def put(self, tweets):
filtered_tweets = filter(lambda t: not tweet.RETWEETED_STATUS in t, tweets)
normalized_tweets = map(lambda t: tweet.to_date(tweet.to_ascii(t)), filtered_tweets)
if len(normalized_tweets) == 0:
return
try:
self._tweet_coll.insert(normalized_tweets, continue_on_error = True)
except DuplicateKeyError:
pass # Ignored.
classifier = "classifier.pickle"
opts, args = getopt.getopt(sys.argv[1:], "hc:d:k:s:e:")
for o, a in opts:
if o == "-d":
db = a
elif o == "-c":
classifier = a
elif o == "-k":
keywords.append(a)
elif o == "-s":
start = datetime.strptime(a, "%Y-%M-%d")
elif o == "-e":
end = datetime.strptime(a, "%Y-%M-%d")
else:
usage()
sys.exit(0)
classifier = Classifier.load(classifier)
aggregator = RetweetWeightedAggregator()
ts = TweetStore(db)
for t in ts.get(keywords, start, end):
s = classifier.classify(t)
print("%s -- sentiment: %s" % (tweet.to_ascii(t)[tweet.TEXT], "positive" if (s == 1) else "negative"))
aggregator.add(t, s)
print("Aggregated sentiment: %f" % aggregator.get_sentiment())
print("ID of last tweet: %d" % aggregator.get_last_id())
print("Total number of tweets: %d" % aggregator.get_num())