ebooks.py

Python based multiple-account-using, markov-tweeting, hamburger-eating CHAMPION OF THE WORLDDDD!

Todo

• Add Py3k compatibility with Flask's _compat.py cheatorama.
• Get gevent-based tweeting actually going. (we're no longer using gevent)
• Move configuration into some sneaky attributes on the subclass (then use inspect.getmembers() to find them)
• Tidy shit up.
• Add example use file.

Example

The general idea is that you subclass ebooks.Ebooks, then you can create an instance and call e.loop() to get into the tweet loop.

You inform ebooks.py what accounts to pull from and send to via attributes on the class that start with t_, as such:

from ebooks import Ebooks

class MyBooks(Ebooks):
    t_someguy = {
        "auth": (OAUTH_TOKEN,
                 OAUTH_SECRET,
                 CONSUMER_KEY,
                 CONSUMER_SECRET),
        "chain_length": 2,  # optional
        "dry": False,  # optional
    }

    t_some_other_guy = {
        # ...as above...
    }

    # These are all optional, with defaults shown:
    dry = False
    verbose = False

    chain_length = 2
    max_words = 30
    length_cap = 140

    timer_fetch = 60 * 5
    timer_tweet = 60 * 10

if __name__ == "__main__":
    e = MyBooks()
    e.loop()

In this example, you're pulling tweets from @someguy (as shown by t_someguy) and sending them to whatever account your OAuth tokens provide access to. You can add multiple accounts (e.g. this will also ebooksify @some_other_guy) by simply adding more attributes.

Other settings are fairly self-explanatory, but here's some explanation regardless:

• dry - this will do everything normally except for interacting with Twitter
• verbose - debug info will be printed to stdout
• chain_length - markov chain length (2 is fairly average)
• max_words - how many words we're aiming for in the tweet (an upper bound).
• length_cap - how long tweets can be
• timer_fetch - how often (in seconds) we should fetch tweets
• timer_tweet - how often (in seconds) we should send tweets