Space-Project/dependencies/glm/gtx/component_wise.inl

129 lines
4.7 KiB
Plaintext
Raw Normal View History

2021-01-16 17:11:39 +01:00
/// @ref gtx_component_wise
/// @file glm/gtx/component_wise.inl
#include <limits>
namespace glm{
namespace detail
{
template <typename T, typename floatType, precision P, template <typename, precision> class vecType, bool isInteger, bool signedType>
struct compute_compNormalize
{};
template <typename T, typename floatType, precision P, template <typename, precision> class vecType>
struct compute_compNormalize<T, floatType, P, vecType, true, true>
{
GLM_FUNC_QUALIFIER static vecType<floatType, P> call(vecType<T, P> const & v)
{
floatType const Min = static_cast<floatType>(std::numeric_limits<T>::min());
floatType const Max = static_cast<floatType>(std::numeric_limits<T>::max());
return (vecType<floatType, P>(v) - Min) / (Max - Min) * static_cast<floatType>(2) - static_cast<floatType>(1);
}
};
template <typename T, typename floatType, precision P, template <typename, precision> class vecType>
struct compute_compNormalize<T, floatType, P, vecType, true, false>
{
GLM_FUNC_QUALIFIER static vecType<floatType, P> call(vecType<T, P> const & v)
{
return vecType<floatType, P>(v) / static_cast<floatType>(std::numeric_limits<T>::max());
}
};
template <typename T, typename floatType, precision P, template <typename, precision> class vecType>
struct compute_compNormalize<T, floatType, P, vecType, false, true>
{
GLM_FUNC_QUALIFIER static vecType<floatType, P> call(vecType<T, P> const & v)
{
return v;
}
};
template <typename T, typename floatType, precision P, template <typename, precision> class vecType, bool isInteger, bool signedType>
struct compute_compScale
{};
template <typename T, typename floatType, precision P, template <typename, precision> class vecType>
struct compute_compScale<T, floatType, P, vecType, true, true>
{
GLM_FUNC_QUALIFIER static vecType<T, P> call(vecType<floatType, P> const & v)
{
floatType const Max = static_cast<floatType>(std::numeric_limits<T>::max()) + static_cast<floatType>(0.5);
vecType<floatType, P> const Scaled(v * Max);
vecType<T, P> const Result(Scaled - static_cast<floatType>(0.5));
return Result;
}
};
template <typename T, typename floatType, precision P, template <typename, precision> class vecType>
struct compute_compScale<T, floatType, P, vecType, true, false>
{
GLM_FUNC_QUALIFIER static vecType<T, P> call(vecType<floatType, P> const & v)
{
return vecType<T, P>(vecType<floatType, P>(v) * static_cast<floatType>(std::numeric_limits<T>::max()));
}
};
template <typename T, typename floatType, precision P, template <typename, precision> class vecType>
struct compute_compScale<T, floatType, P, vecType, false, true>
{
GLM_FUNC_QUALIFIER static vecType<T, P> call(vecType<floatType, P> const & v)
{
return v;
}
};
}//namespace detail
template <typename floatType, typename T, precision P, template <typename, precision> class vecType>
GLM_FUNC_QUALIFIER vecType<floatType, P> compNormalize(vecType<T, P> const & v)
{
GLM_STATIC_ASSERT(std::numeric_limits<floatType>::is_iec559, "'compNormalize' accepts only floating-point types for 'floatType' template parameter");
return detail::compute_compNormalize<T, floatType, P, vecType, std::numeric_limits<T>::is_integer, std::numeric_limits<T>::is_signed>::call(v);
}
template <typename T, typename floatType, precision P, template <typename, precision> class vecType>
GLM_FUNC_QUALIFIER vecType<T, P> compScale(vecType<floatType, P> const & v)
{
GLM_STATIC_ASSERT(std::numeric_limits<floatType>::is_iec559, "'compScale' accepts only floating-point types for 'floatType' template parameter");
return detail::compute_compScale<T, floatType, P, vecType, std::numeric_limits<T>::is_integer, std::numeric_limits<T>::is_signed>::call(v);
}
template <typename T, precision P, template <typename, precision> class vecType>
GLM_FUNC_QUALIFIER T compAdd(vecType<T, P> const & v)
{
T Result(0);
for(length_t i = 0, n = v.length(); i < n; ++i)
Result += v[i];
return Result;
}
template <typename T, precision P, template <typename, precision> class vecType>
GLM_FUNC_QUALIFIER T compMul(vecType<T, P> const & v)
{
T Result(1);
for(length_t i = 0, n = v.length(); i < n; ++i)
Result *= v[i];
return Result;
}
template <typename T, precision P, template <typename, precision> class vecType>
GLM_FUNC_QUALIFIER T compMin(vecType<T, P> const & v)
{
T Result(v[0]);
for(length_t i = 1, n = v.length(); i < n; ++i)
Result = min(Result, v[i]);
return Result;
}
template <typename T, precision P, template <typename, precision> class vecType>
GLM_FUNC_QUALIFIER T compMax(vecType<T, P> const & v)
{
T Result(v[0]);
for(length_t i = 1, n = v.length(); i < n; ++i)
Result = max(Result, v[i]);
return Result;
}
}//namespace glm