/// \file
// Range v3 library
//
// Copyright Casey Carter 2016
//
// Use, modification and distribution is 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)
//
// Project home: https://github.com/ericniebler/range-v3
//
/*
* Random-Number Utilities (randutil)
* Addresses common issues with C++11 random number generation.
* Makes good seeding easier, and makes using RNGs easy while retaining
* all the power.
*
* The MIT License (MIT)
*
* Copyright (c) 2015 Melissa E. O'Neill
*
* 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.
*/
#ifndef RANGES_V3_UTILITY_RANDOM_HPP
#define RANGES_V3_UTILITY_RANDOM_HPP
#include <array>
#include <cstddef>
#include <cstdint>
#include <initializer_list>
#include <new>
#include <random>
#include <meta/meta.hpp>
#include <concepts/concepts.hpp>
#include <range/v3/range_fwd.hpp>
#include <range/v3/algorithm/copy.hpp>
#include <range/v3/algorithm/generate.hpp>
#include <range/v3/functional/invoke.hpp>
#include <range/v3/functional/reference_wrapper.hpp>
#include <range/v3/iterator/concepts.hpp>
#if !RANGES_CXX_THREAD_LOCAL
#include <mutex>
#endif
RANGES_DIAGNOSTIC_PUSH
RANGES_DIAGNOSTIC_IGNORE_CXX17_COMPAT
namespace ranges
{
/// \addtogroup group-numerics
/// @{
// clang-format off
CPP_def
(
template(typename Gen)
concept uniform_random_bit_generator,
requires(int)
(
Gen::min(),
Gen::max(),
concepts::requires_<same_as<invoke_result_t<Gen&>, decltype(Gen::min())>>,
concepts::requires_<same_as<invoke_result_t<Gen&>, decltype(Gen::max())>>
) &&
invocable<Gen &> &&
unsigned_integral<invoke_result_t<Gen&>>
);
// clang-format on
/// @}
/// \cond
namespace detail
{
namespace randutils
{
inline std::array<std::uint32_t, 8> get_entropy()
{
std::array<std::uint32_t, 8> seeds;
// Hopefully high-quality entropy from random_device.
#if defined(__GLIBCXX__) && defined(RANGES_WORKAROUND_VALGRIND_RDRAND)
std::random_device rd{"/dev/urandom"};
#else
std::random_device rd;
#endif
std::uniform_int_distribution<std::uint32_t> dist{};
ranges::generate(seeds, [&] { return dist(rd); });
return seeds;
}
template<typename I>
constexpr auto fast_exp(I x, I power, I result = I{1}) -> CPP_ret(I)( //
requires unsigned_integral<I>)
{
return power == I{0}
? result
: randutils::fast_exp(
x * x, power >> 1, result * (power & I{1} ? x : 1));
}
//////////////////////////////////////////////////////////////////////////////
//
// seed_seq_fe
//
//////////////////////////////////////////////////////////////////////////////
/*
* seed_seq_fe implements a fixed-entropy seed sequence; it conforms to all
* the requirements of a Seed Sequence concept.
*
* seed_seq_fe<N> implements a seed sequence which seeds based on a store of
* N * 32 bits of entropy. Typically, it would be initialized with N or more
* integers.
*
* seed_seq_fe128 and seed_seq_fe256 are provided as convenience typedefs for
* 128- and 256-bit entropy stores respectively. These variants outperform
* std::seed_seq, while being better mixing the bits it is provided as
* entropy. In almost all common use cases, they serve as better drop-in
* replacements for seed_seq.
*
* Technical details
*
* Assuming it constructed with M seed integers as input, it exhibits the
* following properties
*
* * Diffusion/Avalanche: A single-bit change in any of the M inputs has a
* 50% chance of flipping every bit in the bitstream produced by generate.
* Initializing the N-word entropy store with M words requires O(N * M)
* time precisely because of the avalanche requirements. Once constructed,
* calls to generate are linear in the number of words generated.
*
* * Bias freedom/Bijection: If M == N, the state of the entropy store is a
* bijection from the M inputs (i.e., no states occur twice, none are
* omitted). If M > N the number of times each state can occur is the same
* (each state occurs 2**(32*(M-N)) times, where ** is the power function).
* If M < N, some states cannot occur (bias) but no state occurs more
* than once (it's impossible to avoid bias if M < N; ideally N should not
* be chosen so that it is more than M).
*
* Likewise, the generate function has similar properties (with the entropy
* store as the input data). If more outputs are requested than there is
* entropy, some outputs cannot occur. For example, the Mersenne Twister
* will request 624 outputs, to initialize its 19937-bit state, which is
* much larger than a 128-bit or 256-bit entropy pool. But in practice,
* limiting the Mersenne Twister to 2**128 possible initializations gives
* us enough initializations to give a unique initialization to trillions
* of computers for billions of years. If you really have 624 words of
* *real* high-quality entropy you want to use, you probably don't need
* an entropy mixer like this class at all. But if you *really* want to,
* nothing is stopping you from creating a randutils::seed_seq_fe<624>.
*
* * As a consequence of the above properties, if all parts of the provided
* seed data are kept constant except one, and the remaining part is varied
* through K different states, K different output sequences will be
* produced.
*
* * Also, because the amount of entropy stored is fixed, this class never
* performs dynamic allocation and is free of the possibility of generating
* an exception.
*
* Ideas used to implement this code include hashing, a simple PCG generator
* based on an MCG base with an XorShift output function and permutation
* functions on tuples.
*
* More detail at
* http://www.pcg-random.org/posts/developing-a-seed_seq-alternative.html
*/
template<std::size_t count, typename IntRep = std::uint32_t>
struct seed_seq_fe
{
public:
CPP_assert(unsigned_integral<IntRep>);
typedef IntRep result_type;
private:
static constexpr std::size_t mix_rounds = 1 + (count <= 2);
static constexpr std::uint32_t INIT_A = 0x43b0d7e5;
static constexpr std::uint32_t MULT_A = 0x931e8875;
static constexpr std::uint32_t INIT_B = 0x8b51f9dd;
static constexpr std::uint32_t MULT_B = 0x58f38ded;
static constexpr std::uint32_t MIX_MULT_L = 0xca01f9dd;
static constexpr std::uint32_t MIX_MULT_R = 0x4973f715;
static constexpr std::uint32_t XSHIFT = sizeof(IntRep) * 8 / 2;
std::array<IntRep, count> mixer_;
template<typename I, typename S>
auto mix_entropy(I first, S last) -> CPP_ret(void)( //
requires input_iterator<I> && sentinel_for<S, I> &&
convertible_to<iter_reference_t<I>, IntRep>)
{
auto hash_const = INIT_A;
auto hash = [&](IntRep value) RANGES_INTENDED_MODULAR_ARITHMETIC {
value ^= hash_const;
hash_const *= MULT_A;
value *= hash_const;
value ^= value >> XSHIFT;
return value;
};
auto mix = [](IntRep x, IntRep y) RANGES_INTENDED_MODULAR_ARITHMETIC {
IntRep result = MIX_MULT_L * x - MIX_MULT_R * y;
result ^= result >> XSHIFT;
return result;
};
for(auto & elem : mixer_)
{
if(first != last)
{
elem = hash(static_cast<IntRep>(*first));
++first;
}
else
elem = hash(IntRep{0});
}
for(auto & src : mixer_)
for(auto & dest : mixer_)
if(&src != &dest)
dest = mix(dest, hash(src));
for(; first != last; ++first)
for(auto & dest : mixer_)
dest = mix(dest, hash(static_cast<IntRep>(*first)));
}
public:
seed_seq_fe(const seed_seq_fe &) = delete;
void operator=(const seed_seq_fe &) = delete;
template<typename T>
CPP_ctor(seed_seq_fe)(std::initializer_list<T> init)( //
requires convertible_to<T const &, IntRep>)
{
seed(init.begin(), init.end());
}
template<typename I, typename S>
CPP_ctor(seed_seq_fe)(I first, S last)( //
requires input_iterator<I> && sentinel_for<S, I> &&
convertible_to<iter_reference_t<I>, IntRep>)
{
seed(first, last);
}
// generating functions
template<typename I, typename S>
RANGES_INTENDED_MODULAR_ARITHMETIC auto generate(I first,
S const last) const
-> CPP_ret(void)( //
requires random_access_iterator<I> && sentinel_for<S, I>)
{
auto src_begin = mixer_.begin();
auto src_end = mixer_.end();
auto src = src_begin;
auto hash_const = INIT_B;
for(; first != last; ++first)
{
auto dataval = *src;
if(++src == src_end)
src = src_begin;
dataval ^= hash_const;
hash_const *= MULT_B;
dataval *= hash_const;
dataval ^= dataval >> XSHIFT;
*first = dataval;
}
}
constexpr std::size_t size() const
{
return count;
}
template<typename O>
RANGES_INTENDED_MODULAR_ARITHMETIC auto param(O dest) const
-> CPP_ret(void)( //
requires weakly_incrementable<O> &&
indirectly_copyable<decltype(mixer_.begin()), O>)
{
constexpr IntRep INV_A = randutils::fast_exp(MULT_A, IntRep(-1));
constexpr IntRep MIX_INV_L =
randutils::fast_exp(MIX_MULT_L, IntRep(-1));
auto mixer_copy = mixer_;
for(std::size_t round = 0; round < mix_rounds; ++round)
{
// Advance to the final value. We'll backtrack from that.
auto hash_const =
INIT_A * randutils::fast_exp(MULT_A, IntRep(count * count));
for(auto src = mixer_copy.rbegin(); src != mixer_copy.rend();
++src)
for(auto rdest = mixer_copy.rbegin();
rdest != mixer_copy.rend();
++rdest)
if(src != rdest)
{
IntRep revhashed = *src;
auto mult_const = hash_const;
hash_const *= INV_A;
revhashed ^= hash_const;
revhashed *= mult_const;
revhashed ^= revhashed >> XSHIFT;
IntRep unmixed = *rdest;
unmixed ^= unmixed >> XSHIFT;
unmixed += MIX_MULT_R * revhashed;
unmixed *= MIX_INV_L;
*rdest = unmixed;
}
for(auto i = mixer_copy.rbegin(); i != mixer_copy.rend(); ++i)
{
IntRep unhashed = *i;
unhashed ^= unhashed >> XSHIFT;
unhashed *= randutils::fast_exp(hash_const, IntRep(-1));
hash_const *= INV_A;
unhashed ^= hash_const;
*i = unhashed;
}
}
ranges::copy(mixer_copy, dest);
}
template<typename I, typename S>
auto seed(I first, S last) -> CPP_ret(void)( //
requires input_iterator<I> && sentinel_for<S, I> &&
convertible_to<iter_reference_t<I>, IntRep>)
{
mix_entropy(first, last);
// For very small sizes, we do some additional mixing. For normal
// sizes, this loop never performs any iterations.
for(std::size_t i = 1; i < mix_rounds; ++i)
stir();
}
seed_seq_fe & stir()
{
mix_entropy(mixer_.begin(), mixer_.end());
return *this;
}
};
using seed_seq_fe128 = seed_seq_fe<4, std::uint32_t>;
using seed_seq_fe256 = seed_seq_fe<8, std::uint32_t>;
//////////////////////////////////////////////////////////////////////////////
//
// auto_seeded
//
//////////////////////////////////////////////////////////////////////////////
/*
* randutils::auto_seeded
*
* Extends a seed sequence class with a nondeterministic default
* constructor. Uses a variety of local sources of entropy to portably
* initialize any seed sequence to a good default state.
*
* In normal use, it's accessed via one of the following type aliases, which
* use seed_seq_fe128 and seed_seq_fe256 above.
*
* randutils::auto_seed_128
* randutils::auto_seed_256
*
* It's discussed in detail at
* http://www.pcg-random.org/posts/simple-portable-cpp-seed-entropy.html
* and its motivation (why you can't just use std::random_device) here
* http://www.pcg-random.org/posts/cpps-random_device.html
*/
template<typename SeedSeq>
struct auto_seeded : public SeedSeq
{
auto_seeded()
: auto_seeded(randutils::get_entropy())
{}
template<std::size_t N>
auto_seeded(std::array<std::uint32_t, N> const & seeds)
: SeedSeq(seeds.begin(), seeds.end())
{}
using SeedSeq::SeedSeq;
const SeedSeq & base() const
{
return *this;
}
SeedSeq & base()
{
return *this;
}
};
using auto_seed_128 = auto_seeded<seed_seq_fe128>;
using auto_seed_256 = auto_seeded<seed_seq_fe256>;
} // namespace randutils
using default_URNG = meta::if_c<(sizeof(void *) >= sizeof(long long)),
std::mt19937_64, std::mt19937>;
#if !RANGES_CXX_THREAD_LOCAL
template<typename URNG>
class sync_URNG : private URNG
{
mutable std::mutex mtx_;
public:
using URNG::URNG;
sync_URNG() = default;
using typename URNG::result_type;
result_type operator()()
{
std::lock_guard<std::mutex> guard{mtx_};
return static_cast<URNG &>(*this)();
}
using URNG::max;
using URNG::min;
};
using default_random_engine = sync_URNG<default_URNG>;
#else
using default_random_engine = default_URNG;
#endif
template<typename T = void>
default_random_engine & get_random_engine()
{
using Seeder = meta::if_c<(sizeof(default_URNG) > 16),
randutils::auto_seed_256,
randutils::auto_seed_128>;
#if RANGES_CXX_THREAD_LOCAL >= RANGES_CXX_THREAD_LOCAL_11
static thread_local default_random_engine engine{Seeder{}.base()};
#elif RANGES_CXX_THREAD_LOCAL
static __thread bool initialized = false;
static __thread meta::_t<std::aligned_storage<sizeof(default_random_engine),
alignof(default_random_engine)>>
storage;
if(!initialized)
{
::new(static_cast<void *>(&storage))
default_random_engine{Seeder{}.base()};
initialized = true;
}
auto & engine = reinterpret_cast<default_random_engine &>(storage);
#else
static default_random_engine engine{Seeder{}.base()};
#endif // RANGES_CXX_THREAD_LOCAL
return engine;
}
} // namespace detail
/// \endcond
} // namespace ranges
RANGES_DIAGNOSTIC_POP
#endif