nucleotide_base.hpp

Go to the documentation of this file.

// SPDX-FileCopyrightText: 2006-2025 Knut Reinert & Freie Universität Berlin
// SPDX-FileCopyrightText: 2016-2025 Knut Reinert & MPI für molekulare Genetik
// SPDX-License-Identifier: BSD-3-Clause

 
#pragma once
 
#include <seqan3/alphabet/alphabet_base.hpp>
#include <seqan3/alphabet/detail/concept.hpp>
#include <seqan3/alphabet/detail/convert.hpp>
#include <seqan3/alphabet/nucleotide/concept.hpp>
#include <seqan3/utility/char_operations/transform.hpp>
 
namespace seqan3
{

template <typename derived_type, auto size>
class nucleotide_base : public alphabet_base<derived_type, size, char>
{
private:
    using base_t = alphabet_base<derived_type, size, char>;

    constexpr nucleotide_base() noexcept = default;                                    
    constexpr nucleotide_base(nucleotide_base const &) noexcept = default;             
    constexpr nucleotide_base(nucleotide_base &&) noexcept = default;                  
    constexpr nucleotide_base & operator=(nucleotide_base const &) noexcept = default; 
    constexpr nucleotide_base & operator=(nucleotide_base &&) noexcept = default;      
    ~nucleotide_base() noexcept = default;                                             


    friend derived_type;
 
protected:
    // Import from base:
    using typename base_t::char_type;
    using typename base_t::rank_type;
 
public:
    using base_t::alphabet_size;
    using base_t::to_rank;

    // This constructor needs to be public, because constructor templates are not inherited otherwise
    template <typename other_nucl_type>
        requires (!std::same_as<nucleotide_base, other_nucl_type>) && (!std::same_as<derived_type, other_nucl_type>)
              && nucleotide_alphabet<other_nucl_type>
              && detail::convertable_to_through_char_representation<other_nucl_type, derived_type>
    explicit constexpr nucleotide_base(other_nucl_type const & other) noexcept
    {
        static_cast<derived_type &>(*this) =
            detail::convert_through_char_representation<other_nucl_type, derived_type>[seqan3::to_rank(other)];
    }



    constexpr derived_type complement() const noexcept
    {
        return derived_type{}.assign_rank(derived_type{}.rank_complement(to_rank()));
    }


    static constexpr bool char_is_valid(char_type const c) noexcept
    {
        return valid_char_table[static_cast<uint8_t>(c)];
    }
 
private:
    static constexpr std::array<bool, 256> valid_char_table{
        []() constexpr
        {
            std::array<bool, 256> ret{};
 
            // Value-initialisation of std::array does usually initialise. `fill` is explicit.
            ret.fill(false);
 
            // the original valid chars and their lower cases
            for (size_t rank = 0u; rank < derived_type::alphabet_size; ++rank)
            {
                uint8_t c = derived_type::rank_to_char(rank);
                ret[c] = true;
                ret[to_lower(c)] = true;
            }
 
            // U and T shall be accepted for all
            ret['U'] = true;
            ret['T'] = true;
            ret['u'] = true;
            ret['t'] = true;
 
            return ret;
        }()};
};
 
} // namespace seqan3