Difference between revisions of "cpp/utility/tuple size"
Andreas Krug (Talk | contribs) m (fmt) |
m (-`inline`: CWG2387; fmt.) |
||
(3 intermediate revisions by 2 users not shown) | |||
Line 2: | Line 2: | ||
{{cpp/utility/navbar}} | {{cpp/utility/navbar}} | ||
{{dcl begin}} | {{dcl begin}} | ||
− | |||
{{dcl header|array}} | {{dcl header|array}} | ||
+ | {{dcl header|tuple}} | ||
{{dcl header|utility}} | {{dcl header|utility}} | ||
{{dcl header|ranges|notes={{mark since c++20}}}} | {{dcl header|ranges|notes={{mark since c++20}}}} | ||
+ | {{dcl header|complex|notes={{mark since c++26}}}} | ||
{{dcl|num=1|since=c++11|1= | {{dcl|num=1|since=c++11|1= | ||
template< class T > | template< class T > | ||
Line 31: | Line 32: | ||
@1@ The primary template is not defined. An explicit (full) or partial specialization is required to make a type tuple-like. | @1@ The primary template is not defined. An explicit (full) or partial specialization is required to make a type tuple-like. | ||
− | @2-4@ Specializations for a cv-qualified types reuse the {{ | + | @2-4@ Specializations for a cv-qualified types reuse the {{c|value}} from the corresponding cv-unqualified versions by default. |
{{rrev|since=c++17| | {{rrev|since=c++17| | ||
Line 37: | Line 38: | ||
{{v|2-4}} are SFINAE-friendly: if {{c|std::tuple_size<T>::value}} is | {{v|2-4}} are SFINAE-friendly: if {{c|std::tuple_size<T>::value}} is | ||
− | ill-formed when treated as an unevaluated operand, they do not provide the member {{ | + | ill-formed when treated as an unevaluated operand, they do not provide the member {{c|value}}. Access checking is performed as if in a context unrelated to {{tt|tuple_size}} and {{tt|T}}. Only the validity of the immediate context of the expression is considered. This allows |
{{source|1= | {{source|1= | ||
#include <utility> | #include <utility> | ||
Line 55: | Line 56: | ||
{{dsc inc|cpp/container/array/dsc tuple_size}} | {{dsc inc|cpp/container/array/dsc tuple_size}} | ||
{{dsc inc|cpp/ranges/subrange/dsc tuple_size}} | {{dsc inc|cpp/ranges/subrange/dsc tuple_size}} | ||
+ | {{dsc inc|cpp/numeric/complex/dsc tuple_size}} | ||
{{dsc end}} | {{dsc end}} | ||
Line 66: | Line 68: | ||
{{ddcl|header=tuple|since=c++17|1= | {{ddcl|header=tuple|since=c++17|1= | ||
template< class T > | template< class T > | ||
− | + | constexpr std::size_t tuple_size_v = tuple_size<T>::value; | |
}} | }} | ||
Line 81: | Line 83: | ||
template<class T, std::size_t Size> struct Arr { T data[Size]; }; | template<class T, std::size_t Size> struct Arr { T data[Size]; }; | ||
− | // | + | |
+ | // Program-defined specialization of std::tuple_size: | ||
template<class T, std::size_t Size> struct std::tuple_size<Arr<T, Size>> | template<class T, std::size_t Size> struct std::tuple_size<Arr<T, Size>> | ||
: public integral_constant<std::size_t, Size> {}; | : public integral_constant<std::size_t, Size> {}; |
Latest revision as of 09:02, 24 September 2024
Defined in header <array>
|
||
Defined in header <tuple>
|
||
Defined in header <utility>
|
||
Defined in header <ranges>
|
(since C++20) |
|
Defined in header <complex>
|
(since C++26) |
|
template< class T > struct tuple_size; // not defined |
(1) | (since C++11) |
template< class T > struct tuple_size< const T > |
(2) | (since C++11) |
template< class T > struct tuple_size< volatile T > |
(3) | (since C++11) (deprecated in C++20) |
template< class T > struct tuple_size< const volatile T > |
(4) | (since C++11) (deprecated in C++20) |
Provides access to the number of elements in a tuple-like type as a compile-time constant expression.
(2-4) are SFINAE-friendly: if std::tuple_size<T>::value is
ill-formed when treated as an unevaluated operand, they do not provide the member value. Access checking is performed as if in a context unrelated to #include <utility> struct X { int a, b; }; const auto [x, y] = X(); // structured binding declaration first attempts // tuple_size<const X> which attempts to use tuple_size<X>::value, // then soft error encountered, binds to public data members |
(since C++17) |
Contents |
[edit] Specializations
The standard library provides following specializations for standard library types:
(C++11) |
obtains the size of
a |
(C++11) |
obtains the size of a pair (class template specialization) |
(C++11) |
obtains the size of an array (class template specialization) |
obtains the size of a std::ranges::subrange (class template specialization) | |
obtains the size of a std::complex (class template specialization) |
All specializations of std::tuple_size
satisfy UnaryTypeTrait with base characteristic std::integral_constant<std::size_t, N> for some N
.
Users may specialize std::tuple_size
for program-defined types to make them tuple-like. Program-defined specializations must meet the requirements above.
Usually only specialization for cv-unqualified types are needed to be customized.
[edit] Helper variable template
Defined in header <tuple>
|
||
template< class T > constexpr std::size_t tuple_size_v = tuple_size<T>::value; |
(since C++17) | |
Inherited from std::integral_constant
Member constants
value [static] |
for a standard specialization, the number of elements in the tuple-like type T (public static member constant) |
Member functions
operator std::size_t |
converts the object to std::size_t, returns value (public member function) |
operator() (C++14) |
returns value (public member function) |
Member types
Type | Definition |
value_type
|
std::size_t |
type
|
std::integral_constant<std::size_t, value> |
[edit] Example
#include <array> #include <cstddef> #include <ranges> #include <tuple> #include <utility> template<class T, std::size_t Size> struct Arr { T data[Size]; }; // Program-defined specialization of std::tuple_size: template<class T, std::size_t Size> struct std::tuple_size<Arr<T, Size>> : public integral_constant<std::size_t, Size> {}; int main() { using tuple1 = std::tuple<int, char, double>; static_assert(3 == std::tuple_size_v<tuple1>); // uses using template (C++17) using array3x4 = std::array<std::array<int, 3>, 4>; static_assert(4 == std::tuple_size<array3x4>{}); // uses operator std::size_t using pair = std::pair<tuple1, array3x4>; static_assert(2 == std::tuple_size<pair>()); // uses operator() using sub = std::ranges::subrange<char*, char*>; static_assert(2 == std::tuple_size<sub>::value); using Arr5 = Arr<int, 5>; static_assert(5 == std::tuple_size_v<Arr5>); }
[edit] Defect reports
The following behavior-changing defect reports were applied retroactively to previously published C++ standards.
DR | Applied to | Behavior as published | Correct behavior |
---|---|---|---|
LWG 2212 | C++11 | specializations for cv types were not required in some headers, which led to ambiguity | required |
[edit] See also
Structured binding (C++17) | binds the specified names to sub-objects or tuple elements of the initializer |
(C++11) |
obtains the element types of a tuple-like type (class template) |
(C++11) |
creates a tuple by concatenating any number of tuples (function template) |