Surname.java

//Dstl (c) Crown Copyright 2017
package uk.gov.dstl.baleen.annotators.cleaners;

import java.util.HashMap;
import java.util.Map;
import java.util.Map.Entry;
import java.util.regex.Matcher;
import java.util.regex.Pattern;

import org.apache.uima.analysis_engine.AnalysisEngineProcessException;
import org.apache.uima.fit.util.JCasUtil;
import org.apache.uima.jcas.JCas;

import com.google.common.collect.ImmutableSet;

import uk.gov.dstl.baleen.core.pipelines.orderers.AnalysisEngineAction;
import uk.gov.dstl.baleen.types.common.Person;
import uk.gov.dstl.baleen.types.semantic.ReferenceTarget;
import uk.gov.dstl.baleen.uima.BaleenAnnotator;

/**
 * From Person entities found in the document, extract surnames (taken to be the last
 * word in multi-word names) and identify other occurrences. If only one person has
 * that surname in the document, then coreference it.
 * 
 * @baleen.javadoc
 */
public class Surname extends BaleenAnnotator {

	@Override
	protected void doProcess(JCas jCas) throws AnalysisEngineProcessException {
		Map<String, Person> surnames = new HashMap<>();
		
		for(Person p : JCasUtil.select(jCas, Person.class)){
			String name = p.getCoveredText();
			if(!name.contains(" "))
				continue;
				
			String[] nameParts = name.split(" ");
			String surname = nameParts[nameParts.length - 1].toLowerCase();
			
			if(surnames.containsKey(surname)){
				//Existing surname, so ensure existing entity is removed so we don't do coreference
				//Unless the ReferenceTarget is the same, in which case no need to do anything
				Person pers = surnames.get(surname);
				if(pers == null)
					continue;
				
				ReferenceTarget rt = pers.getReferent();
				if(rt == null || !rt.equals(p.getReferent()))
					surnames.put(surname, null);
			}else{
				//New surname
				surnames.put(surname, p);
			}
		}
		
		for(Entry<String, Person> entry : surnames.entrySet()){
			findSurname(jCas, entry.getKey(), entry.getValue());
		}
	}

	private void findSurname(JCas jCas, String surname, Person original){
		Pattern pSurname = Pattern.compile("\\b" + Pattern.quote(surname) + "\\b", Pattern.CASE_INSENSITIVE);
		Matcher m = pSurname.matcher(jCas.getDocumentText());
		while(m.find()){
			if(!JCasUtil.selectCovering(jCas, Person.class, m.start(), m.end()).isEmpty())
				continue;
			
			Person p = new Person(jCas, m.start(), m.end());
			
			if(original != null){
				ReferenceTarget rt = original.getReferent();
				if(rt == null){
					rt = new ReferenceTarget(jCas);
					original.setReferent(rt);
				}
				p.setReferent(rt);
			}
			
			p.addToIndexes();
		}
	}
	
	@Override
	public AnalysisEngineAction getAction() {
		return new AnalysisEngineAction(ImmutableSet.of(Person.class), ImmutableSet.of(Person.class));
	}

}