Component bslmf_isfloatingpoint [Package bslmf]

Provide a compile-time check for floating-point types. More...

Classes
struct	bsl::is_floating_point< t_TYPE >
struct	bsl::is_floating_point< float >
struct	bsl::is_floating_point< double >
struct	bsl::is_floating_point< long double >
struct	bsl::is_floating_point< const t_TYPE >
struct	bsl::is_floating_point< volatile t_TYPE >
struct	bsl::is_floating_point< const volatile t_TYPE >

---

Detailed Description

Outline

• Purpose
• Classes
• Description
  • Usage
    • Example 1: Verify Floating-Point Types

Purpose:

Provide a compile-time check for floating-point types.

Classes:

bsl::is_floating_point	meta-function for determining floating-point types
bsl::is_floating_point_v	the result value of bsl::is_floating_point

See also:

Component bslmf_integralconstant

Description:

This component defines a meta-function, bsl::is_floating_point and a template variable bsl::is_floating_point_v, that represents the result value of the bsl::is_floating_point meta-function, that may be used to query whether a type is a (possibly cv-qualified) floating-point type as defined in section 3.9.1.8 of the C++11 standard [basic.fundamental].

bsl::is_floating_point meets the requirements of the is_floating_point template defined in the C++11 standard [meta.unary.cat].

Note that the template variable is_floating_point_v is defined in the C++17 standard as an inline variable. If the current compiler supports the inline variable C++17 compiler feature, bsl::is_floating_point_v is defined as an inline constexpr bool variable. Otherwise, if the compiler supports the variable templates C++14 compiler feature, bsl::is_floating_point_v is defined as a non-inline constexpr bool variable. See BSLS_COMPILERFEATURES_SUPPORT_INLINE_VARIABLES and BSLS_COMPILERFEATURES_SUPPORT_VARIABLE_TEMPLATES macros in bsls_compilerfeatures component for details.

Usage:

In this section we show intended use of this component.

Example 1: Verify Floating-Point Types:

Suppose that we want to assert whether a particular type is a floating-point type.

First, we create two typedefs -- a floating-point type and a non-floating-point type:

  typedef void  MyType;
  typedef float MyFloatingPointType;

Now, we instantiate the bsl::is_floating_point template for each of the typedefs and assert the value static data member of each instantiation:

  assert(false == bsl::is_floating_point<MyType>::value);
  assert(true  == bsl::is_floating_point<MyFloatingPointType>::value);

Note that if the current compiler supports the variable templates C++14 feature, then we can re-write the snippet of code above using the 'bslis_floating_point_v<T> as follows:

#ifdef BSLS_COMPILERFEATURES_SUPPORT_VARIABLE_TEMPLATES
  assert(false == bsl::is_floating_point_v<MyType>);
  assert(true  == bsl::is_floating_point_v<MyFloatingPointType>);
#endif