Namespaces
Variants
Views
Actions

std::ranges::min_max_result

From cppreference.com
< cpp‎ | algorithm‎ | ranges
 
C++
 
Constrained algorithms
Non-modifying sequence operations
Modifying sequence operations
Operations on uninitialized storage
Partitioning operations
Sorting operations
Binary search operations
Set operations (on sorted ranges)
Heap operations
Minimum/maximum operations
Permutations
 
Return types of constrained algorithms
 
Defined in header <algorithm>
template <class T>
struct min_max_result;
 (since C++20)

ranges::min_max_result is a class template that provides a way to store two objects as a single unit.

This class template has no base classes or declared members other than those shown below. Thus it is suitable for use with structured bindings.

All special member functions of this class template are implicitly declared, which makes specializations be aggregate classes, and propagate triviality, potentially-throwing-ness, and constexpr-ness of corresponding operations on data members.

Contents

[edit] Template parameters

T - the type of the objects that the ranges::min_max_result stores.

[edit] Data members

std::ranges::min_max_result::min

[[no_unique_address]] T min;

may be a referencing object that is supposedly points to a minimum element in a range

std::ranges::min_max_result::max

[[no_unique_address]] T max;

may be a referencing object that is supposedly points to a maximum element in a range

[edit] Member functions

std::ranges::min_max_result::operator min_max_result<T2>

template<class T2>

requires std::convertible_to<const T&, T2>

constexpr operator min_max_result<T2>() const &;
 (1)
template<class T2>

requires std::convertible_to<T, T2>

constexpr operator min_max_result<T2>() &&;
 (2)

Converts *this to the result by constructing every data member of the result from the corresponding member of *this.

1) Equivalent to return {min, max};.
2) Equivalent to return {std::move(min), std::move(max)};.

[edit] Standard library

The following standard library functions use ranges::min_max_result as the return type:

Algorithm functions
(C++20)
 returns the smaller and larger of two elements
(niebloid) [edit]
(C++20)
 returns the smallest and the largest elements in a range
(niebloid) [edit]

[edit] Synopsis 

namespace std::ranges {
    template<class T>
    struct min_max_result {
        [[no_unique_address]] T min;
        [[no_unique_address]] T max;
 
        template<class T2>
        requires std::convertible_to<const T&, T2>
        constexpr operator min_max_result<T2>() const & {
            return {min, max};
        }
 
        template<class T2>
        requires std::convertible_to<T, T2>
        constexpr operator min_max_result<T2>() && {
            return {std::move(min), std::move(max)};
        }
    };
}

[edit] Notes

Each standard library algorithm that uses this family of return types declares a new alias type, e.g. using merge_result = in_in_out_result<I1, I2, O>;.

The names for such aliases are formed by adding the suffix "_result" to the algorithm's name. So, the return type of std::ranges::merge can be named as std::ranges::merge_result.

Unlike std::pair and std::tuple, this class template has data members of meaningful names.

[edit] Example

Run this code
#include <algorithm>
#include <ranges>
 
int main()
{
    constexpr static auto v = {3, 1, 4, 1, 5, 9, 2};
    {
        constexpr auto result = std::ranges::minmax(v);
        static_assert(1 == result.min && 9 == result.max);
    }
    {
        constexpr auto result = std::ranges::minmax_element(v);
        static_assert(1 == *result.min && 9 == *result.max);
    }
}

[edit] See also

 implements binary tuple, i.e. a pair of values
(class template) [edit]
(C++11)
 implements fixed size container, which holds elements of possibly different types
(class template) [edit]
Retrieved from "https://en.cppreference.com/mwiki/index.php?title=cpp/algorithm/ranges/return_types/min_max_result&oldid=125634"