[PATCH] D76896: [clangd] Color dependent names based on their heuristic target if they have one

[Nathan Ridge via Phabricator via cfe-commits]

nridge added inline comments.


================
Comment at: clang-tools-extra/clangd/SemanticHighlighting.cpp:157
+resolveConflict(ArrayRef<HighlightingToken> Tokens) {
+  if (Tokens.size() != 2)
+    return llvm::None;
----------------
sammccall wrote:
> why a limit of 2, vs a loop?
I'm not aware of a scenario where we'd get more than two conflicting highlightings, except when macros are involved in which case we don't have a good way to disambiguate anyways.


================
Comment at: clang-tools-extra/clangd/SemanticHighlighting.cpp:211
         NonConflicting.push_back(TokRef.front());
+      else if (auto Resolved = resolveConflict(Conflicting))
+        NonConflicting.push_back(*Resolved);
----------------
sammccall wrote:
> out of curiosity, *why* is the same token being highlighted as dependentname and resolved? Is it being traversed twice?
It's for the reason you described in [this comment](https://github.com/clangd/clangd/issues/297#issuecomment-593736036): the resolved highlighting comes from `findExplicitReferences()`, and the dependent one form `CollectExtraHighlightings`.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D76896/new/

https://reviews.llvm.org/D76896