@Override
public void process(JCas jcas) throws AnalysisEngineProcessException {
UIMAProfiler.getProfiler("AnalysisEngine").start(this, "process");
Lemmatizer mateLemmatizer = mateLemmatizerModel.getEngine();
Tagger mateTagger = mateTaggerModel.getEngine();
/*
* keeps an array of annotations in memory so as to be able
* to access them by index.
*/
List<WordAnnotation> annotations = Lists.newArrayList();
FSIterator<Annotation> it = jcas.getAnnotationIndex(WordAnnotation.type).iterator();
while (it.hasNext()) {
WordAnnotation a = (WordAnnotation) it.next();
annotations.add(a);
}
String[] tokens = new String[annotations.size() + 2];
// preprends to fake words to prevent Mate from bugging on the two first words
tokens[0] = "<root>";
tokens[1] = "<root2>";
for (int i = 0; i < annotations.size(); i++)
tokens[i + 2] = annotations.get(i).getCoveredText();
SentenceData09 mateSentence = new SentenceData09();
mateSentence.init(tokens);
// Run POS tagging
mateSentence = mateTagger.apply(mateSentence);
// Run lemmatization
mateSentence = mateLemmatizer.apply(mateSentence);
WordAnnotation wordAnnotation;
for (int j = 1; j < mateSentence.length(); j++) {
wordAnnotation = annotations.get(j - 1);
wordAnnotation.setTag(mateSentence.ppos[j]);
wordAnnotation.setLemma(mateSentence.plemmas[j]);
}
UIMAProfiler.getProfiler("AnalysisEngine").stop(this, "process");
}
public synchronized List<StringAnnotation> annotate(IAnnotation<String> sentence)
throws IncompatibleAnnotationException {
if (!modelStarted()) {
startModel();
}
SentenceData09 sent = new SentenceData09();
ArrayList<StringAnnotation> annotations = new ArrayList<StringAnnotation>();
ArrayList<String> forms = new ArrayList<String>();
forms.add("<root>");
List<? extends IAnnotation<String>> tokens =
StringAnnotatorEnum.TOKEN.getAnnotator().annotate(sentence);
for (IAnnotation<String> token : tokens) {
forms.add(token.getAnnotation());
}
sent.init(forms.toArray(new String[forms.size()]));
lemmatizer.apply(sent);
for (int i = 0; i < tokens.size(); i++) {
StringAnnotation ann =
new StringAnnotation(
sent.plemmas[i + 1], tokens.get(i).getStart(), tokens.get(i).getEnd());
annotations.add(ann);
}
return annotations;
}
// how to parse a sentences and call the tools
public static void main(String[] args) throws IOException {
// Create a data container for a sentence
SentenceData09 i = new SentenceData09();
if (args.length == 1) { // input might be a sentence: "This is another test ."
StringTokenizer st = new StringTokenizer(args[0]);
ArrayList<String> forms = new ArrayList<>();
forms.add("<root>");
while (st.hasMoreTokens()) {
forms.add(st.nextToken());
}
i.init(forms.toArray(new String[0]));
} else {
// provide a default sentence
i.init(new String[] {"<root>", "This", "is", "a", "test", "."});
}
// print the forms
for (String l : i.forms) {
Parser.out.println("form : " + l);
}
// tell the lemmatizer the location of the model
is2.lemmatizer.Options optsLemmatizer =
new is2.lemmatizer.Options(new String[] {"-model", "models/lemma-eng.model"});
// create a lemmatizer
Lemmatizer lemmatizer = new Lemmatizer(optsLemmatizer);
// lemmatize a sentence; the result is stored in the stenenceData09 i
i = lemmatizer.apply(i);
// output the lemmata
for (String l : i.plemmas) {
Parser.out.println("lemma : " + l);
}
// tell the tagger the location of the model
is2.tag.Options optsTagger =
new is2.tag.Options(new String[] {"-model", "models/tag-eng.model"});
Tagger tagger = new Tagger(optsTagger);
// String pos[] =tagger.tag(i.forms, i.lemmas);
// i.setPPos(pos);
SentenceData09 tagged = tagger.tag(i);
for (String p : tagged.ppos) {
Parser.out.println("pos " + p);
}
// initialize the options
Options optsParser = new Options(new String[] {"-model", "models/prs-eng-x.model"});
// create a parser
Parser parser = new Parser(optsParser);
// parse the sentence (you get a copy of the input i)
SentenceData09 parse = parser.apply(tagged);
Parser.out.println(parse.toString());
// create some trash on the hard drive :-)
is2.io.CONLLWriter09 writer = new is2.io.CONLLWriter09("example-out.txt");
writer.write(i);
writer.finishWriting();
}