Source code for ngram.WordModels

# -*- coding: utf-8 -*-
# 
# (C) முத்தையா அண்ணாமலை 2015
# 
# N-gram language model for Tamil words

import tamil
def get_ngram_groups( word, n=1):    
    # makes no sense - but can happen
    
    letters = tamil.utf8.get_letters( word )
    L = len( letters )
    if n >= L:
        return [word]
    
    if n == 1:
        return letters
    
    result = []
    for i in range(0,L-n+1):
        result.append(u"".join(letters[i:min(i+n,L)]))
    return result