Examples with StatefulTokenizer edu.illinois.cs.cogcomp.nlp.tokenizer.StatefulTokenizer used on opensource projects

Example 6 with StatefulTokenizer

use of edu.illinois.cs.cogcomp.nlp.tokenizer.StatefulTokenizer in project cogcomp-nlp by CogComp.

the class PipelineFactory method buildPipeline.

/**
     * create an AnnotatorService with the given view names in the argument. The names are supposed
     * be strings, separated by space.
     *
     * @return AnnotatorService with specified NLP components
     * @throws IOException
     * @throws AnnotatorException
     */
public static BasicAnnotatorService buildPipeline(Boolean disableCache, String... views) throws IOException, AnnotatorException {
    List<String> allViewNames = ViewNames.getAllViewNames();
    Map<String, String> nonDefaultValues = new HashMap<>();
    for (String vu : views) {
        if (allViewNames.contains(vu)) {
            switch(vu) {
                case ViewNames.POS:
                    nonDefaultValues.put(PipelineConfigurator.USE_POS.key, Configurator.TRUE);
                    break;
                case ViewNames.LEMMA:
                    nonDefaultValues.put(PipelineConfigurator.USE_LEMMA.key, Configurator.TRUE);
                    break;
                case ViewNames.NER_CONLL:
                    nonDefaultValues.put(PipelineConfigurator.USE_NER_CONLL.key, Configurator.TRUE);
                    break;
                case ViewNames.NER_ONTONOTES:
                    nonDefaultValues.put(PipelineConfigurator.USE_NER_ONTONOTES.key, Configurator.TRUE);
                    break;
                case ViewNames.QUANTITIES:
                    nonDefaultValues.put(PipelineConfigurator.USE_QUANTIFIER.key, Configurator.TRUE);
                    break;
                case ViewNames.SHALLOW_PARSE:
                    nonDefaultValues.put(PipelineConfigurator.USE_SHALLOW_PARSE.key, Configurator.TRUE);
                    break;
                case ViewNames.SRL_VERB:
                    nonDefaultValues.put(PipelineConfigurator.USE_SRL_VERB.key, Configurator.TRUE);
                    break;
                case ViewNames.DEPENDENCY_STANFORD:
                    nonDefaultValues.put(PipelineConfigurator.USE_STANFORD_DEP.key, Configurator.TRUE);
                    break;
                case ViewNames.DEPENDENCY:
                    nonDefaultValues.put(PipelineConfigurator.USE_DEP.key, Configurator.TRUE);
                    break;
                case ViewNames.PARSE_STANFORD:
                    nonDefaultValues.put(PipelineConfigurator.USE_STANFORD_PARSE.key, Configurator.TRUE);
                    break;
                case ViewNames.SRL_PREP:
                    nonDefaultValues.put(PipelineConfigurator.USE_SRL_PREP.key, Configurator.TRUE);
                    break;
                case ViewNames.SRL_COMMA:
                    nonDefaultValues.put(PipelineConfigurator.USE_SRL_COMMA.key, Configurator.TRUE);
                    break;
                default:
                    logger.warn("View name " + vu + " is not supported yet. Look into the readme of the pipeline to see the list of valid annotators. ");
            }
        } else {
            throw new IllegalArgumentException("The view name " + vu + " is not a valid view name. " + "The possible view names are static members of the class `ViewName`. ");
        }
    }
    if (disableCache) {
        nonDefaultValues.put(AnnotatorServiceConfigurator.DISABLE_CACHE.key, Configurator.TRUE);
    } else {
        nonDefaultValues.put(AnnotatorServiceConfigurator.DISABLE_CACHE.key, Configurator.FALSE);
    }
    // using the default settings and changing the views
    ResourceManager fullRm = (new PipelineConfigurator()).getConfig(new Stanford331Configurator().getConfig(nonDefaultValues));
    boolean splitOnHypen = fullRm.getBoolean(PipelineConfigurator.SPLIT_ON_DASH.key);
    TextAnnotationBuilder taBldr = new TokenizerTextAnnotationBuilder(new StatefulTokenizer(splitOnHypen));
    Map<String, Annotator> annotators = buildAnnotators(fullRm);
    return new SentencePipeline(taBldr, annotators, fullRm);
}

Example 7 with StatefulTokenizer

use of edu.illinois.cs.cogcomp.nlp.tokenizer.StatefulTokenizer in project cogcomp-nlp by CogComp.

the class NerOntonotesTest method testOntonotesNer.

@Test
public void testOntonotesNer() {
    TextAnnotationBuilder tab = new TokenizerTextAnnotationBuilder(new StatefulTokenizer());
    Properties props = new Properties();
    NERAnnotator nerOntonotes = NerAnnotatorManager.buildNerAnnotator(new ResourceManager(props), ViewNames.NER_ONTONOTES);
    TextAnnotation taOnto = tab.createTextAnnotation("", "", TEST_INPUT);
    try {
        nerOntonotes.getView(taOnto);
    } catch (AnnotatorException e) {
        e.printStackTrace();
        fail(e.getMessage());
    }
    View v = taOnto.getView(nerOntonotes.getViewName());
    assertEquals(v.getConstituents().size(), 4);
}

Example 8 with StatefulTokenizer

use of edu.illinois.cs.cogcomp.nlp.tokenizer.StatefulTokenizer in project cogcomp-nlp by CogComp.

the class MainClass method annotate.

private static void annotate(String filepath) throws IOException {
    DepAnnotator annotator = new DepAnnotator();
    TextAnnotationBuilder taBuilder = new TokenizerTextAnnotationBuilder(new StatefulTokenizer(true));
    Preprocessor preprocessor = new Preprocessor();
    Files.lines(Paths.get(filepath)).forEach(line -> {
        TextAnnotation ta = taBuilder.createTextAnnotation(line);
        try {
            preprocessor.annotate(ta);
            annotator.addView(ta);
            System.out.println(ta.getView(annotator.getViewName()).toString());
        } catch (AnnotatorException e) {
            e.printStackTrace();
        }
    });
}

Example 9 with StatefulTokenizer

use of edu.illinois.cs.cogcomp.nlp.tokenizer.StatefulTokenizer in project cogcomp-nlp by CogComp.

the class ExternalAnnotatorServiceFactory method buildPipeline.

/**
     * create an AnnotatorService with components specified by the ResourceManager (to override
     * defaults in {@link ExternalToolsConfigurator}
     *
     * @param rm non-default config options
     * @return AnnotatorService with specified NLP components
     * @throws IOException
     * @throws AnnotatorException
     */
public static BasicAnnotatorService buildPipeline(ResourceManager rm) throws IOException, AnnotatorException {
    // Merges default configuration with the user-specified overrides.
    ResourceManager fullRm = (new ExternalToolsConfigurator()).getConfig(rm);
    Boolean splitOnDash = fullRm.getBoolean(ExternalToolsConfigurator.SPLIT_ON_DASH);
    boolean isSentencePipeline = fullRm.getBoolean(ExternalToolsConfigurator.USE_SENTENCE_PIPELINE.key);
    TextAnnotationBuilder taBldr = new TokenizerTextAnnotationBuilder(new StatefulTokenizer(splitOnDash));
    Map<String, Annotator> annotators = buildAnnotators();
    return isSentencePipeline ? new SentencePipeline(taBldr, annotators, fullRm) : new BasicAnnotatorService(taBldr, annotators, fullRm);
}

Example 10 with StatefulTokenizer

use of edu.illinois.cs.cogcomp.nlp.tokenizer.StatefulTokenizer in project cogcomp-nlp by CogComp.

the class MultiLingualTokenizer method getTokenizer.

public static TextAnnotationBuilder getTokenizer(String lang) {
    if (tokenizerMap == null)
        tokenizerMap = new HashMap<>();
    if (!tokenizerMap.containsKey(lang)) {
        TextAnnotationBuilder tokenizer = null;
        if (lang.equals("en"))
            tokenizer = new TokenizerTextAnnotationBuilder(new StatefulTokenizer());
        else if (lang.equals("es"))
            tokenizer = new TokenizerTextAnnotationBuilder(new StanfordAnalyzer());
        else if (lang.equals("zh"))
            tokenizer = new TokenizerTextAnnotationBuilder(new CharacterTokenizer());
        else if (lang.equals("th"))
            tokenizer = new TokenizerTextAnnotationBuilder(new ThaiTokenizer());
        else
            tokenizer = new TokenizerTextAnnotationBuilder(new WhiteSpaceTokenizer());
        tokenizerMap.put(lang, tokenizer);
    }
    return tokenizerMap.get(lang);
}