedTokenSpans);
  }
View Full Code Here
141
142
143
144
145
146
147
148
149
150
   
    int tokenSpanStart = sample.length();
    sample.append(token);
    int tokenSpanEnd = sample.length();
   
    tokenSpans.add(new Span(tokenSpanStart, tokenSpanEnd));
   
    if (!isNextMerged)
        sample.append(" ");
  }
View Full Code Here
152
153
154
155
156
157
158
159
160
161
162
  public static TokenSample parse(String sampleString, String separatorChars) {
   
    if (sampleString == null || separatorChars == null)
        throw new IllegalArgumentException("arguments must not be null!");
   
    Span whitespaceTokenSpans[] = WhitespaceTokenizer.INSTANCE.tokenizePos(sampleString);
   
    // Pre-allocate 20% for newly created tokens
    List<Span> realTokenSpans = new ArrayList<Span>((int) (whitespaceTokenSpans.length * 1.2d));
   
    StringBuilder untaggedSampleString = new StringBuilder();
View Full Code Here
53
54
55
56
57
58
59
60
61
62
    this.tokenizer = tokenizer;
  }

  @Override
  protected TokenSample processSample(TokenSample reference) {
    Span predictions[] = tokenizer.tokenizePos(reference.getText());
    fmeasure.updateScores(reference.getTokenSpans(), predictions);
   
    return new TokenSample(reference.getText(), predictions);
  }
View Full Code Here
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
126
127
128
129
130
131
132
133
134
135
136
137
138
139
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
  @Override
  protected Iterator<Event> createEvents(TokenSample tokenSample) {

    List<Event> events = new ArrayList<Event>(50);

    Span tokens[] = tokenSample.getTokenSpans();
    String text = tokenSample.getText();

    if (tokens.length > 0) {

      int start = tokens[0].getStart();
      int end = tokens[tokens.length - 1].getEnd();

      String sent = text.substring(start, end);

      Span[] candTokens = WhitespaceTokenizer.INSTANCE.tokenizePos(sent);

      int firstTrainingToken = -1;
      int lastTrainingToken = -1;
      for (int ci = 0; ci < candTokens.length; ci++) {
        Span cSpan = candTokens[ci];
        String ctok = sent.substring(cSpan.getStart(), cSpan.getEnd());
        //adjust cSpan to text offsets
        cSpan = new Span(cSpan.getStart() + start, cSpan.getEnd() + start);
        //should we skip this token
        if (ctok.length() > 1
          && (!skipAlphaNumerics || !alphaNumeric.matcher(ctok).matches())) {

          //find offsets of annotated tokens inside of candidate tokens
          boolean foundTrainingTokens = false;
          for (int ti = lastTrainingToken + 1; ti < tokens.length; ti++) {
            if (cSpan.contains(tokens[ti])) {
              if (!foundTrainingTokens) {
                firstTrainingToken = ti;
                foundTrainingTokens = true;
              }
              lastTrainingToken = ti;
            }
            else if (cSpan.getEnd() < tokens[ti].getEnd()) {
              break;
            }
            else if (tokens[ti].getEnd() < cSpan.getStart()) {
              //keep looking
            }
            else {
              if (logger.isLoggable(Level.WARNING)) {
                logger.warning("Bad training token: " + tokens[ti] + " cand: " + cSpan +
                    " token="+text.substring(tokens[ti].getStart(), tokens[ti].getEnd()));
              }
            }
          }

          // create training data
          if (foundTrainingTokens) {

            for (int ti = firstTrainingToken; ti <= lastTrainingToken; ti++) {
              Span tSpan = tokens[ti];
              int cStart = cSpan.getStart();
              for (int i = tSpan.getStart() + 1; i < tSpan.getEnd(); i++) {
                String[] context = cg.getContext(ctok, i - cStart);
                events.add(new Event(TokenizerME.NO_SPLIT, context));
              }

              if (tSpan.getEnd() != cSpan.getEnd()) {
                String[] context = cg.getContext(ctok, tSpan.getEnd() - cStart);
                events.add(new Event(TokenizerME.SPLIT, context));
              }
            }
          }
        }
View Full Code Here
TOP

Related Classes of opennlp.tools.util.Span

  • com.bericotech.clavin.extractor.ApacheExtractor
  • com.tamingtext.opennlp.NameFinderTest
  • net.sf.nlpshell.CorefParse
  • net.sf.nlpshell.Main
  • opennlp.tools.chunker.ChunkerMETest
  • opennlp.tools.chunker.ChunkSample
  • opennlp.tools.chunker.ChunkSampleTest
  • opennlp.tools.cmdline.EvaluationErrorPrinter
  • opennlp.tools.cmdline.namefind.TokenNameFinderTool
  • opennlp.tools.cmdline.parser.ParserTool
    • 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.