hash_join.hpp

Go to the documentation of this file.

/*
 * SPDX-FileCopyrightText: Copyright (c) 2025-2026, NVIDIA CORPORATION.
 * SPDX-License-Identifier: Apache-2.0
 */
 
#pragma once
 
#include <cudf/hashing.hpp>
#include <cudf/join/join.hpp>
#include <cudf/table/table_view.hpp>
#include <cudf/types.hpp>
#include <cudf/utilities/default_stream.hpp>
#include <cudf/utilities/export.hpp>
#include <cudf/utilities/memory_resource.hpp>
#include <cudf/utilities/span.hpp>
 
#include <rmm/cuda_stream_view.hpp>
#include <rmm/device_uvector.hpp>
 
#include <optional>
#include <utility>
 
namespace CUDF_EXPORT cudf {
 
// forward declaration
namespace hashing::detail {
template <typename T>
class MurmurHash3_x86_32;
}  // namespace hashing::detail
 
namespace detail {
template <typename T>
class hash_join;
}  // namespace detail
 
enum class nullable_join : bool { YES, NO };
 
class hash_join {
 public:
  using impl_type = typename cudf::detail::hash_join<
    cudf::hashing::detail::MurmurHash3_x86_32<cudf::hash_value_type>>;  
 
  hash_join() = delete;
  ~hash_join();
  hash_join(hash_join const&)            = delete;
  hash_join(hash_join&&)                 = delete;
  hash_join& operator=(hash_join const&) = delete;
  hash_join& operator=(hash_join&&)      = delete;
 
  hash_join(cudf::table_view const& right,
            null_equality compare_nulls,
            rmm::cuda_stream_view stream = cudf::get_default_stream());
 
  hash_join(cudf::table_view const& right,
            nullable_join has_nulls,
            null_equality compare_nulls,
            double load_factor,
            rmm::cuda_stream_view stream = cudf::get_default_stream());
 
  [[nodiscard]] std::pair<std::unique_ptr<rmm::device_uvector<size_type>>,
                          std::unique_ptr<rmm::device_uvector<size_type>>>
  inner_join(cudf::table_view const& left,
             std::optional<std::size_t> output_size = {},
             rmm::cuda_stream_view stream           = cudf::get_default_stream(),
             rmm::device_async_resource_ref mr = cudf::get_current_device_resource_ref()) const;
 
  [[nodiscard]] std::pair<std::unique_ptr<rmm::device_uvector<size_type>>,
                          std::unique_ptr<rmm::device_uvector<size_type>>>
  left_join(cudf::table_view const& left,
            std::optional<std::size_t> output_size = {},
            rmm::cuda_stream_view stream           = cudf::get_default_stream(),
            rmm::device_async_resource_ref mr      = cudf::get_current_device_resource_ref()) const;
 
  [[nodiscard]] std::pair<std::unique_ptr<rmm::device_uvector<size_type>>,
                          std::unique_ptr<rmm::device_uvector<size_type>>>
  full_join(cudf::table_view const& left,
            std::optional<std::size_t> output_size = {},
            rmm::cuda_stream_view stream           = cudf::get_default_stream(),
            rmm::device_async_resource_ref mr      = cudf::get_current_device_resource_ref()) const;
 
  [[nodiscard]] std::size_t inner_join_size(
    cudf::table_view const& left, rmm::cuda_stream_view stream = cudf::get_default_stream()) const;
 
  [[nodiscard]] std::size_t left_join_size(
    cudf::table_view const& left, rmm::cuda_stream_view stream = cudf::get_default_stream()) const;
 
  [[nodiscard]] std::size_t full_join_size(
    cudf::table_view const& left,
    rmm::cuda_stream_view stream      = cudf::get_default_stream(),
    rmm::device_async_resource_ref mr = cudf::get_current_device_resource_ref()) const;
 
  [[nodiscard]] cudf::join_match_context inner_join_match_context(
    cudf::table_view const& left,
    rmm::cuda_stream_view stream      = cudf::get_default_stream(),
    rmm::device_async_resource_ref mr = cudf::get_current_device_resource_ref()) const;
 
  [[nodiscard]] cudf::join_match_context left_join_match_context(
    cudf::table_view const& left,
    rmm::cuda_stream_view stream      = cudf::get_default_stream(),
    rmm::device_async_resource_ref mr = cudf::get_current_device_resource_ref()) const;
 
  [[nodiscard]] cudf::join_match_context full_join_match_context(
    cudf::table_view const& left,
    rmm::cuda_stream_view stream      = cudf::get_default_stream(),
    rmm::device_async_resource_ref mr = cudf::get_current_device_resource_ref()) const;
 
  [[nodiscard]] std::pair<std::unique_ptr<rmm::device_uvector<size_type>>,
                          std::unique_ptr<rmm::device_uvector<size_type>>>
  partitioned_inner_join(
    cudf::join_partition_context const& context,
    rmm::cuda_stream_view stream      = cudf::get_default_stream(),
    rmm::device_async_resource_ref mr = cudf::get_current_device_resource_ref()) const;
 
  [[nodiscard]] std::pair<std::unique_ptr<rmm::device_uvector<size_type>>,
                          std::unique_ptr<rmm::device_uvector<size_type>>>
  partitioned_left_join(
    cudf::join_partition_context const& context,
    rmm::cuda_stream_view stream      = cudf::get_default_stream(),
    rmm::device_async_resource_ref mr = cudf::get_current_device_resource_ref()) const;
 
  [[nodiscard]] std::pair<std::unique_ptr<rmm::device_uvector<size_type>>,
                          std::unique_ptr<rmm::device_uvector<size_type>>>
  partitioned_full_join(
    cudf::join_partition_context const& context,
    rmm::cuda_stream_view stream      = cudf::get_default_stream(),
    rmm::device_async_resource_ref mr = cudf::get_current_device_resource_ref()) const;
 
  [[nodiscard]] static std::pair<std::unique_ptr<rmm::device_uvector<size_type>>,
                                 std::unique_ptr<rmm::device_uvector<size_type>>>
  finalize_partitioned_full_join(
    cudf::host_span<cudf::device_span<size_type const> const> left_partials,
    cudf::host_span<cudf::device_span<size_type const> const> right_partials,
    size_type left_table_num_rows,
    size_type right_table_num_rows,
    rmm::cuda_stream_view stream      = cudf::get_default_stream(),
    rmm::device_async_resource_ref mr = cudf::get_current_device_resource_ref());
 
 private:
  std::unique_ptr<impl_type const> _impl;
};
  // end of group
 
}  // namespace CUDF_EXPORT cudf