public void analyze(Document document) {
List<Sentence> sentences = document.getSentences();
for (Sentence sentence : sentences) {
Span[] contractionsSpan;
synchronized (this.contractionFinder) {
contractionsSpan = contractionFinder.find(TextUtils.tokensToString(sentence.getTokens()));
}
List<Token> newTokens = sentence.getTokens();
for (int i = contractionsSpan.length - 1; i >= 0; i--) {
int start = contractionsSpan[i].getStart();
String lexeme = sentence.getTokens().get(start).getLexeme();
String[] contractions = ContractionUtility.expand(lexeme);
Token original = newTokens.remove(start);
if (contractions != null) {
for (int j = contractions.length - 1; j >= 0; j--) {
Token token = new TokenImpl(original.getStart(), original.getEnd(), contractions[j]);
newTokens.add(start, token);
String caze = null;
if (j == 0) caze = "B";
else if (j == contractions.length - 1) caze = "E";
else caze = "I";
token.addContext(Analyzers.CONTRACTION_FINDER, caze);
}
} else {
LOGGER.debug("Missing contraction: " + lexeme);
}
}
sentence.setTokens(newTokens);
}
}
