So I don't know how much effort it would take to make it "calibre-web"-ish and how big the impact to calibre-web would be. I tried to take a look how your code works and what it does, but unfortunately you didn't commit it to github (as far as I could see). I'm more inclined towards the first since that what every Kobo Calibre user is supposedly doing now, however it wouldn't match the behavior of the CalibreWeb web interface. I'm wondering if I should try to replicate that shelf management style here, or if I should rely on the secondary database that CalibreWeb seems to be using to manage collections. I've faked this by simply using the shelf name as the id instead of a uuid string, and this has been working so far. This works for the most part, except that the Sync protocol expects shelves to have unique ids. In my CalibreDb-based implementation, I've also been relying on custom columns to replicate this behavior. A calibre plugins wasdeveloped to keep the device shelves in sync with a given "MyShelves" custom column (see ). Historically, Kobo users have been using Calibre custom columns to manage their devices shelves.
Why are shelves/collections managed in a separate database? Or should I try to get it included into the main project? Should I remain in a fork, since I mostly really care about the db.py abstraction. There's a few open questions that I'm hoping to get some opinions on: I ended up forking this project and have started re-implementing it by relying on db.py instead. I'm also storing books on Backblaze, with only the metadata.db file synced to the server where CalibreWeb is running.
I've already implemented a pretty complete implementation relying on the Calibredb commandline tool, but it's a little fragile and has been hard to work with (CalibreDb messes up the directory structure if different OSes interact with the library :-/ ). Store tracking metadata for each book (completion, reading location, last read, etc).Manage books (delete from library, add to collections, etc).List books and download them to the device.It's trivial to change the endpoint that the Kobo device talks to for its cloud syncing protocol.
0 kommentar(er)
