libosmium2-dev 2.11.4-1 / usr / include / osmium / index / id_set.hpp

#ifndef OSMIUM_INDEX_ID_SET_HPP
#define OSMIUM_INDEX_ID_SET_HPP

/*

This file is part of Osmium (http://osmcode.org/libosmium).

Copyright 2013-2017 Jochen Topf <jochen@topf.org> and others (see README).

Boost Software License - Version 1.0 - August 17th, 2003

Permission is hereby granted, free of charge, to any person or organization
obtaining a copy of the software and accompanying documentation covered by
this license (the "Software") to use, reproduce, display, distribute,
execute, and transmit the Software, and to prepare derivative works of the
Software, and to permit third-parties to whom the Software is furnished to
do so, all subject to the following:

The copyright notices in the Software and this entire statement, including
the above license grant, this restriction and the following disclaimer,
must be included in all copies of the Software, in whole or in part, and
all derivative works of the Software, unless such copies or derivative
works are solely in the form of machine-executable object code generated by
a source language processor.

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, TITLE AND NON-INFRINGEMENT. IN NO EVENT
SHALL THE COPYRIGHT HOLDERS OR ANYONE DISTRIBUTING THE SOFTWARE BE LIABLE
FOR ANY DAMAGES OR OTHER LIABILITY, WHETHER IN CONTRACT, TORT OR OTHERWISE,
ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
DEALINGS IN THE SOFTWARE.

*/

#include <algorithm>
#include <cassert>
#include <cstring>
#include <memory>
#include <type_traits>
#include <unordered_set>
#include <vector>

#include <osmium/osm/item_type.hpp>
#include <osmium/osm/types.hpp>

namespace osmium {

    namespace index {

        /**
         * Virtual parent class for IdSets. Use one of the implementations
         * provided.
         */
        template <typename T>
        class IdSet {

        public:

            virtual ~IdSet() {
            }

            /**
             * Add the given Id to the set.
             */
            virtual void set(T id) = 0;

            /**
             * Is the Id in the set?
             */
            virtual bool get(T id) const noexcept = 0;

            /**
             * Is the set empty?
             */
            virtual bool empty() const = 0;

            /**
             * Clear the set.
             */
            virtual void clear() = 0;

        }; // class IdSet

        template <typename T>
        class IdSetDense;

        /**
         * Const_iterator for iterating over a IdSetDense.
         */
        template <typename T>
        class IdSetDenseIterator {

            static_assert(std::is_unsigned<T>::value, "Needs unsigned type");
            static_assert(sizeof(T) >= 4, "Needs at least 32bit type");

            const IdSetDense<T>* m_set;
            T m_value;
            T m_last;

            void next() noexcept {
                while (m_value != m_last && !m_set->get(m_value)) {
                    const T cid = IdSetDense<T>::chunk_id(m_value);
                    assert(cid < m_set->m_data.size());
                    if (!m_set->m_data[cid]) {
                        m_value = (cid + 1) << (IdSetDense<T>::chunk_bits + 3);
                    } else {
                        const auto slot = m_set->m_data[cid][IdSetDense<T>::offset(m_value)];
                        if (slot == 0) {
                            m_value += 8;
                            m_value &= ~0x7;
                        } else {
                            ++m_value;
                        }
                    }
                }
            }

        public:

            using iterator_category = std::forward_iterator_tag;
            using value_type        = T;
            using pointer           = value_type*;
            using reference         = value_type&;

            IdSetDenseIterator(const IdSetDense<T>* set, T value, T last) noexcept :
                m_set(set),
                m_value(value),
                m_last(last) {
                next();
            }

            IdSetDenseIterator<T>& operator++() noexcept {
                if (m_value != m_last) {
                    ++m_value;
                    next();
                }
                return *this;
            }

            IdSetDenseIterator<T> operator++(int) noexcept {
                IdSetDenseIterator<T> tmp(*this);
                operator++();
                return tmp;
            }

            bool operator==(const IdSetDenseIterator<T>& rhs) const noexcept {
                return m_set == rhs.m_set && m_value == rhs.m_value;
            }

            bool operator!=(const IdSetDenseIterator<T>& rhs) const noexcept {
                return ! (*this == rhs);
            }

            T operator*() const noexcept {
                assert(m_value < m_last);
                return m_value;
            }

        }; // class IdSetDenseIterator

        /**
         * A set of Ids of the given type. Internal storage is in chunks of
         * arrays used as bit fields. Internally those chunks will be allocated
         * as needed, so it works relatively efficiently with both smaller
         * and larger Id sets. If it is not used, no memory is allocated at
         * all.
         */
        template <typename T>
        class IdSetDense : public IdSet<T> {

            static_assert(std::is_unsigned<T>::value, "Needs unsigned type");
            static_assert(sizeof(T) >= 4, "Needs at least 32bit type");

            friend class IdSetDenseIterator<T>;

            // This value is a compromise. For node Ids it could be bigger
            // which would mean less (but larger) memory allocations. For
            // relations Ids it could be smaller, because they would all fit
            // into a smaller allocation.
            constexpr static const size_t chunk_bits = 22;
            constexpr static const size_t chunk_size = 1 << chunk_bits;

            std::vector<std::unique_ptr<unsigned char[]>> m_data;
            T m_size = 0;

            static size_t chunk_id(T id) noexcept {
                return id >> (chunk_bits + 3);
            }

            static size_t offset(T id) noexcept {
                return (id >> 3) & ((1 << chunk_bits) - 1);
            }

            static unsigned char bitmask(T id) noexcept {
                return 1 << (id & 0x7);
            }

            T last() const noexcept {
                return static_cast<T>(m_data.size()) * chunk_size * 8;
            }

            unsigned char& get_element(T id) {
                const auto cid = chunk_id(id);
                if (cid >= m_data.size()) {
                    m_data.resize(cid + 1);
                }

                auto& chunk = m_data[cid];
                if (!chunk) {
                    chunk.reset(new unsigned char[chunk_size]);
                    ::memset(chunk.get(), 0, chunk_size);
                }

                return chunk[offset(id)];
            }

        public:

            using const_iterator = IdSetDenseIterator<T>;

            IdSetDense() = default;

            /**
             * Add the Id to the set if it is not already in there.
             *
             * @param id The Id to set.
             * @returns true if the Id was added, false if it was already set.
             */
            bool check_and_set(T id) {
                auto& element = get_element(id);

                if ((element & bitmask(id)) == 0) {
                    element |= bitmask(id);
                    ++m_size;
                    return true;
                }

                return false;
            }

            /**
             * Add the given Id to the set.
             *
             * @param id The Id to set.
             */
            void set(T id) override final {
                (void)check_and_set(id);
            }

            /**
             * Remove the given Id from the set.
             *
             * @param id The Id to set.
             */
            void unset(T id) {
                auto& element = get_element(id);

                if ((element & bitmask(id)) != 0) {
                    element &= ~bitmask(id);
                    --m_size;
                }
            }

            /**
             * Is the Id in the set?
             *
             * @param id The Id to check.
             */
            bool get(T id) const noexcept override final {
                if (chunk_id(id) >= m_data.size()) {
                    return false;
                }
                auto* r = m_data[chunk_id(id)].get();
                if (!r) {
                    return false;
                }
                return (r[offset(id)] & bitmask(id)) != 0;
            }

            /**
             * Is the set empty?
             */
            bool empty() const noexcept override final {
                return m_size == 0;
            }

            /**
             * The number of Ids stored in the set.
             */
            T size() const noexcept {
                return m_size;
            }

            /**
             * Clear the set.
             */
            void clear() override final {
                m_data.clear();
                m_size = 0;
            }

            IdSetDenseIterator<T> begin() const {
                return IdSetDenseIterator<T>{this, 0, last()};
            }

            IdSetDenseIterator<T> end() const {
                return IdSetDenseIterator<T>{this, last(), last()};
            }

        }; // class IdSetDense

        /**
         * IdSet implementation for small Id sets. It writes the Ids
         * into a vector and uses linear search.
         */
        template <typename T>
        class IdSetSmall : public IdSet<T> {

            std::vector<T> m_data;

        public:

            /**
             * Add the given Id to the set.
             */
            void set(T id) override final {
                m_data.push_back(id);
            }

            /**
             * Is the Id in the set? Uses linear search.
             *
             * @param id The Id to check.
             */
            bool get(T id) const noexcept override final {
                const auto it = std::find(m_data.cbegin(), m_data.cend(), id);
                return it != m_data.cend();
            }

            /**
             * Is the Id in the set? Uses a binary search. For larger sets
             * this might be more efficient than calling get(), the set
             * must be sorted.
             *
             * @param id The Id to check.
             * @pre You must have called sort_unique() before calling this
             *      or be sure there are no duplicates and the Ids have been
             *      set in order.
             */
            bool get_binary_search(T id) const noexcept {
                return std::binary_search(m_data.cbegin(), m_data.cend(), id);
            }

            /**
             * Is the set empty?
             */
            bool empty() const noexcept override final {
                return m_data.empty();
            }

            /**
             * Clear the set.
             */
            void clear() override final {
                m_data.clear();
            }

            /**
             * Sort the internal vector and remove any duplicates. Call this
             * before using size(), get_binary_search() or using an iterator.
             */
            void sort_unique() {
                std::sort(m_data.begin(), m_data.end());
                const auto last = std::unique(m_data.begin(), m_data.end());
                m_data.erase(last, m_data.end());

            }

            /**
             * The number of Ids stored in the set.
             *
             * @pre You must have called sort_unique() before calling this
             *      or be sure there are no duplicates.
             */
            size_t size() const noexcept {
                return m_data.size();
            }

            /// Iterator type. There is no non-const iterator.
            using const_iterator = typename std::vector<T>::const_iterator;

            const_iterator begin() const noexcept {
                return m_data.cbegin();
            }

            const_iterator end() const noexcept {
                return m_data.cend();
            }

            const_iterator cbegin() const noexcept {
                return m_data.cbegin();
            }

            const_iterator cend() const noexcept {
                return m_data.cend();
            }

        }; // class IdSetSmall

        template <template<typename> class IdSetType>
        class NWRIdSet {

            using id_set_type = IdSetType<osmium::unsigned_object_id_type>;

            id_set_type m_sets[3];

        public:

            id_set_type& operator()(osmium::item_type type) noexcept {
                return m_sets[osmium::item_type_to_nwr_index(type)];
            }

            const id_set_type& operator()(osmium::item_type type) const noexcept {
                return m_sets[osmium::item_type_to_nwr_index(type)];
            }

        }; // class NWRIdSet

    } // namespace index

} // namespace osmium

#endif // OSMIUM_INDEX_ID_SET_HPP