Twitlib is a python library that helps automate the management of a Twitter account. It was specifically designed to receive and manipulate tweets from Twitter's streaming API using a thread pool design pattern.
Some of the current features include:
- Dumping streamed tweets to JSON files
- Downloading media from streamed tweets
- Mirroring of streamed tweets to another account
- Scalability thanks to a thread pool design pattern
- Extensible through user defined thread subclasses
Please see TODO for API documentation. An example script is provided.
In order to facilitate control of bot accounts without programming or command line knowledge, modules are provided to allow for bot control via direct messages. This control system will authenticate DMs from the bot's owner using the owner's twitter ID. When the bot receives a direct message from the owner it will parse the message content and take a predefined action based on message content.
Ideally, the module would follow functional programming practices as follows
- The parser is constructed with a list of functions that map tweet text content to either function calls that perform a desired response action (if tweet matches some regex) or a noop/None function (if there is no match).
This system could also be used to provide automated responses to users who DM the bot, perhaps directing them to the owner's real account.
Through DM control the bot can perform actions with the full power of a command line, using DMs as a way to exchange input and output data over Twitter. One potential use case is the watermarking of images. Suppose that the bot's host machine has a script that can watermark an image. The bot owner could DM the bot with an image and some trigger phrase ('watermark'), at which point the bot would watermark the image and send the result back to the owner via twitter DM.
Scheduling of account actions can easily be added.
- Tweets could be pre-written and scheduled for release at a later time.
- An account could be set to private or public on a scheduled interval, with additional logic that accounts for when the user last tweeted.