def run(self, hour_range=HOUR_RANGE):
date_range = kuzuha.build_date_filter_by_range({'hours': hour_range})
posts = kuzuha.search('http', _filter=date_range, sort=[])
tweet = ''
for (url, count) in self._count_url(posts).most_common():
if url.startswith('https://twitter.com/'):
tweet_id = self.extract_tweet_id(url)
if tweet_id:
logger.info('RT: id=%s (%s)' % (tweet_id, url))
if not self.debug:
try:
self.twitter.api.statuses.retweet(id=tweet_id)
except TwitterHTTPError as e:
logger.warn('%s %s' % (type(e), str(e)))
continue
title = self._get_title(url)
new_url_info = TWEET_FORMAT % (title, url, count)
expected_length = self.calc_tweet_length(tweet, title, count)
if expected_length < (MAX_TWEET_LENGTH - len(HASH_TAG)):
tweet += new_url_info
else:
tweet = tweet[:-len(DELIMITER)] + HASH_TAG
if tweet != HASH_TAG:
tweet = tweet.replace('\n', '').replace('\r', '')
yield tweet
tweet = new_url_info
if tweet:
if tweet.endswith(DELIMITER):
tweet = tweet[:-len(DELIMITER)]
tweet = tweet.replace('\n', '').replace('\r', '')
yield tweet + HASH_TAG
def run(self, hour_range=HOUR_RANGE):
date_range = kuzuha.build_date_filter_by_range({'hours': hour_range})
posts = kuzuha.search('http', _filter=date_range, sort=[])
tweet = ''
for (url, count) in self._count_url(posts).most_common():
if url.startswith('https://twitter.com/'):
tweet_id = self.extract_tweet_id(url)
if tweet_id:
logger.info('RT: id=%s (%s)' % (tweet_id, url))
if not self.debug:
try:
self.twitter.api.statuses.retweet(id=tweet_id)
except TwitterHTTPError as e:
logger.warn('%s %s' % (type(e), str(e)))
continue
title = self._get_title(url)
new_url_info = TWEET_FORMAT % (title, url, count)
expected_length = self.calc_tweet_length(tweet, title, count)
if expected_length < (MAX_TWEET_LENGTH - len(HASH_TAG)):
tweet += new_url_info
else:
tweet = tweet[:-len(DELIMITER)] + HASH_TAG
if tweet != HASH_TAG:
tweet = tweet.replace('\n', '').replace('\r', '')
yield tweet
tweet = new_url_info
if tweet:
if tweet.endswith(DELIMITER):
tweet = tweet[:-len(DELIMITER)]
tweet = tweet.replace('\n', '').replace('\r', '')
yield tweet + HASH_TAG
def run(self, interval=20):
posts = kuzuha.search('',
_filter=kuzuha.build_date_filter_by_range(
{'minutes': interval}))
pairs = self.get_post_res_pairs(posts)
for (parent, responses) in pairs.items():
if len(responses) >= 2:
ome_posts = set()
logger.info('MENTIONED POST: %s' % parent)
for (lhs, rhs) in itertools.combinations(responses, 2):
logger.info('Compare "%s" with "%s"' % (lhs, rhs))
if lhs and rhs and self.is_ome(lhs, rhs):
logger.info('"%s" and "%s" are OME' % (lhs, rhs))
ome_posts |= {lhs, rhs}
if len(ome_posts) > 1:
num_posts = len(ome_posts) + 1 # childs + parent
max_length = (body_length - num_posts * 2) // num_posts
parent = self.shorten(parent, max_length)
message = '%s『%s』' % (PREFIX, parent)
for ome_post in sorted(ome_posts):
ome_post = self.shorten(ome_post, max_length)
message += '「%s」' % ome_post
message += HASH_TAG
yield message
def run(self, hour_range=24):
date_range = kuzuha.build_date_filter_by_range({'hours': hour_range})
result = self.find('しこる', 'でしこ', date_range)
result += self.find('抜く', 'で抜', date_range)
result += self.find('オナニュ', 'でオナニュす', date_range)
result += self.find('オナニュ', 'でオナニュし', date_range)
if len(result) < MAX_TWEET_LENGTH:
return result
def run(self, hour_range=24):
date_range = kuzuha.build_date_filter_by_range({'hours': hour_range})
result = self.find('しこる', 'でしこ', date_range)
result += self.find('抜く', 'で抜', date_range)
result += self.find('オナニュ', 'でオナニュす', date_range)
result += self.find('オナニュ', 'でオナニュし', date_range)
if len(result) < MAX_TWEET_LENGTH:
return result
def run(self, interval=60, min_length=40):
m_generator = markov.MarkovGenerater()
m = mecab.MeCabWrapper()
posts = kuzuha.search('', _filter=kuzuha.build_date_filter_by_range({'minutes': interval}), sort=[])
words = []
for post in posts:
if 'text' not in post:
continue
text = regex.re_a_tag.sub('', post['text'])
text = normalize.normalize(text)
if 'アニメ時報' in text:
continue
words.append([])
for line in text.splitlines():
words[-1].append(BOS)
for n in m.parse_to_node(line):
words[-1].append('%s,%s' % (n.surface, ''.join(n.feature.split(',')[:5])))
words[-1].append(EOS)
return m_generator.generate(words, min_length)
def run(self, interval=20):
posts = kuzuha.search('', _filter=kuzuha.build_date_filter_by_range({'minutes': interval}))
pairs = self.get_post_res_pairs(posts)
for (parent, responses) in pairs.items():
if len(responses) >= 2:
ome_posts = set()
logger.info('MENTIONED POST: %s' % parent)
for (lhs, rhs) in itertools.combinations(responses, 2):
logger.info('Compare "%s" with "%s"' % (lhs, rhs))
if lhs and rhs and self.is_ome(lhs, rhs):
logger.info('"%s" and "%s" are OME' % (lhs, rhs))
ome_posts |= {lhs, rhs}
if len(ome_posts) > 1:
num_posts = len(ome_posts) + 1 # childs + parent
max_length = (body_length - num_posts*2) // num_posts
parent = self.shorten(parent, max_length)
message = '%s『%s』' % (PREFIX, parent)
for ome_post in sorted(ome_posts):
ome_post = self.shorten(ome_post, max_length)
message += '「%s」' % ome_post
message += HASH_TAG
yield message
def run(self, hour_range=24):
date_range = kuzuha.build_date_filter_by_range({'hours': hour_range})
result = self.find('飲む', 'を飲', date_range)
result += self.find('食べる', 'を食', date_range)
if len(result) < MAX_TWEET_LENGTH:
return result
def run(self, hour_range=24):
date_range = kuzuha.build_date_filter_by_range({'hours': hour_range})
result = self.find('飲む', 'を飲', date_range)
result += self.find('食べる', 'を食', date_range)
if len(result) < MAX_TWEET_LENGTH:
return result
