I left a comment on the issue but I don't think we can use a pattern like this. The 3:CL:12 stuff is indy specific. For the cheqd and the upcoming webvh method these identifiers are much different and can be really almost anything. That's why I changed it to this. I meant to just match anything.

I think every pattern with at least have the did method at the beginning. So we could have a did:<any sting only letters>:<any string> pattern here.

Same with the other object id's.

Ah yes, but a validator that matches anything is not really a validator :-)

If it must start with a qualified did, that's at least something. The pattern after that can be more random, but it assists in understanding the code. So I think it's worth sussing out the true restrictions on what the patterns are -- and then expanding those patterns later if we run into issues

Yes. But I remember the legacy indy(sov) method that still persists (try to change this), doesn't even have the did:sov part at the beginning.

I think I was lost on what to do for the validator and that's why I had it this way. Maybe not having a validator at all would be better for now?

I figure that's an option too. Skipping validation, just to save on regex compute and make the earth greener 😄

Of course it's much of a muchness, doing .+ regex check vs skipping it.
I figured it's at least worth marking this as a to-do, so that it's tracked as an unanswered question: what is the correct validation for anoncreds cred def / schema id / rev reg id. If we as a community can't answer that, then something's wrong 😄 if we're still dragging technical debt of needing to handle unqualified dids, maybe that should be prioritised