def main(keyword):
db_client = MongoClient('localhost', 27017)
db = db_client[db_name]
keywords = keyword
max_id = None
while True:
api = tw.get_api_client()
tweets = get_tweets_by_keyword(api, keywords, max_id)
if len(tweets) > 0:
try:
# Parse the data returned to get max_id to be passed in consequent call.
next_results_url_params = tweets['search_metadata']['next_results']
max_id = next_results_url_params.split('max_id=')[1].split('&')[0]
except:
# No more next pages
break
for status in tweets['statuses']:
if 'media' in status['entities']:
## media = status['entities']['media'][0]['media_url']
db[collection_name+'_'+keyword].insert(status)
def main(users):
db_client = MongoClient('localhost', 27017)
db = db_client[db_name]
users=users
for user in users:
max_id = None
while True:
api = tw.get_api_client()
tweets = get_tweets_by_user(api, user, max_id)
if len(tweets) > 0:
max_id = tweets[-1]['id']
for tweet in tweets:
db[collection_name+'_'+user].insert(tweet)
## db['tweets_'+str(tweet['user']['screen_name'])].insert(tweet)
print 'Inserted', tweet['id']
else:
break