Examples of edu.stanford.nlp.util.CoreMap

edu.stanford.nlp.util.CoreMap

Base type for all annotatable core objects. Should usually be instantiated as {@link ArrayCoreMap}. Many common key definitions live in {@link edu.stanford.nlp.ling.CoreAnnotations}, but others may be defined elsewhere. See {@link edu.stanford.nlp.ling.CoreAnnotations} for details.

Note that implementations of this interface must take care to implement equality correctly: by default, two CoreMaps are .equal if they contain the same keys and all corresponding values are .equal. Subclasses that wish to change this behavior (such as {@link HashableCoreMap}) must make sure that all other CoreMap implementations have a special case in their .equals to use that equality definition when appropriate. Similarly, care must be taken when defining hashcodes. The default hashcode is 37 * sum of all keys' hashcodes plus the sum of all values' hashcodes. However, use of this class as HashMap keys is discouraged because the hashcode can change over time. Consider using a {@link HashableCoreMap}.
@author dramage @author rafferty

    match = m.find();
    assertFalse(match);
  }


  public void testTokenSequenceMatcherABs() throws IOException {
    CoreMap doc = createDocument("A A A A A A A B A A B A C A E A A A A A A A A A A A B A A A");


    // Test sequence with groups
    TokenSequencePattern p = TokenSequencePattern.compile( "/A/+ B");
    TokenSequenceMatcher m = p.getMatcher(doc.get(CoreAnnotations.TokensAnnotation.class));
    boolean match = m.find();
    assertTrue(match);
    assertEquals(0, m.groupCount());
    assertEquals("A A A A A A A B", m.group());
    match = m.find();
    assertTrue(match);
    assertEquals(0, m.groupCount());
    assertEquals("A A B", m.group());
    match = m.find();
    assertTrue(match);
    assertEquals(0, m.groupCount());
    assertEquals("A A A A A A A A A A A B", m.group());
    match = m.find();
    assertFalse(match);


    p = TokenSequencePattern.compile( "(/A/+ B)+");
    m = p.getMatcher(doc.get(CoreAnnotations.TokensAnnotation.class));
    match = m.find();
    assertTrue(match);
    assertEquals(1, m.groupCount());
    assertEquals("A A A A A A A B A A B", m.group());
    match = m.find();
    assertTrue(match);
    assertEquals(1, m.groupCount());
    assertEquals("A A A A A A A A A A A B", m.group());
    match = m.find();
    assertFalse(match);


  /*  p = TokenSequencePattern.compile( "( A+ ( /B/+ )? )*");
    m = p.getMatcher(doc.get(CoreAnnotations.TokensAnnotation.class));
    match = m.find();
    assertTrue(match);
    assertEquals(2, m.groupCount());
    assertEquals("A A A A A A A B A A B A", m.group());
    match = m.find();
    assertTrue(match);
    assertEquals(2, m.groupCount());
    assertEquals("A", m.group());
    match = m.find();
    assertTrue(match);
    assertEquals(2, m.groupCount());
    assertEquals("A A A A A A A A A A A B A A A", m.group());
    match = m.find();
    assertFalse(match);              */


    p = TokenSequencePattern.compile( "(/A/+ /B/+ )+");
    m = p.getMatcher(doc.get(CoreAnnotations.TokensAnnotation.class));
    match = m.find();
    assertTrue(match);
    assertEquals(1, m.groupCount());
    assertEquals("A A A A A A A B A A B", m.group());
    match = m.find();
    assertTrue(match);
    assertEquals(1, m.groupCount());
    assertEquals("A A A A A A A A A A A B", m.group());
    match = m.find();
    assertFalse(match);


    p = TokenSequencePattern.compile( "(/A/+ /C/? /A/* )+");
    m = p.getMatcher(doc.get(CoreAnnotations.TokensAnnotation.class));
    match = m.find();
    assertTrue(match);
    assertEquals(1, m.groupCount());
    assertEquals("A A A A A A A", m.group());
    match = m.find();

View Full Code Here

    match = m.find();
    assertFalse(match);
  }


  public void testTokenSequenceMatcherMultiNodePattern() throws IOException {
    CoreMap doc = createDocument("blah four-years blah blah four - years");


    // Test sequence with groups
    CoreMapNodePattern nodePattern  = CoreMapNodePattern.valueOf("four\\s*-?\\s*years");
    SequencePattern.MultiNodePatternExpr expr = new SequencePattern.MultiNodePatternExpr(
            new MultiCoreMapNodePattern(nodePattern));
    TokenSequencePattern p = TokenSequencePattern.compile(expr);
    TokenSequenceMatcher m = p.getMatcher(doc.get(CoreAnnotations.TokensAnnotation.class));
    boolean match = m.find();
    assertTrue(match);
    assertEquals(0, m.groupCount());
    assertEquals("four-years", m.group());
    match = m.find();
    assertTrue(match);
    assertEquals(0, m.groupCount());
    assertEquals("four - years", m.group());
    match = m.find();
    assertFalse(match);


    p = TokenSequencePattern.compile("(?m) /four\\s*-?\\s*years/");
    m = p.getMatcher(doc.get(CoreAnnotations.TokensAnnotation.class));
    match = m.find();
    assertTrue(match);
    assertEquals(0, m.groupCount());
    assertEquals("four-years", m.group());
    match = m.find();
    assertTrue(match);
    assertEquals(0, m.groupCount());
    assertEquals("four - years", m.group());
    match = m.find();
    assertFalse(match);


    p = TokenSequencePattern.compile("(?m){2,3} /four\\s*-?\\s*years/");
    m = p.getMatcher(doc.get(CoreAnnotations.TokensAnnotation.class));
    match = m.find();
    assertTrue(match);
    assertEquals(0, m.groupCount());
    assertEquals("four - years", m.group());
    match = m.find();
    assertFalse(match);


    p = TokenSequencePattern.compile( "(?m){1,2} /four\\s*-?\\s*years/");
    m = p.getMatcher(doc.get(CoreAnnotations.TokensAnnotation.class));
    match = m.find();
    assertTrue(match);
    assertEquals(0, m.groupCount());
    assertEquals("four-years", m.group());
    match = m.find();
    assertFalse(match);


    p = TokenSequencePattern.compile("(?m){1,3} /four\\s*-?\\s*years/ ==> &annotate( { ner=YEAR } )");
    m = p.getMatcher(doc.get(CoreAnnotations.TokensAnnotation.class));
    match = m.find();
    assertTrue(match);
    assertEquals(0, m.groupCount());
    assertEquals("four-years", m.group());
    p.getAction().apply(m, 0);

View Full Code Here

    match = m.find();
    assertFalse(match);
  }


  public void testTokenSequenceMatcherBackRef() throws IOException {
    CoreMap doc = createDocument("A A A A A A A B A A B A C A E A A A A A A A A A A A B A A A");


    // Test sequence with groups
    TokenSequencePattern p = TokenSequencePattern.compile( "(/A/+) B \\1");
    TokenSequenceMatcher m = p.getMatcher(doc.get(CoreAnnotations.TokensAnnotation.class));
    boolean match = m.find();
    assertTrue(match);
    assertEquals(1, m.groupCount());
    assertEquals("A A B A A", m.group());
    match = m.find();

View Full Code Here

    assertFalse(match);


  }


  public void testMultiPatternMatcher() throws IOException {
    CoreMap doc = createDocument(testText1);


    // Test simple sequence
    TokenSequencePattern p1 = TokenSequencePattern.compile("/Archbishop/ /of/ /Canterbury/");
    p1.setPriority(1);
    TokenSequencePattern p2 = TokenSequencePattern.compile("/[a-zA-Z]+/{1,2}  /of/ /[a-zA-Z]+/+");
    MultiPatternMatcher<CoreMap> m = new MultiPatternMatcher<CoreMap>(p2,p1);
    List<SequenceMatchResult<CoreMap>> matched = m.findNonOverlapping(doc.get(CoreAnnotations.TokensAnnotation.class));
    assertEquals(4, matched.size());
    assertEquals("first Bishop of London", matched.get(0).group());
    assertEquals("Archbishop of Canterbury", matched.get(1).group());
    assertEquals("a member of the Gregorian mission sent to England to convert the", matched.get(2).group());
    assertEquals("as Bishop of London in", matched.get(3).group());

View Full Code Here

    assertEquals("a member of the Gregorian mission sent to England to convert the", matched.get(2).group());
    assertEquals("as Bishop of London in", matched.get(3).group());
  }


  public void testStringPatternMatchCaseInsensitive() throws IOException {
    CoreMap doc = createDocument(testText1);


    // Test simple sequence
    Env env = TokenSequencePattern.getNewEnv();
    env.setDefaultStringPatternFlags(Pattern.CASE_INSENSITIVE);
    TokenSequencePattern p = TokenSequencePattern.compile(env, "/archbishop/ /of/ /canterbury/");
    TokenSequenceMatcher m = p.getMatcher(doc.get(CoreAnnotations.TokensAnnotation.class));
    assertTrue(m.find());
    assertEquals("Archbishop of Canterbury", m.group());
    assertFalse(m.find());


    p = TokenSequencePattern.compile(env, "/ARCHBISHOP/ /OF/ /CANTERBURY/");
    m = p.getMatcher(doc.get(CoreAnnotations.TokensAnnotation.class));
    assertTrue(m.find());
    assertEquals("Archbishop of Canterbury", m.group());
    assertFalse(m.find());
  }

View Full Code Here

    assertEquals("Archbishop of Canterbury", m.group());
    assertFalse(m.find());
  }


  public void testStringMatchCaseInsensitive() throws IOException {
    CoreMap doc = createDocument(testText1);


    // Test simple sequence
    Env env = TokenSequencePattern.getNewEnv();
    env.setDefaultStringMatchFlags(NodePattern.CASE_INSENSITIVE);
    TokenSequencePattern p = TokenSequencePattern.compile(env, "archbishop of canterbury");
    TokenSequenceMatcher m = p.getMatcher(doc.get(CoreAnnotations.TokensAnnotation.class));
    assertTrue(m.find());
    assertEquals("Archbishop of Canterbury", m.group());
    assertFalse(m.find());


    p = TokenSequencePattern.compile(env, "ARCHBISHOP OF CANTERBURY");
    m = p.getMatcher(doc.get(CoreAnnotations.TokensAnnotation.class));
    assertTrue(m.find());
    assertEquals("Archbishop of Canterbury", m.group());
    assertFalse(m.find());
  }

View Full Code Here

  }


  //just to test if a pattern is compiling or not
  public void testCompile() {
    String s = "(?$se \"matching\" \"this\"|\"don't\")";
    CoreMap doc = createDocument("does this do matching this");
    TokenSequencePattern p = TokenSequencePattern.compile(s);
    TokenSequenceMatcher m = p.getMatcher(doc.get(CoreAnnotations.TokensAnnotation.class));
    boolean match = m.find();
    assertTrue(match);
    //assertEquals(m.group(), "matching this");
  }

View Full Code Here

  public void testCaseInsensitive1(){
    Env env = TokenSequencePattern.getNewEnv();
    env.setDefaultStringPatternFlags(Pattern.CASE_INSENSITIVE);
    env.setDefaultStringMatchFlags(NodePattern.CASE_INSENSITIVE);
    String s = "for /President/";
    CoreMap doc = createDocument("for president");
    TokenSequencePattern p = TokenSequencePattern.compile(env, s);
    TokenSequenceMatcher m = p.getMatcher(doc.get(CoreAnnotations.TokensAnnotation.class));
    boolean match = m.find();
    assertTrue(match);
  }

View Full Code Here

    Env env = TokenSequencePattern.getNewEnv();
    env.setDefaultStringPatternFlags(Pattern.CASE_INSENSITIVE);
    env.setDefaultStringMatchFlags(NodePattern.CASE_INSENSITIVE);


    String s = "for president";
    CoreMap doc = createDocument("for President");


    TokenSequencePattern p = TokenSequencePattern.compile(env, s);
    TokenSequenceMatcher m = p.getMatcher(doc.get(CoreAnnotations.TokensAnnotation.class));
    boolean match = m.find();
    assertTrue(match);
  }

View Full Code Here


        for (CorefChain corefChain : result.values()) {
          if(corefChain.getMentionsInTextualOrder().size() < 2) continue;
          Set<CoreLabel> coreferentTokens = Generics.newHashSet();
          for (CorefMention mention : corefChain.getMentionsInTextualOrder()) {
            CoreMap sentence = annotation.get(CoreAnnotations.SentencesAnnotation.class).get(mention.sentNum - 1);
            CoreLabel token = sentence.get(CoreAnnotations.TokensAnnotation.class).get(mention.headIndex - 1);
            coreferentTokens.add(token);
          }
          for (CoreLabel token : coreferentTokens) {
            token.set(CorefCoreAnnotations.CorefClusterAnnotation.class, coreferentTokens);
          }

View Full Code Here

0 1 2 3 4 5 6 7 8 9

TOP

Related Classes of edu.stanford.nlp.util.CoreMap

edu.stanford.nlp.dcoref.ACEMentionExtractor

edu.stanford.nlp.dcoref.CoNLLMentionExtractor

edu.stanford.nlp.dcoref.Document

edu.stanford.nlp.dcoref.MUCMentionExtractor

edu.stanford.nlp.dcoref.RuleBasedCorefMentionFinder

edu.stanford.nlp.dcoref.SieveCoreferenceSystem

edu.stanford.nlp.ie.crf.TestSequenceModel

edu.stanford.nlp.ie.machinereading.BasicRelationFeatureFactory

edu.stanford.nlp.ie.machinereading.domains.ace.AceReader

edu.stanford.nlp.ie.machinereading.EntityExtractorResultsPrinter

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.