IIRC, Manuel expressed a concern offline that adding another way to configure matchers would make matchers more difficult to develop and use and will also add more complexities wrt dynamic matchers and tools like clang_query. I may be wrong or missing something, so it would be nice if Manuel answered himself. I remember us talking about different alternatives, but don't recall any idea that would work well with memoization etc.
One of the problems I was going to solve with this is avoid matching of typelocs, variables and expressions generated from range-based for loop and implicit assignment operators. http://reviews.llvm.org/D6580 EMAIL PREFERENCES http://reviews.llvm.org/settings/panel/emailpreferences/ _______________________________________________ cfe-commits mailing list [email protected] http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits
