GH-34630: [C++] Second block of refactoring to move acero out of liba…

…rrow (#34575) ### Rationale for this change This is the second and last block of refactoring before moving acero out of libarrow. This PR removes the remaining dependencies from libarrow into the compute/exec directory. ### What changes are included in this PR? In detail: * compute/exec/expression is moved inside compute since it is needed in several kernels * all compute/exec specific include files are removed from compute/api.h * light_array_exec_test.cc i(introduced in the previous refactoring) s merged back into light_array_test * testing generators are refactored to remove the need for exec/options.h * some utilities are moved from compute/exec/util to compute/util * updated python and r libraries usage of moved files ### Are these changes tested? Yes using existing tests (some of which have been updated due to refactoring). ### Are there any user-facing changes? no * Issue: #15280 * Closes: #34630
apache · Mar 18, 2023 · 8524572 · 8524572
1 parent 682c112
commit 8524572
Show file tree

Hide file tree

Showing 62 changed files with 328 additions and 336 deletions.
diff --git a/c_glib/arrow-glib/compute.hpp b/c_glib/arrow-glib/compute.hpp
@@ -20,6 +20,7 @@
 #pragma once
 
 #include <arrow/compute/api.h>
+#include <arrow/compute/exec/exec_plan.h>
 
 #include <arrow-glib/compute.h>
 

diff --git a/c_glib/arrow-glib/expression.hpp b/c_glib/arrow-glib/expression.hpp
@@ -19,7 +19,7 @@
 
 #pragma once
 
-#include <arrow/compute/exec/expression.h>
+#include <arrow/compute/expression.h>
 
 #include <arrow-glib/expression.h>
 

diff --git a/cpp/examples/arrow/compute_register_example.cc b/cpp/examples/arrow/compute_register_example.cc
@@ -18,10 +18,10 @@
 #include <arrow/api.h>
 #include <arrow/compute/api.h>
 #include <arrow/compute/exec/exec_plan.h>
-#include <arrow/compute/exec/expression.h>
 #include <arrow/compute/exec/options.h>
 #include <arrow/util/async_generator.h>
 #include <arrow/util/future.h>
+#include "arrow/compute/expression.h"
 
 #include <cstdlib>
 #include <iostream>

diff --git a/cpp/examples/arrow/dataset_documentation_example.cc b/cpp/examples/arrow/dataset_documentation_example.cc
@@ -20,7 +20,6 @@
 
 #include <arrow/api.h>
 #include <arrow/compute/cast.h>
-#include <arrow/compute/exec/expression.h>
 #include <arrow/dataset/dataset.h>
 #include <arrow/dataset/discovery.h>
 #include <arrow/dataset/file_base.h>
@@ -31,6 +30,7 @@
 #include <arrow/ipc/writer.h>
 #include <arrow/util/iterator.h>
 #include <parquet/arrow/writer.h>
+#include "arrow/compute/expression.h"
 
 #include <iostream>
 #include <vector>

diff --git a/cpp/examples/arrow/dataset_parquet_scan_example.cc b/cpp/examples/arrow/dataset_parquet_scan_example.cc
@@ -16,14 +16,14 @@
 // under the License.
 
 #include <arrow/api.h>
-#include <arrow/compute/exec/expression.h>
 #include <arrow/dataset/dataset.h>
 #include <arrow/dataset/discovery.h>
 #include <arrow/dataset/file_base.h>
 #include <arrow/dataset/file_parquet.h>
 #include <arrow/dataset/scanner.h>
 #include <arrow/filesystem/filesystem.h>
 #include <arrow/filesystem/path_util.h>
+#include "arrow/compute/expression.h"
 
 #include <cstdlib>
 #include <iostream>

diff --git a/cpp/examples/arrow/dataset_skyhook_scan_example.cc b/cpp/examples/arrow/dataset_skyhook_scan_example.cc
@@ -16,14 +16,14 @@
 // under the License.
 
 #include <arrow/api.h>
-#include <arrow/compute/exec/expression.h>
 #include <arrow/dataset/dataset.h>
 #include <arrow/dataset/discovery.h>
 #include <arrow/dataset/file_base.h>
 #include <arrow/dataset/scanner.h>
 #include <arrow/filesystem/filesystem.h>
 #include <arrow/filesystem/path_util.h>
 #include <skyhook/client/file_skyhook.h>
+#include "arrow/compute/expression.h"
 
 #include <cstdlib>
 #include <iostream>

diff --git a/cpp/examples/arrow/join_example.cc b/cpp/examples/arrow/join_example.cc
@@ -21,8 +21,8 @@
 #include <arrow/api.h>
 #include <arrow/compute/api.h>
 #include <arrow/compute/exec/exec_plan.h>
-#include <arrow/compute/exec/expression.h>
 #include <arrow/csv/api.h>
+#include "arrow/compute/expression.h"
 
 #include <arrow/dataset/dataset.h>
 #include <arrow/dataset/plan.h>

diff --git a/cpp/src/arrow/CMakeLists.txt b/cpp/src/arrow/CMakeLists.txt
@@ -395,7 +395,7 @@ list(APPEND
      compute/exec/asof_join_node.cc
      compute/exec/bloom_filter.cc
      compute/exec/exec_plan.cc
-     compute/exec/expression.cc
+     compute/expression.cc
      compute/exec/fetch_node.cc
      compute/exec/filter_node.cc
      compute/exec/hash_join.cc

diff --git a/cpp/src/arrow/compute/CMakeLists.txt b/cpp/src/arrow/compute/CMakeLists.txt
@@ -83,7 +83,8 @@ add_arrow_compute_test(internals_test
                        kernel_test.cc
                        light_array_test.cc
                        registry_test.cc
-                       key_hash_test.cc)
+                       key_hash_test.cc
+                       expression_test.cc)
 
 add_arrow_benchmark(function_benchmark PREFIX "arrow-compute")
 

diff --git a/cpp/src/arrow/compute/api.h b/cpp/src/arrow/compute/api.h
@@ -38,24 +38,16 @@
 /// @{
 /// @}
 
-#include "arrow/compute/exec/expression.h"  // IWYU pragma: export
-
-/// \defgroup execnode-options Concrete option classes for ExecNode options
-/// @{
-/// @}
-
-#include "arrow/compute/exec/options.h"  // IWYU pragma: export
+#include "arrow/compute/expression.h"  // IWYU pragma: export
 
 /// \defgroup execnode-row Utilities for working with data in a row-major format
 /// @{
 /// @}
 
 #include "arrow/compute/row/grouper.h"  // IWYU pragma: export
 
-/// \defgroup execnode-components Components associated with ExecNode
+/// \defgroup execnode-components Components associated with ExecBatch
 /// @{
 /// @}
 
-#include "arrow/compute/exec.h"            // IWYU pragma: export
-#include "arrow/compute/exec/exec_plan.h"  // IWYU pragma: export
-#include "arrow/compute/exec/groupby.h"    // IWYU pragma: export
+#include "arrow/compute/exec.h"  // IWYU pragma: export
diff --git a/cpp/src/arrow/compute/exec.h b/cpp/src/arrow/compute/exec.h
@@ -30,7 +30,7 @@
 #include <vector>
 
 #include "arrow/array/data.h"
-#include "arrow/compute/exec/expression.h"
+#include "arrow/compute/expression.h"
 #include "arrow/compute/type_fwd.h"
 #include "arrow/datum.h"
 #include "arrow/result.h"

diff --git a/cpp/src/arrow/compute/exec/CMakeLists.txt b/cpp/src/arrow/compute/exec/CMakeLists.txt
@@ -17,12 +17,11 @@
 
 arrow_install_all_headers("arrow/compute/exec")
 
-add_arrow_compute_test(expression_test
+add_arrow_compute_test(subtree_test
                        REQUIRE_ALL_KERNELS
                        PREFIX
                        "arrow-compute"
                        SOURCES
-                       expression_test.cc
                        subtree_test.cc)
 
 add_arrow_compute_test(plan_test
@@ -69,11 +68,6 @@ add_arrow_compute_test(util_test
                        SOURCES
                        util_test.cc
                        task_util_test.cc)
-add_arrow_compute_test(light_array_exec_test
-                       PREFIX
-                       "arrow-compute"
-                       SOURCES
-                       light_array_exec_test.cc)
 
 add_arrow_benchmark(expression_benchmark PREFIX "arrow-compute")
 

diff --git a/cpp/src/arrow/compute/exec/exec_plan.cc b/cpp/src/arrow/compute/exec/exec_plan.cc
@@ -24,11 +24,11 @@
 #include <unordered_set>
 
 #include "arrow/compute/exec.h"
-#include "arrow/compute/exec/expression.h"
 #include "arrow/compute/exec/options.h"
 #include "arrow/compute/exec/query_context.h"
 #include "arrow/compute/exec/task_util.h"
 #include "arrow/compute/exec/util.h"
+#include "arrow/compute/expression.h"
 #include "arrow/compute/registry.h"
 #include "arrow/datum.h"
 #include "arrow/record_batch.h"

diff --git a/cpp/src/arrow/compute/exec/exec_plan.h b/cpp/src/arrow/compute/exec/exec_plan.h
@@ -28,8 +28,8 @@
 
 #include "arrow/compute/api_vector.h"
 #include "arrow/compute/exec.h"
+#include "arrow/compute/exec/type_fwd.h"
 #include "arrow/compute/ordering.h"
-#include "arrow/compute/type_fwd.h"
 #include "arrow/type_fwd.h"
 #include "arrow/util/future.h"
 #include "arrow/util/macros.h"

diff --git a/cpp/src/arrow/compute/exec/expression_benchmark.cc b/cpp/src/arrow/compute/exec/expression_benchmark.cc
@@ -20,8 +20,8 @@
 #include <thread>
 
 #include "arrow/compute/cast.h"
-#include "arrow/compute/exec/expression.h"
 #include "arrow/compute/exec/test_util.h"
+#include "arrow/compute/expression.h"
 #include "arrow/dataset/partition.h"
 #include "arrow/testing/generator.h"
 #include "arrow/testing/gtest_util.h"

diff --git a/cpp/src/arrow/compute/exec/fetch_node.cc b/cpp/src/arrow/compute/exec/fetch_node.cc
@@ -21,11 +21,11 @@
 #include "arrow/compute/exec.h"
 #include "arrow/compute/exec/accumulation_queue.h"
 #include "arrow/compute/exec/exec_plan.h"
-#include "arrow/compute/exec/expression.h"
 #include "arrow/compute/exec/map_node.h"
 #include "arrow/compute/exec/options.h"
 #include "arrow/compute/exec/query_context.h"
 #include "arrow/compute/exec/util.h"
+#include "arrow/compute/expression.h"
 #include "arrow/datum.h"
 #include "arrow/result.h"
 #include "arrow/util/checked_cast.h"

diff --git a/cpp/src/arrow/compute/exec/filter_benchmark.cc b/cpp/src/arrow/compute/exec/filter_benchmark.cc
@@ -20,8 +20,8 @@
 #include "benchmark/benchmark.h"
 
 #include "arrow/compute/exec/benchmark_util.h"
-#include "arrow/compute/exec/expression.h"
 #include "arrow/compute/exec/options.h"
+#include "arrow/compute/expression.h"
 #include "arrow/record_batch.h"
 #include "arrow/testing/random.h"
 

diff --git a/cpp/src/arrow/compute/exec/filter_node.cc b/cpp/src/arrow/compute/exec/filter_node.cc
@@ -18,10 +18,10 @@
 #include "arrow/compute/api_vector.h"
 #include "arrow/compute/exec.h"
 #include "arrow/compute/exec/exec_plan.h"
-#include "arrow/compute/exec/expression.h"
 #include "arrow/compute/exec/map_node.h"
 #include "arrow/compute/exec/options.h"
 #include "arrow/compute/exec/query_context.h"
+#include "arrow/compute/expression.h"
 #include "arrow/datum.h"
 #include "arrow/result.h"
 #include "arrow/util/checked_cast.h"