> As stated before, I am not a fan of the pessimistic approach, but if everybody else is OK with always applying brackets in the type hierarchy pass.
Just one question, is there a possibility that there will be a dot inside a foreign name in a supertype name? The splitting logic would fail in such a case.
I have two more remarks. For one, I think we should use something like _unsanitizable_ instead of _insane_ in the log entries.
The second thing is that this code runs a lot on a hot pat