SHOW:
|
|
- or go back to the newest paste.
| 1 | # nurble.py by http://twitter.com/johnradke | |
| 2 | import nltk | |
| 3 | ||
| 4 | ok = ['NN', 'NNP', 'NNS'] | |
| 5 | punc = [',', '.', '!', '?'] | |
| 6 | ||
| 7 | def nurbword(taggedWord): | |
| 8 | - | if taggedWord[1] in ok + punc: return taggedWord[0] |
| 8 | + | if taggedWord[1] in ok + punc: return taggedWord[0].upper() |
| 9 | return 'nurble' | |
| 10 | ||
| 11 | def untok(words): | |
| 12 | return "".join(words[0:1] + [w if w in punc else " " + w for w in words[1:]]) | |
| 13 | ||
| 14 | ############################# | |
| 15 | ||
| 16 | f = open('sotu.txt')
| |
| 17 | sotu = f.read() | |
| 18 | f.close() | |
| 19 | ||
| 20 | sentences = nltk.tokenize.sent_tokenize(sotu) | |
| 21 | ||
| 22 | taggedSentences = [nltk.pos_tag(nltk.word_tokenize(s)) for s in sentences] | |
| 23 | ||
| 24 | nurbled = open('nurbled.txt', 'w')
| |
| 25 | nurbled.write(" ".join(untok([nurbword(w) for w in s]) for s in taggedSentences)) |
