def test_extract_word():
got = mecab.extract_word('環境音楽だ', '名詞')
assert got, [u'環境' == u'音楽']
got = mecab.extract_word('環境音楽だ', '名詞', phrase=True)
assert got, [u'環境', u'音楽' == u'環境音楽']
got = mecab.extract_word('寝ろ', '動詞', rootform=True)
assert got == [u'寝る']
def test_extract_word():
got = mecab.extract_word('環境音楽だ', '名詞')
assert_equals(got, [u'環境', u'音楽'])
got = mecab.extract_word('環境音楽だ', '名詞', phrase=True)
assert_equals(got, [u'環境', u'音楽', u'環境音楽'])
got = mecab.extract_word('寝ろ', '動詞', rootform=True)
assert_equals(got, [u'寝る'])
def respond(text):
"""Extract a past post responding a post similar to given text
"""
for response in _extract_response_by_search([text], False):
if response:
yield response
query = mecab.extract_word(text, 'content_word')
for response in _extract_response_by_search([' '.join(query)], True):
if response:
yield response
query = mecab.extract_word(text, ('名詞,固有名詞',))
for response in _extract_response_by_search([' '.join(query)], True):
if response:
yield response
def respond(text, *args):
"""Extract a past post responding a post similar to given text
"""
for response in _extract_response_by_search([text], False):
if response:
yield response
query = mecab.extract_word(text, 'content_word')
for response in _extract_response_by_search([' '.join(query)], True):
if response:
yield response
query = mecab.extract_word(text, ('名詞,固有名詞',))
for response in _extract_response_by_search([' '.join(query)], True):
if response:
yield response
def count(self, log):
for post in log:
for idx in ('text', 'q1', 'q2'):
if not post.get(idx):
continue
if isinstance(post[idx], list):
post[idx] = '\n'.join(post[idx])
for line in post[idx].splitlines():
line = self.prepare_for_counting(line)
for w in mecab.extract_word(line):
ws = []
ws.append(self.word_ids[w])
if w not in self.words:
self.words.append(w)
cntr = Counter(ws)
for word in cntr.keys():
self.word_counts[word] += cntr
self.unique_wordcounts += cntr
def count(self, log):
for post in log:
for idx in ('text', 'q1', 'q2'):
if not post.get(idx):
continue
if isinstance(post[idx], list):
post[idx] = '\n'.join(post[idx])
for line in post[idx].splitlines():
line = self.prepare_for_counting(line)
for w in mecab.extract_word(line):
if w in NG_WORDS:
continue
ws = []
ws.append(self.word_ids[w])
if w not in self.words:
self.words.append(w)
cntr = Counter(ws)
for word in cntr.keys():
self.word_counts[word] += cntr
self.unique_wordcounts += cntr
def respond(text, *args):
logs = kuzuha.search(mecab.extract_word(text))
logs = [cleansing(log.get('text', '')) for log in logs]
for message in get_longest_common_substring(''.join(logs)):
if message:
yield message + '(;´Д`)'