Ah, yeah - I had forgotten about this, but it's kind of important, otherwise it's not possible to tell apart indentation metadata for the same symbols in different namespaces.

Probably inclojure-ts-mode it makes sense to assume that if something is passed unqualified we should use the current namespace. On the other hand - I don't think we'll be able to deal with things like ns aliases, at least not without parsing everything in thens form.

Probably inclojure-ts-mode it makes sense to assume that if something is passed unqualified we should use the current namespace. On the other hand - I don't think we'll be able to deal with things like ns aliases, at least not without parsing everything in thens form.

Hmm, that's a good point, if macro is defined in the same namespace then this function will be called without ns in current implementation, but macro can also be required using:refer and using current ns will be not valid solution.

@bbatsov I'd probably suggest to merge this as is. I'll try to figure out how to deal with edge cases separately.

I've just tested it with CIDER, if a macro is imported using:refer it's actually being successfully resolved bycider-resolve--get-in, it also works fine if macro is imported with:use keyword, so it looks like that the fix from this PR is sufficient to make it work with CIDER.

Deal!