Tyler Smith wrote:
> iflist = ifstring.replace('\n', ' ').split()
The replace is not needed, split() splits on any whitespace including
newlines.
> # build hash of form (prefix1, prefix2): [word]
> word_hash = {}
> for i in range(len(iflist) - 2):
> pr1 = iflist[i]
> pr2 = iflist[i+1]
> tv = iflist[i+2]
> tk = (pr1, pr2)
or, tk = tuple(iflist[i:i+2])
> if word_hash.has_key(tk) and tv not in word_hash[tk]:
> word_hash[tk].append(tv)
> else:
> word_hash[tk] = [tv]
This could be done much more cleanly with a defaultdict that maps a pair
to a set:
from collections import defaultdict
wordhash = defaultdict(set)
...then just
wordhash[tk].add(tv)
> if word_hash.has_key((iflist[-2], iflist[-1])):
> word_hash[(iflist[-2], iflist[-1])].append('\n')
> else:
> word_hash[(iflist[-2], iflist[-1])] = ['\n']
word_hash[(iflist[-2], iflist[-1])].add('\n')
> # generate new words from hash
> word_num = 0
> while w2 <> '\n' and word_num < word_max:
> w1, w2 = w2, random.choice(word_hash[(w1, w2)])
> print w2,
> word_num += 1
or,
for word_num in range(word_max):
w1, w2 = w2, random.choice(word_hash[(w1, w2)])
print w2,
if w2 == '\n':
break
Kent
_______________________________________________
Tutor maillist - Tutor@python.org
http://mail.python.org/mailman/listinfo/tutor
