[cfe-dev] ConditionalOperator::setCond is gone

Olaf Krzikalla Olaf.Krzikalla at tu-dresden.de
Thu Mar 10 06:03:26 PST 2011


Am 09.03.2011 17:27, schrieb Douglas Gregor:
> It's bad library design to write a library that can't maintain its own invariants, but that's exactly the problem with ConditionalOperator::setCond() and the many other mutators in the AST. I'm fine with having some way of mutating the AST, if it preserves invariants.
OK, agreed. So please can you tell me which invariants in particular are 
violated by inplace mutations. Maybe we can find a satisfying solution.

Best Olaf



More information about the cfe-dev mailing list