refactor: 💥 unit_magnitude moved to detail namespace

This commit is contained in:
Mateusz Pusz
2024-11-21 09:17:07 +01:00
parent 16e816d4cb
commit a673c8d3f3
5 changed files with 655 additions and 632 deletions

View File

@ -0,0 +1,638 @@
// The MIT License (MIT)
//
// Copyright (c) 2018 Mateusz Pusz
//
// Permission is hereby granted, free of charge, to any person obtaining a copy
// of this software and associated documentation files (the "Software"), to deal
// in the Software without restriction, including without limitation the rights
// to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
// copies of the Software, and to permit persons to whom the Software is
// furnished to do so, subject to the following conditions:
//
// The above copyright notice and this permission notice shall be included in all
// copies or substantial portions of the Software.
//
// THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
// IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
// FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
// AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
// LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
// OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
// SOFTWARE.
#pragma once
// IWYU pragma: private, include <mp-units/framework.h>
#include <mp-units/bits/constexpr_math.h>
#include <mp-units/bits/hacks.h>
#include <mp-units/bits/math_concepts.h>
#include <mp-units/bits/module_macros.h>
#include <mp-units/bits/ratio.h>
#include <mp-units/bits/text_tools.h>
#include <mp-units/ext/prime.h>
#include <mp-units/ext/type_traits.h>
#include <mp-units/framework/customization_points.h>
#include <mp-units/framework/expression_template.h>
#include <mp-units/framework/symbol_text.h>
#include <mp-units/framework/unit_magnitude_concepts.h>
#include <mp-units/framework/unit_symbol_formatting.h>
#ifndef MP_UNITS_IN_MODULE_INTERFACE
#ifdef MP_UNITS_IMPORT_STD
import std;
#else
#include <concepts>
#include <cstdint>
#include <cstdlib>
#include <limits>
#include <numbers>
#include <optional>
#endif
#endif
namespace mp_units::detail {
template<typename T>
concept MagArg = std::integral<T> || MagConstant<T>;
/**
* @brief Any type which can be used as a basis vector in a power_v.
*
* We have two categories.
*
* The first is just an integral type (either `int` or `std::intmax_t`). This is for prime number bases.
* These can always be used directly as NTTPs.
*
* The second category is a _custom tag type_, which inherits from `mag_constant` and has a static member variable
* `value` of type `long double` that holds its value. We choose `long double` to get the greatest degree of precision;
* users who need a different type can convert from this at compile time. This category is for any irrational base
* we admit into our representation (on which, more details below).
*/
// TODO Unify with `power` if UTPs (P1985) are accepted by the Committee
template<auto V, int Num, int... Den>
requires(valid_ratio<Num, Den...> && !ratio_one<Num, Den...>)
struct power_v {
static constexpr auto base = V;
static constexpr ratio exponent{Num, Den...};
};
template<typename Element>
[[nodiscard]] consteval auto get_base(Element element)
{
if constexpr (is_specialization_of_v<Element, power_v>)
return Element::base;
else
return element;
}
template<typename Element>
[[nodiscard]] consteval auto get_base_value(Element element)
{
if constexpr (is_specialization_of_v<Element, power_v>)
return get_base_value(Element::base);
else if constexpr (MagConstant<Element>)
return element._value_;
else
return element;
}
template<typename Element>
[[nodiscard]] MP_UNITS_CONSTEVAL ratio get_exponent(Element)
{
if constexpr (is_specialization_of_v<Element, power_v>)
return Element::exponent;
else
return ratio{1};
}
template<auto V, ratio R>
[[nodiscard]] consteval auto power_v_or_T()
{
if constexpr (R.den == 1) {
if constexpr (R.num == 1)
return V;
else
return power_v<V, R.num>{};
} else {
return power_v<V, R.num, R.den>{};
}
}
template<typename M>
[[nodiscard]] consteval auto mag_inverse(M)
{
return power_v_or_T<get_base(M{}), -1 * get_exponent(M{})>();
}
// `widen_t` gives the widest arithmetic type in the same category, for intermediate computations.
template<typename T>
using widen_t = conditional<std::is_arithmetic_v<T>,
conditional<std::is_floating_point_v<T>, long double,
conditional<std::is_signed_v<T>, std::intmax_t, std::uintmax_t>>,
T>;
template<typename T>
[[nodiscard]] consteval widen_t<T> compute_base_power(auto el)
{
// This utility can only handle integer powers. To compute rational powers at compile time, we'll
// need to write a custom function.
//
// Note that since this function should only be called at compile time, the point of these
// terminations is to act as "static_assert substitutes", not to actually terminate at runtime.
const auto exp = get_exponent(el);
if (exp.num < 0) {
if constexpr (std::is_integral_v<T>) {
std::abort(); // Cannot represent reciprocal as integer
} else {
return T{1} / compute_base_power<T>(mag_inverse(el));
}
}
const auto pow_result =
checked_int_pow(static_cast<widen_t<T>>(get_base_value(el)), static_cast<std::uintmax_t>(exp.num));
if (pow_result.has_value()) {
const auto final_result =
(exp.den > 1) ? root(pow_result.value(), static_cast<std::uintmax_t>(exp.den)) : pow_result;
if (final_result.has_value()) {
return final_result.value();
} else {
std::abort(); // Root computation failed.
}
} else {
std::abort(); // Power computation failed.
}
}
[[nodiscard]] consteval bool is_rational_impl(auto element)
{
return std::is_integral_v<decltype(get_base(element))> && get_exponent(element).den == 1;
}
[[nodiscard]] consteval bool is_integral_impl(auto element)
{
return is_rational_impl(element) && get_exponent(element).num > 0;
}
[[nodiscard]] consteval bool is_positive_integral_power_impl(auto element)
{
auto exp = get_exponent(element);
return exp.den == 1 && exp.num > 0;
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
// Magnitude product implementation.
[[nodiscard]] consteval bool mag_less(auto lhs, auto rhs)
{
// clang-arm64 raises "error: implicit conversion from 'long' to 'long double' may lose precision" so we need an
// explicit cast
using ct = std::common_type_t<decltype(get_base_value(lhs)), decltype(get_base_value(rhs))>;
return static_cast<ct>(get_base_value(lhs)) < static_cast<ct>(get_base_value(rhs));
}
// The largest integer which can be extracted from any magnitude with only a single basis vector.
template<auto M>
[[nodiscard]] consteval auto integer_part(unit_magnitude<M>);
[[nodiscard]] consteval std::intmax_t integer_part(ratio r) { return r.num / r.den; }
template<auto M>
[[nodiscard]] consteval auto remove_positive_power(unit_magnitude<M> m);
template<auto M>
[[nodiscard]] consteval auto remove_mag_constants(unit_magnitude<M> m);
template<auto M>
[[nodiscard]] consteval auto only_positive_mag_constants(unit_magnitude<M> m);
template<auto M>
[[nodiscard]] consteval auto only_negative_mag_constants(unit_magnitude<M> m);
template<MagArg auto Base, int Num, int Den = 1>
requires(get_base_value(Base) > 0)
[[nodiscard]] consteval UnitMagnitude auto mag_power_lazy();
template<typename T>
struct magnitude_base {};
template<auto H, auto... T>
struct magnitude_base<unit_magnitude<H, T...>> {
template<auto H2, auto... T2>
[[nodiscard]] friend consteval UnitMagnitude auto _multiply_impl(unit_magnitude<H, T...>, unit_magnitude<H2, T2...>)
{
if constexpr (mag_less(H, H2)) {
if constexpr (sizeof...(T) == 0) {
// Shortcut for the "pure prepend" case, which makes it easier to implement some of the other cases.
return unit_magnitude<H, H2, T2...>{};
} else {
return unit_magnitude<H>{} * (unit_magnitude<T...>{} * unit_magnitude<H2, T2...>{});
}
} else if constexpr (mag_less(H2, H)) {
return unit_magnitude<H2>{} * (unit_magnitude<H, T...>{} * unit_magnitude<T2...>{});
} else {
if constexpr (is_same_v<decltype(get_base(H)), decltype(get_base(H2))>) {
constexpr auto partial_product = unit_magnitude<T...>{} * unit_magnitude<T2...>{};
if constexpr (get_exponent(H) + get_exponent(H2) == 0) {
return partial_product;
} else {
// Make a new power_v with the common base of H and H2, whose power is their powers' sum.
constexpr auto new_head = power_v_or_T<get_base(H), get_exponent(H) + get_exponent(H2)>();
if constexpr (get_exponent(new_head) == 0) {
return partial_product;
} else {
return unit_magnitude<new_head>{} * partial_product;
}
}
}
}
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
// Common Magnitude.
//
// The "common Magnitude" C, of two Magnitudes M1 and M2, is the largest Magnitude such that each of its inputs is
// expressible by only positive basis powers relative to C. That is, both (M1 / C) and (M2 / C) contain only positive
// powers in the expansion on our basis.
//
// For rational Magnitudes (or, more precisely, Magnitudes that are rational _relative to each other_), this reduces
// to the familiar convention from the std::chrono library: it is the largest Magnitude C such that each input
// Magnitude is an _integer multiple_ of C. The connection can be seen by considering the definition in the above
// paragraph, and recognizing that both the bases and the powers are all integers for rational Magnitudes.
//
// For relatively _irrational_ Magnitudes (whether from irrational bases, or fractional powers of integer bases), the
// notion of a "common type" becomes less important, because there is no way to preserve pure integer multiplication.
// When we go to retrieve our value, we'll be stuck with a floating point approximation no matter what choice we make.
// Thus, we make the _simplest_ choice which reproduces the correct convention in the rational case: namely, taking
// the minimum power for each base (where absent bases implicitly have a power of 0).
template<auto H2, auto... T2>
[[nodiscard]] friend consteval auto _common_magnitude(unit_magnitude<H, T...>, unit_magnitude<H2, T2...>)
{
if constexpr (get_base_value(H) < get_base_value(H2)) {
// When H1 has the smaller base, prepend to result from recursion.
return remove_positive_power(unit_magnitude<H>{}) *
_common_magnitude(unit_magnitude<T...>{}, unit_magnitude<H2, T2...>{});
} else if constexpr (get_base_value(H2) < get_base_value(H)) {
// When H2 has the smaller base, prepend to result from recursion.
return remove_positive_power(unit_magnitude<H2>{}) *
_common_magnitude(unit_magnitude<H, T...>{}, unit_magnitude<T2...>{});
} else {
// When the bases are equal, pick whichever has the lower power.
constexpr auto common_tail = _common_magnitude(unit_magnitude<T...>{}, unit_magnitude<T2...>{});
if constexpr (get_exponent(H) < get_exponent(H2)) {
return unit_magnitude<H>{} * common_tail;
} else {
return unit_magnitude<H2>{} * common_tail;
}
}
}
};
template<auto... Ms>
[[nodiscard]] consteval std::size_t magnitude_list_size(unit_magnitude<Ms...>)
{
return sizeof...(Ms);
}
template<typename CharT, std::output_iterator<CharT> Out>
constexpr Out print_separator(Out out, const unit_symbol_formatting& fmt)
{
if (fmt.separator == unit_symbol_separator::half_high_dot) {
if (fmt.char_set != character_set::utf8)
MP_UNITS_THROW(
std::invalid_argument("'unit_symbol_separator::half_high_dot' can be only used with 'character_set::utf8'"));
const std::string_view dot = "" /* U+22C5 DOT OPERATOR */;
out = detail::copy(dot.begin(), dot.end(), out);
} else {
*out++ = ' ';
}
return out;
}
template<typename CharT, std::output_iterator<CharT> Out, auto... Ms>
requires(sizeof...(Ms) == 0)
[[nodiscard]] constexpr auto mag_constants_text(Out out, unit_magnitude<Ms...>, const unit_symbol_formatting&, bool)
{
return out;
}
template<typename CharT, std::output_iterator<CharT> Out, auto M, auto... Rest>
[[nodiscard]] constexpr auto mag_constants_text(Out out, unit_magnitude<M, Rest...>, const unit_symbol_formatting& fmt,
bool negative_power)
{
auto to_symbol = [&]<typename T>(T v) {
out = copy_symbol<CharT>(get_base(v)._symbol_, fmt.char_set, negative_power, out);
constexpr ratio r = get_exponent(T{});
return copy_symbol_exponent<CharT, abs(r.num), r.den>(fmt.char_set, negative_power, out);
};
return (to_symbol(M), ..., (print_separator<CharT>(out, fmt), to_symbol(Rest)));
}
template<typename CharT, UnitMagnitude auto Num, UnitMagnitude auto Den, UnitMagnitude auto NumConstants,
UnitMagnitude auto DenConstants, std::intmax_t Exp10, std::output_iterator<CharT> Out>
constexpr Out magnitude_symbol_impl(Out out, const unit_symbol_formatting& fmt)
{
bool numerator = false;
constexpr auto num_value = _get_value<std::intmax_t>(Num);
if constexpr (num_value != 1) {
constexpr auto num = regular<num_value>();
out = copy_symbol<CharT>(num, fmt.char_set, false, out);
numerator = true;
}
constexpr auto num_constants_size = magnitude_list_size(NumConstants);
if constexpr (num_constants_size) {
if (numerator) out = print_separator<CharT>(out, fmt);
out = mag_constants_text<CharT>(out, NumConstants, fmt, false);
numerator = true;
}
using enum unit_symbol_solidus;
bool denominator = false;
constexpr auto den_value = _get_value<std::intmax_t>(Den);
constexpr auto den_constants_size = magnitude_list_size(DenConstants);
constexpr auto den_size = (den_value != 1) + den_constants_size;
auto start_denominator = [&]() {
if (fmt.solidus == always || (fmt.solidus == one_denominator && den_size == 1)) {
if (!numerator) *out++ = '1';
*out++ = '/';
if (den_size > 1) *out++ = '(';
} else if (numerator) {
out = print_separator<CharT>(out, fmt);
}
};
const bool negative_power = fmt.solidus == never || (fmt.solidus == one_denominator && den_size > 1);
if constexpr (den_value != 1) {
constexpr auto den = regular<den_value>();
start_denominator();
out = copy_symbol<CharT>(den, fmt.char_set, negative_power, out);
denominator = true;
}
if constexpr (den_constants_size) {
if (denominator)
out = print_separator<CharT>(out, fmt);
else
start_denominator();
out = mag_constants_text<CharT>(out, DenConstants, fmt, negative_power);
if (fmt.solidus == always && den_size > 1) *out++ = ')';
denominator = true;
}
if constexpr (Exp10 != 0) {
if (numerator || denominator) {
constexpr auto mag_multiplier = symbol_text(u8" × " /* U+00D7 MULTIPLICATION SIGN */, " x ");
out = copy_symbol<CharT>(mag_multiplier, fmt.char_set, negative_power, out);
}
constexpr auto exp = symbol_text("10") + superscript<Exp10>();
out = copy_symbol<CharT>(exp, fmt.char_set, negative_power, out);
}
return out;
}
/**
* @brief A representation for positive real numbers which optimizes taking products and rational powers.
*
* Magnitudes can be treated as values. Each type encodes exactly one value. Users can multiply, divide, raise to
* rational powers, and compare for equality.
*/
template<auto... Ms>
struct unit_magnitude : magnitude_base<unit_magnitude<Ms...>> {
template<UnitMagnitude M>
[[nodiscard]] friend consteval UnitMagnitude auto operator*(unit_magnitude lhs, M rhs)
{
if constexpr (sizeof...(Ms) == 0)
return rhs;
else if constexpr (is_same_v<M, unit_magnitude<>>)
return lhs;
else
return _multiply_impl(lhs, rhs);
}
[[nodiscard]] friend consteval auto operator/(unit_magnitude lhs, UnitMagnitude auto rhs)
{
return lhs * _pow<-1>(rhs);
}
template<UnitMagnitude Rhs>
[[nodiscard]] friend consteval bool operator==(unit_magnitude, Rhs)
{
return is_same_v<unit_magnitude, Rhs>;
}
private:
// all below functions should in fact be in a `detail` namespace but are placed here to benefit from the ADL
[[nodiscard]] friend consteval bool _is_integral(const unit_magnitude&) { return (is_integral_impl(Ms) && ...); }
[[nodiscard]] friend consteval bool _is_rational(const unit_magnitude&) { return (is_rational_impl(Ms) && ...); }
[[nodiscard]] friend consteval bool _is_positive_integral_power(const unit_magnitude&)
{
return (is_positive_integral_power_impl(Ms) && ...);
}
/**
* @brief The value of a Magnitude in a desired type T.
*/
template<typename T>
requires((is_integral_impl(Ms) && ...)) || treat_as_floating_point<T>
[[nodiscard]] friend consteval T _get_value(const unit_magnitude&)
{
// Force the expression to be evaluated in a constexpr context, to catch, e.g., overflow.
constexpr T result = checked_static_cast<T>((compute_base_power<T>(Ms) * ... * T{1}));
return result;
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
// Magnitude rational powers implementation.
template<int Num, int Den = 1>
[[nodiscard]] friend consteval auto _pow(unit_magnitude)
{
if constexpr (Num == 0) {
return unit_magnitude<>{};
} else {
return unit_magnitude<power_v_or_T<get_base(Ms), get_exponent(Ms) * ratio{Num, Den}>()...>{};
}
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
// Magnitude numerator and denominator implementation.
[[nodiscard]] friend consteval auto _numerator(unit_magnitude)
{
return (integer_part(unit_magnitude<Ms>{}) * ... * unit_magnitude<>{});
}
[[nodiscard]] friend consteval auto _denominator(unit_magnitude) { return _numerator(_pow<-1>(unit_magnitude{})); }
[[nodiscard]] friend consteval auto _remove_positive_powers(unit_magnitude)
{
return (unit_magnitude<>{} * ... * remove_positive_power(unit_magnitude<Ms>{}));
}
[[nodiscard]] friend consteval auto _common_magnitude_type_impl(unit_magnitude)
{
return (std::intmax_t{} * ... * decltype(get_base_value(Ms)){});
}
[[nodiscard]] friend consteval auto _extract_components(unit_magnitude)
{
constexpr auto ratio = (unit_magnitude<>{} * ... * remove_mag_constants(unit_magnitude<Ms>{}));
if constexpr (ratio == unit_magnitude{})
return std::tuple{ratio, unit_magnitude<>{}, unit_magnitude<>{}};
else {
constexpr auto num_constants = (unit_magnitude<>{} * ... * only_positive_mag_constants(unit_magnitude<Ms>{}));
constexpr auto den_constants = (unit_magnitude<>{} * ... * only_negative_mag_constants(unit_magnitude<Ms>{}));
return std::tuple{ratio, num_constants, den_constants};
}
}
template<typename T>
[[nodiscard]] friend consteval ratio _get_power([[maybe_unused]] T base, unit_magnitude)
{
return ((get_base_value(Ms) == base ? get_exponent(Ms) : ratio{0}) + ... + ratio{0});
}
[[nodiscard]] friend consteval std::intmax_t _extract_power_of_10(unit_magnitude m)
{
const auto power_of_2 = _get_power(2, m);
const auto power_of_5 = _get_power(5, m);
if ((power_of_2 * power_of_5).num <= 0) return 0;
return integer_part((abs(power_of_2) < abs(power_of_5)) ? power_of_2 : power_of_5);
}
template<typename CharT, std::output_iterator<CharT> Out>
friend constexpr Out _magnitude_symbol(Out out, unit_magnitude, const unit_symbol_formatting& fmt)
{
if constexpr (unit_magnitude{} == unit_magnitude<1>{}) {
return out;
} else {
constexpr auto extract_res = _extract_components(unit_magnitude{});
constexpr UnitMagnitude auto ratio = std::get<0>(extract_res);
constexpr UnitMagnitude auto num_constants = std::get<1>(extract_res);
constexpr UnitMagnitude auto den_constants = std::get<2>(extract_res);
constexpr std::intmax_t exp10 = _extract_power_of_10(ratio);
if constexpr (abs(exp10) < 3) {
// print the value as a regular number (without exponent)
constexpr UnitMagnitude auto num = _numerator(unit_magnitude{});
constexpr UnitMagnitude auto den = _denominator(unit_magnitude{});
// TODO address the below
static_assert(ratio == num / den, "Printing rational powers not yet supported");
return magnitude_symbol_impl<CharT, num, den, num_constants, den_constants, 0>(out, fmt);
} else {
// print the value as a number with exponent
// if user wanted a regular number for this magnitude then probably a better scaled unit should be used
constexpr UnitMagnitude auto base = ratio / mag_power_lazy<10, exp10>();
constexpr UnitMagnitude auto num = _numerator(base);
constexpr UnitMagnitude auto den = _denominator(base);
// TODO address the below
static_assert(base == num / den, "Printing rational powers not yet supported");
return magnitude_symbol_impl<CharT, num, den, num_constants, den_constants, exp10>(out, fmt);
}
}
}
};
[[nodiscard]] consteval auto _common_magnitude(unit_magnitude<>, UnitMagnitude auto m)
{
return _remove_positive_powers(m);
}
[[nodiscard]] consteval auto _common_magnitude(UnitMagnitude auto m, unit_magnitude<>)
{
return _remove_positive_powers(m);
}
[[nodiscard]] consteval auto _common_magnitude(unit_magnitude<> m, unit_magnitude<>) { return m; }
// The largest integer which can be extracted from any magnitude with only a single basis vector.
template<auto M>
[[nodiscard]] consteval auto integer_part(unit_magnitude<M>)
{
constexpr auto power_num = get_exponent(M).num;
constexpr auto power_den = get_exponent(M).den;
if constexpr (std::is_integral_v<decltype(get_base(M))> && (power_num >= power_den)) {
// largest integer power
return unit_magnitude<power_v_or_T<get_base(M), power_num / power_den>()>{}; // Note: integer division intended
} else {
return unit_magnitude<>{};
}
}
template<auto M>
[[nodiscard]] consteval auto remove_positive_power(unit_magnitude<M> m)
{
if constexpr (get_exponent(M).num < 0) {
return m;
} else {
return unit_magnitude<>{};
}
}
template<auto M>
[[nodiscard]] consteval auto remove_mag_constants(unit_magnitude<M> m)
{
if constexpr (MagConstant<decltype(get_base(M))>)
return unit_magnitude<>{};
else
return m;
}
template<auto M>
[[nodiscard]] consteval auto only_positive_mag_constants(unit_magnitude<M> m)
{
if constexpr (MagConstant<decltype(get_base(M))> && get_exponent(M) >= 0)
return m;
else
return unit_magnitude<>{};
}
template<auto M>
[[nodiscard]] consteval auto only_negative_mag_constants(unit_magnitude<M> m)
{
if constexpr (MagConstant<decltype(get_base(M))> && get_exponent(M) < 0)
return m;
else
return unit_magnitude<>{};
}
// Returns the most precise type to express the magnitude factor
template<UnitMagnitude auto M>
using common_magnitude_type = decltype(_common_magnitude_type_impl(M));
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
// `mag()` implementation.
// Helper to perform prime factorization at compile time.
template<std::intmax_t N>
requires gt_zero<N>
struct prime_factorization {
[[nodiscard]] static consteval std::intmax_t get_or_compute_first_factor()
{
return static_cast<std::intmax_t>(find_first_factor(N));
}
static constexpr std::intmax_t first_base = get_or_compute_first_factor();
static constexpr std::intmax_t first_power = multiplicity(first_base, N);
static constexpr std::intmax_t remainder = remove_power(first_base, first_power, N);
static constexpr auto value =
unit_magnitude<power_v_or_T<first_base, ratio{first_power}>()>{} * prime_factorization<remainder>::value;
};
// Specialization for the prime factorization of 1 (base case).
template<>
struct prime_factorization<1> {
static constexpr unit_magnitude<> value{};
};
template<std::intmax_t N>
constexpr auto prime_factorization_v = prime_factorization<N>::value;
template<MagArg auto V>
[[nodiscard]] consteval UnitMagnitude auto make_magnitude()
{
if constexpr (MagConstant<MP_UNITS_REMOVE_CONST(decltype(V))>)
return unit_magnitude<V>{};
else
return prime_factorization_v<V>;
}
} // namespace mp_units::detail

View File

@ -265,7 +265,7 @@ struct scaled_unit_impl : detail::unit_interface, detail::propagate_point_origin
* instantiate this type automatically based on the unit arithmetic equation provided by the user.
*/
template<UnitMagnitude auto M, Unit U>
requires(M != unit_magnitude<>{} && M != mag<1>)
requires(M != detail::unit_magnitude<>{} && M != mag<1>)
struct scaled_unit final : detail::scaled_unit_impl<M, U> {};
namespace detail {

View File

@ -46,7 +46,7 @@ MP_UNITS_EXPORT template<typename T>
concept Unit = detail::SymbolicConstant<T> && std::derived_from<T, detail::unit_interface>;
template<UnitMagnitude auto M, Unit U>
requires(M != unit_magnitude<>{} && M != mag<1>)
requires(M != detail::unit_magnitude<>{} && M != mag<1>)
struct scaled_unit;
MP_UNITS_EXPORT template<symbol_text Symbol, auto...>

View File

@ -23,44 +23,35 @@
#pragma once
// IWYU pragma: private, include <mp-units/framework.h>
#include <mp-units/bits/constexpr_math.h>
#include <mp-units/bits/math_concepts.h>
#include <mp-units/bits/hacks.h>
#include <mp-units/bits/module_macros.h>
#include <mp-units/bits/ratio.h>
#include <mp-units/bits/text_tools.h>
#include <mp-units/ext/prime.h>
#include <mp-units/ext/type_traits.h>
#include <mp-units/framework/customization_points.h>
#include <mp-units/framework/expression_template.h>
#include <mp-units/bits/unit_magnitude.h>
#include <mp-units/framework/symbol_text.h>
#include <mp-units/framework/unit_magnitude_concepts.h>
#include <mp-units/framework/unit_symbol_formatting.h>
#ifndef MP_UNITS_IN_MODULE_INTERFACE
#ifdef MP_UNITS_IMPORT_STD
import std;
#else
#include <concepts>
#include <cstdint>
#include <cstdlib>
#include <limits>
#include <numbers>
// TODO remove when deprecated known_first_factor is removed
#include <optional>
#endif
#endif
namespace mp_units {
MP_UNITS_EXPORT_BEGIN
#if defined MP_UNITS_COMP_CLANG || MP_UNITS_COMP_CLANG < 18
MP_UNITS_EXPORT template<symbol_text Symbol>
template<symbol_text Symbol>
struct mag_constant {
static constexpr auto _symbol_ = Symbol;
};
#else
MP_UNITS_EXPORT template<symbol_text Symbol, long double Value>
template<symbol_text Symbol, long double Value>
requires(Value > 0)
struct mag_constant {
static constexpr auto _symbol_ = Symbol;
@ -68,625 +59,15 @@ struct mag_constant {
};
#endif
namespace detail {
template<typename T>
concept MagArg = std::integral<T> || MagConstant<T>;
}
/**
* @brief Any type which can be used as a basis vector in a power_v.
*
* We have two categories.
*
* The first is just an integral type (either `int` or `std::intmax_t`). This is for prime number bases.
* These can always be used directly as NTTPs.
*
* The second category is a _custom tag type_, which inherits from `mag_constant` and has a static member variable
* `value` of type `long double` that holds its value. We choose `long double` to get the greatest degree of precision;
* users who need a different type can convert from this at compile time. This category is for any irrational base
* we admit into our representation (on which, more details below).
*/
// TODO Unify with `power` if UTPs (P1985) are accepted by the Committee
template<auto V, int Num, int... Den>
requires(detail::valid_ratio<Num, Den...> && !detail::ratio_one<Num, Den...>)
struct power_v {
static constexpr auto base = V;
static constexpr detail::ratio exponent{Num, Den...};
};
namespace detail {
template<typename Element>
[[nodiscard]] consteval auto get_base(Element element)
{
if constexpr (is_specialization_of_v<Element, power_v>)
return Element::base;
else
return element;
}
template<typename Element>
[[nodiscard]] consteval auto get_base_value(Element element)
{
if constexpr (is_specialization_of_v<Element, power_v>)
return get_base_value(Element::base);
else if constexpr (MagConstant<Element>)
return element._value_;
else
return element;
}
template<typename Element>
[[nodiscard]] MP_UNITS_CONSTEVAL ratio get_exponent(Element)
{
if constexpr (is_specialization_of_v<Element, power_v>)
return Element::exponent;
else
return ratio{1};
}
template<auto V, ratio R>
[[nodiscard]] consteval auto power_v_or_T()
{
if constexpr (R.den == 1) {
if constexpr (R.num == 1)
return V;
else
return power_v<V, R.num>{};
} else {
return power_v<V, R.num, R.den>{};
}
}
template<typename M>
[[nodiscard]] consteval auto mag_inverse(M)
{
return power_v_or_T<get_base(M{}), -1 * get_exponent(M{})>();
}
// `widen_t` gives the widest arithmetic type in the same category, for intermediate computations.
template<typename T>
using widen_t = conditional<std::is_arithmetic_v<T>,
conditional<std::is_floating_point_v<T>, long double,
conditional<std::is_signed_v<T>, std::intmax_t, std::uintmax_t>>,
T>;
template<typename T>
[[nodiscard]] consteval widen_t<T> compute_base_power(auto el)
{
// This utility can only handle integer powers. To compute rational powers at compile time, we'll
// need to write a custom function.
//
// Note that since this function should only be called at compile time, the point of these
// terminations is to act as "static_assert substitutes", not to actually terminate at runtime.
const auto exp = get_exponent(el);
if (exp.num < 0) {
if constexpr (std::is_integral_v<T>) {
std::abort(); // Cannot represent reciprocal as integer
} else {
return T{1} / compute_base_power<T>(mag_inverse(el));
}
}
const auto pow_result =
checked_int_pow(static_cast<widen_t<T>>(get_base_value(el)), static_cast<std::uintmax_t>(exp.num));
if (pow_result.has_value()) {
const auto final_result =
(exp.den > 1) ? root(pow_result.value(), static_cast<std::uintmax_t>(exp.den)) : pow_result;
if (final_result.has_value()) {
return final_result.value();
} else {
std::abort(); // Root computation failed.
}
} else {
std::abort(); // Power computation failed.
}
}
[[nodiscard]] consteval bool is_rational_impl(auto element)
{
return std::is_integral_v<decltype(get_base(element))> && get_exponent(element).den == 1;
}
[[nodiscard]] consteval bool is_integral_impl(auto element)
{
return is_rational_impl(element) && get_exponent(element).num > 0;
}
[[nodiscard]] consteval bool is_positive_integral_power_impl(auto element)
{
auto exp = get_exponent(element);
return exp.den == 1 && exp.num > 0;
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
// Magnitude product implementation.
[[nodiscard]] consteval bool mag_less(auto lhs, auto rhs)
{
// clang-arm64 raises "error: implicit conversion from 'long' to 'long double' may lose precision" so we need an
// explicit cast
using ct = std::common_type_t<decltype(get_base_value(lhs)), decltype(get_base_value(rhs))>;
return static_cast<ct>(get_base_value(lhs)) < static_cast<ct>(get_base_value(rhs));
}
// The largest integer which can be extracted from any magnitude with only a single basis vector.
template<auto M>
[[nodiscard]] consteval auto integer_part(unit_magnitude<M>);
[[nodiscard]] consteval std::intmax_t integer_part(ratio r) { return r.num / r.den; }
template<auto M>
[[nodiscard]] consteval auto remove_positive_power(unit_magnitude<M> m);
template<auto M>
[[nodiscard]] consteval auto remove_mag_constants(unit_magnitude<M> m);
template<auto M>
[[nodiscard]] consteval auto only_positive_mag_constants(unit_magnitude<M> m);
template<auto M>
[[nodiscard]] consteval auto only_negative_mag_constants(unit_magnitude<M> m);
template<MagArg auto Base, int Num, int Den = 1>
requires(detail::get_base_value(Base) > 0)
[[nodiscard]] consteval UnitMagnitude auto mag_power_lazy();
template<typename T>
struct magnitude_base {};
template<auto H, auto... T>
struct magnitude_base<unit_magnitude<H, T...>> {
template<auto H2, auto... T2>
[[nodiscard]] friend consteval UnitMagnitude auto _multiply_impl(unit_magnitude<H, T...>, unit_magnitude<H2, T2...>)
{
if constexpr (mag_less(H, H2)) {
if constexpr (sizeof...(T) == 0) {
// Shortcut for the "pure prepend" case, which makes it easier to implement some of the other cases.
return unit_magnitude<H, H2, T2...>{};
} else {
return unit_magnitude<H>{} * (unit_magnitude<T...>{} * unit_magnitude<H2, T2...>{});
}
} else if constexpr (mag_less(H2, H)) {
return unit_magnitude<H2>{} * (unit_magnitude<H, T...>{} * unit_magnitude<T2...>{});
} else {
if constexpr (is_same_v<decltype(get_base(H)), decltype(get_base(H2))>) {
constexpr auto partial_product = unit_magnitude<T...>{} * unit_magnitude<T2...>{};
if constexpr (get_exponent(H) + get_exponent(H2) == 0) {
return partial_product;
} else {
// Make a new power_v with the common base of H and H2, whose power is their powers' sum.
constexpr auto new_head = power_v_or_T<get_base(H), get_exponent(H) + get_exponent(H2)>();
if constexpr (get_exponent(new_head) == 0) {
return partial_product;
} else {
return unit_magnitude<new_head>{} * partial_product;
}
}
}
}
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
// Common Magnitude.
//
// The "common Magnitude" C, of two Magnitudes M1 and M2, is the largest Magnitude such that each of its inputs is
// expressible by only positive basis powers relative to C. That is, both (M1 / C) and (M2 / C) contain only positive
// powers in the expansion on our basis.
//
// For rational Magnitudes (or, more precisely, Magnitudes that are rational _relative to each other_), this reduces
// to the familiar convention from the std::chrono library: it is the largest Magnitude C such that each input
// Magnitude is an _integer multiple_ of C. The connection can be seen by considering the definition in the above
// paragraph, and recognizing that both the bases and the powers are all integers for rational Magnitudes.
//
// For relatively _irrational_ Magnitudes (whether from irrational bases, or fractional powers of integer bases), the
// notion of a "common type" becomes less important, because there is no way to preserve pure integer multiplication.
// When we go to retrieve our value, we'll be stuck with a floating point approximation no matter what choice we make.
// Thus, we make the _simplest_ choice which reproduces the correct convention in the rational case: namely, taking
// the minimum power for each base (where absent bases implicitly have a power of 0).
template<auto H2, auto... T2>
[[nodiscard]] friend consteval auto _common_magnitude(unit_magnitude<H, T...>, unit_magnitude<H2, T2...>)
{
using detail::remove_positive_power;
if constexpr (detail::get_base_value(H) < detail::get_base_value(H2)) {
// When H1 has the smaller base, prepend to result from recursion.
return remove_positive_power(unit_magnitude<H>{}) *
_common_magnitude(unit_magnitude<T...>{}, unit_magnitude<H2, T2...>{});
} else if constexpr (detail::get_base_value(H2) < detail::get_base_value(H)) {
// When H2 has the smaller base, prepend to result from recursion.
return remove_positive_power(unit_magnitude<H2>{}) *
_common_magnitude(unit_magnitude<H, T...>{}, unit_magnitude<T2...>{});
} else {
// When the bases are equal, pick whichever has the lower power.
constexpr auto common_tail = _common_magnitude(unit_magnitude<T...>{}, unit_magnitude<T2...>{});
if constexpr (detail::get_exponent(H) < detail::get_exponent(H2)) {
return unit_magnitude<H>{} * common_tail;
} else {
return unit_magnitude<H2>{} * common_tail;
}
}
}
};
template<auto... Ms>
[[nodiscard]] consteval std::size_t magnitude_list_size(unit_magnitude<Ms...>)
{
return sizeof...(Ms);
}
template<typename CharT, std::output_iterator<CharT> Out>
constexpr Out print_separator(Out out, const unit_symbol_formatting& fmt)
{
if (fmt.separator == unit_symbol_separator::half_high_dot) {
if (fmt.char_set != character_set::utf8)
MP_UNITS_THROW(
std::invalid_argument("'unit_symbol_separator::half_high_dot' can be only used with 'character_set::utf8'"));
const std::string_view dot = "" /* U+22C5 DOT OPERATOR */;
out = detail::copy(dot.begin(), dot.end(), out);
} else {
*out++ = ' ';
}
return out;
}
template<typename CharT, std::output_iterator<CharT> Out, auto... Ms>
requires(sizeof...(Ms) == 0)
[[nodiscard]] constexpr auto mag_constants_text(Out out, unit_magnitude<Ms...>, const unit_symbol_formatting&, bool)
{
return out;
}
template<typename CharT, std::output_iterator<CharT> Out, auto M, auto... Rest>
[[nodiscard]] constexpr auto mag_constants_text(Out out, unit_magnitude<M, Rest...>, const unit_symbol_formatting& fmt,
bool negative_power)
{
auto to_symbol = [&]<typename T>(T v) {
out = copy_symbol<CharT>(get_base(v)._symbol_, fmt.char_set, negative_power, out);
constexpr ratio r = get_exponent(T{});
return copy_symbol_exponent<CharT, abs(r.num), r.den>(fmt.char_set, negative_power, out);
};
return (to_symbol(M), ..., (print_separator<CharT>(out, fmt), to_symbol(Rest)));
}
template<typename CharT, UnitMagnitude auto Num, UnitMagnitude auto Den, UnitMagnitude auto NumConstants,
UnitMagnitude auto DenConstants, std::intmax_t Exp10, std::output_iterator<CharT> Out>
constexpr Out magnitude_symbol_impl(Out out, const unit_symbol_formatting& fmt)
{
bool numerator = false;
constexpr auto num_value = _get_value<std::intmax_t>(Num);
if constexpr (num_value != 1) {
constexpr auto num = detail::regular<num_value>();
out = copy_symbol<CharT>(num, fmt.char_set, false, out);
numerator = true;
}
constexpr auto num_constants_size = magnitude_list_size(NumConstants);
if constexpr (num_constants_size) {
if (numerator) out = print_separator<CharT>(out, fmt);
out = mag_constants_text<CharT>(out, NumConstants, fmt, false);
numerator = true;
}
using enum unit_symbol_solidus;
bool denominator = false;
constexpr auto den_value = _get_value<std::intmax_t>(Den);
constexpr auto den_constants_size = magnitude_list_size(DenConstants);
constexpr auto den_size = (den_value != 1) + den_constants_size;
auto start_denominator = [&]() {
if (fmt.solidus == always || (fmt.solidus == one_denominator && den_size == 1)) {
if (!numerator) *out++ = '1';
*out++ = '/';
if (den_size > 1) *out++ = '(';
} else if (numerator) {
out = print_separator<CharT>(out, fmt);
}
};
const bool negative_power = fmt.solidus == never || (fmt.solidus == one_denominator && den_size > 1);
if constexpr (den_value != 1) {
constexpr auto den = detail::regular<den_value>();
start_denominator();
out = copy_symbol<CharT>(den, fmt.char_set, negative_power, out);
denominator = true;
}
if constexpr (den_constants_size) {
if (denominator)
out = print_separator<CharT>(out, fmt);
else
start_denominator();
out = mag_constants_text<CharT>(out, DenConstants, fmt, negative_power);
if (fmt.solidus == always && den_size > 1) *out++ = ')';
denominator = true;
}
if constexpr (Exp10 != 0) {
if (numerator || denominator) {
constexpr auto mag_multiplier = symbol_text(u8" × " /* U+00D7 MULTIPLICATION SIGN */, " x ");
out = copy_symbol<CharT>(mag_multiplier, fmt.char_set, negative_power, out);
}
constexpr auto exp = symbol_text("10") + detail::superscript<Exp10>();
out = copy_symbol<CharT>(exp, fmt.char_set, negative_power, out);
}
return out;
}
} // namespace detail
/**
* @brief A representation for positive real numbers which optimizes taking products and rational powers.
*
* Magnitudes can be treated as values. Each type encodes exactly one value. Users can multiply, divide, raise to
* rational powers, and compare for equality.
*/
template<auto... Ms>
struct unit_magnitude : detail::magnitude_base<unit_magnitude<Ms...>> {
template<UnitMagnitude M>
[[nodiscard]] friend consteval UnitMagnitude auto operator*(unit_magnitude lhs, M rhs)
{
if constexpr (sizeof...(Ms) == 0)
return rhs;
else if constexpr (is_same_v<M, unit_magnitude<>>)
return lhs;
else
return _multiply_impl(lhs, rhs);
}
[[nodiscard]] friend consteval auto operator/(unit_magnitude lhs, UnitMagnitude auto rhs)
{
return lhs * _pow<-1>(rhs);
}
template<UnitMagnitude Rhs>
[[nodiscard]] friend consteval bool operator==(unit_magnitude, Rhs)
{
return is_same_v<unit_magnitude, Rhs>;
}
private:
// all below functions should in fact be in a `detail` namespace but are placed here to benefit from the ADL
[[nodiscard]] friend consteval bool _is_integral(const unit_magnitude&)
{
return (detail::is_integral_impl(Ms) && ...);
}
[[nodiscard]] friend consteval bool _is_rational(const unit_magnitude&)
{
return (detail::is_rational_impl(Ms) && ...);
}
[[nodiscard]] friend consteval bool _is_positive_integral_power(const unit_magnitude&)
{
return (detail::is_positive_integral_power_impl(Ms) && ...);
}
/**
* @brief The value of a Magnitude in a desired type T.
*/
template<typename T>
requires((detail::is_integral_impl(Ms) && ...)) || treat_as_floating_point<T>
[[nodiscard]] friend consteval T _get_value(const unit_magnitude&)
{
// Force the expression to be evaluated in a constexpr context, to catch, e.g., overflow.
constexpr T result = detail::checked_static_cast<T>((detail::compute_base_power<T>(Ms) * ... * T{1}));
return result;
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
// Magnitude rational powers implementation.
template<int Num, int Den = 1>
[[nodiscard]] friend consteval auto _pow(unit_magnitude)
{
if constexpr (Num == 0) {
return unit_magnitude<>{};
} else {
return unit_magnitude<
detail::power_v_or_T<detail::get_base(Ms), detail::get_exponent(Ms) * detail::ratio{Num, Den}>()...>{};
}
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
// Magnitude numerator and denominator implementation.
[[nodiscard]] friend consteval auto _numerator(unit_magnitude)
{
return (detail::integer_part(unit_magnitude<Ms>{}) * ... * unit_magnitude<>{});
}
[[nodiscard]] friend consteval auto _denominator(unit_magnitude) { return _numerator(_pow<-1>(unit_magnitude{})); }
[[nodiscard]] friend consteval auto _remove_positive_powers(unit_magnitude)
{
return (unit_magnitude<>{} * ... * detail::remove_positive_power(unit_magnitude<Ms>{}));
}
[[nodiscard]] friend consteval auto _common_magnitude_type_impl(unit_magnitude)
{
return (std::intmax_t{} * ... * decltype(detail::get_base_value(Ms)){});
}
[[nodiscard]] friend consteval auto _extract_components(unit_magnitude)
{
constexpr auto ratio = (unit_magnitude<>{} * ... * detail::remove_mag_constants(unit_magnitude<Ms>{}));
if constexpr (ratio == unit_magnitude{})
return std::tuple{ratio, unit_magnitude<>{}, unit_magnitude<>{}};
else {
constexpr auto num_constants =
(unit_magnitude<>{} * ... * detail::only_positive_mag_constants(unit_magnitude<Ms>{}));
constexpr auto den_constants =
(unit_magnitude<>{} * ... * detail::only_negative_mag_constants(unit_magnitude<Ms>{}));
return std::tuple{ratio, num_constants, den_constants};
}
}
template<typename T>
[[nodiscard]] friend consteval detail::ratio _get_power([[maybe_unused]] T base, unit_magnitude)
{
return ((detail::get_base_value(Ms) == base ? detail::get_exponent(Ms) : detail::ratio{0}) + ... +
detail::ratio{0});
}
[[nodiscard]] friend consteval std::intmax_t _extract_power_of_10(unit_magnitude m)
{
const auto power_of_2 = _get_power(2, m);
const auto power_of_5 = _get_power(5, m);
if ((power_of_2 * power_of_5).num <= 0) return 0;
return detail::integer_part((detail::abs(power_of_2) < detail::abs(power_of_5)) ? power_of_2 : power_of_5);
}
template<typename CharT, std::output_iterator<CharT> Out>
friend constexpr Out _magnitude_symbol(Out out, unit_magnitude, const unit_symbol_formatting& fmt)
{
if constexpr (unit_magnitude{} == unit_magnitude<1>{}) {
return out;
} else {
constexpr auto extract_res = _extract_components(unit_magnitude{});
constexpr UnitMagnitude auto ratio = std::get<0>(extract_res);
constexpr UnitMagnitude auto num_constants = std::get<1>(extract_res);
constexpr UnitMagnitude auto den_constants = std::get<2>(extract_res);
constexpr std::intmax_t exp10 = _extract_power_of_10(ratio);
if constexpr (detail::abs(exp10) < 3) {
// print the value as a regular number (without exponent)
constexpr UnitMagnitude auto num = _numerator(unit_magnitude{});
constexpr UnitMagnitude auto den = _denominator(unit_magnitude{});
// TODO address the below
static_assert(ratio == num / den, "Printing rational powers not yet supported");
return detail::magnitude_symbol_impl<CharT, num, den, num_constants, den_constants, 0>(out, fmt);
} else {
// print the value as a number with exponent
// if user wanted a regular number for this magnitude then probably a better scaled unit should be used
constexpr UnitMagnitude auto base = ratio / detail::mag_power_lazy<10, exp10>();
constexpr UnitMagnitude auto num = _numerator(base);
constexpr UnitMagnitude auto den = _denominator(base);
// TODO address the below
static_assert(base == num / den, "Printing rational powers not yet supported");
return detail::magnitude_symbol_impl<CharT, num, den, num_constants, den_constants, exp10>(out, fmt);
}
}
}
};
[[nodiscard]] consteval auto _common_magnitude(unit_magnitude<>, UnitMagnitude auto m)
{
return _remove_positive_powers(m);
}
[[nodiscard]] consteval auto _common_magnitude(UnitMagnitude auto m, unit_magnitude<>)
{
return _remove_positive_powers(m);
}
[[nodiscard]] consteval auto _common_magnitude(unit_magnitude<> m, unit_magnitude<>) { return m; }
namespace detail {
// The largest integer which can be extracted from any magnitude with only a single basis vector.
template<auto M>
[[nodiscard]] consteval auto integer_part(unit_magnitude<M>)
{
constexpr auto power_num = get_exponent(M).num;
constexpr auto power_den = get_exponent(M).den;
if constexpr (std::is_integral_v<decltype(get_base(M))> && (power_num >= power_den)) {
// largest integer power
return unit_magnitude<power_v_or_T<get_base(M), power_num / power_den>()>{}; // Note: integer division intended
} else {
return unit_magnitude<>{};
}
}
template<auto M>
[[nodiscard]] consteval auto remove_positive_power(unit_magnitude<M> m)
{
if constexpr (get_exponent(M).num < 0) {
return m;
} else {
return unit_magnitude<>{};
}
}
template<auto M>
[[nodiscard]] consteval auto remove_mag_constants(unit_magnitude<M> m)
{
if constexpr (MagConstant<decltype(get_base(M))>)
return unit_magnitude<>{};
else
return m;
}
template<auto M>
[[nodiscard]] consteval auto only_positive_mag_constants(unit_magnitude<M> m)
{
if constexpr (MagConstant<decltype(get_base(M))> && get_exponent(M) >= 0)
return m;
else
return unit_magnitude<>{};
}
template<auto M>
[[nodiscard]] consteval auto only_negative_mag_constants(unit_magnitude<M> m)
{
if constexpr (MagConstant<decltype(get_base(M))> && get_exponent(M) < 0)
return m;
else
return unit_magnitude<>{};
}
// Returns the most precise type to express the magnitude factor
template<UnitMagnitude auto M>
using common_magnitude_type = decltype(_common_magnitude_type_impl(M));
} // namespace detail
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
// `mag()` implementation.
MP_UNITS_EXPORT template<std::intmax_t N>
template<std::intmax_t N>
[[deprecated("`known_first_factor` is no longer necessary and can simply be removed")]]
constexpr std::optional<std::intmax_t>
known_first_factor = std::nullopt;
namespace detail {
// Helper to perform prime factorization at compile time.
template<std::intmax_t N>
requires gt_zero<N>
struct prime_factorization {
[[nodiscard]] static consteval std::intmax_t get_or_compute_first_factor()
{
return static_cast<std::intmax_t>(find_first_factor(N));
}
static constexpr std::intmax_t first_base = get_or_compute_first_factor();
static constexpr std::intmax_t first_power = multiplicity(first_base, N);
static constexpr std::intmax_t remainder = remove_power(first_base, first_power, N);
static constexpr auto value =
unit_magnitude<power_v_or_T<first_base, ratio{first_power}>()>{} * prime_factorization<remainder>::value;
};
// Specialization for the prime factorization of 1 (base case).
template<>
struct prime_factorization<1> {
static constexpr unit_magnitude<> value{};
};
template<std::intmax_t N>
constexpr auto prime_factorization_v = prime_factorization<N>::value;
template<MagArg auto V>
[[nodiscard]] consteval UnitMagnitude auto make_magnitude()
{
if constexpr (MagConstant<MP_UNITS_REMOVE_CONST(decltype(V))>)
return unit_magnitude<V>{};
else
return prime_factorization_v<V>;
}
} // namespace detail
MP_UNITS_EXPORT_BEGIN
template<detail::MagArg auto V>
requires(detail::get_base_value(V) > 0)
constexpr UnitMagnitude auto mag = detail::make_magnitude<V>();
@ -710,7 +91,7 @@ inline constexpr struct pi final : mag_constant<symbol_text{u8"π" /* U+03C0 GRE
static constexpr auto _value_ = std::numbers::pi_v<long double>;
#else
inline constexpr struct pi final :
mag_constant<symbol_text{u8"π" /* U+03C0 GREEK SMALL LETTER PI */, "pi"}, std::numbers::pi_v<long double>> {
mag_constant<symbol_text{u8"π" /* U+03C0 GREEK SMALL LETTER PI */, "pi"}, std::numbers::pi_v<long double> > {
#endif
} pi;
inline constexpr auto π /* U+03C0 GREEK SMALL LETTER PI */ = pi;
@ -723,7 +104,7 @@ namespace detail {
// This is introduced to break the dependency cycle between `unit_magnitude::_magnitude_text` and `prime_factorization`
template<MagArg auto Base, int Num, int Den>
requires(detail::get_base_value(Base) > 0)
requires(get_base_value(Base) > 0)
[[nodiscard]] consteval UnitMagnitude auto mag_power_lazy()
{
return _pow<Num, Den>(mag<Base>);

View File

@ -49,13 +49,17 @@ struct mag_constant;
MP_UNITS_EXPORT template<typename T>
concept MagConstant = detail::SymbolicConstant<T> && is_derived_from_specialization_of_v<T, mag_constant>;
namespace detail {
template<auto... Ms>
struct unit_magnitude;
}
/**
* @brief Concept to detect whether T is a valid UnitMagnitude.
*/
MP_UNITS_EXPORT template<typename T>
concept UnitMagnitude = is_specialization_of_v<T, unit_magnitude>;
concept UnitMagnitude = is_specialization_of_v<T, detail::unit_magnitude>;
} // namespace mp_units