FWIW, my experiment seems to be working very well, generating the c-k-c connector doc pages directly from the .json files under connectors without any symlinks or copying. It relies on a lot of unpublished Antora extensions so I’m working on getting them ready and publishing them. It would be great to get a definitive answer about what should be present. There’s still the question of whether the catalog json files that aren’t copied from under connectors should be there.
BTW it turns out there is one connector with non-generated documentation, camel-syslog-kafka-connector. David Jencks > On Nov 2, 2021, at 2:28 AM, Zoran Regvart <zo...@regvart.com> wrote: > > Hi David, > > On Sun, Oct 31, 2021 at 4:02 PM David Jencks <david.a.jen...@gmail.com> wrote: >> TL/DR >> Can I assume that only projects currently under connectors should have pages >> in the website? (These are the pages indexed in reference/index.html and the >> nav) > > I do think so, perhaps Andrea T. can chime in here to help. I think it > would be a worthwhile effort to get those pages generated from the > JSON metadata, and thank you for taking this on. > > [snip] > >> I’m wondering if the processes that copy the adoc files under docs and the >> json files into the catalog should have a cleanup step to remove >> no-longer-current files. As noted above, I haven’t found how this copying is >> done. If my experiment succeeds, the adoc files won’t exist any more, being >> generated by Antora from the json files under connectors, so only the >> catalog would potentially need this cleaning step. > > Yeah, one of things we're not doing that well is cleaning up some of > the old stuff, but the catalog JSON files should be the most relevant > and up to date, so that would scratch that problem... > > zoran > -- > Zoran Regvart
