minimiser_hash.hpp

Go to the documentation of this file.

// -----------------------------------------------------------------------------------------------------
// Copyright (c) 2006-2023, Knut Reinert & Freie Universität Berlin
// Copyright (c) 2016-2023, Knut Reinert & MPI für molekulare Genetik
// This file may be used, modified and/or redistributed under the terms of the 3-clause BSD-License
// shipped with this file and also available at: https://github.com/seqan/seqan3/blob/master/LICENSE.md
// -----------------------------------------------------------------------------------------------------
 
#pragma once
 
#include <seqan3/alphabet/views/complement.hpp>
#include <seqan3/core/detail/strong_type.hpp>
#include <seqan3/search/views/kmer_hash.hpp>
#include <seqan3/search/views/minimiser.hpp>
 
namespace seqan3
{
struct seed : seqan3::detail::strong_type<uint64_t, seed>
{
    using seqan3::detail::strong_type<uint64_t, seed>::strong_type;
};
 
struct window_size : seqan3::detail::strong_type<uint32_t, window_size>
{
    using seqan3::detail::strong_type<uint32_t, window_size>::strong_type;
};
} // namespace seqan3
 
namespace seqan3::detail
{
struct minimiser_hash_fn
{
    constexpr auto operator()(shape const & shape, window_size const window_size) const
    {
        return seqan3::detail::adaptor_from_functor{*this, shape, window_size};
    }
 
    constexpr auto operator()(shape const & shape, window_size const window_size, seed const seed) const
    {
        return seqan3::detail::adaptor_from_functor{*this, shape, window_size, seed};
    }
 
    template <std::ranges::range urng_t>
    constexpr auto operator()(urng_t && urange,
                              shape const & shape,
                              window_size const window_size,
                              seed const seed = seqan3::seed{0x8F3F73B5CF1C9ADE}) const
    {
        static_assert(std::ranges::viewable_range<urng_t>,
                      "The range parameter to views::minimiser_hash cannot be a temporary of a non-view range.");
        static_assert(std::ranges::forward_range<urng_t>,
                      "The range parameter to views::minimiser_hash must model std::ranges::forward_range.");
        static_assert(semialphabet<std::ranges::range_reference_t<urng_t>>,
                      "The range parameter to views::minimiser_hash must be over elements of seqan3::semialphabet.");
 
        if (shape.size() > window_size.get())
            throw std::invalid_argument{"The size of the shape cannot be greater than the window size."};
 
        auto forward_strand = std::forward<urng_t>(urange) | seqan3::views::kmer_hash(shape)
                            | std::views::transform(
                                  [seed](uint64_t i)
                                  {
                                      return i ^ seed.get();
                                  });
 
        auto reverse_strand = std::forward<urng_t>(urange) | seqan3::views::complement | std::views::reverse
                            | seqan3::views::kmer_hash(shape)
                            | std::views::transform(
                                  [seed](uint64_t i)
                                  {
                                      return i ^ seed.get();
                                  })
                            | std::views::reverse;
 
        return seqan3::detail::minimiser_view(forward_strand, reverse_strand, window_size.get() - shape.size() + 1);
    }
};
 
} // namespace seqan3::detail
 
namespace seqan3::views
{
 
inline constexpr auto minimiser_hash = detail::minimiser_hash_fn{};
 
 
} // namespace seqan3::views