fr.univnantes.termsuite.api.TerminoExtractor Maven / Gradle / Ivy

Go to download
Show more of this group Show more artifacts with this name
Show all versions of termsuite-core Show documentation
A Java UIMA-based toolbox for multilingual and efficient terminology extraction an multilingual term alignment
The newest version!
package fr.univnantes.termsuite.api;

import java.util.Optional;

import org.slf4j.Logger;
import org.slf4j.LoggerFactory;

import fr.univnantes.termsuite.engines.TerminologyExtractorEngine;
import fr.univnantes.termsuite.framework.Pipeline;
import fr.univnantes.termsuite.framework.PipelineStats;
import fr.univnantes.termsuite.framework.TermSuiteFactory;
import fr.univnantes.termsuite.model.IndexedCorpus;
import fr.univnantes.termsuite.model.occurrences.EmptyOccurrenceStore;
import fr.univnantes.termsuite.utils.TermHistory;

/**
 * 
 * A builder and launcher class for execute a terminology extraction
 * pipeline from a PreprocessedCorpus or a Terminology.
 * 
 * @author Damien Cram
 * 
 * @see Preprocessor
 *
 */
public class TerminoExtractor {
	private static final Logger LOGGER = LoggerFactory.getLogger(TerminoExtractor.class);
	
	private Optional history = Optional.empty();
	private ExtractorOptions options;
	private Optional resourceConfig = Optional.empty();
	private Optional stats = Optional.empty();

	private Optional listener = Optional.empty();

	public TerminoExtractor setResourceConfig(ResourceConfig resourceConfig) {
		this.resourceConfig = Optional.of(resourceConfig);
		return this;
	}
	
	public TerminoExtractor watch(String termString) {
		if(!this.history.isPresent()) {
			this.history = Optional.of(new TermHistory());
		}
		this.history.get().addWatchedTermString(termString);
		return this;
	}

	public TerminoExtractor setHistory(TermHistory history) {
		this.history = Optional.of(history);
		return this;
	}

	public TerminoExtractor setListener(PipelineListener listener) {
		this.listener  = Optional.of(listener);
		return this;
	}

	public TerminoExtractor setOptions(ExtractorOptions options) {
		this.options = options;
		return this;
	}
	
	public IndexedCorpus execute(XMICorpus corpus, int maxSize) {
		IndexedCorpus iCorpus = TermSuite.toIndexedCorpus(corpus, maxSize);
		execute(iCorpus);
		return iCorpus;
	}
	
	/**
	 * Return the {@link PipelineStats} of the terminology
	 * extraction pipeline. 
	 * 
	 * @return
	 * 		the {@link Optional} {@link PipelineStats}, {@link Optional#empty()}
	 * 		if pipeline not finished.
	 */
	public Optional getStats() {
		return stats;
	}
	
	public PipelineStats execute(IndexedCorpus corpus) {
		if(options == null)
			options = TermSuite.getDefaultExtractorConfig(corpus.getTerminology().getLang());
		checkConfig(corpus);
		
		Pipeline pipeline = TermSuiteFactory.createPipeline(
				TerminologyExtractorEngine.class, 
				corpus,
				resourceConfig.orElse(null),
				history.orElse(null),
				listener.orElse(null),
				options
				);
		PipelineStats stats = pipeline.run();
		this.stats = Optional.of(stats);
		return stats;
	}

	private void checkConfig(IndexedCorpus corpus) {
		if(options.getContextualizerOptions().isEnabled()) {
			if(corpus.getOccurrenceStore() instanceof EmptyOccurrenceStore) 
				throw new ConfigurationException("Cannot contextualize with no occurrence. Got occurrence store class: " + EmptyOccurrenceStore.class);
			
			if(options.getPostProcessorConfig().isEnabled()) {
				LOGGER.warn("Deactivating post-processor for contextualizer, otherwise too many terms would be removed by post-porcessor.");
				options.getPostProcessorConfig().setEnabled(false);
			}
		}
		
		if(options.getGathererConfig().isSemanticEnabled() && !options.getContextualizerOptions().isEnabled()) {
			LOGGER.warn("Enabling contextualizer for semantic alignment");
			options.getContextualizerOptions().setEnabled(true);
		}
	}
}