In an IRC conversation with dgregor he offered the following opinion: The current state is bad (the Ignore* functions are not well specified). We'll want to go towards clearer specified functions for ignoring all implicit magic and getting to what was written, and for getting through redundant syntax to a semantically equivalent expression (and perhaps others in the future).
The first would be what we'll want to do with IgnoreImplicit - so basically we'll want to build IgnoreImplicit towards ignoring *all* implicit magic and getting us what was written syntactiacally. The only place IgnoreImplicit seems to be used in currently is ARCMigrate, and we should see failing tests if we break its invariants... http://llvm-reviews.chandlerc.com/D720 _______________________________________________ cfe-commits mailing list [email protected] http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits
