T9 python

gistfile1.py

import re, sys

keys = ('abc', 'def', 'ghi', 'jkl', 'mno', 'pqrs', 'tuv', 'wxyz')
m = dict((l, str(n)) for n, letters in enumerate(keys, start=2) for l in letters)
data, ocurrences = {}, {}
wmatch = re.compile('[^a-z]+')

def learn(word):
    num = ''.join(m[c] for c in word)
    for i in xrange(1, len(word) + 1):
        inp = num[:i]

        if inp not in data:
            data[inp] = set([word])
        else:
            data[inp].add(word)

def train(f):
    for word in wmatch.split(f.read().lower()):
        if word in ocurrences:
            ocurrences[word] += 1
        else:
            learn(word)
            ocurrences[word] = 1

def search(n):
    if not n in data:
        sys.exit(0)
    results = ([], [])
    for match in data[n]:
        results[int(len(match) != len(n))].append((match, ocurrences[match]))

    k = lambda m: m[1]
    return (sorted(results[0], key=k, reverse=True),
            sorted(results[1], key=k, reverse=True))

if __name__ == '__main__':
    with open('alice_in_wonderland.txt', 'r') as f:
        train(f)

    results = search(sys.argv[1])

    for i, t in ((0, 'Exact'), (1, 'Prefix')):
        print "%s matches for %s:" % (t, sys.argv[1])
        for match, oc in results[i]:
            print match