Add overload constructor and add docs

cpp/include/cudf/detail/join.hpp

@@ -90,12 +90,12 @@ struct hash_join {
    *
    * @param build The build table, from which the hash table is built.
    * @param has_nulls Flag to indicate if the there exists any nulls in the `build` table or
-   *        any table that will be used later for join.
+   *        any probe table that will be used later for join.
    * @param compare_nulls Controls whether null join-key values should match or not.
    * @param stream CUDA stream used for device memory operations and kernel launches.
    */
   hash_join(cudf::table_view const& build,
-            std::optional<bool> has_nulls,
+            bool has_nulls,
             cudf::null_equality compare_nulls,
             rmm::cuda_stream_view stream);
 

cpp/include/cudf/join.hpp

@@ -282,13 +282,20 @@ class hash_join {
    * undefined.
    *
    * @param build The build table, from which the hash table is built
-   * @param has_nulls Flag to indicate if the there exists any nulls in the `build` table or
-   *        any table that will be used later for join
    * @param compare_nulls Controls whether null join-key values should match or not
    * @param stream CUDA stream used for device memory operations and kernel launches
    */
   hash_join(cudf::table_view const& build,
-            std::optional<bool> has_nulls,
+            null_equality compare_nulls,
+            rmm::cuda_stream_view stream = cudf::get_default_stream());
+  /**
+   * @copydoc hash_join
+   *
+   * @param has_nulls Flag to indicate if the there exists any nulls in the `build` table or
+   *        any probe table that will be used later for join
+   */
+  hash_join(cudf::table_view const& build,
+            bool has_nulls,
             null_equality compare_nulls,
             rmm::cuda_stream_view stream = cudf::get_default_stream());
 
@@ -303,6 +310,9 @@ class hash_join {
    * @param mr Device memory resource used to allocate the returned table and columns' device
    * memory.
    *
+   * @throw cudf::logic_error If the input probe table has nulls while this hash_join class was not
+   * constructed with null check.
+   *
    * @return A pair of columns [`left_indices`, `right_indices`] that can be used to construct
    * the result of performing an inner join between two tables with `build` and `probe`
    * as the the join keys .
@@ -325,6 +335,9 @@ class hash_join {
    * @param mr Device memory resource used to allocate the returned table and columns' device
    * memory.
    *
+   * @throw cudf::logic_error If the input probe table has nulls while this hash_join class was not
+   * constructed with null check.
+   *
    * @return A pair of columns [`left_indices`, `right_indices`] that can be used to construct
    * the result of performing a left join between two tables with `build` and `probe`
    * as the the join keys .
@@ -347,6 +360,9 @@ class hash_join {
    * @param mr Device memory resource used to allocate the returned table and columns' device
    * memory.
    *
+   * @throw cudf::logic_error If the input probe table has nulls while this hash_join class was not
+   * constructed with null check.
+   *
    * @return A pair of columns [`left_indices`, `right_indices`] that can be used to construct
    * the result of performing a full join between two tables with `build` and `probe`
    * as the the join keys .
@@ -365,6 +381,9 @@ class hash_join {
    * @param probe The probe table, from which the tuples are probed
    * @param stream CUDA stream used for device memory operations and kernel launches
    *
+   * @throw cudf::logic_error If the input probe table has nulls while this hash_join class was not
+   * constructed with null check.
+   *
    * @return The exact number of output when performing an inner join between two tables with
    * `build` and `probe` as the the join keys .
    */
@@ -378,6 +397,9 @@ class hash_join {
    * @param probe The probe table, from which the tuples are probed
    * @param stream CUDA stream used for device memory operations and kernel launches
    *
+   * @throw cudf::logic_error If the input probe table has nulls while this hash_join class was not
+   * constructed with null check.
+   *
    * @return The exact number of output when performing a left join between two tables with `build`
    * and `probe` as the the join keys .
    */
@@ -393,6 +415,9 @@ class hash_join {
    * @param mr Device memory resource used to allocate the intermediate table and columns' device
    * memory.
    *
+   * @throw cudf::logic_error If the input probe table has nulls while this hash_join class was not
+   * constructed with null check.
+   *
    * @return The exact number of output when performing a full join between two tables with `build`
    * and `probe` as the the join keys .
    */

cpp/src/join/hash_join.cu

@@ -359,10 +359,10 @@ std::size_t get_full_join_size(
 
 template <typename Hasher>
 hash_join<Hasher>::hash_join(cudf::table_view const& build,
-                             std::optional<bool> has_nulls,
+                             bool has_nulls,
                              cudf::null_equality compare_nulls,
                              rmm::cuda_stream_view stream)
-  : _has_nulls(has_nulls.value_or(true)),
+  : _has_nulls(has_nulls),
     _is_empty{build.num_rows() == 0},
     _nulls_equal{compare_nulls},
     _hash_table{compute_hash_table_size(build.num_rows()),
@@ -381,8 +381,8 @@ hash_join<Hasher>::hash_join(cudf::table_view const& build,
 
   if (_is_empty) { return; }
 
-  auto const row_bitmask = std::move(
-    cudf::detail::bitmask_and(build, stream, rmm::mr::get_current_device_resource()).first);
+  auto const row_bitmask =
+    cudf::detail::bitmask_and(build, stream, rmm::mr::get_current_device_resource()).first;
   cudf::detail::build_join_hash_table(_build,
                                       _preprocessed_build,
                                       _hash_table,
@@ -586,7 +586,15 @@ hash_join<Hasher>::compute_hash_join(cudf::table_view const& probe,
 hash_join::~hash_join() = default;
 
 hash_join::hash_join(cudf::table_view const& build,
-                     std::optional<bool> has_nulls,
+                     null_equality compare_nulls,
+                     rmm::cuda_stream_view stream)
+  // If we cannot know beforehand about null existence then let's assume that there are nulls.
+  : hash_join(build, true /*has_nulls*/, compare_nulls, stream)
+{
+}
+
+hash_join::hash_join(cudf::table_view const& build,
+                     bool has_nulls,
                      null_equality compare_nulls,
                      rmm::cuda_stream_view stream)
   : _impl{std::make_unique<const impl_type>(build, has_nulls, compare_nulls, stream)}