We define custom type traits for number types we use. More...
Collaboration diagram for Type Traits:
Modules | |
| is_field_type | |
All types that represent a field are marked with is_field_type. | |
| is_finite_type | |
All types that can represent only numbers from a finite domain are marked with is_finite_type. | |
| is_float_type | |
All types that represent floating point numbers are marked with is_float_type. | |
| is_integer_type | |
All integral types that can (in theory) represent all integers are marked with is_integer_type. | |
| is_number_type | |
All types that represent any kind of number are marked with is_number_type. | |
| is_rational_type | |
All integral types that can (in theory) represent all rationals are marked with is_rational_type. | |
| IntegralType | |
| The associated integral type of any type can be defined with IntegralType. | |
| UnderlyingNumberType | |
| The number type that some type is built upon can be defined with UnderlyingNumberType. | |
Files | |
| file | typetraits.h |
| file | typetraits.h |
| file | typetraits.h |
| file | typetraits.h |
Detailed Description
We define custom type traits for number types we use.
We use the notation conventions of the STL, being lower cases with underscores.
We define the following type traits:
is_field_type: Types that represent elements from a field.is_finite_type: Types that represent only a finite domain.is_float_type: Types that represent real numbers using a floating point representation.is_integer_type: Types that represent the set of integral numbers.is_subset_of_integers_type: Types that may represent some integral numbers.is_number_type: Types that represent numbers.is_rational_type: Types that may represent any rational number.is_subset_of_rationals_type: Types that may represent some rational numbers.
A more exact definition for each of these type traits can be found in their own documentation.
Additionally, we define related types in a type traits like manner:
IntegralType: Integral type, that the given type is based on. For fractions, this would be the type of the numerator and denominator.UnderlyingNumberType: Number type that is used within a more complex type. For polynomials, this would be the number type of the coefficients.
Note that we keep away from similar type traits defined in the standard [4] (20.9) (like std::is_integral or std::is_floating_point, as they are not meant to be specialized for custom types.
