/** Annotation remove event */
public void annotationRemoved(AnnotationSetEvent ase) {
if (!disableListener && ase.getSourceDocument() == this) {
AnnotationSet as = (AnnotationSet) ase.getSource();
Annotation annot = ase.getAnnotation();
FeatureMap features = Factory.newFeatureMap();
features.putAll(annot.getFeatures());
boolean defaultAS = as.getName() == null;
for (String docID : combinedDocumentIds) {
Document aDoc = compoundDocument.getDocument(docID);
// find out the details which refer to the deleted annotation
OffsetDetails od = getOffsetDetails(docID, as.getName(), annot);
if (od == null) continue;
if (defaultAS) {
aDoc.getAnnotations().remove(od.getOriginalAnnotation());
} else {
aDoc.getAnnotations(as.getName()).remove(od.getOriginalAnnotation());
}
removeOffsetDetails(docID, od);
break;
}
}
}
/** Test the default tokeniser */
public void testHashGazetteer() throws Exception {
// get a document
Document doc =
Factory.newDocument(new URL(TestDocument.getTestServerName() + "tests/doc0.html"));
System.out.println(doc.getFeatures().get("gate.SourceURL"));
// create a default gazetteer
FeatureMap params = Factory.newFeatureMap();
HashGazetteer gaz =
(HashGazetteer) Factory.createResource("com.ontotext.gate.gazetteer.HashGazetteer", params);
// runtime stuff
gaz.setDocument(doc);
gaz.setAnnotationSetName(GAZ_AS);
gaz.execute();
assertTrue(
"the Annotation set resulting of the execution of the OntoText "
+ "Natural Gazetteer is empty.",
!doc.getAnnotations(GAZ_AS).isEmpty());
// check whether the annotations are as expected
assertEquals("wrong number of lookup annotations found", 76, doc.getAnnotations(GAZ_AS).size());
} // testHashGazetteer();
public void tokenize() {
AnnotationSet tokenizationAs = gateDocument.getAnnotations("Tokenization");
AnnotationSet defaultAs = gateDocument.getAnnotations("");
for (Iterator<Annotation> it = tokenizationAs.iterator(); it.hasNext(); ) {
Annotation currentTokenAnnotation = it.next();
FeatureMap tokenFeaturesMap = currentTokenAnnotation.getFeatures();
FeatureMap curFeaturesMap = Factory.newFeatureMap();
if ("Token".compareToIgnoreCase(currentTokenAnnotation.getType()) == 0) {
curFeaturesMap.put("string", tokenFeaturesMap.get("string"));
curFeaturesMap.put("root", tokenFeaturesMap.get("lemma"));
curFeaturesMap.put("category", tokenFeaturesMap.get("POS"));
// Add the new Token to the Annotation Set
defaultAs.add(
currentTokenAnnotation.getStartNode(),
currentTokenAnnotation.getEndNode(),
currentTokenAnnotation.getType(),
curFeaturesMap);
}
}
gateDocument.removeAnnotationSet("Tokenization");
}
@SuppressWarnings("unchecked")
public Set<String> processDoc(String str) throws Exception {
Set<String> toReturn = new HashSet<String>();
Corpus c = null;
Document aDoc = null;
try {
c = Factory.newCorpus("sample");
aDoc = Factory.newDocument(str);
c.add(aDoc);
controller.setCorpus(c);
controller.execute();
AnnotationSet aSet = aDoc.getAnnotations("StockSymbols");
for (Annotation annot : aSet) {
String symbol = (String) annot.getFeatures().get("sym");
toReturn.add(symbol);
}
} catch (Exception e) {
throw e;
} finally {
if (aDoc != null) {
Factory.deleteResource(aDoc);
}
if (c != null) {
Factory.deleteResource(c);
}
}
return toReturn;
}
/** Annotation added event */
public void annotationAdded(AnnotationSetEvent ase) {
if (!disableListener && ase.getSourceDocument() == this) {
AnnotationSet as = (AnnotationSet) ase.getSource();
Annotation annot = ase.getAnnotation();
annot.addAnnotationListener(this);
FeatureMap features = Factory.newFeatureMap();
features.putAll(annot.getFeatures());
boolean defaultAS = as.getName() == null;
for (String docID : combinedDocumentIds) {
Document aDoc = compoundDocument.getDocument(docID);
long stOffset = getOffsetInSrcDocument(docID, annot.getStartNode().getOffset().longValue());
if (stOffset == -1) continue;
long enOffset = getOffsetInSrcDocument(docID, annot.getEndNode().getOffset().longValue());
if (enOffset == -1) continue;
Annotation originalAnnot = null;
try {
Integer id = annot.getId();
if (defaultAS) {
aDoc.getAnnotations()
.add(id, new Long(stOffset), new Long(enOffset), annot.getType(), features);
originalAnnot = aDoc.getAnnotations().get(id);
} else {
aDoc.getAnnotations(as.getName())
.add(id, new Long(stOffset), new Long(enOffset), annot.getType(), features);
originalAnnot = aDoc.getAnnotations(as.getName()).get(id);
}
} catch (InvalidOffsetException ioe) {
System.out.println(aDoc.getName() + "=" + stOffset + "=" + enOffset);
throw new GateRuntimeException(ioe);
}
OffsetDetails od = new OffsetDetails();
od.setOldStartOffset(stOffset);
od.setOldEndOffset(enOffset);
od.setNewStartOffset(annot.getStartNode().getOffset().longValue());
od.setNewEndOffset(annot.getEndNode().getOffset().longValue());
od.setOriginalAnnotation(originalAnnot);
od.setNewAnnotation(annot);
addNewOffsetDetails(docID, od);
break;
}
}
}
public void splitter() {
AnnotationSet sDetectionAS = gateDocument.getAnnotations("SentenceDetection");
AnnotationSet defaultAs = gateDocument.getAnnotations("");
for (Iterator<Annotation> it = sDetectionAS.iterator(); it.hasNext(); ) {
Annotation currentSentenceAnnotation = it.next();
// Add the Sentence to the Annotation Set
defaultAs.add(
currentSentenceAnnotation.getStartNode(),
currentSentenceAnnotation.getEndNode(),
"Sentence",
null);
}
gateDocument.removeAnnotationSet("SentenceDetection");
}
/**
* Generation of a GATE document from a Behemoth one
*
* @param key URL of the input doc
* @param inputDoc
* @return
* @throws ResourceInstantiationException
* @throws InvalidOffsetException
* @throws IOException
*/
public gate.Document generateGATEDoc(BehemothDocument inputDoc)
throws ResourceInstantiationException, InvalidOffsetException, IOException {
gate.Document gatedocument = null;
// if no text is available (e.g. Tika has not extracted it)
// let GATE do the parsing itself from the binary content
if (inputDoc.getText() == null) {
try {
gatedocument = generateGATEDocFromLocalDump(inputDoc);
// transfer the text from GATE to Behemoth
String textContent = gatedocument.getContent().toString();
inputDoc.setText(textContent);
return gatedocument;
} catch (Exception e) {
LOG.error("Can't generate GATE doc from byte dump", e);
}
}
// if the input document does not have any text -> create a doc with an
// empty text
String text = inputDoc.getText();
if (inputDoc.getText() == null) text = "";
else text = inputDoc.getText();
gatedocument = Factory.newDocument(text);
// then the metadata as document features
FeatureMap docFeatures = gatedocument.getFeatures();
String docUrl = inputDoc.getUrl();
if (docUrl != null) docFeatures.put("gate.SourceURL", docUrl);
if (inputDoc.getMetadata() != null) {
Iterator<Entry<Writable, Writable>> iter = inputDoc.getMetadata().entrySet().iterator();
while (iter.hasNext()) {
Entry<Writable, Writable> entry = iter.next();
String skey = entry.getKey().toString().trim();
String svalue = null;
if (entry.getValue() != null) svalue = entry.getValue().toString().trim();
docFeatures.put(skey, svalue);
}
}
// finally the annotations as original markups
// TODO change the name of the annotation set via config
AnnotationSet outputAS = gatedocument.getAnnotations("Original markups");
for (Annotation annot : inputDoc.getAnnotations()) {
// add to outputAS as a GATE annotation
FeatureMap features = Factory.newFeatureMap();
features.putAll(annot.getFeatures());
outputAS.add(annot.getStart(), annot.getEnd(), annot.getType(), features);
}
return gatedocument;
}
protected void doExecute(Document theDocument) throws ExecutionException {
interrupted = false;
if (theDocument == null) {
throw new ExecutionException("No document to process!");
}
AnnotationSet outputAS = theDocument.getAnnotations(getOutputAnnotationSet());
if (containingType == null || containingType.isEmpty()) {
annotateText(document, outputAS, 0, document.getContent().size());
} else {
AnnotationSet inputAS = null;
if (inputASName == null || inputASName.isEmpty()) {
inputAS = theDocument.getAnnotations();
} else {
inputAS = theDocument.getAnnotations(inputASName);
}
AnnotationSet containingAnns = inputAS.get(containingType);
for (Annotation containingAnn : containingAnns) {
annotateText(
document, outputAS, gate.Utils.start(containingAnn), gate.Utils.end(containingAnn));
}
}
}
@Override
public void execute() throws ExecutionException {
Document doc = getDocument();
AnnotationSet as = doc.getAnnotations(getAnnotationSetName());
AnnotationSet tocs = as.get(getTokenAnnotationTypeName());
try {
for (Annotation t : tocs) {
String content = Utils.stringFor(doc, t);
String val = getOrthographyValue(content);
if (val != null) t.getFeatures().put("orth", val);
}
} catch (Exception e) {
throw new ExecutionException(e);
}
}
/**
* Checks two documents for equality.
*
* @param doc1 a document
* @param doc2 another document
* @return a boolean.
*/
public static boolean documentsEqual(Document doc1, Document doc2) {
message = "";
if (doc1 == null ^ doc2 == null) {
message = "Documents not equal: null<>non-null!";
return false;
}
if (doc1 == null) return true;
if (!check(doc1.getContent(), doc2.getContent())) {
message = "Document contents different!";
return false;
}
if (!check(doc1.getAnnotations(), doc2.getAnnotations())) {
message = "Documents default AS not equal!";
return false;
}
if (doc1 instanceof TextualDocument) {
if (doc2 instanceof TextualDocument) {
if (!check(
((TextualDocument) doc1).getEncoding(), ((TextualDocument) doc2).getEncoding())) {
message = "Textual documents with different encodings!";
return false;
}
} else {
message = "Documents not equal: textual<>non-textual!";
return false;
}
}
if (!check(doc1.getFeatures(), doc2.getFeatures())) {
message = "Documents features not equal!";
return false;
}
// needs friend declaration :(
// if(!markupAware.equals(doc.markupAware)) return false;
if (!check(doc1.getNamedAnnotationSets(), doc2.getNamedAnnotationSets())) {
message = "Documents named annots not equal!";
return false;
}
// if(doc1 instanceof DocumentImpl){
// if(doc2 instanceof DocumentImpl){
// if(! check(((DocumentImpl)doc1).getNextNodeId(),
// ((DocumentImpl)doc2).getNextNodeId())){
// message = "Documents next nodeID not equal!";
// return false;
// }
// if(! check(((DocumentImpl)doc1).getNextAnnotationId(),
// ((DocumentImpl)doc2).getNextAnnotationId())){
// message = "Documents next annotationIDs not equal!";
// return false;
// }
// }else{
// message = "Documents not equal: DocumentImpl<>non-DocumentImpl!";
// return false;
// }
// }
if (!check(doc1.getSourceUrl(), doc2.getSourceUrl())) {
message = "Documents sourceURLs not equal!";
return false;
}
if (!(check(doc1.getSourceUrlStartOffset(), doc2.getSourceUrlStartOffset())
&& check(doc1.getSourceUrlEndOffset(), doc2.getSourceUrlEndOffset()))) {
message = "Documents sourceURLOffsets not equal!";
return false;
}
return true;
}
public void annotationSetRemoved(DocumentEvent de) {
Document doc = (Document) de.getSource();
if (this == doc) {
doc.getAnnotations(de.getAnnotationSetName()).removeAnnotationSetListener(this);
}
}
/**
* The main entry point. First we parse the command line options (see usage() method for details),
* then we take all remaining command line parameters to be file names to process. Each file is
* loaded, processed using the application and the results written to the output file
* (inputFile.out.xml).
*/
public static void main(String[] args) throws Exception {
parseCommandLine(args);
// initialise GATE - this must be done before calling any GATE APIs
Gate.init();
// load the saved application
CorpusController application =
(CorpusController) PersistenceManager.loadObjectFromFile(gappFile);
// Create a Corpus to use. We recycle the same Corpus object for each
// iteration. The string parameter to newCorpus() is simply the
// GATE-internal name to use for the corpus. It has no particular
// significance.
ArrayList<String> files = getFilesFromDir(inputDir);
gate.Corpus corpus = createCorpus(files);
// Corpus corpus = Factory.newCorpus("BatchProcessApp Corpus");
application.setCorpus(corpus);
System.out.println("Processing " + files.size() + " files");
// process the files one by one
for (int i = 0; i < files.size(); i++) {
// load the document (using the specified encoding if one was given)
File docFile = new File(files.get(i));
System.out.print("Processing document " + docFile + " (" + i + ") ...");
Document doc = Factory.newDocument(docFile.toURL(), encoding);
// put the document in the corpus
corpus.add(doc);
// run the application
application.execute();
// remove the document from the corpus again
corpus.clear();
String docXMLString = null;
// if we want to just write out specific annotation types, we must
// extract the annotations into a Set
if (annotTypesToWrite != null) {
// Create a temporary Set to hold the annotations we wish to write out
Set annotationsToWrite = new HashSet();
// we only extract annotations from the default (unnamed) AnnotationSet
// in this example
AnnotationSet defaultAnnots = doc.getAnnotations("Output");
Iterator annotTypesIt = annotTypesToWrite.iterator();
while (annotTypesIt.hasNext()) {
// extract all the annotations of each requested type and add them to
// the temporary set
AnnotationSet annotsOfThisType = defaultAnnots.get((String) annotTypesIt.next());
if (annotsOfThisType != null) {
annotationsToWrite.addAll(annotsOfThisType);
}
}
// create the XML string using these annotations
docXMLString = doc.toXml(annotationsToWrite, true);
}
// otherwise, just write out the whole document as GateXML
else {
docXMLString = doc.toXml();
}
// Release the document, as it is no longer needed
Factory.deleteResource(doc);
// output the XML to <inputFile>.out.xml
System.out.println("Writing file " + docFile.getName());
String outputFileName = docFile.getName() + ".out.xml";
// File outputFile = new File(docFile.getParentFile(), outputFileName);
File outputFile = new File(new File(outputDir).getAbsolutePath(), outputFileName);
// Write output files using the same encoding as the original
FileOutputStream fos = new FileOutputStream(outputFile);
BufferedOutputStream bos = new BufferedOutputStream(fos);
OutputStreamWriter out;
if (encoding == null) {
out = new OutputStreamWriter(bos);
} else {
out = new OutputStreamWriter(bos, encoding);
}
out.write(docXMLString);
out.close();
System.out.println("done");
} // for each file
System.out.println("All done");
} // void main(String[] args)
private RetObj ProcessRecords() throws Exception {
// Create a Corpus to use. We recycle the same Corpus object for each
// iteration.
Corpus corpus = Factory.newCorpus("BatchProcessApp Corpus");
this.application.setCorpus(corpus);
// object for returned data
List<String> processedlines = new ArrayList<String>();
List<String> processedText = new ArrayList<String>();
for (int record_num = 0; record_num < this.recs.size(); ++record_num) {
/*if( record_num % Math.ceil(((double) this.recs.size())/10.0) == 0)
System.out.println("Thread " + this.threadID + ": "+ ((int) ((double)record_num)/((double) this.recs.size())*100.0 ) +"% complete.");
*/
// first, split title from body and get embedded age in title..
String title_age = "-1";
String sep = "..THIS IS MY SEPARATION STRING..";
String title = "";
String body = this.recs.get(record_num);
Boolean trimmed = false;
int age_end = body.indexOf(",> ");
if (age_end >= 0 && age_end < body.length()) {
int age_start = body.lastIndexOf("-", age_end);
if (age_start >= 0 && age_start < age_end) {
title_age = body.substring(age_start + 1, age_end).trim();
if (!isInteger(title_age)) title_age = "-1";
else {
title = body.substring(0, age_start);
body = body.substring(age_end + 2, body.length());
body = title + sep + body;
trimmed = true;
}
}
if (!trimmed) {
title = body.substring(0, age_end);
body = body.substring(age_end + 2, body.length());
body = title + sep + body;
trimmed = true;
}
}
// --------------------
org.jsoup.nodes.Document htmldoc =
Jsoup.parseBodyFragment(body.replaceAll("COMMA_GOES_HERE", ","));
Elements links = htmldoc.select("a[href]");
Elements media = htmldoc.select("[src]");
Elements imports = htmldoc.select("link[href]");
processedText.add(htmldoc.text().replace(sep, " "));
Document doc = Factory.newDocument(htmldoc.text());
// put the document in the corpus
corpus.add(doc);
// run the application
this.application.execute();
// remove the document from the corpus again
corpus.clear();
// extract annotations
String line = "";
AnnotationSet Annots = doc.getAnnotations("");
Integer FirstPersonCount = 0, ThirdPersonCount = 0;
AnnotationSet FirstPerson = Annots.get("FirstPerson");
if (FirstPerson != null) FirstPersonCount = FirstPerson.size();
AnnotationSet ThirdPerson = Annots.get("ThirdPerson");
if (ThirdPerson != null) ThirdPersonCount = ThirdPerson.size();
line += FirstPersonCount.toString() + "," + ThirdPersonCount.toString() + ",";
AnnotationSet Names = Annots.get("Name");
if (Names == null || Names.size() < 1) line += ",";
else {
Iterator<Annotation> Iter = Names.inDocumentOrder().iterator();
while (Iter.hasNext()) {
Annotation Ann = Iter.next();
Object Feat = Ann.getFeatures().get("name");
if (Feat != null) line += Feat.toString();
if (Iter.hasNext()) line += ";";
}
line += ",";
}
AnnotationSet Age = Annots.get("Age");
if (Age == null || Age.size() < 1) line += title_age + ",";
else {
Iterator<Annotation> Iter = Age.inDocumentOrder().iterator();
line += title_age + ";";
while (Iter.hasNext()) {
Annotation Ann = Iter.next();
Object Feat = Ann.getFeatures().get("age");
if (Feat != null) line += Feat.toString();
if (Iter.hasNext()) line += ";";
}
line += ",";
}
AnnotationSet Cost = Annots.get("Cost");
if (Cost == null || Cost.size() < 1) line += ",";
else {
Iterator<Annotation> Iter = Cost.inDocumentOrder().iterator();
while (Iter.hasNext()) {
Annotation Ann = Iter.next();
Object Feat = Ann.getFeatures().get("value");
if (Feat != null) line += Feat.toString();
else line += "none";
line += "/";
Feat = Ann.getFeatures().get("target_value");
if (Feat != null) line += Feat.toString();
else line += "none";
line += "/";
Feat = Ann.getFeatures().get("target_type");
if (Feat != null) line += Feat.toString();
else line += "none";
if (Iter.hasNext()) line += ";";
}
line += ",";
}
AnnotationSet height = Annots.get("height");
if (height == null || height.size() < 1) line += ",,";
else {
String ft = "";
String inch = "";
Iterator<Annotation> Iter = height.inDocumentOrder().iterator();
while (Iter.hasNext()) {
Annotation Ann = Iter.next();
Object Feat = Ann.getFeatures().get("feet");
if (Feat != null) ft += Feat.toString();
else ft += "none";
Feat = Ann.getFeatures().get("inches");
if (Feat != null) inch += Feat.toString();
else inch += "none";
if (Iter.hasNext()) {
ft += ";";
inch += ";";
}
}
line += ft + "," + inch + ",";
}
AnnotationSet weight = Annots.get("weight");
if (weight == null || weight.size() < 1) line += ",";
else {
Iterator<Annotation> Iter = weight.inDocumentOrder().iterator();
while (Iter.hasNext()) {
Annotation Ann = Iter.next();
Object Feat = Ann.getFeatures().get("pounds");
if (Feat != null) line += Feat.toString();
if (Iter.hasNext()) line += ";";
}
line += ",";
}
AnnotationSet measurement = Annots.get("measurement");
if (measurement == null || measurement.size() < 1) line += ",,,,";
else {
String cup = "";
String chest = "";
String waist = "";
String hip = "";
Iterator<Annotation> Iter = measurement.inDocumentOrder().iterator();
while (Iter.hasNext()) {
Annotation Ann = Iter.next();
Object Feat = Ann.getFeatures().get("cup");
if (Feat != null) cup += Feat.toString();
else cup += "none";
Feat = Ann.getFeatures().get("chest");
if (Feat != null) chest += Feat.toString();
else chest += "none";
Feat = Ann.getFeatures().get("waist");
if (Feat != null) waist += Feat.toString();
else waist += "none";
Feat = Ann.getFeatures().get("hip");
if (Feat != null) hip += Feat.toString();
else hip += "none";
if (Iter.hasNext()) {
cup += ";";
chest += ";";
waist += ";";
hip += ";";
}
}
line += cup + "," + chest + "," + waist + "," + hip + ",";
}
AnnotationSet Ethnicity = Annots.get("Ethnicity");
if (Ethnicity == null || Ethnicity.size() < 1) line += ",";
else {
Iterator<Annotation> Iter = Ethnicity.inDocumentOrder().iterator();
while (Iter.hasNext()) {
Annotation Ann = Iter.next();
Object Feat = Ann.getFeatures().get("ethnicity");
if (Feat != null)
line += Feat.toString().toLowerCase().replaceAll(",", " ").replaceAll(";", " ");
if (Iter.hasNext()) line += ";";
}
line += ",";
}
AnnotationSet SkinColor = Annots.get("SkinColor");
if (SkinColor == null || SkinColor.size() < 1) line += ",";
else {
Iterator<Annotation> Iter = SkinColor.inDocumentOrder().iterator();
while (Iter.hasNext()) {
Annotation Ann = Iter.next();
Object Feat = Ann.getFeatures().get("color");
if (Feat != null)
line += Feat.toString().toLowerCase().replaceAll(",", " ").replaceAll(";", " ");
if (Iter.hasNext()) line += ";";
}
line += ",";
}
AnnotationSet EyeColor = Annots.get("EyeColor");
if (EyeColor == null || EyeColor.size() < 1) line += ",";
else {
Iterator<Annotation> Iter = EyeColor.inDocumentOrder().iterator();
while (Iter.hasNext()) {
Annotation Ann = Iter.next();
Object Feat = Ann.getFeatures().get("color");
if (Feat != null)
line += Feat.toString().toLowerCase().replaceAll(",", " ").replaceAll(";", " ");
if (Iter.hasNext()) line += ";";
}
line += ",";
}
AnnotationSet HairColor = Annots.get("HairColor");
if (HairColor == null || HairColor.size() < 1) line += ",";
else {
Iterator<Annotation> Iter = HairColor.inDocumentOrder().iterator();
while (Iter.hasNext()) {
Annotation Ann = Iter.next();
Object Feat = Ann.getFeatures().get("color");
if (Feat != null)
line += Feat.toString().toLowerCase().replaceAll(",", " ").replaceAll(";", " ");
if (Iter.hasNext()) line += ";";
}
line += ",";
}
AnnotationSet Restriction = Annots.get("Restriction");
if (Restriction == null || Restriction.size() < 1) line += ",,,";
else {
String type = "";
String ethnicity = "";
String age = "";
Iterator<Annotation> Iter = Restriction.inDocumentOrder().iterator();
while (Iter.hasNext()) {
Annotation Ann = Iter.next();
Object Feat = Ann.getFeatures().get("type");
if (Feat != null) type += Feat.toString();
else type += "none";
Feat = Ann.getFeatures().get("ethnicity");
if (Feat != null)
ethnicity += Feat.toString().toLowerCase().replaceAll(",", " ").replaceAll(";", " ");
else ethnicity += "none";
Feat = Ann.getFeatures().get("age");
if (Feat != null) age += Feat.toString();
else age += "none";
if (Iter.hasNext()) {
type += ";";
ethnicity += ";";
age += ";";
}
}
line += type + "," + ethnicity + "," + age + ",";
}
AnnotationSet Phone = Annots.get("PhoneNumber");
if (Phone == null || Phone.size() < 1) line += ",,,";
else {
String value = "";
String state = "";
String city = "";
Iterator<Annotation> Iter = Phone.inDocumentOrder().iterator();
while (Iter.hasNext()) {
Annotation Ann = Iter.next();
Object Feat = Ann.getFeatures().get("value");
if (Feat != null) value += Feat.toString();
else value += "none";
Feat = Ann.getFeatures().get("state");
if (Feat != null) state += Feat.toString();
else state += "none";
Feat = Ann.getFeatures().get("area");
if (Feat != null)
city += Feat.toString().toLowerCase().replaceAll(",", " ").replaceAll(";", " ");
else city += "none";
if (Iter.hasNext()) {
value += ";";
state += ";";
city += ";";
}
}
line += value + "," + state + "," + city + ",";
}
String Emails = "";
AnnotationSet Email = Annots.get("Email");
if (Email == null || Email.size() < 1) Emails = "";
else {
Iterator<Annotation> Iter = Email.inDocumentOrder().iterator();
while (Iter.hasNext()) {
Annotation Ann = Iter.next();
Object Feat = Ann.getFeatures().get("email");
if (Feat != null)
Emails += Feat.toString().toLowerCase().replaceAll(",", " ").replaceAll(";", " ") + ";";
}
}
if (links != null) {
for (Element l : links) {
String href = l.attr("abs:href");
if (href == null) continue;
if (href.length() > 7 && href.substring(0, 7).toLowerCase().equals("mailto:")) {
Emails +=
href.substring(7, href.length()).replaceAll(",", " ").replaceAll(";", " ") + ";";
}
}
}
if (Emails.length() > 0 && Emails.substring(Emails.length() - 1, Emails.length()).equals(";"))
Emails = Emails.substring(0, Emails.length() - 1);
line += Emails + ",";
String Urls = "";
AnnotationSet Url = Annots.get("Url");
if (Url == null || Url.size() < 1) Urls = "";
else {
Iterator<Annotation> Iter = Url.inDocumentOrder().iterator();
while (Iter.hasNext()) {
Annotation Ann = Iter.next();
Object Feat = Ann.getFeatures().get("url");
if (Feat != null)
Urls += Feat.toString().toLowerCase().replaceAll(",", " ").replaceAll(";", " ") + ";";
}
}
if (links != null) {
for (Element l : links) {
String href = l.attr("abs:href");
if (href == null) continue;
if (href.length() <= 7 || !href.substring(0, 7).toLowerCase().equals("mailto:")) {
Urls += href.replaceAll(",", " ").replaceAll(";", " ") + ";";
}
}
}
if (imports != null) {
for (Element l : imports) {
String href = l.attr("abs:href");
if (href == null) continue;
Urls += href.replaceAll(",", " ").replaceAll(";", " ") + ";";
}
}
if (Urls.length() > 0 && Urls.substring(Urls.length() - 1, Urls.length()).equals(";"))
Urls = Urls.substring(0, Urls.length() - 1);
line += Urls + ",";
String Medias = "";
if (media != null) {
for (Element l : media) {
String src = l.attr("abs:src");
if (src == null) continue;
Medias += src.replaceAll(",", " ").replaceAll(";", " ") + ";";
}
}
if (Medias.length() > 0 && Medias.substring(Medias.length() - 1, Medias.length()).equals(";"))
Medias = Medias.substring(0, Medias.length() - 1);
line += Medias;
processedlines.add(line);
// Release the document, as it is no longer needed
Factory.deleteResource(doc);
}
Factory.deleteResource(corpus);
RetObj out = new RetObj(processedlines, processedText);
return out;
}
/**
* Run from the command-line, with a list of URLs as argument.
*
* <p><B>NOTE:</B><br>
* This code will run with all the documents in memory - if you want to unload each from memory
* after use, add code to store the corpus in a DataStore.
*/
public static void main(String args[]) throws GateException, IOException {
// initialise the GATE library
Out.prln("Initialising GATE...");
Gate.init();
Out.prln("...GATE initialised");
// initialise ANNIE (this may take several minutes)
StandAloneAnnie annie = new StandAloneAnnie();
annie.initAnnie();
// create a GATE corpus and add a document for each command-line
// argument
Corpus corpus = Factory.newCorpus("StandAloneAnnie corpus");
for (int i = 0; i < args.length; i++) {
URL u = new URL(args[i]);
FeatureMap params = Factory.newFeatureMap();
params.put("sourceUrl", u);
params.put("preserveOriginalContent", new Boolean(true));
params.put("collectRepositioningInfo", new Boolean(true));
Out.prln("Creating doc for " + u);
Document doc = (Document) Factory.createResource("gate.corpora.DocumentImpl", params);
corpus.add(doc);
} // for each of args
// tell the pipeline about the corpus and run it
annie.setCorpus(corpus);
annie.execute();
// for each document, get an XML document with the
// person and location names added
Iterator iter = corpus.iterator();
int count = 0;
String startTagPart_1 = "<span GateID=\"";
String startTagPart_2 = "\" title=\"";
String startTagPart_3 = "\" style=\"background:Red;\">";
String endTag = "</span>";
while (iter.hasNext()) {
Document doc = (Document) iter.next();
AnnotationSet defaultAnnotSet = doc.getAnnotations();
Set annotTypesRequired = new HashSet();
annotTypesRequired.add("Person");
annotTypesRequired.add("Location");
Set<Annotation> peopleAndPlaces =
new HashSet<Annotation>(defaultAnnotSet.get(annotTypesRequired));
FeatureMap features = doc.getFeatures();
String originalContent =
(String) features.get(GateConstants.ORIGINAL_DOCUMENT_CONTENT_FEATURE_NAME);
RepositioningInfo info =
(RepositioningInfo) features.get(GateConstants.DOCUMENT_REPOSITIONING_INFO_FEATURE_NAME);
++count;
File file = new File("StANNIE_" + count + ".HTML");
Out.prln("File name: '" + file.getAbsolutePath() + "'");
if (originalContent != null && info != null) {
Out.prln("OrigContent and reposInfo existing. Generate file...");
Iterator it = peopleAndPlaces.iterator();
Annotation currAnnot;
SortedAnnotationList sortedAnnotations = new SortedAnnotationList();
while (it.hasNext()) {
currAnnot = (Annotation) it.next();
sortedAnnotations.addSortedExclusive(currAnnot);
} // while
StringBuffer editableContent = new StringBuffer(originalContent);
long insertPositionEnd;
long insertPositionStart;
// insert anotation tags backward
Out.prln("Unsorted annotations count: " + peopleAndPlaces.size());
Out.prln("Sorted annotations count: " + sortedAnnotations.size());
for (int i = sortedAnnotations.size() - 1; i >= 0; --i) {
currAnnot = (Annotation) sortedAnnotations.get(i);
insertPositionStart = currAnnot.getStartNode().getOffset().longValue();
insertPositionStart = info.getOriginalPos(insertPositionStart);
insertPositionEnd = currAnnot.getEndNode().getOffset().longValue();
insertPositionEnd = info.getOriginalPos(insertPositionEnd, true);
if (insertPositionEnd != -1 && insertPositionStart != -1) {
editableContent.insert((int) insertPositionEnd, endTag);
editableContent.insert((int) insertPositionStart, startTagPart_3);
editableContent.insert((int) insertPositionStart, currAnnot.getType());
editableContent.insert((int) insertPositionStart, startTagPart_2);
editableContent.insert((int) insertPositionStart, currAnnot.getId().toString());
editableContent.insert((int) insertPositionStart, startTagPart_1);
} // if
} // for
FileWriter writer = new FileWriter(file);
writer.write(editableContent.toString());
writer.close();
} // if - should generate
else if (originalContent != null) {
Out.prln("OrigContent existing. Generate file...");
Iterator it = peopleAndPlaces.iterator();
Annotation currAnnot;
SortedAnnotationList sortedAnnotations = new SortedAnnotationList();
while (it.hasNext()) {
currAnnot = (Annotation) it.next();
sortedAnnotations.addSortedExclusive(currAnnot);
} // while
StringBuffer editableContent = new StringBuffer(originalContent);
long insertPositionEnd;
long insertPositionStart;
// insert anotation tags backward
Out.prln("Unsorted annotations count: " + peopleAndPlaces.size());
Out.prln("Sorted annotations count: " + sortedAnnotations.size());
for (int i = sortedAnnotations.size() - 1; i >= 0; --i) {
currAnnot = (Annotation) sortedAnnotations.get(i);
insertPositionStart = currAnnot.getStartNode().getOffset().longValue();
insertPositionEnd = currAnnot.getEndNode().getOffset().longValue();
if (insertPositionEnd != -1 && insertPositionStart != -1) {
editableContent.insert((int) insertPositionEnd, endTag);
editableContent.insert((int) insertPositionStart, startTagPart_3);
editableContent.insert((int) insertPositionStart, currAnnot.getType());
editableContent.insert((int) insertPositionStart, startTagPart_2);
editableContent.insert((int) insertPositionStart, currAnnot.getId().toString());
editableContent.insert((int) insertPositionStart, startTagPart_1);
} // if
} // for
FileWriter writer = new FileWriter(file);
writer.write(editableContent.toString());
writer.close();
} else {
Out.prln("Content : " + originalContent);
Out.prln("Repositioning: " + info);
}
String xmlDocument = doc.toXml(peopleAndPlaces, false);
String fileName = new String("StANNIE_toXML_" + count + ".HTML");
FileWriter writer = new FileWriter(fileName);
writer.write(xmlDocument);
writer.close();
} // for each doc
} // main
// Process an input document with GATE and a Reporter
public synchronized BehemothDocument[] process(BehemothDocument inputDoc, Reporter reporter) {
if (reporter != null) reporter.setStatus("GATE : " + inputDoc.getUrl().toString());
boolean clearBehemothAnnotations = config.getBoolean("gate.deleteBehemothAnnotations", false);
// process the text passed as value with the application
// a) create a GATE document based on the text value
gate.Document gatedocument = null;
try {
gatedocument = generateGATEDoc(inputDoc);
// add it to the current corpus
corpus.add(gatedocument);
// get the application and assign the corpus to it
this.GATEapplication.setCorpus(corpus);
// process it with GATE
this.GATEapplication.execute();
AnnotationSet annots = null;
if ("".equals(filters.getAnnotationSetName())) annots = gatedocument.getAnnotations();
else annots = gatedocument.getAnnotations(filters.getAnnotationSetName());
// enrich the input doc with the annotations from
// the GATE application
// transfer the annotations from the GATE document
// to the Behemoth one using the filters
List<com.digitalpebble.behemoth.Annotation> beheannotations =
convertGATEAnnotationsToBehemoth(annots, inputDoc);
// sort the annotations before adding them?
Collections.sort(beheannotations);
// clear the existing behemoth annotations
if (clearBehemothAnnotations) {
inputDoc.getAnnotations().clear();
}
inputDoc.getAnnotations().addAll(beheannotations);
// add counters about num of annotations added
if (reporter != null)
for (com.digitalpebble.behemoth.Annotation annot : beheannotations) {
reporter.incrCounter("GATE", annot.getType(), 1);
}
// Add the document features from GATE to Behemoth
Set<String> docFeatFilter = this.filters.getDocFeaturesFilter();
MapWritable beheMD = inputDoc.getMetadata(true);
if (docFeatFilter.size() > 0) {
for (String docFeatName : docFeatFilter) {
Object featValue = gatedocument.getFeatures().get(docFeatName);
if (featValue != null) {
beheMD.put(new Text(docFeatName), new Text(featValue.toString()));
}
}
}
if (reporter != null) reporter.incrCounter("GATE", "Document", 1);
} catch (Exception e) {
LOG.error(inputDoc.getUrl().toString(), e);
if (reporter != null) reporter.incrCounter("GATE", "Exceptions", 1);
} finally {
// remove the document from the corpus again
corpus.clear();
// and from memory
if (gatedocument != null) Factory.deleteResource(gatedocument);
}
// currently returns only the input document
return new BehemothDocument[] {inputDoc};
}
/**
* This method annotates paragraphs in a GATE document. The investigated text spans beetween start
* and end offsets and the paragraph annotations are created in the annotSetName. If annotSetName
* is null then they are creted in the default annotation set.
*
* @param aDoc is the gate document on which the paragraph detection would be performed.If it is
* null or its content it's null then the method woul simply return doing nothing.
* @param startOffset is the index form the document content from which the paragraph detection
* will start
* @param endOffset is the offset where the detection will end.
* @param annotSetName is the name of the set in which paragraph annotation would be created.The
* annotation type created will be "paragraph"
*/
public void annotateParagraphs(Document aDoc, int startOffset, int endOffset, String annotSetName)
throws DocumentFormatException {
// Simply return if the document is null or its content
if (aDoc == null || aDoc.getContent() == null) return;
// Simply return if the start is > than the end
if (startOffset > endOffset) return;
// Decide where to put the newly detected annotations
AnnotationSet annotSet = null;
if (annotSetName == null) annotSet = aDoc.getAnnotations();
else annotSet = aDoc.getAnnotations(annotSetName);
// Extract the document content
String content = aDoc.getContent().toString();
// This is the offset marking the start of a para
int startOffsetPara = startOffset;
// This marks the ned of a para
int endOffsetPara = endOffset;
// The initial sate of the FSA
int state = 1;
// This field marks that a BR entity was read
// A BR entity can be NL or NL CR, depending on the operating system (UNIX
// or DOS)
boolean readBR = false;
int index = startOffset;
while (index < endOffset) {
// Read the current char
char ch = content.charAt(index);
// Test if a BR entity was read
if (ch == '\n') {
readBR = true;
// If \n is followed by a \r then advance the index in order to read a
// BR entity
while ((index + 1 < endOffset) && (content.charAt(index + 1) == '\r')) index++;
} // End if
switch (state) {
// It is the initial and also a final state
// Stay in state 1 while it reads whitespaces
case 1:
{
// If reads a non whitespace char then move to state 2 and record
// the beggining of a paragraph
if (!Character.isWhitespace(ch)) {
state = 2;
startOffsetPara = index;
} // End if
}
break;
// It can be also a final state.
case 2:
{
// Stay in state 2 while reading chars != BR entities
if (readBR) {
// If you find a BR char go to state 3. The possible end of the para
// can be index. This will be confirmed by state 3. So, this is why
// the end of a para is recorded here.
readBR = false;
endOffsetPara = index;
state = 3;
} // End if
}
break;
// It can be also a final state
// From state 3 there are only 2 possible ways: (state 2 or state1)
// In state 1 it needs to read a BR
// For state 2 it nead to read something different then a BR
case 3:
{
if (readBR) {
// A BR was read. Go to state 1
readBR = false;
state = 1;
// Create an annotation type paragraph
try {
annotSet.add(
new Long(startOffsetPara),
new Long(endOffsetPara),
"paragraph",
Factory.newFeatureMap());
} catch (gate.util.InvalidOffsetException ioe) {
throw new DocumentFormatException(
"Coudn't create a paragraph" + " annotation", ioe);
} // End try
} else {
// Go to state 2 an keep reading chars
state = 2;
} // End if
}
break;
} // End switch
// Prepare to read the next char.
index++;
} // End while
endOffsetPara = index;
// Investigate where the finite automata has stoped
if (state == 2 || state == 3) {
// Create an annotation type paragraph
try {
annotSet.add(
new Long(startOffsetPara),
// Create the final annotation using the endOffset
new Long(endOffsetPara),
"paragraph",
Factory.newFeatureMap());
} catch (gate.util.InvalidOffsetException ioe) {
throw new DocumentFormatException("Coudn't create a paragraph" + " annotation", ioe);
} // End try
} // End if
} // End annotateParagraphs();