package gate.opennlp; import java.io.DataInputStream; import java.io.IOException; import java.net.MalformedURLException; import java.net.URL; import java.util.Iterator; import java.util.zip.GZIPInputStream; import org.apache.log4j.Logger; import opennlp.maxent.MaxentModel; import opennlp.maxent.io.BinaryGISModelReader; import opennlp.tools.sentdetect.SentenceDetectorME; import opennlp.tools.tokenize.TokenizerME; import opennlp.tools.util.Span; import gate.Annotation; import gate.AnnotationSet; import gate.Factory; import gate.FeatureMap; import gate.Resource; import gate.creole.AbstractLanguageAnalyser; import gate.creole.ExecutionException; import gate.creole.ResourceInstantiationException; import gate.util.InvalidOffsetException; /** * Wrapper for the opennlp tokenizer * @author <A HREF="mailto:georgiev@ontotext.com>georgi.georgiev@ontotext.com</A> * Created: Thu Dec 11 16:25:59 EET 2008 */ public @SuppressWarnings("all") class OpenNlpTokenizer extends AbstractLanguageAnalyser { public static final long serialVersionUID = 1L; private static final Logger logger = Logger.getLogger(OpenNlpTokenizer.class); String inputASName; public String getInputASName() { return inputASName; } public void setInputASName(String inputASName) { this.inputASName = inputASName; } // private members private String annotationSetName = null; TokenizerME tokenizer = null; URL model; @Override public void execute() throws ExecutionException { // text doc annotations AnnotationSet annotations; if (annotationSetName != null && annotationSetName.length() > 0) annotations = document.getAnnotations(annotationSetName); else annotations = document.getAnnotations(); // get sentence annotations //AnnotationSet sentences = document.getAnnotations("Sentence"); // getdoc.get text String text = document.getContent().toString(); // run tokenizer Span[] spans = tokenizer.tokenizePos(text); // compare the resulting // spans and add annotations for (int i = 0; i < spans.length; i++) { FeatureMap fm = Factory.newFeatureMap(); // type fm.put("source", "openNLP"); fm.put("string", text.substring(spans[i].getStart(), spans[i] .getEnd())); // source // fm.put("type", "urn:lsid:ontotext.com:kim:iextraction:Token"); try { annotations.add(Long.valueOf(spans[i].getStart()), Long .valueOf(spans[i].getEnd()), "Token", fm); } catch (InvalidOffsetException e) { e.printStackTrace(); throw new RuntimeException(e); } //startSpan = spans[i].getEnd() + countSpaces(spans[i].getEnd()); } } int countSpaces(int lastSpan) { int ws = 0; String text = document.getContent().toString(); char[] context = text.substring(lastSpan - 1, text.length() >= lastSpan + 50 ? lastSpan + 50 : text.length()) .toCharArray(); for (int i = 0; i < context.length; i++) { if (Character.isWhitespace(context[i])) ws++; else break; } return ws; } @Override public Resource init() throws ResourceInstantiationException { //logger.info("The string of Tokenizer file is: "+model); tokenizer = new TokenizerME( getModel(model)); logger.warn("OpenNLP Tokenizer initialized!");//System.out.println("OpenNLP Tokenizer initialized!"); return this; } @Override public void reInit() throws ResourceInstantiationException { init(); } /** * @author georgiev * @return MaxentModel * @param String * path to MaxentModel */ public static MaxentModel getModel(URL name) { try { return new BinaryGISModelReader(new DataInputStream( new GZIPInputStream(name.openStream()))).getModel(); } catch (IOException E) { E.printStackTrace(); logger.error("OpenNLP Tokenizer can not be initialized!"); throw new RuntimeException("OpenNLP Tokenizer can not be initialized!", E); } } /* getters and setters for the PR */ /* public members */ public void setAnnotationSetName(String a) { annotationSetName = a; } public String getAnnotationSetName() { return annotationSetName; } public URL getModel() { return model; } public void setModel(URL model) { this.model = model; }/* getters and setters for the PR */ }