For now, use the token kind Unknown. We may be able to improve on this
using HeuristicResolver.
Details
Diff Detail
- Repository
- rG LLVM Github Monorepo
Event Timeline
clang-tools-extra/clangd/unittests/SemanticHighlightingTests.cpp | ||
---|---|---|
656 | Not sure how I feel about this being a _decl. I know it is in the technical sense, but as a user I think of it more as a reference to Base::member. |
clang-tools-extra/clangd/unittests/SemanticHighlightingTests.cpp | ||
---|---|---|
656 | Yeah, I agree, and the combination of decl and dependentName is pretty funny. We have non-dependent equivalent cases (using ::foo, not the renaming versions) and they don't seem to be marked as decl. (I think what's happening is that findExplicitReferences yields the *UsingShadowDecl*s rather than the *UsingDecl*, as only the former are directly bound to a specific e.g. overload being referenced) So with that as precedent, it seems like it would be legitimate to special-case UnresolvedUsingValueDecl when adding the decl modifier. |
Not sure how I feel about this being a _decl.
I know it is in the technical sense, but as a user I think of it more as a reference to Base::member.