def test_stream():
t = Twarc()
count = 0
for tweet in t.stream("obama"):
assert tweet['id_str']
assert tweet['text']
count += 1
if count == 50:
break
assert count == 50
class TwitterStreamKafka(object):
# WORKING TWITTER HOSE
def __init__(self, search_terms):
logging.info("initializing TwitterStream Kafka")
# globals to all instances
self.t = Twarc(localConfig.client_key, localConfig.client_secret, localConfig.access_token, localConfig.access_token_secret)
self.search_terms = search_terms
# method to capture twitter stream
def captureStream(self):
for tweet in self.t.stream(",".join(self.search_terms)):
result = producer.send_messages("betweezered", json.dumps(tweet))
class TwitterStreamKafka(object):
# WORKING TWITTER HOSE
def __init__(self, search_terms):
logging.info("initializing TwitterStream Kafka")
# globals to all instances
self.t = Twarc(localConfig.client_key, localConfig.client_secret,
localConfig.access_token,
localConfig.access_token_secret)
self.search_terms = search_terms
# method to capture twitter stream
def captureStream(self):
for tweet in self.t.stream(",".join(self.search_terms)):
result = producer.send_messages("betweezered", json.dumps(tweet))
class TwitterHarvester(BaseHarvester):
def __init__(self, process_interval_secs=1200, mq_config=None, debug=False):
BaseHarvester.__init__(self, mq_config=mq_config, process_interval_secs=process_interval_secs, debug=debug)
self.twarc = None
def harvest_seeds(self):
# Create a twarc
self._create_twarc()
# Dispatch message based on type.
harvest_type = self.message.get("type")
log.debug("Harvest type is %s", harvest_type)
if harvest_type == "twitter_search":
self.search()
elif harvest_type == "twitter_filter":
self.filter()
else:
raise KeyError
def _create_twarc(self):
self.twarc = Twarc(self.message["credentials"]["consumer_key"],
self.message["credentials"]["consumer_secret"],
self.message["credentials"]["access_token"],
self.message["credentials"]["access_token_secret"])
def search(self):
incremental = self.message.get("options", {}).get("incremental", False)
for seed in self.message.get("seeds", []):
query = seed.get("token")
# Get since_id from state_store
since_id = self.state_store.get_state(__name__, "{}.since_id".format(query)) if incremental else None
max_tweet_id = self._process_tweets(self.twarc.search(query, since_id=since_id))
log.debug("Searching on %s since %s returned %s tweets.", query,
since_id, self.harvest_result.summary.get("tweet"))
# Update state store
if incremental and max_tweet_id:
self.state_store.set_state(__name__, "{}.since_id".format(query), max_tweet_id)
def filter(self):
assert len(self.message.get("seeds", [])) == 1
track = self.message["seeds"][0]["token"]
self._process_tweets(self.twarc.stream(track))
def _process_tweets(self, tweets):
max_tweet_id = None
for count, tweet in enumerate(tweets):
if not count % 100:
log.debug("Processed %s tweets", count)
if self.stop_event.is_set():
log.debug("Stopping since stop event set.")
break
if "text" in tweet:
with self.harvest_result_lock:
max_tweet_id = max(max_tweet_id, tweet.get("id"))
self.harvest_result.increment_summary("tweet")
if "urls" in tweet["entities"]:
for url in tweet["entities"]["urls"]:
self.harvest_result.urls.append(url["expanded_url"])
if "media" in tweet["entities"]:
for media in tweet["entities"]["media"]:
self.harvest_result.urls.append(media["media_url"])
return max_tweet_id
