/**
* Initialize resources and processors that has been config, and register them in Resource Manager
*
* @param propertiesManager
* @return
*/
public boolean initialize(PropertiesManager propertiesManager) {
assert propertiesManager != null;
this.propertiesManager = propertiesManager;
// this.configurables = new ArrayList<Configurable>();
// read the config and initialize those resources and processors that has been config
// TODO; only initialize the resources and processors by reflection from the feature sets, has
// to change the framework to start
try {
String srcLang = this.getSrcLang();
String trgLang = this.getTrgLang();
// slang dict only for the target
SlangDictionary trgSlangDict = new SlangDictionary(trgLang);
if (trgSlangDict.isConfigured(propertiesManager)) {
trgSlangDict.load(propertiesManager);
trgSlangDict.register();
VariantsSlangProcessor p1 = new VariantsSlangProcessor(trgSlangDict);
this.trgResourceProcessors.add(p1);
}
// abbreviation
AbbreviationDictionary abbrevDict = new AbbreviationDictionary(trgLang);
if (abbrevDict.isConfigured(propertiesManager)) {
abbrevDict.load(propertiesManager);
abbrevDict.register();
AbbreviationsProcessor p2 = new AbbreviationsProcessor(abbrevDict);
this.trgResourceProcessors.add(p2);
}
// other features
} catch (Exception e) {
isInitialized = false;
return false;
}
isInitialized = true;
return true;
}
@Override
public void processNextSentence(Sentence sentence) {
assert abbreviationDictionary != null;
String strLine = sentence.getText();
int abbrevConflicts = 0;
Set<String> abbrevs = abbreviationDictionary.getAbbrevSet();
for (String abbrev : abbrevs) {
int pos = 0;
for (String word : sentence.getTokens()) {
if (word.equals(abbrev)) {
String position = sentence.getIndex() + "-" + pos;
for (Map.Entry<String, String> entry : position2abbrev.entrySet()) {
String aPos = entry.getKey();
String aAbbrev = entry.getValue();
if (aAbbrev != abbrev) { // not the same one
// find how close they are by meaning
Set<String> meaningSetA =
new HashSet<String>(abbreviationDictionary.getMeaningSetOfAbbreviation(aAbbrev));
Set<String> meaningSetB = abbreviationDictionary.getMeaningSetOfAbbreviation(abbrev);
meaningSetA.retainAll(meaningSetB);
if (meaningSetA.size() > 0) {
abbrevConflicts++;
}
}
}
}
}
}
sentence.setValue("abbrev_conflicts", abbrevConflicts); // number of conflicts
if (position2abbrev.size() > 0) {
sentence.setValue(
"abbrev_conflicts_divided_by_count", abbrevConflicts * 1.0 / position2abbrev.size());
} else {
sentence.setValue("abbrev_conflicts_divided_by_count", 0.0);
}
}
@Override
public void globalProcessing(Context context) {
assert abbreviationDictionary != null;
Set<String> abbrevs = abbreviationDictionary.getAbbrevSet();
BufferedReader br = null;
try {
br =
new BufferedReader(
new InputStreamReader(new FileInputStream(context.getTargetFilePath())));
String strLine;
int lineCount = 0;
while ((strLine = br.readLine()) != null) {
strLine = strLine.trim();
for (String abbrev : abbrevs) {
int pos = 0;
for (String word : strLine.split("\\s+")) {
if (word.equals(abbrev)) {
String position = lineCount + "-" + pos;
position2abbrev.put(position, abbrev);
}
pos++;
}
}
lineCount++;
}
} catch (Exception e) {
e.printStackTrace();
} finally {
if (br != null) {
try {
br.close();
} catch (IOException e) {
e.printStackTrace();
}
}
}
}