mastodon-archive-static-site

A Python script to convert a Mastodon user archive into a static site.

Purpose

This script will take the .tar.gz archive that a Mastodon server can generate for one of its users (requested via Settings → Import and export → Data export → Request your archive) and, for each public toot in there, create a subdirectory with an index.html file and all of the toot’s attachments, so that you can seamlessly replace the original toot permalinks with these static files.

Status

This is an unfinished prototype.

The main thing that’s missing at this time is a beautiful (but self-contained) design for the generated pages. Hit me up if you’re interested in contributing that, I’m not good at it.

Usage

Requires Python 3.6 or higher with Jinja2 and (optionally) requests installed.

Run ./generate.py --help for usage information.

By default, the generated pages are written to ./output/. You can use -o DIR to change that.

A simple example call would be ./generate.py archive-20220414144106-fbdf204a1bdf9375618d8a032d5a64d8.tar.gz.

Archiving to the Wayback Machine

This whole feature is experimental and kinda sucks right now.

If the Mastodon instance still exists, you can use the -w command-line switch to ask the Wayback Machine aka web.archive.org to add each toot to its archive.

You’ll need to have requests installed.

This flag will result in one API call to the Wayback Machine for each public toot in the archive. Note that each of these calls usually takes significantly more than 30 seconds, and I have not yet tested whether archive.org is happy about thousands of requests, even though it’s only one at a time.

There is currently no way to resume these calls if your generate.py invocation gets terminated or something, you’ll have to start over. Try to avoid this, I guess.

Contact

This project lives at https://codeberg.org/scy/mastodon-archive-static-site. You can find the author on Mastodon: @scy@chaos.social.

License

MIT.