def fb_get_postlinks_from_timeline(fbbrowser,
url="https://facebook.com",
count=10,
logger=astra.baselogger):
fbbrowser.get(url)
karma.wait()
postlinks = []
while len(postlinks) < count:
posts = []
postcontents = []
posts = posts + fbbrowser.find_elements_by_class_name(
"userContentWrapper")
for post in posts:
postcontents.append(BeautifulSoup(post.get_property("innerHTML")))
for postcontent in postcontents:
pc = list(postcontent.find_all("a", {"class": "_5pcq"}))
if len(pc) > 0:
pc = pc[0]
else:
continue
url = "https://facebook.com" + pc.get("href")
postlinks.append(url)
logger.info("Got %s posts " % (len(postlinks)))
postlinks = list(set(postlinks))
if "https://facebook.com#" in postlinks:
postlinks.pop(postlinks.index("https://facebook.com#"))
karma.scroll_page(fbbrowser)
karma.wait()
return postlinks[:count]
def poll(botname, botstop):
while True:
if not os.path.exists(botstop):
botname.logger.info(
u"{} waiting for messages. Touch {} to stop".format(
botname.name, botstop))
#botname.updater.start_polling()
#botname.updater.stop()
karma.wait(waittime="medium", logger=botname.logger)
try:
p = botname.get_latest_updates()
botname.logger.info(u"{} got {} messages".format(
botname.name, len(p)))
if len(p) > 0:
for update in p:
botname.updater.dispatcher.process_update(update)
#messageparser(botname,update,handler,logger=botname.logger)
except Exception as e:
botname.logger.warning(
"Timed out, will try again in a bit {}".format(repr(e)))
else:
botname.logger.info(
u"Stopfile found, {} exiting bot update loop".format(
botname.name))
break
def twython_get_ntweets_for_search(tw,
search,
tcount,
geocode=None,
maxtries=0,
logger=astra.baselogger):
#tweets=[]
p = pandas.DataFrame(
tw.search(q=search, count=tcount, tweet_mode="extended")['statuses'])
if len(p) > 0:
tweetdf = p.drop_duplicates(subset="full_text",
keep="first").reset_index()
else:
return pandas.DataFrame()
tries = 0
if maxtries == 0:
maxtries = (tcount / 20) * 3
while len(tweetdf) < tcount:
logger.info("Currently have %s tweets" % len(tweetdf))
if "id" in p.columns:
p = pandas.DataFrame(
tw.search(q=search,
count=tcount,
tweet_mode="extended",
max_id=p['id'].iloc[-1])['statuses'])
else:
logger.info(p.columns)
logger.info("Got %s more tweets" % len(p))
if len(p) > 1:
p = p.drop_duplicates(subset="full_text",
keep="first").reset_index(drop=True)
else:
logger.info("Empty tweetset")
break
logger.info("%s tweets are unique" % len(p))
tweetdf = pandas.concat([tweetdf, p])
tweetdf = tweetdf.drop_duplicates(subset="full_text",
keep="first").reset_index(drop=True)
logger.info("Sleeping...")
karma.wait(logger=logger)
logger.info("Got %s tweets" % len(tweetdf))
tries += 1
if (tries > maxtries):
break
return tweetdf.head(tcount)
def twython_search(tw,
searchstring,
logger=astra.baselogger,
tcount=100,
maxtries=10):
results = []
logger.info("Searching Twitter for " + searchstring)
results = tw.search(q=searchstring, count=tcount,
tweet_mode="extended")['statuses']
#logger.info("Got " + str(len(results)))
tries = 0
while len(results) < tcount:
if tries > maxtries:
break
maxid = results[-1]['id']
results = results + tw.search(
q=searchstring, count=tcount, max_id=maxid)['statuses']
karma.wait(logger=logger)
tries += 1
logger.info("Got " + str(len(results)) + " for search query " +
searchstring)
return results[:tcount]
