Gemini protocol support (fetching & parsing)

[sjehuda]

This code could be relevant.

https://git.xmpp-it.net/sch/Slixfeed/src/branch/master/src/slixfeed/retriever/gemini.py

https://git.xmpp-it.net/sch/Slixfeed/src/branch/master/src/slixfeed/parser/gmi.py

[LeXofLeviafan]

…?

[sjehuda]

buku retrieves metadata from HTML files over HTTP.

I was wondering, whether it would be acceptable to support GemText of Gemini also.

[LeXofLeviafan]

This seems like a… very niche thing 🤔

One would need special software to even access such links (even just programmatically); and there doesn't seem to be many of those to begin with.

…For now, I'd suggest writing a standalone CLI script (and leaving a link here in case anyone else is interested), which fetches necessary data for the provided URL then invokes Buku to add/update a bookmark for that URL 🤷‍♂

[sjehuda]

LeXofLeviafan. The code, to which I refer to, handles the Gemini protocol and the markup file GemText.

[LeXofLeviafan]

No, a 3rd-party library handles the Gemini protocol in that code. One which has multiple dependencies, and cannot be used for anything other than handling said protocol.

Like I said: it's a feature that requires special software to work, and is very unlikely to ever be invoked by the vast majority of the users. It makes little sense to implement something which is both obscure and dependency-heavy like this into Buku.

An external CLI script, however, is perfectly doable… and it would be best for it to be implemented/maintained by someone who's actually planning to use it. (If/when Buku implements plugin support, it can be repurposed to be used that way, but, again, that's up to whoever would use it.)