std::numeric_limits<T>::traps

From cppreference.com
 
 
Utilities library
General utilities
Date and time
Function objects
Formatting library (C++20)
(C++11)
Relational operators (deprecated in C++20)
Integer comparison functions
(C++20)(C++20)(C++20)   
(C++20)
Swap and type operations
(C++14)
(C++11)
(C++11)
(C++11)
(C++17)
Common vocabulary types
(C++11)
(C++17)
(C++17)
(C++17)
(C++11)
(C++17)
(C++23)
Elementary string conversions
(C++17)
(C++17)
 
 
 
static const bool traps;
(until C++11)
static constexpr bool traps;
(since C++11)

The value of std::numeric_limits<T>::traps is true for all arithmetic types T that have at least one value at the start of the program that, if used as an argument to an arithmetic operation, will generate a trap.

Standard specializations

T value of std::numeric_limits<T>::traps
/* non-specialized */ false
bool false
char usually true
signed char usually true
unsigned char usually true
wchar_t usually true
char8_t (since C++20) usually true
char16_t (since C++11) usually true
char32_t (since C++11) usually true
short usually true
unsigned short usually true
int usually true
unsigned int usually true
long usually true
unsigned long usually true
long long (since C++11) usually true
unsigned long long (since C++11) usually true
float usually false
double usually false
long double usually false

Notes

On most platforms integer division by zero always traps, and std::numeric_limits<T>::traps is true for all integer types that support the value 0. The exception is the type bool: even though division by false traps due to integral promotion from bool to int, it is the zero-valued int that traps. Zero is not a value of type bool.

On most platforms, floating-point exceptions may be turned on and off at run time (e.g. feenableexcept() on Linux or _controlfp on Windows), in which case the value of std::numeric_limits<T>::traps for floating-point types reflects the state of floating-point trapping facility at the time of program startup, which is false on most modern systems. An exception would be a DEC Alpha program, where it is true if compiled without -ieee.

Example

#include <iostream>
#include <limits>
 
int main()
{
    std::cout << std::boolalpha
              << "bool:     traps = " << std::numeric_limits<bool>::traps << '\n'
              << "char:     traps = " << std::numeric_limits<char>::traps << '\n'
              << "char16_t: traps = " << std::numeric_limits<char16_t>::traps << '\n'
              << "long:     traps = " << std::numeric_limits<long>::traps << '\n'
              << "float:    traps = " << std::numeric_limits<float>::traps << '\n';
}

Possible output:

// GCC output:
bool:     traps = true
char:     traps = true
char16_t: traps = true
long:     traps = true
float:    traps = false
 
// Clang output:
bool:     traps = false
char:     traps = true
char16_t: traps = true
long:     traps = true
float:    traps = false

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 497 C++98 it was unclear what is returned if trapping
is enabled or disabled at runtime
returns the enable status
at the start of the program

See also

Floating-point environment
identifies floating-point types that detect tinyness before rounding
(public static member constant)
identifies the floating-point types that detect loss of precision as denormalization loss rather than inexact result
(public static member constant)