public synchronized String processNative(BehemothDocument inputDoc, Reporter reporter) {
if (reporter != null) reporter.setStatus("GATE : " + inputDoc.getUrl().toString());
// 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();
// transfer the annotations from the GATE document
// to the Behemoth one using the filters
if (reporter != null) reporter.incrCounter("GATE", "Document", 1);
return gatedocument.toXml();
} 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);
}
return null;
}
@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;
}
public JSONObject persian_sentiment(String text) throws Exception {
oncreate();
File PersianGapp = new File("C:/Users/mohammad/Desktop/New folder/Gate/application.xgapp");
// initialise GATE - this must be done before calling any GATE APIs
Gate.init();
// load the saved application
CorpusController application =
(CorpusController) PersistenceManager.loadObjectFromFile(PersianGapp);
// 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.
Corpus corpus = Factory.newCorpus("BatchProcessApp Corpus");
application.setCorpus(corpus);
// process the files one by one
// load the document (using the specified encoding if one was given)
Document doc = Factory.newDocument(text);
// put the document in the corpus
corpus.add(doc);
// run the application
application.execute();
String featureName = "Doc_sentiment";
FeatureMap features = doc.getFeatures();
// remove the document from the corpus again
corpus.clear();
// doc.getFeatures().
// Release the document, as it is no longer needed
Factory.deleteResource(doc);
LinkedHashMap originalContent = (LinkedHashMap) features.get(featureName);
String obj = (String) originalContent.get("sentiment");
// BigDecimal pos =(BigDecimal) originalContent.get("positive");
// BigDecimal neg =(BigDecimal) originalContent.get("negative");
// System.out.println(obj);
// create Json for response to user
JSONObject obj1 = new JSONObject();
obj1.put("sentiment", obj);
/*obj1.put("positive",pos);
//obj1.put("negative",neg);
System.out.print("----------");
System.out.print(obj1);
System.out.print("----------");*/
// application.cleanup();
return obj1;
}
/**
* Call the closure once for each document in this corpus, loading and unloading documents as
* appropriate in the case of a persistent corpus, and adding the return values of each call to
* the given collection.
*
* @param self the corpus to traverse
* @param closure the closure to call
* @return a list of the return values from each closure call.
*/
public static <T> Collection<T> collect(Corpus self, Collection<T> coll, Closure<T> closure) {
for (int i = 0; i < self.size(); i++) {
boolean docWasLoaded = self.isDocumentLoaded(i);
Document doc = self.get(i);
coll.add(closure.call(doc));
if (!docWasLoaded) {
self.unloadDocument(doc);
Factory.deleteResource(doc);
}
}
return coll;
}
/**
* Call the closure once for each document in this corpus, loading and unloading documents as
* appropriate in the case of a persistent corpus.
*
* @param self the corpus to traverse
* @param closure the closure to call
* @return the corpus.
*/
public static <T> Object each(Corpus self, Closure<T> closure) {
for (int i = 0; i < self.size(); i++) {
boolean docWasLoaded = self.isDocumentLoaded(i);
Document doc = self.get(i);
closure.call(doc);
if (!docWasLoaded) {
self.unloadDocument(doc);
Factory.deleteResource(doc);
}
}
return self;
}
private static gate.Corpus createCorpus(ArrayList<String> files) throws GateException {
gate.Corpus corpus = Factory.newCorpus("Transient Gate Corpus");
for (String file : files) {
System.out.print("\t " + file);
try {
corpus.add(Factory.newDocument(new File(file).toURL()));
System.out.println(" -- success");
} catch (gate.creole.ResourceInstantiationException e) {
System.out.println(" -- failed (" + e.getMessage() + ")");
} catch (Exception e) {
System.out.println(" -- " + e.getMessage());
}
}
return corpus;
}
/** Clear up the resources used after one test. */
private void clearOneTest() {
corpus.clear();
Factory.deleteResource(corpus);
Factory.deleteResource(learningApi);
controller.remove(learningApi);
controller.cleanup();
Factory.deleteResource(controller);
}
/**
* Constructor to create a SerialCorpus from a transient one. This is called by adopt() to store
* the transient corpus and re-route the methods calls to it, until the corpus is sync-ed on disk.
* After that, the transientCorpus will always be null, so the new functionality will be used
* instead.
*/
protected SerialCorpusImpl(Corpus tCorpus) {
// copy the corpus name and features from the one in memory
this.setName(tCorpus.getName());
this.setFeatures(tCorpus.getFeatures());
docDataList = new ArrayList<DocumentData>();
// now cache the names of all docs for future use
List<String> docNames = tCorpus.getDocumentNames();
for (int i = 0; i < docNames.size(); i++) {
Document doc = tCorpus.get(i);
docDataList.add(new DocumentData(docNames.get(i), null, doc.getClass().getName()));
}
// copy all the documents from the transient corpus
documents = new ArrayList<Document>();
documents.addAll(tCorpus);
// make sure we fire events when docs are added/removed/etc
Gate.getCreoleRegister().addCreoleListener(this);
}
public void setTransientSource(Object source) {
if (!(source instanceof Corpus)) return;
// the following initialisation is only valid when we're
// constructing
// this object from a transient one. If it has already been stored
// in
// a datastore, then the initialisation is done in readObject()
// since
// this method is the one called by serialisation, when objects
// are restored.
if (this.dataStore != null && this.lrPersistentId != null) return;
Corpus tCorpus = (Corpus) source;
// copy the corpus name and features from the one in memory
this.setName(tCorpus.getName());
this.setFeatures(tCorpus.getFeatures());
docDataList = new ArrayList<DocumentData>();
// now cache the names of all docs for future use
List<String> docNames = tCorpus.getDocumentNames();
for (int i = 0; i < docNames.size(); i++) {
Document aDoc = tCorpus.get(i);
docDataList.add(new DocumentData(docNames.get(i), null, aDoc.getClass().getName()));
}
// copy all the documents from the transient corpus
documents = new ArrayList<Document>();
documents.addAll(tCorpus);
this.addedDocs = new Vector<Document>();
this.removedDocIDs = new Vector<String>();
this.changedDocs = new Vector<Document>();
// make sure we fire events when docs are added/removed/etc
Gate.getCreoleRegister().addCreoleListener(this);
}
public static void main(String[] args) throws Exception {
// Logger.getLogger(DocumentFeaturesDiff.class).setLevel(Level.ALL);
GateUtils.initGateKeepLog();
GateUtils.registerCzsemPlugin();
ProcessingResource eval =
new PRSetup.SinglePRSetup(LearningEvaluator.class)
.putFeature("keyASName", ":-)")
// .putFeature("responseASName", "lemma_flex")
.putFeature("responseASName", "flex")
.putFeature("keyAnnotationsAreInDocumentFeatures", true)
.putFeatureList("annotationTypes", "Lookup")
.putFeatureList("featureNames", "meshID")
.createPR();
SerialAnalyserController controller =
(SerialAnalyserController)
Factory.createResource(SerialAnalyserController.class.getCanonicalName());
controller.add(eval);
Corpus corpus = Factory.newCorpus(null);
corpus.populate(
new File("C:\\Users\\dedek\\Desktop\\bmc\\experiment\\analyzed").toURI().toURL(),
// new File("C:\\Users\\dedek\\Desktop\\bmca_devel").toURI().toURL(),
null,
"utf8",
false);
System.err.println("populated");
controller.setCorpus(corpus);
controller.execute();
}
/**
* Loading the configurationg file and corpus for testing. And make settings as in the GATE Gui.
*/
void loadSettings(String configFileName, String corpusDirName, String inputasN, String outputasN)
throws GateException, IOException {
LogService.minVerbosityLevel = 0;
if (LogService.minVerbosityLevel > 0)
System.out.println("Learning Home : " + learningHome.getAbsolutePath());
FeatureMap parameters = Factory.newFeatureMap();
URL configFileURL = new File(configFileName).toURI().toURL();
parameters.put("configFileURL", configFileURL);
learningApi =
(LearningAPIMain) Factory.createResource("gate.learning.LearningAPIMain", parameters);
// Load the corpus
corpus = Factory.newCorpus("DataSet");
ExtensionFileFilter fileFilter = new ExtensionFileFilter();
fileFilter.addExtension("xml");
File[] xmlFiles = new File(corpusDirName).listFiles(fileFilter);
Arrays.sort(
xmlFiles,
new Comparator<File>() {
public int compare(File a, File b) {
return a.getName().compareTo(b.getName());
}
});
for (File f : xmlFiles) {
if (!f.isDirectory()) {
Document doc = Factory.newDocument(f.toURI().toURL(), "UTF-8");
doc.setName(f.getName());
corpus.add(doc);
}
}
// URL tempURL = new File(corpusDirName).toURI().toURL();
// corpus.populate(tempURL, fileFilter, "UTF-8", false);
// Set the inputAS
learningApi.setInputASName(inputasN);
learningApi.setOutputASName(outputasN);
controller =
(gate.creole.SerialAnalyserController)
Factory.createResource("gate.creole.SerialAnalyserController");
controller.setCorpus(corpus);
controller.add(learningApi);
}
/**
* @param args
* @throws Exception
*/
public static void main(String[] args) throws Exception {
/* Parse command line arguments */
Getopt g = new Getopt("gateExtractor", args, "i:r:te");
g.setOpterr(false);
String inputPath = "";
String outputPath = "";
boolean train = false;
boolean eval = false;
boolean run = false;
int c;
String arg;
while ((c = g.getopt()) != -1) {
switch (c) {
case 'i':
arg = g.getOptarg();
if (arg == null || arg.isEmpty()) {
usage("Please provide an input path");
}
inputPath = arg;
break;
case 'r':
run = true;
arg = g.getOptarg();
if (arg == null || arg.isEmpty()) {
usage("Please provide an output path");
}
outputPath = arg;
break;
case 't':
train = true;
break;
case 'e':
eval = true;
break;
case '?':
default:
usage(null);
}
}
if (args.length == 0 || (!run && !train && !eval)) {
usage("Nothing to do.");
}
if (inputPath == null || inputPath.isEmpty()) {
usage("Please provide an input path");
}
if (run && (outputPath == null || outputPath.isEmpty())) {
usage("Please provide an output directory!");
}
if (train && eval) {
usage("Only one mode allowed at a time");
}
if (train && run) {
usage("Only one mode allowed at a time");
}
if (eval && run) {
usage("Only one mode allowed at a time");
}
/* Initialize GATE */
String location =
new File(Main.class.getProtectionDomain().getCodeSource().getLocation().getPath())
.getParent();
String resourcesFolder = location + "/resources";
Gate.setGateHome(new File(resourcesFolder));
/* Create ml-config.xml with threads */
createConfig(resourcesFolder + File.separator);
Gate.init();
/* Load Corpus */
log.info("Loading Corpus ... ");
Corpus corpus = Factory.newCorpus("Training Corpus");
File directory = new File(inputPath);
URL url = directory.toURI().toURL();
corpus.populate(url, null, null, true);
log.info("Done loading Corpus!");
Pipeline pipeline = null;
/* Do Tagging */
pipeline = new Tagger();
pipeline.run(corpus, resourcesFolder);
/* Train */
if (train) {
pipeline = new Trainer();
pipeline.run(corpus, resourcesFolder);
}
/* Apply learned rules */
if (run) {
pipeline = new Extractor();
pipeline.run(corpus, resourcesFolder);
ExecutorService executorService = Executors.newFixedThreadPool(20);
for (int i = 0; i < corpus.size(); i++) {
executorService.execute(new OutputGenerator(outputPath, corpus.get(i)));
}
executorService.shutdown();
executorService.awaitTermination(Long.MAX_VALUE, TimeUnit.DAYS);
}
/* Evaluate results */
if (eval) {
pipeline = new Evaluator();
pipeline.run(corpus, resourcesFolder);
}
/* Clean up */
Factory.deleteResource(corpus);
outputFile_mlConfigThreads.delete();
}
/**
* 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;
}
// 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};
}
/**
* 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
public static void main(String[] args) throws Exception {
GateUtils.initGate();
ie =
new SpcInteractionExport(
"SpcInteractionExport/SpcInteractionExport_" + ProjectSetup.makeTimeStamp() + ".csv");
/**/
initController();
/**/
CsvReader in =
new CsvReader("C:/Users/dedek/Desktop/DATLOWE/LP_SPC.csv", ';', Charset.forName("cp1250"));
in.readHeaders();
StopRequestDetector srd = new StopRequestDetector();
srd.startDetector();
int num = 0;
while (!srd.stop_requested) {
if (!in.readRecord()) break;
String spcCode = in.get(0);
ie.setSpcCode(spcCode);
String doc = in.get(3);
String pdf = in.get(4);
String fileStr = pdf;
File file = null;
if (!fileStr.isEmpty()) {
file = new File("C:/Users/dedek/Desktop/DATLOWE/SPC_all/" + fileStr);
if (!file.exists()) fileStr = "";
}
if (fileStr.isEmpty()) {
fileStr = doc;
}
if (!fileStr.isEmpty()) {
file = new File("C:/Users/dedek/Desktop/DATLOWE/SPC_all/" + fileStr);
if (!file.exists()) fileStr = "";
}
System.err.format(
"%4d %s record: %s %s %s\n",
++num, ProjectSetup.makeTimeStamp(), spcCode, in.get(1), fileStr);
if (fileStr.isEmpty()) {
System.err.format("WARNING ommiting record: %s %s\n", in.get(0), in.get(1));
continue;
}
Document gateDoc = Factory.newDocument(file.toURI().toURL(), "utf8");
corpus.add(gateDoc);
try {
controller.execute();
} catch (ExecutionException e) {
System.err.println("----------------------- EXECUTION INTERUPTED -------------------");
e.printStackTrace();
initController();
System.err.println("----------------------- EXECUTION RESTARTED -------------------");
}
corpus.clear();
// GateUtils.saveGateDocumentToXML(gateDoc,
// "C:/Users/dedek/Desktop/DATLOWE/SPC_ananlyzed/"+spcCode+".xml");
Factory.deleteResource(gateDoc);
}
in.close();
ie.close();
GateUtils.deleteAllPublicGateResources();
}
