std::chrono::treat_as_floating_point
From cppreference.com
Defined in header <chrono>
|
||
template <class Rep> struct treat_as_floating_point : std::is_floating_point<Rep> {}; |
(since C++11) | |
The std::chrono::treat_as_floating_point
trait helps determine if a duration can be converted to another duration with a different tick period.
Implicit conversions between two durations normally depends on the tick period of the durations. However, implicit conversions can happen regardless of tick period if std::chrono::treat_as_floating_point<Rep>::value == true.
Helper variable template
template< class Rep > inline constexpr bool treat_as_floating_point_v = treat_as_floating_point<Rep>::value; |
(since C++17) | |
Example
Run this code
#include <iostream> #include <chrono> #include <thread> void timed_piece_of_code() { std::chrono::milliseconds simulated_work(2); std::this_thread::sleep_for(simulated_work); } int main() { auto start = std::chrono::high_resolution_clock::now(); std::cout << "Running some timed piece of code..." << '\n'; timed_piece_of_code(); auto stop = std::chrono::high_resolution_clock::now(); // A floating point milliseconds type using FpMilliseconds = std::chrono::duration<float, std::chrono::milliseconds::period>; static_assert(std::chrono::treat_as_floating_point<FpMilliseconds::rep>::value, "Rep required to be floating point"); // Note that implicit conversion is not allowed here auto i_ms = std::chrono::duration_cast<std::chrono::milliseconds>(stop - start); // Note that implicit conversion is allowed here auto f_ms = FpMilliseconds(stop - start); std::cout << "Time in milliseconds, using default rep: " << i_ms.count() << '\n'; std::cout << "Time in milliseconds, using floating point rep: " << f_ms.count() << '\n'; }
Possible output:
Running some timed piece of code... Timing stats: Time in milliseconds, using default rep: 2 Time in milliseconds, using floating point rep: 2.57307