massberg marked 2 inline comments as done. massberg added inline comments.
================ Comment at: clang-tools-extra/clangd/SemanticHighlighting.cpp:969 + if (ATL.isConstrained()) { + StartLoc = ATL.getConceptNameInfo().getEndLoc().getLocWithOffset(1); + } ---------------- sammccall wrote: > getLocWithOffset(1) is just jumping one byte in the source code > why is this correct? > > IIUC the intention here is to jump to the `auto` token, and the AutoTypeLoc > doesn't actually expose this > > I'd suggest: > - for now just bailing out - this is not important enough to hack around > (and the current hack doesn't seem robust) > - if you have the appetite, adding this location to AutoTypeLoc seems like a > good thing It seems that we can directly get the correct location, see above. The test had been wrong before as the auto in `auto` in `Fooable auto f = ..` should have a class deduced token as it would have without the concept. This is now fixed. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D154580/new/ https://reviews.llvm.org/D154580 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits