/**
* Parse a sentence into a list of ConLLWord objects
*
* @param sentence list of line strings from the ConLL file
* @return
*/
private List<ConLLWord> transformSentenceToObjects(List<String> sentence, int startOffset) {
String sentenceString = "";
List<ConLLWord> sentenceObjects = new ArrayList<ConLLWord>();
int offset = startOffset;
boolean addSpaceToOffset = false;
for (String line : sentence) {
ConLLWord word = getWordFromCorpusLine(line);
// TODO: check if word is punctuation class word instead of this static check
if (word.getWordString().matches("([\\.!?,;:)]|'+)")) {
// remove whitespace before punctuation mark and decrement offset
if (sentenceString.endsWith(" ")) {
sentenceString = sentenceString.substring(0, sentenceString.length() - 1);
offset = offset - 1;
}
sentenceString += word.getWordString();
// no space here
} else if (word.getWordString().matches("([(]|`+)")) {
sentenceString += word.getWordString();
// add a space after the word
} else {
sentenceString += word.getWordString() + " ";
addSpaceToOffset = true;
}
word.setStart(offset);
int wordEnd = offset + word.getWordString().length();
word.setEnd(wordEnd);
offset = wordEnd;
if (addSpaceToOffset) {
offset++;
addSpaceToOffset = false;
}
sentenceObjects.add(word);
}
this.tempSentence = sentenceString.trim();
// return the length of the sentence added to the offset
return sentenceObjects;
}
