feat: Change catalog to produce asciidoc

[bjchambers]

No description provided.

[bjchambers]

This is part of #352. This produces separate AsciiDoc pages for each catalog entry. Let's discuss steps for connecting it. I think I see:

1. Run the existing binary built in the main CI to produce a zip that we archive. We can then use https://gitlab.com/antora/antora-collector-extension to download the zip and extract it. This potentially has issues with versioning.
2. Use https://gitlab.com/antora/antora-collector-extension with a script that builds the catalog binary and uses it. This will currently be slow. We could improve it by creating a version of the catalog binary that doesn't include sparrow (eg., a "generate-only" binary). That would have far fewer dependencies and be much faster to build.
3. Some other approach, possibly having the binary push the generated stuff to a separate "kaskada-catalog" repo, keeping a main branch in sync, as well as tagged branches for each release.

One benefit of that last option is that it makes it possible to manually edit old catalogs.