Feature Request: Twitter Archive support

[rclough]

Twitter has an option to download your archive of tweets all the way back, and it generally takes only a few minutes. It gives you the tweets in a big CSV, as well as JSON files divided by month. It would be nice to be able to use this as the corpus. I havent had the time to delve into the code to see how the current corpus is generated, but it probably wouldnt be too hard to parse it and generate a corpus in your current format.