Add a parser for tweet archives

[serin-delaunay]

User ebooks accounts are a common type of bot. They generate (presumably Markov) text based on a user's tweet archives. CMTQ can parse uploaded files, but its current line-separation method is not suited to extracting tweet text from tweet archive files.

• Add a way to upload a tweet archive file (a second file upload button?)
• Parse tweet archive files to strip out metadata
• Use each tweet's text as a line of source text.