Although rarely a prolific one, I first started using Twitter in 2010.
It's unlikely that I've tweeted too much of value, however I am a bit of a datahoarder, and felt that it'd be a shame to lose that content just because Elon Musk buys and tries to run businesses without any understanding for how they works, or earn their money.
So, I decided I to build an archive of my tweets.
The script in this repo consumes the JSON file created by twitter-dump (I have a dockerised version) and generates a static HTML archive.
It's a hastily hacked together script, originally built for some other analysis.
Usage is simple, but you will need dominate installed:
pip3 install dominateThen, call the script and pass it the path to your twitter-dump JSON file as the first and only argument
./build_mirror.py bentasker.jsonThe biggest limitation is that only Tweet text is mirrored - videos and images are not pulled into the archive.
This is because Twitter heavily obfuscates media URLs in the front-end.
That information is available via the API, but to access that you need an API token, which requires you to have provided a verified phone number (and I'm not doing that, especially now).
As noted above, media links are heavily obfuscated in Twitter's front-end, so user avatars cannot automatically be fetched.
However, tweets really do look quite weird without an avatar attached to them, so there is some support built into the eventual archive.
A directory called avatar will be created.
If you put a JPG in there, using the user's handle as the filename (e.g. bentasker.jpg) then it will be used/displayed in the archive.
Copyright (c) 2022 B Tasker
Released under BSD 3-clause license, see LICENSE