Readwise import dropped in v0.9?

[mAAdhaTTah]

Looks like this was dropped in the migration from v0.7 -> v0.9. Was this intentional? Would this be something you'd be willing to re-accept? I'm investigating upgrading to v0.9 and realized this isn't part of it anymore so debating what approach to time (I might just write a service to curl it myself and have archivebox ingest a jsonl, but tbd).

[pirate]

Yeah I cleaned house on the parsers a bit, @tk1475 is correct in that they have to be moved to plugins.

I am definitely open to addding it back as a plugin, but the new architecture makes it slightly trickier. Now instead of running imported files through a parser in a dedicated phase during ingestion, instead sources are treated like normal pages and crawled with depth=1 to get all the urls out, and plugins that can get urls out of a page report them as normal URLs like urls discovered during recursive depth>=1 parsing.

The goal is to move towards [url-like-source] -> [... outputs of any type] as the common architecture everywhere, as this will let us do things like importing emails from a url like imap://{user}:{pass}@{host}:{port}/{mailbox} and recursively archiving links discovered within, importing photos from a url like applephotos://{appleid}/{album}?after=2026 and recursively archiving OCR'd links within them, etc. in the future.

I'm open to porting over the old readwise parser into a new plugin, but for now that intermediate approach generating JSONL is a great stop-gap.

Note --overwrite/--update are no longer recommended, instead I recommend using archivebox config --set ONLY_NEW=False or using the [x] Only New checkbox in the UI.

[mAAdhaTTah]

Yeah, the url-like-source was how that was implemented in v0.7, so if that isn't available yet, then I'll proceed with the JSONL implementation for now and will keep an eye out for that architecture change.

[pirate]

Opened a new issue to track it here: ArchiveBox/abx-plugins#75