This is a script to use node.js and CouchDB to archive your tweets. You can then use CouchDB views to analyze your tweets or just keep them around.

IT'S JAVASCRIPT ALL THE WAY DOWN.

• Install node.js, version 0.3 or later. (It's probably possible to be compatible with version 0.2 with only minor code changes.) Then, install npm. Do npm install underscore oauth to get the needed modules.

• Install CouchDB. I think version 0.8 or later should be okay, but haven't tried with anything before 1.0. Create a CouchDB database (it can be empty) where your tweets will be stored.

• Create a Twitter app.

• Copy config.json.sample to config.json and fill in the values appropriately.

• Run node couch-tweet-archiver.js and hope for the best. The initial run should fetch all your tweets (up to Twitter's paging limit of 3200); future runs should fetch only newer tweets.

• Use futon or your own code to create views and query the database. Or just keep it forever.

Here are some things that you could work on, if you're so inclined.

• There's no actual reason that this is restricted to archiving your own tweets; you could easily modify the code to archive someone else's.

• Similarly, you could modify the trim_user, include_rts, and include_entities parameters to your liking.

• There could be a library of useful standard views, or even some CouchDB shows and lists for a usable web UI.

• The code could be extended to work even when CouchDB is not in “Admin Party” mode.

• You could make this all somehow actually useful.

This program is free software. It comes without any warranty, to the extent permitted by applicable law. You can redistribute it and/or modify it under the terms of the Do What The Fuck You Want To Public License, Version 2, as published by Sam Hocevar. See http://sam.zoy.org/wtfpl/COPYING for more details.