Java POS.addToIndexes方法代码示例

import de.tudarmstadt.ukp.dkpro.core.api.lexmorph.type.pos.POS; //导入方法依赖的package包/类
private void convertPos(JCas aJCas, TextCorpus aCorpusData, Map<String, Token> aTokens)
{
    if (aCorpusData.getPosTagsLayer() == null) {
        return;
    }
    for (int i = 0; i < aCorpusData.getPosTagsLayer().size(); i++) {
        eu.clarin.weblicht.wlfxb.tc.api.Token[] posTokens = aCorpusData.getPosTagsLayer()
                .getTokens(aCorpusData.getPosTagsLayer().getTag(i));
        String value = aCorpusData.getPosTagsLayer().getTag(i).getString();

        POS outPos = new POS(aJCas);

        outPos.setBegin(aTokens.get(posTokens[0].getID()).getBegin());
        outPos.setEnd(aTokens.get(posTokens[0].getID()).getEnd());
        outPos.setPosValue(value);
        outPos.addToIndexes();

        // Set the POS to the token
        aTokens.get(posTokens[0].getID()).setPos(outPos);
    }
}
 

import de.tudarmstadt.ukp.dkpro.core.api.lexmorph.type.pos.POS; //导入方法依赖的package包/类
@SuppressWarnings("unused")
private void getCas1(JCasBuilder aBuilder, JCas aJCas)
{
    Token token1 = aBuilder.add("Hallo", Token.class);
    POS pos1 = new POS(aJCas, token1.getBegin(), token1.getEnd());
    pos1.setPosValue("I");
    pos1.addToIndexes();
    token1.setPos(pos1);
    Token token2 = aBuilder.add("Welt", Token.class);
    POS pos2 = new POS(aJCas, token2.getBegin(), token2.getEnd());
    pos2.setPosValue("N");
    pos2.addToIndexes();
    token2.setPos(pos2);
    Token token3 = aBuilder.add("!", Token.class);
    POS pos3 = new POS(aJCas, token3.getBegin(), token3.getEnd());
    pos3.setPosValue("SENT");
    token3.setPos(pos3);
    pos3.addToIndexes();
}
 

import de.tudarmstadt.ukp.dkpro.core.api.lexmorph.type.pos.POS; //导入方法依赖的package包/类
@SuppressWarnings("unused")
private void getCas2(JCasBuilder aBuilder, JCas aJCas)
{
    Token token1 = aBuilder.add("Hallo", Token.class);
    POS pos1 = new POS(aJCas, token1.getBegin(), token1.getEnd());
    pos1.setPosValue("O");
    pos1.addToIndexes();
    token1.setPos(pos1);
    Token token2 = aBuilder.add("Welt", Token.class);
    POS pos2 = new POS(aJCas, token2.getBegin(), token2.getEnd());
    pos2.setPosValue("N");
    pos2.addToIndexes();
    token2.setPos(pos2);
    Token token3 = aBuilder.add("!", Token.class);
    POS pos3 = new POS(aJCas, token3.getBegin(), token3.getEnd());
    pos3.setPosValue("SENT");
    token3.setPos(pos3);
    pos3.addToIndexes();
}
 

import de.tudarmstadt.ukp.dkpro.core.api.lexmorph.type.pos.POS; //导入方法依赖的package包/类
private POS createPartOfSpeechAnnotationFromOutcome(JCas aJCas, int begin, int end,
        String aOutcome)
{
    Type posTag = mappingProvider.getTagType(aOutcome);
    POS posAnno = (POS) aJCas.getCas().createAnnotation(posTag, begin, end);
    posAnno.setPosValue(aOutcome);
    posAnno.addToIndexes();

    return posAnno;

}
 

import de.tudarmstadt.ukp.dkpro.core.api.lexmorph.type.pos.POS; //导入方法依赖的package包/类
/**
 * Create {@link Token} in the {@link CAS}. If the lemma and pos columns are not empty it will
 * create {@link Lemma} and {@link POS} annotations
 */
private void createToken(JCas aJCas, StringBuilder text, Map<Integer, String> tokens,
        Map<Integer, String> pos, Map<Integer, String> lemma, Map<String, Token> tokensStored)
{
    int tokenBeginPosition = 0;
    int tokenEndPosition = 0;

    for (int i = 1; i <= tokens.size(); i++) {
        tokenBeginPosition = text.indexOf(tokens.get(i), tokenBeginPosition);
        Token outToken = new Token(aJCas, tokenBeginPosition, text.indexOf(tokens.get(i),
                tokenBeginPosition) + tokens.get(i).length());
        tokenEndPosition = text.indexOf(tokens.get(i), tokenBeginPosition)
                + tokens.get(i).length();
        tokenBeginPosition = tokenEndPosition;
        outToken.addToIndexes();

        // Add pos to CAS if exist
        if (!pos.get(i).equals("_")) {
            POS outPos = new POS(aJCas, outToken.getBegin(), outToken.getEnd());
            outPos.setPosValue(pos.get(i));
            outPos.addToIndexes();
            outToken.setPos(outPos);
        }

        // Add lemma if exist
        if (!lemma.get(i).equals("_")) {
            Lemma outLemma = new Lemma(aJCas, outToken.getBegin(), outToken.getEnd());
            outLemma.setValue(lemma.get(i));
            outLemma.addToIndexes();
            outToken.setLemma(outLemma);
        }
        tokensStored.put("t_" + i, outToken);
    }
}
 

import de.tudarmstadt.ukp.dkpro.core.api.lexmorph.type.pos.POS; //导入方法依赖的package包/类
@Test
public void testTokenAttachedAnnotationsWithValues() throws Exception
{
    JCas jcas = makeJCasOneSentence();
    
    List<Token> tokens = new ArrayList<>(select(jcas, Token.class));
    
    Token t1 = tokens.get(0);
    
    Lemma l1 = new Lemma(jcas, t1.getBegin(), t1.getEnd());
    l1.setValue("lemma1");
    l1.addToIndexes();
    t1.setLemma(l1);
    
    MorphologicalFeatures m1 = new MorphologicalFeatures(jcas, t1.getBegin(), t1.getEnd());
    m1.setValue("morph");
    m1.setTense("tense1");
    m1.addToIndexes();
    t1.setMorph(m1);
    
    POS p1 = new POS(jcas, t1.getBegin(), t1.getEnd());
    p1.setPosValue("pos1");
    p1.addToIndexes();
    t1.setPos(p1);
    
    Stem s1 = new Stem(jcas, t1.getBegin(), t1.getEnd());
    s1.setValue("stem1");
    s1.addToIndexes();
    t1.setStem(s1);
    
    writeAndAssertEquals(jcas, WebannoTsv3Writer.PARAM_SPAN_LAYERS,
            asList(MorphologicalFeatures.class, POS.class, Lemma.class, Stem.class));
}
 

import de.tudarmstadt.ukp.dkpro.core.api.lexmorph.type.pos.POS; //导入方法依赖的package包/类
@Test
public void testDependencyWithValues() throws Exception 
{
    JCas jcas = makeJCasOneSentence();
    
    List<Token> tokens = new ArrayList<>(select(jcas, Token.class));
    Token t1 = tokens.get(0);
    Token t2 = tokens.get(1);
    
    POS p1 = new POS(jcas, t1.getBegin(), t1.getEnd());
    p1.setPosValue("POS1");
    p1.addToIndexes();
    t1.setPos(p1);

    POS p2 = new POS(jcas, t2.getBegin(), t2.getEnd());
    p2.setPosValue("POS2");
    p2.addToIndexes();
    t2.setPos(p2);
    
    Dependency dep1 = new Dependency(jcas);
    dep1.setGovernor(t1);
    dep1.setDependent(t2);
    dep1.setBegin(dep1.getDependent().getBegin());
    dep1.setEnd(dep1.getDependent().getEnd());
    dep1.addToIndexes();
    
    writeAndAssertEquals(jcas, 
            WebannoTsv3Writer.PARAM_SPAN_LAYERS, asList(POS.class),
            WebannoTsv3Writer.PARAM_RELATION_LAYERS, asList(Dependency.class));
}
 

import de.tudarmstadt.ukp.dkpro.core.api.lexmorph.type.pos.POS; //导入方法依赖的package包/类
@Test
public void testSimpleSymbolicRules()
    throws Exception
{
    ConstraintsGrammar parser = new ConstraintsGrammar(new FileInputStream(
            "src/test/resources/rules/symbolic1.rules"));
    Parse p = parser.Parse();

    ParsedConstraints constraints = p.accept(new ParserVisitor());

    JCas jcas = JCasFactory.createJCas();

    CollectionReader reader = createReader(Conll2006Reader.class,
            Conll2006Reader.PARAM_SOURCE_LOCATION, "src/test/resources/text/1.conll");
    
    reader.getNext(jcas.getCas());

    POS pos = new POS(jcas, 8, 9);
    pos.setPosValue("pronoun");
    pos.addToIndexes();
    
    Evaluator constraintsEvaluator = new ValuesGenerator();

    Lemma lemma = select(jcas, Lemma.class).iterator().next();
    
    List<PossibleValue> possibleValues = constraintsEvaluator.generatePossibleValues(lemma,
            "value", constraints);

    List<PossibleValue> expectedOutput = new ArrayList<>();
    expectedOutput.add(new PossibleValue("good", true));

    assertEquals(expectedOutput, possibleValues);
}
 

import de.tudarmstadt.ukp.dkpro.core.api.lexmorph.type.pos.POS; //导入方法依赖的package包/类
@Test
public void testSimpleSymbolicRules2()
    throws Exception
{
    ConstraintsGrammar parser = new ConstraintsGrammar(new FileInputStream(
            "src/test/resources/rules/symbolic2.rules"));
    Parse p = parser.Parse();

    ParsedConstraints constraints = p.accept(new ParserVisitor());

    JCas jcas = JCasFactory.createJCas();

    CollectionReader reader = createReader(Conll2006Reader.class,
            Conll2006Reader.PARAM_SOURCE_LOCATION, "src/test/resources/text/1.conll");
    
    reader.getNext(jcas.getCas());

    POS pos = new POS(jcas, 8, 9);
    pos.setPosValue("pronoun");
    pos.addToIndexes();
    
    Evaluator constraintsEvaluator = new ValuesGenerator();

    Lemma lemma = select(jcas, Lemma.class).iterator().next();
    
    List<PossibleValue> possibleValues = constraintsEvaluator.generatePossibleValues(lemma,
            "value", constraints);

    List<PossibleValue> expectedOutput = new ArrayList<>();
    expectedOutput.add(new PossibleValue("good", true));

    assertEquals(expectedOutput, possibleValues);
}
 

import de.tudarmstadt.ukp.dkpro.core.api.lexmorph.type.pos.POS; //导入方法依赖的package包/类
@Test
public void singleNoDifferencesWithAdditionalCas1Test()
    throws Exception
{
    JCas user1 = JCasFactory.createJCas();
    user1.setDocumentText("test");

    JCas user2 = JCasFactory.createJCas();
    user2.setDocumentText("test");
    
    JCas user3 = JCasFactory.createJCas();
    user3.setDocumentText("test");
    POS pos3 = new POS(user3, 0, 4);
    pos3.setPosValue("test");
    pos3.addToIndexes();
    
    Map<String, List<JCas>> casByUser = new LinkedHashMap<>();
    casByUser.put("user1", asList(user1));
    casByUser.put("user2", asList(user2));
    casByUser.put("user3", asList(user3));
    
    List<String> entryTypes = asList(POS.class.getName());

    List<SpanDiffAdapter> diffAdapters = asList(SpanDiffAdapter.POS);

    DiffResult result = CasDiff2.doDiff(entryTypes, diffAdapters,
            LinkCompareBehavior.LINK_TARGET_AS_LABEL, casByUser);

    result.print(System.out);
    
    casByUser.remove("user3");
    
    AgreementResult agreement = AgreementUtils.getAgreement(
            ConcreteAgreementMeasure.KRIPPENDORFF_ALPHA_NOMINAL_AGREEMENT, false, result,
            entryTypes.get(0), "PosValue", casByUser);
    
    assertEquals(1, agreement.getTotalSetCount());
    assertEquals(1, agreement.getIrrelevantSets().size());
    assertEquals(0, agreement.getRelevantSetCount());
}
 

import de.tudarmstadt.ukp.dkpro.core.api.lexmorph.type.pos.POS; //导入方法依赖的package包/类
@Test
public void singleNoDifferencesWithAdditionalCas2Test()
    throws Exception
{
    JCas user1 = JCasFactory.createJCas();
    user1.setDocumentText("test");

    JCas user2 = JCasFactory.createJCas();
    user2.setDocumentText("test");
    
    JCas user3 = JCasFactory.createJCas();
    user3.setDocumentText("test");
    POS pos3 = new POS(user3, 0, 4);
    pos3.addToIndexes();
    
    Map<String, List<JCas>> casByUser = new LinkedHashMap<>();
    casByUser.put("user1", asList(user1));
    casByUser.put("user2", asList(user2));
    casByUser.put("user3", asList(user3));
    
    List<String> entryTypes = asList(POS.class.getName());

    List<SpanDiffAdapter> diffAdapters = asList(SpanDiffAdapter.POS);

    DiffResult result = CasDiff2.doDiff(entryTypes, diffAdapters,
            LinkCompareBehavior.LINK_TARGET_AS_LABEL, casByUser);

    result.print(System.out);
    
    casByUser.remove("user3");
    
    AgreementResult agreement = AgreementUtils.getAgreement(
            ConcreteAgreementMeasure.KRIPPENDORFF_ALPHA_NOMINAL_AGREEMENT, false, result,
            entryTypes.get(0), "PosValue", casByUser);
    
    assertEquals(1, agreement.getTotalSetCount());
    assertEquals(1, agreement.getIrrelevantSets().size());
    assertEquals(0, agreement.getRelevantSetCount());
}
 

import de.tudarmstadt.ukp.dkpro.core.api.lexmorph.type.pos.POS; //导入方法依赖的package包/类
@Override
public void process(JCas jcas) throws AnalysisEngineProcessException {
	try {
		CAS cas = jcas.getCas();
		mappingProvider.configure(cas);
		
		for (Sentence sentenceAnno : JCasUtil.select(jcas, Sentence.class)) {
			List<Token> tokens = JCasUtil.selectCovered(jcas, Token.class, sentenceAnno);
			List<String> tokenStrings = JCasUtil.toText(tokens);
			List<PosTaggedToken> taggedTokens;
			
			synchronized (innerTool) {
				innerTool.setTokenizedSentence(tokenStrings);
				innerTool.process();
				taggedTokens = innerTool.getPosTaggedTokens();
			}
			
			if (taggedTokens.size() != tokens.size()) {
				throw new PosTaggerException("Got pos tagging for " + taggedTokens.size() +
						" tokens, should have gotten according to the total number of tokens in the sentence: " + tokens.size());
			}
			
			Iterator<Token> tokenIter = tokens.iterator();
			for (PosTaggedToken taggedToken : taggedTokens) {
				Token tokenAnno = tokenIter.next();
				String tagString = taggedToken.getPartOfSpeech().getStringRepresentation();
				
				// Get an annotation with the appropriate UIMA type via the mappingProvider
				Type posTag = mappingProvider.getTagType(tagString);
				POS posAnnotation = (POS) cas.createAnnotation(posTag, tokenAnno.getBegin(), tokenAnno.getEnd());
				posAnnotation.setPosValue(tagString);
				posAnnotation.addToIndexes();
				
				tokenAnno.setPos(posAnnotation);
			}
		}
	} catch (PosTaggerException e) {
		throw new AnalysisEngineProcessException(AnalysisEngineProcessException.ANNOTATOR_EXCEPTION, null, e);
	}
}
 

import de.tudarmstadt.ukp.dkpro.core.api.lexmorph.type.pos.POS; //导入方法依赖的package包/类
public void getNext(JCas aJCas)
    throws IOException, CollectionException
{

    DocumentMetaData md = new DocumentMetaData(aJCas);
    md.setDocumentTitle("");
    md.setDocumentId("" + (instanceId++));
    md.setLanguage(language);
    md.addToIndexes();

    try {
        posMappingProvider.configure(aJCas.getCas());
    }
    catch (AnalysisEngineProcessException e1) {
        throw new CollectionException(e1);
    }

    StringBuilder documentText = new StringBuilder();

    int seqStart = 0;
    for (int k = 0; k < sequences.size(); k++) {

        List<String> sequence = sequences.get(k);

        for (int i = 0; i < sequence.size(); i++) {
            String pairs = sequence.get(i).replaceAll(" +", " ");

            int idxLastSpace = pairs.lastIndexOf(" ");
            String token = pairs.substring(0, idxLastSpace);
            String tag = pairs.substring(idxLastSpace+1);

            int tokenLen = token.length();
            if(lowerCase){
                token = token.toLowerCase();
            }
            documentText.append(token);

            int tokStart = documentText.length() - tokenLen;
            int tokEnd = documentText.length();
            Token t = new Token(aJCas, tokStart, tokEnd);
            t.addToIndexes();

            if (i + 1 < sequence.size()) {
                documentText.append(" ");
            }

            Type posTag = posMappingProvider.getTagType(tag);
            POS pos = (POS) aJCas.getCas().createAnnotation(posTag, t.getBegin(),
                    t.getEnd());
            pos.setPosValue(tag);
            pos.addToIndexes();
            t.setPos(pos);
        }
        Sentence sentence = new Sentence(aJCas, seqStart, documentText.length());
        sentence.addToIndexes();

        if (k + 1 < sequences.size()) {
            documentText.append(" ");
        }
        seqStart = documentText.length();

    }
    aJCas.setDocumentText(documentText.toString());
}
 

import de.tudarmstadt.ukp.dkpro.core.api.lexmorph.type.pos.POS; //导入方法依赖的package包/类
@Test
public void testSimplePath()
    throws Exception
{
    ConstraintsGrammar parser = new ConstraintsGrammar(new FileInputStream(
            "src/test/resources/rules/10.rules"));
    Parse p = parser.Parse();

    ParsedConstraints constraints = p.accept(new ParserVisitor());

    JCas jcas = JCasFactory.createJCas();
    jcas.setDocumentText("The sun.");

    // Add token annotations
    Token t_the = new Token(jcas, 0, 3);
    t_the.addToIndexes();
    Token t_sun = new Token(jcas, 0, 3);
    t_sun.addToIndexes();

    // Add POS annotations and link them to the tokens
    POS p_the = new POS(jcas, t_the.getBegin(), t_the.getEnd());
    p_the.setPosValue("DET");
    p_the.addToIndexes();
    t_the.setPos(p_the);
    POS p_sun = new POS(jcas, t_sun.getBegin(), t_sun.getEnd());
    p_sun.setPosValue("NN");
    p_sun.addToIndexes();
    t_sun.setPos(p_sun);

    // Add dependency annotations
    Dependency dep_the_sun = new Dependency(jcas);
    dep_the_sun.setGovernor(t_sun);
    dep_the_sun.setDependent(t_the);
    dep_the_sun.setDependencyType("det");
    dep_the_sun.setBegin(dep_the_sun.getGovernor().getBegin());
    dep_the_sun.setEnd(dep_the_sun.getGovernor().getEnd());
    dep_the_sun.addToIndexes();

    Evaluator constraintsEvaluator = new ValuesGenerator();

    List<PossibleValue> possibleValues = constraintsEvaluator.generatePossibleValues(
            dep_the_sun, "DependencyType", constraints);

    List<PossibleValue> expectedOutput = new LinkedList<>();
    expectedOutput.add(new PossibleValue("det", false));

    assertEquals(expectedOutput, possibleValues);
}
 

import de.tudarmstadt.ukp.dkpro.core.api.lexmorph.type.pos.POS; //导入方法依赖的package包/类
@Test
public void twoWithoutLabelTest()
    throws Exception
{
    JCas user1 = JCasFactory.createJCas();
    user1.setDocumentText("test");
    new POS(user1, 0, 1).addToIndexes();
    new POS(user1, 1, 2).addToIndexes();
    POS p1 = new POS(user1, 3, 4);
    p1.setPosValue("A");
    p1.addToIndexes();

    JCas user2 = JCasFactory.createJCas();
    user2.setDocumentText("test");
    new POS(user2, 0, 1).addToIndexes();
    new POS(user2, 2, 3).addToIndexes();
    POS p2 = new POS(user2, 3, 4);
    p2.setPosValue("B");
    p2.addToIndexes();
    
    
    Map<String, List<JCas>> casByUser = new LinkedHashMap<>();
    casByUser.put("user1", asList(user1));
    casByUser.put("user2", asList(user2));
    
    List<String> entryTypes = asList(POS.class.getName());

    List<SpanDiffAdapter> diffAdapters = asList(SpanDiffAdapter.POS);

    DiffResult result = CasDiff2.doDiff(entryTypes, diffAdapters,
            LinkCompareBehavior.LINK_TARGET_AS_LABEL, casByUser);

    result.print(System.out);
    
    AgreementResult agreement = AgreementUtils.getAgreement(
            ConcreteAgreementMeasure.KRIPPENDORFF_ALPHA_NOMINAL_AGREEMENT, false, result,
            entryTypes.get(0), "PosValue", casByUser);
    
    assertEquals(4, agreement.getTotalSetCount());
    assertEquals(0, agreement.getIrrelevantSets().size());
    // the following two counts are zero because the incomplete sets are not excluded!
    assertEquals(2, agreement.getIncompleteSetsByPosition().size());
    assertEquals(0, agreement.getIncompleteSetsByLabel().size());
    assertEquals(3, agreement.getSetsWithDifferences().size());
    assertEquals(4, agreement.getRelevantSetCount());
    assertEquals(0.4, agreement.getAgreement(), 0.01);
    ICodingAnnotationItem item1 = agreement.getStudy().getItem(0);
    ICodingAnnotationItem item2 = agreement.getStudy().getItem(1);
    ICodingAnnotationItem item3 = agreement.getStudy().getItem(2);
    assertEquals("", item1.getUnit(0).getCategory());
    assertEquals("", item1.getUnit(1).getCategory());
    assertEquals("", item2.getUnit(0).getCategory());
    assertEquals(null, item2.getUnit(1).getCategory());
    assertEquals(null, item3.getUnit(0).getCategory());
    assertEquals("", item3.getUnit(1).getCategory());
}