Code snippets for module 6

1. JAPE

1.1. Slide 17

Rule: HelloWorld
(
  {Token.string == "Hello"}
  {Token.string == "World"}
):hello
-->
{
  System.out.println("Hello world");
}

1.2. Exercise 1 (Slide 23)

Rule: ListEntities
({Person}|{Organization}|{Location}):ent
-->
{
  AnnotationSet ents = bindings.get("ent");
  for(Annotation e : ents) {
    System.out.println("Found " + e.getType()
        + " annotation");
    System.out.println("  features: "
        + e.getFeatures());
  }
}

1.3. Slide 25

-->
:uniTown{
  uniTownAnnots.iterator().next().getFeatures()
    .put("hasUniversity", Boolean.TRUE);
}

1.4. Slide 27

-->
:uniTown{
  String townString = doc.getContent().getContent(
        uniTownAnnots.firstNode().getOffset(),
        uniTownAnnots.lastNode().getOffset())
      .toString();
  // don't add an annotation if this town has been seen before.  If we
  // return, the UniversityTown annotation will not be created.
  if(!((Set)doc.getFeatures().get("knownTowns"))
      .add(townString)) return;
},
:uniTown.UniversityTown = {}

1.5. Slide 29

Rule: LcString
({Token}):tok
-->
:tok {
  for(Annotation a : tokAnnots) {
    // get the FeatureMap for the annotation
    FeatureMap fm = a.getFeatures();
    // get the ``string'' feature
    String str = (String)fm.get("string");
    // convert it to lower case and store
    fm.put("lcString", str.toLowerCase());
  }
}

1.6. Exercise 2 (Slide 30)

Rule: GeneralizePOSTag
({Token}):tok
-->
:tok {
  for(Annotation t : tokAnnots) {
    String pos = (String)t.getFeatures()
                          .get("category");
    if(pos != null) {
      int gpLen = pos.length();
      if(gpLen > 2) gpLen = 2;
      t.getFeatures().put("generalCategory",
          pos.substring(0, gpLen));
    }
  }
}

1.7. Slide 31

Rule: Location
({Lookup.majorType = "location"}):loc
-->
:loc.Location = { kind = :loc.Lookup.minorType,
      rule = "Location"},
:loc {
  inputAS.removeAll(locAnnots);
}

1.8. Slide 32

Rule: Location
({Lookup.majorType = "location"}):loc
-->
:loc {
  try {
    String str = doc.getContent().getContent(
        locAnnots.firstNode().getOffset(),
        locAnnots.lastNode().getOffset())
      .toString();
  }
  catch(InvalidOffsetException e) {
    // can't happen, but won't compile without the catch
  }
}

1.9. Slide 34

Rule: NPTokens
({NounPhrase}):np
-->
:np {
  List<String> posTags = new ArrayList<String>();
  for(Annotation tok : inputAS.get("Token")
      .getContained(start(npAnnots), end(npAnnots))) {
    posTags.add(
        (String)tok.getFeatures().get("category"));
  }
  FeatureMap fm =
    npAnnots.iterator().next().getFeatures();
  fm.put("posTags", posTags);
  fm.put("numTokens", (long)posTags.size());
}

1.10. Exercise 3 (Slide 35)

Imports: { import static gate.Utils.*; }
Phase: NumNouns
Input: Sentence
Options: control = appelt

Rule: CountNouns
({Sentence}):sent
-->
:sent {
  AnnotationSet tokens = inputAS.get("Token")
    .getContained(start(sentAnnots), end(sentAnnots));
  long numNouns = 0;
  for(Annotation t : tokens) {
    if("NN".equals(t.getFeatures()
                    .get("generalCategory"))) {
      numNouns++;
    }
  }
  sentAnnots.iterator().next().getFeatures()
    .put("numNouns", numNouns);
}

1.11. Slide 36

Rule: Section
({SectionHeading}):sect
-->
:sect {
  doc.getFeatures().put("currentSection",
      stringFor(doc, sectAnnots));
}

Rule: Entity
({Entity}):ent
-->
:ent {
  entAnnots.iterator().next().getFeatures()
    .put("inSection",
         doc.getFeatures().get("currentSection"));
}

2. Ontologies

2.1. Slide 46

Gate.init();
// load the Ontology plugin
Gate.getCreoleRegister().registerDirectories(
    new File(Gate.getPluginsHome(), "Ontology")
      .toURI().toURL());

Ontology emptyOnto = (Ontology)Factory.createResource(
    "gate.creole.ontology.impl.sesame.OWLIMOntology");

2.2. Slide 47

// init GATE and load plugin as before...

URL owl = new File("ontology.owl").toURI().toURL();
FeatureMap params = Factory.newFeatureMap();
params.put("rdfXmlURL", owl);

Ontology theOntology = (Ontology)Factory.createResource(
    "gate.creole.ontology.impl.sesame.OWLIMOntology",
    params);

2.3. Slide 50

// get all the `top' classes
Set<OClass> tops = ontology.getOClasses(true);

// list them along with their labels
for(OClass c : tops) {
  System.out.println(c.getONodeID() +
      " (" + c.getLabels() + ")");
}

// find a class by URI
OURI uri = ontology.createOURIForName("Person");
OClass personClass = ontology.getOClass(uri);

2.4. Slide 51

// get direct instances of a class
Set<OInstance> people = ontology.getOInstances(
    personClass, OConstants.Closure.DIRECT_CLOSURE);

// get instances of a class or any of its subclasses
Set<OInstance> allPeople = ontology.getOInstances(
    personClass, OConstants.Closure.TRANSITIVE_CLOSURE);

2.5. Slide 52

// get a datatype property
OURI namePropURI = ontology.createOURI(
    "http://example.org/stuff/1.0/hasName");
DatatypeProperty nameProp = ontology
    .getDatatypeProperty(namePropURI);

// find property values for an instance
for(OInstance person : allPeople) {
  List<Literal> names =
      ontology.getDatatypePropertyValues(nameProp);
  for(Literal name : names) {
    System.out.println("Person " + person.getONodeID()
        + " hasName " + name.toTurtle());
  }
}

2.6. Slide 53

// University of Sheffield instance
OURI uosURI = ontology.createOURIForName(
    "UniversityOfSheffield");
OInstance uosInstance = ontology.getOInstance(uosURI);

// worksFor property
OURI worksForURI = ontology.createOURIForName(
    "worksFor");
ObjectProperty worksFor = ontology.getObjectProperty(
    worksForURI);

// find all the people who work for the University of Sheffield
List<OResource> uniEmployees = 
    ontology.getOResourcesWith(worksFor, uosInstance);

2.7. Slide 55

OURI personURI = ontology.createOURIForName("Person");
OClass personClass = ontology.getOClass(personURI);

// create a new class as a subclass of an existing class
OURI empURI = ontology.createOURIForName("Employee");
OClass empClass = ontology.addOClass(empURI);
personClass.addSubClass(empClass);

// create an instance
OURI fredURI = ontology.createOURIForName("FredSmith");
OInstance fred = ontology.addOInstance(fredURI,
                                       empClass);

// Fred works for the University of Sheffield
fred.addObjectPropertyValue(worksFor, uosInstance);

2.8. Exercise 1 (Slide 59)

// Create an instance of the City class representing Sheffield
OURI cityURI = ontology.createOURIForName("City");
OClass cityClass = ontology.getOClass(cityURI);
OURI sheffieldURI = ontology.generateOURI("Sheffield");
OInstance sheffield = ontology.addOInstance(sheffieldURI,
                                            cityClass);

// Create a new class named "University" as a subclass of Organization
OURI orgURI = ontology.createOURIForName("Organization");
OURI uniURI = ontology.createOURIForName("University");
OClass orgClass = ontology.getOClass(orgURI);
OClass uniClass = ontology.addOClass(uniURI);
orgClass.addSubClass(uniClass);

// Create an instance of the University class representing the University of Sheffield
// note OUtils.toResourceName to escape spaces
OURI unishefURI = ontology.generateOURI(
    OUtils.toResourceName("University of Sheffield"));
OInstance unishef = ontology.addOInstance(unishefURI,
                                          uniClass);

// Create an object property basedAt with domain Organization and range Location
OURI locationURI = ontology.createOURIForName("Location");
OClass locationClass = ontology.getOClass(locationURI);
OURI basedAtURI = ontology.createOURIForName("basedAt");
ObjectProperty basedAt = ontology.addObjectProperty(
    basedAtURI, Collections.singleton(orgClass),
    Collections.singleton(locationClass));
//TIP: Collections.singleton returns an immutable set containing only the given 
object

// Specify that the University of Sheffield is basedAt Sheffield
try {
  unishef.addObjectPropertyValue(basedAt, sheffield);
} catch(InvalidValueException e) {
  throw new JapeException(e);
}

2.9. Exercise 2 (Slide 66)

// Create some useful objects - rdfs:label property and a Literal for the covered text.
AnnotationProperty rdfsLabel = ontology.getAnnotationProperty(
    ontology.createOURI(OConstants.RDFS.LABEL));
Literal text = new Literal(stringFor(doc, locAnnots));

for(Annotation m: locAnnots) {
  // determine the right class
  OURI classUri = ontology.createOURI(
      (String)m.getFeatures().get("class"));
  OClass clazz = ontology.getOClass(classUri);

  // get all existing instances of that class
  Set<OInstance> instances = ontology.getOInstances(clazz,
      OConstants.DIRECT_CLOSURE);

  // see if any of them have the right label -- if so, we assume they're the same
  OInstance inst = null;
  for(OInstance candidate : instances) {
    if(candidate.getAnnotationPropertyValues(
          rdfsLabel).contains(text)) {
      // found it!
      inst = candidate;
      break;
    }
  }

  if(inst == null) {
    // not found an existing instance, create one with a generated name
    String instName = OUtils.toResourceName(text.getValue());
    OURI instURI = ontology.generateOURI(instName + "_");
    inst = ontology.addOInstance(instURI, clazz);
    // and label it with the covered text
    inst.addAnnotationPropertyValue(rdfsLabel, text);
  }

  // finally, store the URI of the (new or existing) instance on the annotation
  m.getFeatures().put("inst",
      inst.getONodeID().toString());
}