On Wed, Aug 8, 2018 at 5:00 AM Ilya Biryukov via Phabricator < revi...@reviews.llvm.org> wrote:
> ilya-biryukov added a comment. > > In https://reviews.llvm.org/D50154#1191002, @dblaikie wrote: > > > What's the motivation for clangd to differ from clang here? > > > The presentation of diagnostics to the users is different between clang > and clangd: > > - clang diagnostics are always prefixed with the file, location and > severity of the diagnostic, e.g. `main.cpp:1:2: error: expected '}'` > *nod* That's a fairly natural one - and I assume the information is provided semantically - source, line, column, and text. A client can stitch them together as clang does or it can render the information some other way (by placing the text as a mouseover/popup at the source location). I don't see that as a divergence - but providing the full fidelity of the information (rather than prematurely stringifying it all together) & leaving it up to the client to decide how to render it to the user. > - In LSP clients (VSCode in particular), the diagnostic message is the > first thing the user sees and it looks a little nicer (subjectively) if the > first letter is capitalized. > It looks nicer in VSCode because that's the UI convention for mouseover text (in Windows, at least) - first letter capitalized. But it seems hard to me to generalize from there to all or most LSP clients. (& I also worry that such a mechanical transformation might be erroneous in some messages) (well, there's at least one case where reversing this would fail - there's a diagnostic in Clang that starts with "ISO" - so uppercasing does nothing, but lowercasing it to try to get back to the original would result in "iSO" - several others like "Objective-C", "Neon", "GCC", etc) (also, there's one that this will mangle: "ms_struct may not produce Microsoft-compatible layouts" - I assume capitalizing "ms_struct" would be incorrect (though arguably/mostly identifiers like that are quoted in error messages - and that may be reasonable/correct/better to do that here). Similar use of 'sizeof' at the start of a diagnostic, 'os_log()', - those'll be an issue no matter where the implementation is (in clangd or its client(s)) though - short of having a more customized thing in the diagnostics table itself to record which things can be capitalized/how they should be capitalized - but, yeah, arguably quoting might be the right thing in many, maybe all, of those cases) > > See the screenshots from VSCode on how diagnostics are presented: > F6901986: image.png <https://reviews.llvm.org/F6901986> F6901990: > image.png <https://reviews.llvm.org/F6901990> > > > Repository: > rCTE Clang Tools Extra > > https://reviews.llvm.org/D50154 > > > >
_______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
