Any archival project worth its salt would never store just the tweet text, but also all associated metadata. The best way I've seen to do this is to just save t

Any archival project worth its salt would never store just the tweet text, but also all associated metadata. The best way I've seen to do this is to just save the JSON returned by an API as it's easily parseable. You also have to consider that the utility of Twitter isn't in the single tweets, but rather the whole comment chains. Lastly, Twitter has images as well.