================
Comment at: clang-tidy/ClangTidyDiagnosticConsumer.h:50
@@ -48,3 +49,3 @@
struct ClangTidyError {
- ClangTidyError(const ClangTidyMessage &Message);
+ ClangTidyError(const std::string &CheckName, const ClangTidyMessage
&Message);
----------------
Daniel Jasper wrote:
> Why not StringRef?
Don't know why. Changed to StringRef.
================
Comment at: clang-tidy/ClangTidyDiagnosticConsumer.h:91
@@ -87,1 +90,3 @@
+ /// \brief Returns the name of clang-tidy check, which produced this
+ /// diagnostic ID.
----------------
Daniel Jasper wrote:
> .. name of the clang-tidy check which .. (add "the", remove comma).
Done.
================
Comment at: clang-tidy/ClangTidy.h:86
@@ -78,3 +85,3 @@
protected:
ClangTidyContext *Context;
----------------
Daniel Jasper wrote:
> Do the individual check implementation still need access to the Context? If
> not, making it private might be a good step towards better layering.
Done.
================
Comment at: clang-tidy/ClangTidy.h:79
@@ +78,3 @@
+ /// \brief Add a diagnostic with the check's name.
+ DiagnosticBuilder diag(SourceLocation Loc, StringRef Message);
+
----------------
Daniel Jasper wrote:
> All the other methods in Clang are call "Diag". Consistency with that might
> trump local consistency here (or consistency with the Coding Standards), but
> I am not sure.
I think, having a coding style-compliant name is better (as it's also local
style-compliant), as being compliant with the naming style of similar methods
in completely different classes. Plus, creating more style violations will make
fixing them harder, when/if someone decides to do this ;)
http://llvm-reviews.chandlerc.com/D2534
_______________________________________________
cfe-commits mailing list
[email protected]
http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits
