scraper.py

import json
from twython import TwythonStreamer

#Fill these in with the values Twitter give you when you register as a developer
APP_KEY = ''
APP_SECRET = ''

consumer_key = ''
consumer_secret = ''
access_token = ''
access_token_secret = ''


class TweetStreamer(TwythonStreamer):

    tweets = []

    def on_success(self, data):
        # If you get a tweet back (and not an error code)
        if 'text' in data:
            self.tweets.append(data)
        # I only wanted 500
        if len(self.tweets) == 500:
            self.disconnect()
            # Because I only wanted 500, I did not write to disk until I had them in memory
            with open('tweets.json', 'w') as f:
                json.dump(self.tweets, f)
        # In reality, this is a great spot to pipe the tweets to a MongoDB
        
    # If we get an error, just give up.
    def on_error(self, status_code, data):
        print(status_code)
        self.disconnect()

streamer = TweetStreamer(consumer_key, consumer_secret,
                         access_token, access_token_secret)
# Filter to get spatially enabled (no necessarily lat/lon encoded)
streamer.statuses.filter(locations='-112.5,33,-111.5,34', lang='en')