std::projected - cppreference.com (original) (raw)

Defined in header
(1)
template< std::indirectly_readable I, std::indirectly_regular_unary_invocable<I> Proj > struct projected { using value_type = std::remove_cvref_t <std::indirect_result_t<Proj&, I>>; std::indirect_result_t<Proj&, I> operator*() const; // not defined }; (since C++20) (until C++26)
template< std::indirectly_readable I, std::indirectly_regular_unary_invocable<I> Proj > using projected = /*projected-impl*/<I, Proj>::/*type*/; (since C++26)
template< std::weakly_incrementable I, class Proj > struct incrementable_traits<std::projected<I, Proj>> { using difference_type = std::iter_difference_t<I>; }; (2) (since C++20) (until C++26)
Helper templates
template< class I, class Proj > struct /*projected-impl*/ { struct /*type*/ { using value_type = std::remove_cvref_t <std::indirect_result_t<Proj&, I>>; using difference_type = std::iter_difference_t<I>; // conditionally present std::indirect_result_t<Proj&, I> operator*() const; // not defined }; }; (3) (since C++26) (exposition only*)
  1. Class(until C++26)Alias(since C++26) template projected combines an indirectly_readable type I and a callable object type Proj into a new indirectly_readable type whose reference type is the result of applying Proj to the std::iter_reference_t<I>.

For the exposition-only nested class /*type*/, the nested type difference_type exists only if I models weakly_incrementable.

projected is used only to constrain algorithms that accept callable objects and projections, and hence its operator*() is not defined.

[edit] Template parameters

I - an indirectly readable type
Proj - projection applied to a dereferenced I

[edit] Notes

The indirect layer prevents I and Proj to be associated classes of projected. When an associated class of I or Proj is an incomplete class type, the indirect layer avoids the unnecessary attempt to inspect the definition of that type that results in hard error.

[edit] Example

[edit] See also