support/segment.py

   1 import sys
   2 from functools import lru_cache
   3 from support.language_models import Pwords
   4
   5 sys.setrecursionlimit(1000000)
   6
   7 @lru_cache()
   8 def segment(text):
   9     """Return a list of words that is the best segmentation of text.
  10     """
  11     if not text: return []
  12     candidates = ([first]+segment(rest) for first,rest in splits(text))
  13     return max(candidates, key=Pwords)
  14
  15 def splits(text, L=20):
  16     """Return a list of all possible (first, rest) pairs, len(first)<=L.
  17     """
  18     return [(text[:i+1], text[i+1:])
  19             for i in range(min(len(text), L))]
  20