alphabet_variant.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 <algorithm>
#include <array>
#include <cassert>
#include <span>
#include <utility>
#include <variant>
 
#include <seqan3/alphabet/alphabet_base.hpp>
#include <seqan3/alphabet/composite/detail.hpp>
#include <seqan3/utility/type_list/traits.hpp>
#include <seqan3/utility/type_traits/lazy_conditional.hpp>
 
namespace seqan3::detail
{
 
template <typename t, size_t extent = std::dynamic_extent>
struct is_span : std::false_type
{};
 
template <typename t, size_t extent>
struct is_span<std::span<t, extent>> : std::true_type
{};
 
template <typename other_t, typename... alternative_types>
inline constexpr bool variant_general_guard =
    (!std::same_as<other_t, alphabet_variant<alternative_types...>>)
    && (!std::is_base_of_v<alphabet_variant<alternative_types...>, other_t>)
    && (!(std::same_as<other_t, alternative_types> || ...))
    && (!list_traits::contains<alphabet_variant<alternative_types...>, recursive_required_types_t<other_t>>);
 
template <typename lhs_t, typename rhs_t, bool lhs_rhs_switched, typename... alternative_types>
inline constexpr bool variant_comparison_guard =
    (instantiate_if_v<lazy<weakly_equality_comparable_with_trait, rhs_t, alternative_types>,
                      (std::same_as<lhs_t, alphabet_variant<alternative_types...>>)
                          && (variant_general_guard<rhs_t, alternative_types...>)
                          && !(lhs_rhs_switched && is_type_specialisation_of_v<rhs_t, alphabet_variant>)>
     || ...);
} // namespace seqan3::detail
 
namespace seqan3
{
 
template <typename... alternative_types>
    requires (detail::writable_constexpr_alphabet<alternative_types> && ...) && (std::regular<alternative_types> && ...)
          && (sizeof...(alternative_types) >= 2)
class alphabet_variant :
    public alphabet_base<alphabet_variant<alternative_types...>,
                         (static_cast<size_t>(alphabet_size<alternative_types>) + ...),
                         char>
{
private:
    using base_t = alphabet_base<alphabet_variant<alternative_types...>,
                                 (static_cast<size_t>(alphabet_size<alternative_types>) + ...),
                                 char>;
 
    static_assert((std::is_same_v<alphabet_char_t<alternative_types>, char> && ...),
                  "The alphabet_variant is currently only tested for alphabets with char_type char. "
                  "Contact us on GitHub if you have a different use case: https://github.com/seqan/seqan3 .");
 
    friend base_t;
 
    using alternatives = seqan3::type_list<alternative_types...>;
 
    static_assert(((seqan3::list_traits::count<alternative_types, alternatives> == 1) && ... && true),
                  "All types in a alphabet_variant must be distinct.");
 
    using typename base_t::char_type;
    using typename base_t::rank_type;
 
public:
    using base_t::alphabet_size;
    using base_t::assign_rank;
    using base_t::to_char;
    using base_t::to_rank;
 
    using seqan3_required_types = type_list<alternative_types...>;
    using seqan3_recursive_required_types = list_traits::concat<
        seqan3_required_types,
        detail::transformation_trait_or_t<detail::recursive_required_types<alternative_types>, type_list<>>...>;
 
    template <typename alternative_t>
    static constexpr bool is_alternative() noexcept
    {
        return seqan3::pack_traits::contains<alternative_t, alternative_types...>;
    }
 
    constexpr alphabet_variant() noexcept = default;                                     
    constexpr alphabet_variant(alphabet_variant const &) noexcept = default;             
    constexpr alphabet_variant(alphabet_variant &&) noexcept = default;                  
    constexpr alphabet_variant & operator=(alphabet_variant const &) noexcept = default; 
    constexpr alphabet_variant & operator=(alphabet_variant &&) noexcept = default;      
    ~alphabet_variant() noexcept = default;                                              
 
    template <typename alternative_t>
        requires (!std::same_as<alternative_t, alphabet_variant>) && (!detail::is_span<alternative_t>::value)
              && (!std::is_base_of_v<alphabet_variant, alternative_t>)
              && (!list_traits::contains<
                  alphabet_variant,
                  detail::transformation_trait_or_t<detail::recursive_required_types<alternative_t>, type_list<>>>)
              && (is_alternative<alternative_t>())
    constexpr alphabet_variant(alternative_t const alternative) noexcept
    {
        assign_rank(rank_by_type_(alternative));
    }
 
    template <typename indirect_alternative_t>
        requires (!detail::is_span<indirect_alternative_t>::value)
              && ((
                  detail::instantiate_if_v<detail::lazy<std::is_convertible, indirect_alternative_t, alternative_types>,
                                           detail::variant_general_guard<indirect_alternative_t, alternative_types...>>
                  || ...))
    constexpr alphabet_variant(indirect_alternative_t const rhs) noexcept
    {
        using alternative_predicate = detail::implicitly_convertible_from<indirect_alternative_t>;
        constexpr auto alternative_position =
            seqan3::list_traits::find_if<alternative_predicate::template invoke, alternatives>;
        using alternative_t = seqan3::list_traits::at<alternative_position, alternatives>;
        assign_rank(rank_by_type_(alternative_t(rhs)));
    }
 
    template <typename indirect_alternative_t>
        requires (
            (!(detail::instantiate_if_v<detail::lazy<std::is_convertible, indirect_alternative_t, alternative_types>,
                                        detail::variant_general_guard<indirect_alternative_t, alternative_types...>>
               || ...))
            && (detail::instantiate_if_v<detail::lazy<std::is_constructible, alternative_types, indirect_alternative_t>,
                                         detail::variant_general_guard<indirect_alternative_t, alternative_types...>>
                || ...))
    constexpr explicit alphabet_variant(indirect_alternative_t const rhs) noexcept
    {
        using alternative_predicate = detail::constructible_from<indirect_alternative_t>;
        constexpr auto alternative_position =
            seqan3::list_traits::find_if<alternative_predicate::template invoke, alternatives>;
        using alternative_t = seqan3::list_traits::at<alternative_position, alternatives>;
        assign_rank(rank_by_type_(alternative_t(rhs)));
    }
 
    template <typename indirect_alternative_t>
        requires (detail::variant_general_guard<indirect_alternative_t, alternative_types...>
                  && (weakly_assignable_from<alternative_types, indirect_alternative_t> || ...))
    constexpr alphabet_variant & operator=(indirect_alternative_t const & rhs) noexcept
    {
        using alternative_predicate = detail::assignable_from<indirect_alternative_t>;
        constexpr auto alternative_position =
            seqan3::list_traits::find_if<alternative_predicate::template invoke, alternatives>;
        using alternative_t = seqan3::list_traits::at<alternative_position, alternatives>;
        alternative_t alternative{};
        alternative = rhs;
        assign_rank(rank_by_type_(alternative));
        return *this;
    }
 
    template <size_t index>
    constexpr bool holds_alternative() const noexcept
    {
        static_assert(index < alphabet_size, "The alphabet_variant contains less alternatives than you are checking.");
        return (to_rank() >= partial_sum_sizes[index]) && (to_rank() < partial_sum_sizes[index + 1]);
    }
 
    template <size_t index>
    constexpr auto convert_to() const
    {
        return convert_impl<index, true>();
    }
 
    template <size_t index>
    constexpr auto convert_unsafely_to() const noexcept
    {
        return convert_impl<index, false>();
    }
 
    template <typename alternative_t>
    constexpr bool holds_alternative() const noexcept
        requires (is_alternative<alternative_t>())
    {
        constexpr size_t index = seqan3::list_traits::find<alternative_t, alternatives>;
        return holds_alternative<index>();
    }
 
    template <typename alternative_t>
    constexpr alternative_t convert_to() const
        requires (is_alternative<alternative_t>())
    {
        constexpr size_t index = seqan3::list_traits::find<alternative_t, alternatives>;
        return convert_impl<index, true>();
    }
 
    template <typename alternative_t>
    constexpr alternative_t convert_unsafely_to() const noexcept
        requires (is_alternative<alternative_t>())
    {
        constexpr size_t index = seqan3::list_traits::find<alternative_t, alternatives>;
        return convert_impl<index, false>();
    }
 
    template <typename alphabet_variant_t, typename indirect_alternative_type>
        requires (!detail::is_span<alphabet_variant_t>::value)
              && (detail::variant_comparison_guard<alphabet_variant_t,
                                                   indirect_alternative_type,
                                                   false,
                                                   alternative_types...>)
    friend constexpr bool operator==(alphabet_variant_t const lhs, indirect_alternative_type const rhs) noexcept
    {
        using alternative_predicate = detail::weakly_equality_comparable_with_<indirect_alternative_type>;
        constexpr auto alternative_position =
            seqan3::list_traits::find_if<alternative_predicate::template invoke, alternatives>;
        using alternative_t = seqan3::list_traits::at<alternative_position, alternatives>;
        return lhs.template holds_alternative<alternative_t>()
            && (lhs.template convert_unsafely_to<alternative_t>() == rhs);
    }
 
    template <typename alphabet_variant_t, typename indirect_alternative_type>
        requires (!detail::is_span<alphabet_variant_t>::value)
              && (detail::variant_comparison_guard<alphabet_variant_t,
                                                   indirect_alternative_type,
                                                   false,
                                                   alternative_types...>)
    friend constexpr bool operator!=(alphabet_variant_t const lhs, indirect_alternative_type const rhs) noexcept
    {
        return !(lhs == rhs);
    }
 
    template <typename alphabet_variant_t, typename indirect_alternative_type, typename = void>
        requires (!detail::is_span<alphabet_variant_t>::value)
              && (detail::variant_comparison_guard<alphabet_variant_t,
                                                   indirect_alternative_type,
                                                   true,
                                                   alternative_types...>)
    friend constexpr bool operator==(indirect_alternative_type const lhs, alphabet_variant_t const rhs) noexcept
    {
        return rhs == lhs;
    }
 
    template <typename alphabet_variant_t, typename indirect_alternative_type, typename = void>
        requires (!detail::is_span<alphabet_variant_t>::value)
              && (detail::variant_comparison_guard<alphabet_variant_t,
                                                   indirect_alternative_type,
                                                   true,
                                                   alternative_types...>)
    friend constexpr bool operator!=(indirect_alternative_type const lhs, alphabet_variant_t const rhs) noexcept
    {
        return rhs != lhs;
    }
 
    static constexpr bool char_is_valid(char_type const chr) noexcept
    {
        using index_t = std::make_unsigned_t<char_type>;
        return first_valid_char_table[static_cast<index_t>(chr)] < sizeof...(alternative_types);
    }
 
protected:
 
    template <size_t index, bool throws>
    constexpr auto convert_impl() const noexcept(!throws) -> seqan3::list_traits::at<index, alternatives>
    {
        static_assert(index < alphabet_size, "The alphabet_variant contains less alternatives than you are checking.");
        using alternative_t = seqan3::list_traits::at<index, alternatives>;
 
        if constexpr (throws)
        {
            if (!holds_alternative<index>()) // [[unlikely]]
            {
                throw std::bad_variant_access{};
            }
        }
 
        return seqan3::assign_rank_to(to_rank() - partial_sum_sizes[index], alternative_t{});
    }
 
    template <size_t index, typename alternative_t>
        requires (is_alternative<alternative_t>())
    static constexpr rank_type rank_by_index_(alternative_t const & alternative) noexcept
    {
        return partial_sum_sizes[index] + static_cast<rank_type>(seqan3::to_rank(alternative));
    }
 
    template <typename alternative_t>
        requires (is_alternative<alternative_t>())
    static constexpr rank_type rank_by_type_(alternative_t const & alternative) noexcept
    {
        constexpr size_t index = seqan3::list_traits::find<alternative_t, alternatives>;
        return rank_by_index_<index>(alternative);
    }
 
    static constexpr char_type rank_to_char(rank_type const rank)
    {
        return rank_to_char_table[rank];
    }
 
    static constexpr rank_type char_to_rank(char_type const chr)
    {
        using index_t = std::make_unsigned_t<char_type>;
        return char_to_rank_table[static_cast<index_t>(chr)];
    }
 
    static constexpr std::array<rank_type, sizeof...(alternative_types) + 1> partial_sum_sizes{
        []() constexpr
        {
            constexpr size_t N = sizeof...(alternative_types) + 1;
 
            std::array<rank_type, N> partial_sum{0, seqan3::alphabet_size<alternative_types>...};
 
            for (size_t i = 1u; i < N; ++i)
                partial_sum[i] += partial_sum[i - 1];
 
            return partial_sum;
        }()};
 
    static constexpr std::array<char_type, alphabet_size> rank_to_char_table{
        []() constexpr
        {
            auto assign_value_to_char = [](auto alternative, auto & value_to_char, auto & value) constexpr
            {
                using alternative_t = std::decay_t<decltype(alternative)>;
                for (size_t i = 0u; i < seqan3::alphabet_size<alternative_t>; ++i, ++value)
                    value_to_char[value] = seqan3::to_char(seqan3::assign_rank_to(i, alternative));
            };
 
            size_t value{};
            std::array<char_type, alphabet_size> value_to_char{};
 
            // initializer lists guarantee sequencing;
            // the following expression behaves as:
            // for(auto alternative: alternative_types)
            //    assign_rank_to_char(alternative, rank_to_char, value);
            ((assign_value_to_char(alternative_types{}, value_to_char, value)), ...);
 
            return value_to_char;
        }()};
 
    static constexpr auto first_valid_char_table{
        []() constexpr
        {
            constexpr size_t alternative_size = sizeof...(alternative_types);
            constexpr size_t table_size = detail::size_in_values_v<char_type>;
            using first_alphabet_t = detail::min_viable_uint_t<alternative_size>;
 
            std::array<first_alphabet_t, table_size> lookup_table{};
 
            for (size_t i = 0u; i < table_size; ++i)
            {
                char_type chr = static_cast<char_type>(i);
 
                std::array<bool, alternative_size> valid_chars{char_is_valid_for<alternative_types>(chr)...};
 
                auto found_it = std::find(valid_chars.begin(), valid_chars.end(), true);
                lookup_table[i] = found_it - valid_chars.begin();
            }
 
            return lookup_table;
        }()};
 
    static constexpr std::array<rank_type, detail::size_in_values_v<char_type>> char_to_rank_table{
        []() constexpr
        {
            constexpr size_t alternative_size = sizeof...(alternative_types);
            constexpr size_t table_size = detail::size_in_values_v<char_type>;
 
            std::array<rank_type, table_size> char_to_rank{};
 
            for (size_t i = 0u; i < table_size; ++i)
            {
                char_type chr = static_cast<char_type>(i);
 
                std::array<rank_type, alternative_size> ranks{
                    rank_by_type_(assign_char_to(chr, alternative_types{}))...};
 
                // if no char_is_valid_for any alternative, use the rank of the first alternative
                char_to_rank[i] = first_valid_char_table[i] < alternative_size ? ranks[first_valid_char_table[i]] : 0;
            }
 
            return char_to_rank;
        }()};
};
 
} // namespace seqan3