[BUG]: unicode handling maybe broken

Describe the bug

In a tweet someone posted https://godbolt.org/z/fdWx983sa which tries to std::wcout some unicode things.

Steps to reproduce

Click the link and see the output is SomeText F F

Expected behavior

The output should contain unicode chars, something like SomeText Φ Φ"

Reproduction link

https://godbolt.org/z/fdWx983sa

Screenshots

Not applicable

Operating System

No response

Browser version

No response

[partouf]

Dug a bit into cppreference and the GCC testsuite, and they basically do this too when testing std::wcout with unicode characters

    std::locale loc("en_US.UTF-8");
    std::locale::global(loc);

e.g. https://github.com/gcc-mirror/gcc/blob/16e2427f50c208dfe07d07f18009969502c25dc8/libstdc%2B%2B-v3/testsuite/27_io/objects/wchar_t/11.cc

They only seem to not use it in situations with just ASCII.

There’s also examples on the internet calling std::imbue(std::wcout, loc), but that only applies to language related formatting as far as I can tell. The global() call also affects the encoding/decoding.

It might be that it’s a GCC bug that they haven’t noticed because it’s never tested without global() - but that’s just speculation from my perspective. I don’t know what the standard says on this subject, but I bet it’s one of those “implementation defined” kind of things.

[deanturpin]

fmt::print also starts working with the addition of setlocale.

int main() {

  setlocale(LC_ALL, "en_US.utf8");

  const auto out = std::wstring{L"cya \u03A6"};
  fmt::print(L"{}\n", out);
}

https://godbolt.org/z/xKWjTs6z7