Example usage for edu.stanford.nlp.parser.common ParserGrammar parserQuery

Introduction

In this page you can find the example usage for edu.stanford.nlp.parser.common ParserGrammar parserQuery.

Prototype

public abstract ParserQuery parserQuery();

Source Link

Usage

From source file:de.tudarmstadt.ukp.dkpro.core.stanfordnlp.StanfordParser.java

License:Open Source License

/**
 * Processes the given text using the StanfordParser.
 *
 * @param aJCas// ww  w  .  j a v a2 s.co  m
 *            the {@link JCas} to process
 * @see org.apache.uima.analysis_component.JCasAnnotator_ImplBase#process(org.apache.uima.jcas.JCas)
 */
@Override
public void process(JCas aJCas) throws AnalysisEngineProcessException {
    modelProvider.configure(aJCas.getCas());
    posMappingProvider.configure(aJCas.getCas());
    constituentMappingProvider.configure(aJCas.getCas());

    Type typeToParse;
    if (annotationTypeToParse != null) {
        typeToParse = aJCas.getCas().getTypeSystem().getType(annotationTypeToParse);
    } else {
        typeToParse = JCasUtil.getType(aJCas, Sentence.class);
    }
    FSIterator<Annotation> typeToParseIterator = aJCas.getAnnotationIndex(typeToParse).iterator();

    // Iterator each Sentence or whichever construct to parse

    while (typeToParseIterator.hasNext()) {
        Annotation currAnnotationToParse = typeToParseIterator.next();
        List<HasWord> tokenizedSentence = new ArrayList<HasWord>();
        List<Token> tokens = new ArrayList<Token>();

        // Split sentence to tokens for annotating indexes
        for (Token token : JCasUtil.selectCovered(Token.class, currAnnotationToParse)) {
            tokenizedSentence.add(tokenToWord(token));
            tokens.add(token);
        }

        getContext().getLogger().log(FINE, tokenizedSentence.toString());
        ParserGrammar parser = modelProvider.getResource();

        Tree parseTree;
        try {
            if (tokenizedSentence.size() > maxTokens) {
                continue;
            }

            if (ptb3Escaping) {
                tokenizedSentence = CoreNlpUtils.applyPtbEscaping(tokenizedSentence, quoteBegin, quoteEnd);
            }

            // Get parse
            ParserQuery query = parser.parserQuery();
            query.parse(tokenizedSentence);
            parseTree = query.getBestParse();
        } catch (Exception e) {
            throw new AnalysisEngineProcessException(e);
        }

        // Create new StanfordAnnotator object
        StanfordAnnotator sfAnnotator = null;
        try {
            sfAnnotator = new StanfordAnnotator(new TreeWithTokens(parseTree, tokens));
            sfAnnotator.setPosMappingProvider(posMappingProvider);
            sfAnnotator.setConstituentMappingProvider(constituentMappingProvider);
        } catch (CASException e) {
            throw new AnalysisEngineProcessException(e);
        }

        // Create Penn bracketed structure annotations
        if (writePennTree) {
            sfAnnotator.createPennTreeAnnotation(currAnnotationToParse.getBegin(),
                    currAnnotationToParse.getEnd());
        }

        // Create dependency annotations
        if (writeDependency) {
            doCreateDependencyTags(parser, sfAnnotator, parseTree, tokens);
        }

        // Create constituent annotations
        if (writeConstituent) {
            sfAnnotator.createConstituentAnnotationFromTree(parser.getTLPParams().treebankLanguagePack(),
                    writePos);
        }
    }
}

---