[PATCH] D18919: [Clang-tidy] Add check "modernize use using"
Piotr Padlewski via cfe-commits
cfe-commits at lists.llvm.org
Tue May 3 03:01:37 PDT 2016
Prazek added inline comments.
================
Comment at: clang-tidy/modernize/UseUsingCheck.cpp:22
@@ +21,3 @@
+void UseUsingCheck::registerMatchers(MatchFinder *Finder) {
+ if (!getLangOpts().CPlusPlus)
+ return;
----------------
hokein wrote:
> Prazek wrote:
> > hokein wrote:
> > > Should be CplusPlus11 here.
> > BTW is there any policy about that? I see that some checks from modernize require C++11 (e.g. modernize-make-unique which is in C++14) and other require just C++ (modernize-loop-convert), and it even has a comment
> >
> > // Only register the matchers for C++. Because this checker is used for
> > // modernization, it is reasonable to run it on any C++ standard with the
> > // assumption the user is trying to modernize their codebase.
> > if (!getLangOpts().CPlusPlus)
> > return;
> >
> >
> > I have 2 thoughts for this:
> > 1. there should be note in documentation about it, so the user won't spend time debuging why the check doesn't do anything. e.g. "This check requires to compile code with C++11 or higher"
> > 2. I would suggest modernize checks to require standard version the same or higher for C++ standars that doesn't break backwards compatibility:
> > e.g. loop-convert should require C++11, make-shared C++14, this check also C++11,
> > but for modernize-increment-bool, that is deprecated in C++17, it should require just C++ (because if someone need it, the he wont be able to compile his code with C++17).
> This is a good point.
>
> As far as I know, we don't have detailed policy about the modernized checks. It depends on the check author. Basically the modern words means "C++11" feature.
>
> I'm +1 on adding a note in each modernized check's document.
Cool. I sent email to mailing list to ask other devs for opinion.
Krystyna, please add note at the end of documentation
"This check requires using C++11 or higher to run."
Repository:
rL LLVM
http://reviews.llvm.org/D18919
More information about the cfe-commits
mailing list