[thelamest]

I hope you’ll get a better answer, but a minimal solution is to use snscrape (https://github.com/JustAnotherArchivist/snscrape) to download a raw JSON dump of a user timeline. Some caveats: 1. This doesn’t work for all users (e.g. some accounts seem to remain deindexed after unsuspensions). 2. While you can get rid of 90% of the file size by removing fluff columns (e.g. related to processing media and emojis), keep a backup. When importing such JSON, I unwittingly did a lossy data type conversion, and it can be irreparable once the tweets are deleted.

$ snscrape --progress --jsonl twitter-user jack > jack.json

[PS With courtesy rate limiting, depending on the account size, you may then script a way to send each tweet URL for archival in the Wayback Machine.]