195 lines
5.7 KiB
C++
195 lines
5.7 KiB
C++
// (C) Copyright John Maddock 2006.
|
|
// (C) Copyright Johan Rade 2006.
|
|
// (C) Copyright Paul A. Bristow 2011 (added changesign).
|
|
|
|
// Use, modification and distribution are subject to the
|
|
// Boost Software License, Version 1.0. (See accompanying file
|
|
// LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
|
|
|
|
#ifndef BOOST_MATH_TOOLS_SIGN_HPP
|
|
#define BOOST_MATH_TOOLS_SIGN_HPP
|
|
|
|
#ifdef _MSC_VER
|
|
#pragma once
|
|
#endif
|
|
|
|
#include <boost/math/tools/config.hpp>
|
|
#include <boost/math/special_functions/math_fwd.hpp>
|
|
#include <boost/math/special_functions/detail/fp_traits.hpp>
|
|
|
|
namespace boost{ namespace math{
|
|
|
|
namespace detail {
|
|
|
|
// signbit
|
|
|
|
#ifdef BOOST_MATH_USE_STD_FPCLASSIFY
|
|
template<class T>
|
|
inline int signbit_impl(T x, native_tag const&)
|
|
{
|
|
return (std::signbit)(x);
|
|
}
|
|
#endif
|
|
|
|
// Generic versions first, note that these do not handle
|
|
// signed zero or NaN.
|
|
|
|
template<class T>
|
|
inline int signbit_impl(T x, generic_tag<true> const&)
|
|
{
|
|
return x < 0;
|
|
}
|
|
|
|
template<class T>
|
|
inline int signbit_impl(T x, generic_tag<false> const&)
|
|
{
|
|
return x < 0;
|
|
}
|
|
|
|
#if defined(__GNUC__) && (LDBL_MANT_DIG == 106)
|
|
//
|
|
// Special handling for GCC's "double double" type,
|
|
// in this case the sign is the same as the sign we
|
|
// get by casting to double, no overflow/underflow
|
|
// can occur since the exponents are the same magnitude
|
|
// for the two types:
|
|
//
|
|
inline int signbit_impl(long double x, generic_tag<true> const&)
|
|
{
|
|
return (boost::math::signbit)(static_cast<double>(x));
|
|
}
|
|
inline int signbit_impl(long double x, generic_tag<false> const&)
|
|
{
|
|
return (boost::math::signbit)(static_cast<double>(x));
|
|
}
|
|
#endif
|
|
|
|
template<class T>
|
|
inline int signbit_impl(T x, ieee_copy_all_bits_tag const&)
|
|
{
|
|
typedef BOOST_DEDUCED_TYPENAME fp_traits<T>::type traits;
|
|
|
|
BOOST_DEDUCED_TYPENAME traits::bits a;
|
|
traits::get_bits(x,a);
|
|
return a & traits::sign ? 1 : 0;
|
|
}
|
|
|
|
template<class T>
|
|
inline int signbit_impl(T x, ieee_copy_leading_bits_tag const&)
|
|
{
|
|
typedef BOOST_DEDUCED_TYPENAME fp_traits<T>::type traits;
|
|
|
|
BOOST_DEDUCED_TYPENAME traits::bits a;
|
|
traits::get_bits(x,a);
|
|
|
|
return a & traits::sign ? 1 : 0;
|
|
}
|
|
|
|
// Changesign
|
|
|
|
// Generic versions first, note that these do not handle
|
|
// signed zero or NaN.
|
|
|
|
template<class T>
|
|
inline T (changesign_impl)(T x, generic_tag<true> const&)
|
|
{
|
|
return -x;
|
|
}
|
|
|
|
template<class T>
|
|
inline T (changesign_impl)(T x, generic_tag<false> const&)
|
|
{
|
|
return -x;
|
|
}
|
|
#if defined(__GNUC__) && (LDBL_MANT_DIG == 106)
|
|
//
|
|
// Special handling for GCC's "double double" type,
|
|
// in this case we need to change the sign of both
|
|
// components of the "double double":
|
|
//
|
|
inline long double (changesign_impl)(long double x, generic_tag<true> const&)
|
|
{
|
|
double* pd = reinterpret_cast<double*>(&x);
|
|
pd[0] = boost::math::changesign(pd[0]);
|
|
pd[1] = boost::math::changesign(pd[1]);
|
|
return x;
|
|
}
|
|
inline long double (changesign_impl)(long double x, generic_tag<false> const&)
|
|
{
|
|
double* pd = reinterpret_cast<double*>(&x);
|
|
pd[0] = boost::math::changesign(pd[0]);
|
|
pd[1] = boost::math::changesign(pd[1]);
|
|
return x;
|
|
}
|
|
#endif
|
|
|
|
template<class T>
|
|
inline T changesign_impl(T x, ieee_copy_all_bits_tag const&)
|
|
{
|
|
typedef BOOST_DEDUCED_TYPENAME fp_traits<T>::sign_change_type traits;
|
|
|
|
BOOST_DEDUCED_TYPENAME traits::bits a;
|
|
traits::get_bits(x,a);
|
|
a ^= traits::sign;
|
|
traits::set_bits(x,a);
|
|
return x;
|
|
}
|
|
|
|
template<class T>
|
|
inline T (changesign_impl)(T x, ieee_copy_leading_bits_tag const&)
|
|
{
|
|
typedef BOOST_DEDUCED_TYPENAME fp_traits<T>::sign_change_type traits;
|
|
|
|
BOOST_DEDUCED_TYPENAME traits::bits a;
|
|
traits::get_bits(x,a);
|
|
a ^= traits::sign;
|
|
traits::set_bits(x,a);
|
|
return x;
|
|
}
|
|
|
|
|
|
} // namespace detail
|
|
|
|
template<class T> int (signbit)(T x)
|
|
{
|
|
typedef typename detail::fp_traits<T>::type traits;
|
|
typedef typename traits::method method;
|
|
// typedef typename boost::is_floating_point<T>::type fp_tag;
|
|
typedef typename tools::promote_args_permissive<T>::type result_type;
|
|
return detail::signbit_impl(static_cast<result_type>(x), method());
|
|
}
|
|
|
|
template <class T>
|
|
inline int sign BOOST_NO_MACRO_EXPAND(const T& z)
|
|
{
|
|
return (z == 0) ? 0 : (boost::math::signbit)(z) ? -1 : 1;
|
|
}
|
|
|
|
template <class T> typename tools::promote_args_permissive<T>::type (changesign)(const T& x)
|
|
{ //!< \brief return unchanged binary pattern of x, except for change of sign bit.
|
|
typedef typename detail::fp_traits<T>::sign_change_type traits;
|
|
typedef typename traits::method method;
|
|
// typedef typename boost::is_floating_point<T>::type fp_tag;
|
|
typedef typename tools::promote_args_permissive<T>::type result_type;
|
|
|
|
return detail::changesign_impl(static_cast<result_type>(x), method());
|
|
}
|
|
|
|
template <class T, class U>
|
|
inline typename tools::promote_args_permissive<T, U>::type
|
|
copysign BOOST_NO_MACRO_EXPAND(const T& x, const U& y)
|
|
{
|
|
BOOST_MATH_STD_USING
|
|
typedef typename tools::promote_args_permissive<T, U>::type result_type;
|
|
return (boost::math::signbit)(static_cast<result_type>(x)) != (boost::math::signbit)(static_cast<result_type>(y))
|
|
? (boost::math::changesign)(static_cast<result_type>(x)) : static_cast<result_type>(x);
|
|
}
|
|
|
|
} // namespace math
|
|
} // namespace boost
|
|
|
|
|
|
#endif // BOOST_MATH_TOOLS_SIGN_HPP
|
|
|
|
|