aaron.ballman added a comment.

> Which solution do you prefer?

If I understand the issue properly: both. :-)

Having the AST track information that's been folded away is still useful -- 
some users are using the AST for purposes other than codegen, and the fact that 
a construct has been folded away is good to know about while still retaining as 
much AST fidelity as possible.

On the other hand, from an AST matcher perspective, I think it's natural for 
users to write `ifStmt(isConstexpr())` and so that seems like a useful 
extension to the matcher. Further, it is extensible if the committee adds other 
constexpr foo statements.

As for which solution gets used by this check to fix the PR, I don't have a 
strong opinion at this time (currently at WG14 meetings and a bit distracted).


Repository:
  rCTE Clang Tools Extra

https://reviews.llvm.org/D46027



_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
  • [PATCH] D46027: [clang-tid... Gábor Horváth via Phabricator via cfe-commits
    • [PATCH] D46027: [clan... Aaron Ballman via Phabricator via cfe-commits

Reply via email to