def __init__(self,
username=None,
password=None,
delay=10,
log='monitor.log'):
#logging
logfile = path.join(paths.log, log)
self.logger = logging.getLogger('monitor')
self.logger.addHandler(logging.FileHandler(logfile))
self.logger.setLevel(logging.INFO)
#init
self.delay = Delay(delay)
self.spider = SpiderBase(username=username,
password=password,
delay=self.delay)
def __init__(self, username=None, password=None, delay=10, log='monitor.log'):
#logging
logfile = path.join(paths.log, log)
self.logger = logging.getLogger('monitor')
self.logger.addHandler(logging.FileHandler(logfile))
self.logger.setLevel(logging.INFO)
#init
self.delay = Delay(delay)
self.spider = SpiderBase(username=username, password=password, delay=self.delay)
class Monitor(SpiderBase):
def __init__(self, username=None, password=None, delay=10, log='monitor.log'):
#logging
logfile = path.join(paths.log, log)
self.logger = logging.getLogger('monitor')
self.logger.addHandler(logging.FileHandler(logfile))
self.logger.setLevel(logging.INFO)
#init
self.delay = Delay(delay)
self.spider = SpiderBase(username=username, password=password, delay=self.delay)
def cleandb(self):
for k in self._keywords():
if k.accessed < (datetime.utcnow() - timedelta(30)):
session.query(Mention)\
.filter(Mention.keyword_uid==k.uid)\
.delete(synchronize_session=False)
session.query(Monitoring)\
.filter(Monitoring.keyword_uid==k.uid)\
.delete(synchronize_session=False)
session.query(Keyword)\
.filter(Keyword.uid==k.uid)\
.delete(synchronize_session=False)
elif len(k.mentions) > 50:
m = session.query(Mention)\
.filter(Mention.keyword_uid==k.uid)\
.order_by(Mention.created.desc())\
.offset(49).first()
session.query(Mention)\
.filter(Mention.keyword_uid==k.uid)\
.filter(Mention.uid<m.uid)\
.delete(synchronize_session=False)
session.commit()
def monitor_posts(self):
url = 'http://www.reddit.com/r/all/new.json?sort=new'
newest = None
while True:
seen, after, next_newest = self._scan_posts(url, newest)
if seen > 20:
self.delay.more()
elif seen < 10:
self.delay.less()
if seen == 0:
time.sleep(5)
after_url = '%s&count=50&after=%s' % (url, after)
seen = self._scan_posts(after_url, newest)[0]
newest = next_newest
self.delay.sleep()
def _scan_posts(self, url, newest):
seen = 0
data = self.spider._get_json(url)
posts = data['data']['children']
after = data['data']['after']
for i, c in enumerate(posts):
post = c['data']
if i == 0:
next_newest = post['id']
if post['id'] <= newest:
seen = len(posts) - i
break
title = post['title'].lower()
selftext = post['selftext'].lower()
for k in self._mentioned_keywords(title, text2=selftext):
mention = Mention()
mention.keyword_uid = k.uid
mention.thread_id = post['id']
mention.author = post['author']
mention.subreddit = post['subreddit']
mention.created = unix_string(int(post['created_utc']))
session.commit()
return (seen, after, next_newest)
def monitor_comments(self):
url = 'http://www.reddit.com/comments.json'
newest = None
while True:
seen, after, next_newest = self._scan_comments(url, newest)
if seen > 20:
self.delay.more()
elif seen < 10:
self.delay.less()
if seen == 0:
time.sleep(5)
after_url = '%s?count=50&after=%s' % (url, after)
seen = self._scan_comments(after_url, newest)[0]
newest = next_newest
self.delay.sleep()
def _scan_comments(self, url, newest):
seen = 0
data = self.spider._get_json(url)
comments = data['data']['children']
after = data['data']['after']
for i, c in enumerate(comments):
comment = c['data']
if i == 0:
next_newest = comment['id']
if comment['id'] <= newest:
seen = len(comments) - i - 1
break
body = comment['body'].lower()
for k in self._mentioned_keywords(body):
mention = Mention()
mention.keyword_uid = k.uid
mention.thread_id = comment['link_id'][3:]
mention.comment_id = comment['id']
mention.author = comment['author']
mention.subreddit = comment['subreddit']
mention.created = unix_string(int(comment['created_utc']))
session.commit()
return (seen, after, next_newest)
def _mentioned_keywords(self, text, text2=None):
for k in self._keywords():
regex = re.compile('(^|.*\W)%s($|\W|s$|s\W)' % re.escape(k.keyword), re.IGNORECASE)
if regex.match(text) or (text2 is not None and regex.match(text2)):
yield k
@memoize('monitor_keywords', time=60)
def _keywords(self):
keywords = session.query(Keyword).order_by(Keyword.keyword.asc()).all()
return keywords
def __init__(self, username=None, password=None):
SpiderBase.__init__(self, username=username, password=password)
self.delay = Delay(5)
class Monitor(SpiderBase):
def __init__(self,
username=None,
password=None,
delay=10,
log='monitor.log'):
#logging
logfile = path.join(paths.log, log)
self.logger = logging.getLogger('monitor')
self.logger.addHandler(logging.FileHandler(logfile))
self.logger.setLevel(logging.INFO)
#init
self.delay = Delay(delay)
self.spider = SpiderBase(username=username,
password=password,
delay=self.delay)
def cleandb(self):
for k in self._keywords():
if k.accessed < (datetime.utcnow() - timedelta(30)):
session.query(Mention)\
.filter(Mention.keyword_uid==k.uid)\
.delete(synchronize_session=False)
session.query(Monitoring)\
.filter(Monitoring.keyword_uid==k.uid)\
.delete(synchronize_session=False)
session.query(Keyword)\
.filter(Keyword.uid==k.uid)\
.delete(synchronize_session=False)
elif len(k.mentions) > 50:
m = session.query(Mention)\
.filter(Mention.keyword_uid==k.uid)\
.order_by(Mention.created.desc())\
.offset(49).first()
session.query(Mention)\
.filter(Mention.keyword_uid==k.uid)\
.filter(Mention.uid<m.uid)\
.delete(synchronize_session=False)
session.commit()
def monitor_posts(self):
url = 'http://www.reddit.com/r/all/new.json?sort=new'
newest = None
while True:
seen, after, next_newest = self._scan_posts(url, newest)
if seen > 20:
self.delay.more()
elif seen < 10:
self.delay.less()
if seen == 0:
time.sleep(5)
after_url = '%s&count=50&after=%s' % (url, after)
seen = self._scan_posts(after_url, newest)[0]
newest = next_newest
self.delay.sleep()
def _scan_posts(self, url, newest):
seen = 0
data = self.spider._get_json(url)
posts = data['data']['children']
after = data['data']['after']
for i, c in enumerate(posts):
post = c['data']
if i == 0:
next_newest = post['id']
if post['id'] <= newest:
seen = len(posts) - i
break
title = post['title'].lower()
selftext = post['selftext'].lower()
for k in self._mentioned_keywords(title, text2=selftext):
mention = Mention()
mention.keyword_uid = k.uid
mention.thread_id = post['id']
mention.author = post['author']
mention.subreddit = post['subreddit']
mention.created = unix_string(int(post['created_utc']))
session.commit()
return (seen, after, next_newest)
def monitor_comments(self):
url = 'http://www.reddit.com/comments.json'
newest = None
while True:
seen, after, next_newest = self._scan_comments(url, newest)
if seen > 20:
self.delay.more()
elif seen < 10:
self.delay.less()
if seen == 0:
time.sleep(5)
after_url = '%s?count=50&after=%s' % (url, after)
seen = self._scan_comments(after_url, newest)[0]
newest = next_newest
self.delay.sleep()
def _scan_comments(self, url, newest):
seen = 0
data = self.spider._get_json(url)
comments = data['data']['children']
after = data['data']['after']
for i, c in enumerate(comments):
comment = c['data']
if i == 0:
next_newest = comment['id']
if comment['id'] <= newest:
seen = len(comments) - i - 1
break
body = comment['body'].lower()
for k in self._mentioned_keywords(body):
mention = Mention()
mention.keyword_uid = k.uid
mention.thread_id = comment['link_id'][3:]
mention.comment_id = comment['id']
mention.author = comment['author']
mention.subreddit = comment['subreddit']
mention.created = unix_string(int(comment['created_utc']))
session.commit()
return (seen, after, next_newest)
def _mentioned_keywords(self, text, text2=None):
for k in self._keywords():
regex = re.compile(
'(^|.*\W)%s($|\W|s$|s\W)' % re.escape(k.keyword),
re.IGNORECASE)
if regex.match(text) or (text2 is not None and regex.match(text2)):
yield k
@memoize('monitor_keywords', time=60)
def _keywords(self):
keywords = session.query(Keyword).order_by(Keyword.keyword.asc()).all()
return keywords
def __init__(self, username=None, password=None):
SpiderBase.__init__(self, username=username, password=password)
self.delay = Delay(5)
