Open lenary opened 9 years ago
It's the same situation with simplified expressions and the cil code generator has some rather fragile support for that.
It would be much nicer not to have to handle multiple SDefaultCase
clauses as well.
This probably occurs when SucCase is compiled into DefaultCase.
In the defunctionalised form of
Prelude.Classes.Eq.==
for Nat, there are twoDDefaultCase
clauses.I think this violates the (unwritten) agreement that code generators will only get one default case in a case statement. This isn't an issue for my code generator, but I could see it becoming an issue for the C compiler, for example.