anteater: src/de/mpiwg/anteater/ml/impl/ICUTextParser.java comparison

anteater

comparison

equal deleted inserted replaced

--1:000000000000
+:036535fcd179
+package de.mpiwg.anteater.ml.impl;
+import java.util.ArrayList;
+import java.util.List;
+import java.util.Locale;
+import com.ibm.icu.text.BreakIterator;
+import de.mpiwg.anteater.ml.ITextParser;
+public class ICUTextParser implements ITextParser {
+	@Override
+	public List<String> getSentences(String text) {
+		BreakIterator iterator = BreakIterator.getSentenceInstance(Locale.US);
+		iterator.setText(text);
+		List<String> sentences = new ArrayList<String>();
+		int start = iterator.first();
+		for (int end = iterator.next();
+		    end != BreakIterator.DONE;
+		    start = end, end = iterator.next()) {
+		  	sentences.add(text.substring(start,end));
+		}
+		return sentences;
+	}
+	@Override
+	public List<Word> getSubjects(String sentence) {
+		// TODO Auto-generated method stub
+		return null;
+	}
+	@Override
+	public List<Word> getAbbreviations(String sentence) {
+		// TODO Auto-generated method stub
+		return null;
+	}
+}

Mercurial > hg > anteater