Support for pointers to incomplete types

[mgieseki]

First of all, thank you for providing this implementation of std::variant. It's great to have a lightweight alternative to boost::variant available in C++11. I just stumbled over an issue with recursive type definitions, and I'm not sure whether it's a limitation of C++11 or the variant implementation. The following struct definition compiles fine with GCC's and Clang's C++17 support:

struct S {
    std::variant<
        std::map<int, S>*
    > var;
};

When replacing std::variant with mpark::variant, I get an error message regarding the incomplete type S of std::pair<int, S>::second. Is it possible to enhance your implementation so that this kind of recursive type definition compiles with mpark::variant under C++11 as well?

[mpark]

Thanks for the report! I just tried it out and confirmed the issue. It either looks like my fallback implementation of is_swappable is instantiating too eagerly, or I'm checking for is_swappable where I shouldn't be. I'm going to be attending CppCon this week, so I'll be sure to fix this after the conference!

[mgieseki]

Awesome. Thanks for the confirmation and for having a look into the issue. Have a great time at CppCon.

[mpark]

The following commits address this issue:

• https://github.com/mpark/variant/commit/e567a710ea6429b921fd42fa415d12232eab7e46
• https://github.com/mpark/variant/commit/9e06a9d6a17fb31f2e4aabd42e59f48bd0ee0ed8
• https://github.com/mpark/variant/commit/ea542a0fef86fb8c1696238b95bdaddb8ff2337c

[mgieseki]

Great! Thanks for fixing this issue. It works flawlessly now.

[cor3ntin]

I'm experiencing something similar with gcc 7, seems to work with 8 https://godbolt.org/z/0fb-bV (also works if I use standard variant) Thanks!

[mpark]

The core of the issue here was that my implementation of invoke was instantiating too much. This is fixed by 5713032c68c6dad70a6a1d9f72b80c737c249a28