RFC: TypeTraits Variables

[Vicente J. Botet Escriba]

Hi,

please see the attached proposal that presents an alternative to the not adopted section 4 of “N3546 - TransformationTraits Redux , v2” that makes use of the new template variables C++14 feature to provide a shorter name to access associated template struct traits ::value.
For example

template<typename T>
constexpr bool is_reference_c = is_reference<T>::value;

Comments welcome,
Vicente

[Akira Takahashi]

HI, Vicente

2013/5/26 Vicente J. Botet Escriba <vicent...@wanadoo.fr>

I totally agree!

I think need more variable.

<functional> header

namespace std {

  template <class T> constexpr bool is_bind_expression_c = is_bind_expression<T>::value;

  template <class T> constexpr int is_placeholder_c = is_placeholder<T>::value;

}

Thanks,

Akira

>>========================
Akira Takahashi
mailto:faitha...@gmail.com
site: https://sites.google.com/site/faithandbrave/about/en

[Jonathan Wakely]

Concepts Lite will add constexpr functions that allow Is_default_constructible<T>() and Is_same<T, U>(), do we need variables if we have those?  Is there any advantage to having both?  I would expect a proposal to include some discussion of that.

[toma...@gmail.com]

W dniu wtorek, 28 maja 2013 12:37:43 UTC+2 użytkownik Jonathan Wakely napisał:

Concepts Lite will add constexpr functions that allow Is_default_constructible<T>() and Is_same<T, U>(), do we need variables if we have those?  Is there any advantage to having both?  I would expect a proposal to include some discussion of that.

With the acceptance of the operator() for integral_constant, the both proposal may be merged by declaring:
template<typename T>
constexpr is_bind_expression<T> Is_bind_expression;

Then Is_bind_expression<T> will be both convertible to bool (via constexpr conversion operator) and expression Is_bind_expression<T>() used by Concepts-lite will be well-formed.

[Vicente J. Botet Escriba]

Le 28/05/13 12:37, Jonathan Wakely a �crit :

I would expect that concept lite accept template bool variables in the
same way it accepts constexpr nullary boolean functions. As variables
don't need the "()" I think it is enough to support the variable form.

The problem on the current experimentation with ConceptLite is that we
don't have yet template variables.

Vicente

[Vicente J. Botet Escriba]

Le 28/05/13 14:50, toma...@gmail.com a �crit�:

W dniu wtorek, 28 maja 2013 12:37:43 UTC+2 u�ytkownik Jonathan Wakely napisa�:

Concepts Lite will add constexpr functions that allow Is_default_constructible<T>() and Is_same<T, U>(), do we need variables if we have those?ďż˝ Is there any advantage to having both?ďż˝ I would expect a proposal to include some discussion of that.

With the acceptance of the operator() for integral_constant, the both proposal may be merged by declaring:
template<typename T>
constexpr is_bind_expression<T> Is_bind_expression;

Then Is_bind_expression<T> will be both convertible to bool (via constexpr conversion operator) and expression Is_bind_expression<T>() used by Concepts-lite will be well-formed.

Interesting, but I'm not sure ConceptLite would accept it as shortcut

ďż˝ template <Is_bind_expression T> f(T);

Otherwise I would expect that this could be used in the requires clause, even without "()"

ďż˝ template <typename T>ďż˝
ďż˝ requites Is_bind_expression<T>
ďż˝ f(T);

Vicente