kuhar marked 2 inline comments as done.
kuhar added inline comments.
================
Comment at: clang-tidy/modernize/UseEmplaceCheck.cpp:93
+ to(functionDecl(hasName("::std::make_pair"))))
+ ))))
+ .bind("make_pair"));
----------------
Prazek wrote:
> JonasToth wrote:
> > is the new line here necessary? i think it looks better if the `.bind` is
> > on this line.
> Better question is "is it clang formated?"
It was clang-formatted, but I do agree that it looked a bit weird. I removed
the extra newline.
================
Comment at: test/clang-tidy/modernize-use-emplace.cpp:284
+ // CHECK-FIXES: v.emplace_back(42LL, 13);
+
+ v.push_back(std::make_pair<char, char>(0, 3));
----------------
Prazek wrote:
> I would add here test like:
>
> class X {
> X(std:;pair<int, int> a) {}
> };
>
> std::vector<X> v;
> v.push_back(make_pair(42, 42));
>
> I guess as long as X ctor is not explicit this can happen, and we can't
> transform it to
> emplace.back(42, 42)
Nice idea for a test case, added.
https://reviews.llvm.org/D32395
_______________________________________________
cfe-commits mailing list
[email protected]
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
