sammccall accepted this revision. sammccall added a comment. This revision is now accepted and ready to land.
This looks OK, the asymmetry still seems a little odd to me and could be reduced a little. (Please also resolve Ilya's question about references with him, I don't have a strong opinion) ================ Comment at: clangd/index/SymbolCollector.cpp:303 + // canonical declaration in the index. + if (D->getFriendObjectKind() != Decl::FriendObjectKind::FOK_None) + D = ASTNode.OrigD; ---------------- Ah, now I understand - we're picking another to use as canonical. But the exception for definitions should apply here too. And there's nothing special about *OrigD* that makes it a good pick for canonical. For determinism, can we instead iterate over the redecls of D and pick the first one that's not a friend or is a definition? (I'd pull that check out into a function `shouldSkipDecl` and rename the existing one `shouldSkipSymbol`, but up to you) Repository: rCTE Clang Tools Extra https://reviews.llvm.org/D47623 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
