In this issue [1] we want to have some better early parsing of CG metadata before starting the import. JSON schema looks to be a good way to do that. I've experimented with it a bit and it raised few questions. It is not a short file (basic variant about 230 lines and I would like to inject it with more constraints) so it doesn't seem to be a good idea to put it into the code directly. Also if it will not be in the code itself (but separate json) others can use it directly in their CG implementations to precheck the json before sending it to koji.
My question is what would be the ways you would like to use it? /from that it should be decided where it should be placed, packaging, etc./
Feel free to comment here or in the issue.
[1] https://pagure.io/koji/issue/2690
koji-devel@lists.fedorahosted.org