markov.py

import re
from random import randint

RANK = 10

RE = re.compile("[\w']+")

dct = dict()

def _tuples(arr, r):
    return [(tuple([arr[i + t] for t in range(0, r - 1)]), arr[i + r - 1]) for i in range(len(arr) - (r - 1))]
	
#open("h2g2.txt").read().lower()
for k, v in _tuples(RE.findall(open("h2g2.txt").read().lower()), RANK):
	dct.setdefault(k, []).append(v)

#print dct
	
k = dct.keys()
start = k[randint(0, len(k) - 1)]
for i in range(32):
	lst = dct[start]
	next = lst[randint(0, len(lst) - 1)]
	print next,
	sl = list(start)
	sl.pop(0)
	sl.append(next)
	start = tuple(sl)