On December 1, 2024 7:57:45 PM Avi Kivity <avi_at_[hidden]> wrote:

> On Sun, 2024-12-01 at 19:11 +0300, Andrey Semashev via Std-Proposals
> wrote:
>> On December 1, 2024 6:57:16 PM Avi Kivity via Std-Proposals
>> <std-proposals_at_[hidden]> wrote:
>>
>>> Functions and member functions are invocable, but constructors are
>>> not.
>>>
>>> I propose to add
>>>
>>> template <typename T, typename... Args>
>>> T std::construct(Args&&... args) {
>>> return T(std::forward<decltype(Args)>(args)...);
>>> }
>>>
>>> With this, we can pass a constructor where other functions can be
>>> passed.
>>>
>>> // build a callback that creates and returns a Type1 thing
>>> std::function<Type1 (Arg3)> make_somthing =
>>> std::bind_front(std::construct<Type1>, arg1, std::ref(arg2));
>>>
>>> // transform a vector of ints to a vector of some other type
>>> auto foo = some_container |
>>> std::views::transform(std::construct<AnotherType>) |
>>> std::ranges::to<std::vector>();
>>
>> Binding a function pointer is less efficient than a function object.
>
> Typically, inlining will remove the difference.

The function pointer must still increase the size of the binder, which may
make it too large to fit in e. g. std::function small buffer.

And whether the compiler is able to devirtualize the call is more of a
question than with a function object with an inline operator().

> Or we can make it a
> niebloid.

Sorry, I don't know what this means.

>
>
>> And
>> given that lambdas are an easy way to create function objects, I
>> don't
>> really see the benefits of the proposal.
>
>
> I believe that writing
>
> std::views::transform(std::construct<AnotherType>)
>
> is more readable than
>
> std::views::transform([] (auto&& x) { return
> AnotherType<std::forward<decltype(x)>(x); });

I think this is not equivalent to the original example. Is AnotherType a
template?