2015-10-09 07:44:21 +00:00
|
|
|
from __future__ import unicode_literals, print_function
|
2014-12-23 04:16:57 +00:00
|
|
|
from os import path
|
2015-01-04 19:01:32 +00:00
|
|
|
import codecs
|
2014-12-23 04:16:57 +00:00
|
|
|
|
2015-08-27 07:16:11 +00:00
|
|
|
try:
|
|
|
|
import ujson as json
|
|
|
|
except ImportError:
|
|
|
|
import json
|
|
|
|
|
2015-10-09 07:44:21 +00:00
|
|
|
from .parts_of_speech import NOUN, VERB, ADJ, PUNCT
|
2016-01-13 18:46:17 +00:00
|
|
|
from .util import get_package
|
2015-08-27 07:16:11 +00:00
|
|
|
|
2014-12-23 04:16:57 +00:00
|
|
|
|
|
|
|
class Lemmatizer(object):
|
2015-08-27 07:16:11 +00:00
|
|
|
@classmethod
|
2016-01-15 17:01:02 +00:00
|
|
|
def load(cls, via):
|
|
|
|
pkg = get_package(via)
|
2015-08-27 07:16:11 +00:00
|
|
|
index = {}
|
|
|
|
exc = {}
|
2015-09-12 04:03:44 +00:00
|
|
|
for pos in ['adj', 'noun', 'verb']:
|
2015-12-29 15:56:02 +00:00
|
|
|
with pkg.open(('wordnet', 'index.%s' % pos), default=None) as file_:
|
|
|
|
index[pos] = read_index(file_) if file_ is not None else set()
|
|
|
|
with pkg.open(('wordnet', '%s.exc' % pos), default=None) as file_:
|
|
|
|
exc[pos] = read_exc(file_) if file_ is not None else {}
|
2015-12-31 03:13:15 +00:00
|
|
|
rules = pkg.load_json(('vocab', 'lemma_rules.json'), default={})
|
2015-08-27 07:16:11 +00:00
|
|
|
return cls(index, exc, rules)
|
2014-12-23 04:16:57 +00:00
|
|
|
|
2015-08-27 07:16:11 +00:00
|
|
|
def __init__(self, index, exceptions, rules):
|
|
|
|
self.index = index
|
|
|
|
self.exc = exceptions
|
|
|
|
self.rules = rules
|
2015-08-25 13:46:19 +00:00
|
|
|
|
2015-08-27 07:16:11 +00:00
|
|
|
def __call__(self, string, pos):
|
|
|
|
if pos == NOUN:
|
|
|
|
pos = 'noun'
|
|
|
|
elif pos == VERB:
|
|
|
|
pos = 'verb'
|
|
|
|
elif pos == ADJ:
|
|
|
|
pos = 'adj'
|
2015-10-09 07:44:21 +00:00
|
|
|
elif pos == PUNCT:
|
|
|
|
pos = 'punct'
|
2015-09-10 12:51:39 +00:00
|
|
|
lemmas = lemmatize(string, self.index.get(pos, {}), self.exc.get(pos, {}), self.rules.get(pos, []))
|
2015-09-08 13:38:03 +00:00
|
|
|
return lemmas
|
2014-12-23 04:16:57 +00:00
|
|
|
|
|
|
|
def noun(self, string):
|
2015-08-25 13:46:19 +00:00
|
|
|
return self(string, 'noun')
|
2014-12-23 04:16:57 +00:00
|
|
|
|
|
|
|
def verb(self, string):
|
2015-08-25 13:46:19 +00:00
|
|
|
return self(string, 'verb')
|
2014-12-23 04:16:57 +00:00
|
|
|
|
|
|
|
def adj(self, string):
|
2015-08-25 13:46:19 +00:00
|
|
|
return self(string, 'adj')
|
2014-12-23 04:16:57 +00:00
|
|
|
|
2015-10-09 07:44:21 +00:00
|
|
|
def punct(self, string):
|
|
|
|
return self(string, 'punct')
|
|
|
|
|
2014-12-23 04:16:57 +00:00
|
|
|
|
|
|
|
def lemmatize(string, index, exceptions, rules):
|
|
|
|
string = string.lower()
|
|
|
|
forms = []
|
|
|
|
if string in index:
|
|
|
|
forms.append(string)
|
|
|
|
forms.extend(exceptions.get(string, []))
|
|
|
|
for old, new in rules:
|
|
|
|
if string.endswith(old):
|
|
|
|
form = string[:len(string) - len(old)] + new
|
2015-10-09 07:44:21 +00:00
|
|
|
if form in index or not form.isalpha():
|
2014-12-23 04:16:57 +00:00
|
|
|
forms.append(form)
|
|
|
|
if not forms:
|
|
|
|
forms.append(string)
|
|
|
|
return set(forms)
|
|
|
|
|
|
|
|
|
2015-12-07 05:01:28 +00:00
|
|
|
def read_index(fileobj):
|
2014-12-23 04:16:57 +00:00
|
|
|
index = set()
|
2015-12-07 05:01:28 +00:00
|
|
|
for line in fileobj:
|
2014-12-23 04:16:57 +00:00
|
|
|
if line.startswith(' '):
|
|
|
|
continue
|
|
|
|
pieces = line.split()
|
|
|
|
word = pieces[0]
|
|
|
|
if word.count('_') == 0:
|
|
|
|
index.add(word)
|
|
|
|
return index
|
|
|
|
|
|
|
|
|
2015-12-07 05:01:28 +00:00
|
|
|
def read_exc(fileobj):
|
2014-12-23 04:16:57 +00:00
|
|
|
exceptions = {}
|
2015-12-07 05:01:28 +00:00
|
|
|
for line in fileobj:
|
2014-12-23 04:16:57 +00:00
|
|
|
if line.startswith(' '):
|
|
|
|
continue
|
|
|
|
pieces = line.split()
|
|
|
|
exceptions[pieces[0]] = tuple(pieces[1:])
|
|
|
|
return exceptions
|