From cppreference.com
template<classT>structhas_unique_object_representations;(since C++17)std::has_unique_object_representations is a
.
If T is
and if any two objects of type T with the same value have the same
, provides the member constant value equal true. For any other type, value is false.
For the purpose of this trait, two arrays have the same value if their elements have the same values, two non-union classes have the same value if their direct subobjects have the same value, and two unions have the same value if they have the same active member and the value of that member is the same.
It is implementation-defined which scalar types satisfy this trait, but unsigned(until C++20) integer types that do not use padding bits are guaranteed to have unique object representations.
If std::remove_all_extents_t<T> is an incomplete type other than (possibly cv-qualified) void, the behavior is undefined.
If the program adds specializations for std::has_unique_object_representations or std::has_unique_object_representations_v, the behavior is undefined.
Template parameters
T - a type to check Helper variable template
template<classT>constexprboolhas_unique_object_representations_v=has_unique_object_representations<T>::value;(since C++17)Inherited from
Member constants
value
[static]
true if T has unique object representations, false otherwise
(public static member constant)Member functions
operator bool
converts the object to bool, returns value
(public member function)operator()
(C++14)
returns value
(public member function)Member types
Type Definition value_typebooltypestd::integral_constant<bool,value>Notes
This trait was introduced to make it possible to determine whether a type can be correctly hashed by hashing its object representation as a byte array.
macroValueStdFeature
__cpp_lib_has_unique_object_representations
(C++17)std::has_unique_object_representationsExample
Run this code
#include<cstdint>#include<type_traits>structunpadded{std::uint32_ta,b;};structlikely_padded{std::uint8_tc;std::uint16_tst;std::uint32_ti;};intmain(){// Every value of a char corresponds to exactly one object representation.static_assert(std::has_unique_object_representations_v<char>);// For IEC 559 floats, assertion passes because the value NaN has// multiple object representations.static_assert(!std::has_unique_object_representations_v<float>);// Should succeed in any sane implementation because unpadded// is typically not padded, and std::uint32_t cannot contain padding bits.static_assert(std::has_unique_object_representations_v<unpadded>);// Fails in most implementations because padding bits are inserted// between the data members c and st for the purpose of aligning st to 16 bits.static_assert(!std::has_unique_object_representations_v<likely_padded>);// Notable architectural divergence:static_assert(std::has_unique_object_representations_v<bool>);// x86// static_assert(!std::has_unique_object_representations_v<bool>); // ARM}Defect reports
The following behavior-changing defect reports were applied retroactively to previously published C++ standards.
DR Applied to Behavior as published Correct behavior
C++17 T could be an array of unknown bound
even if its element type is incomplete required the element
type to be complete See also