-
Notifications
You must be signed in to change notification settings - Fork 447
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
[GLUTEN-6805][CH] support function array_remove/array_repeat (#7210)
* support array_repeat * support function arrayRemove * finish dev * remove logs * fix failed uts
- Loading branch information
1 parent
5e5b8d9
commit 804f08c
Showing
8 changed files
with
206 additions
and
98 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
100 changes: 100 additions & 0 deletions
100
cpp-ch/local-engine/Parser/scalar_function_parser/arrayRemove.cpp
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,100 @@ | ||
/* | ||
* Licensed to the Apache Software Foundation (ASF) under one or more | ||
* contributor license agreements. See the NOTICE file distributed with | ||
* this work for additional information regarding copyright ownership. | ||
* The ASF licenses this file to You under the Apache License, Version 2.0 | ||
* (the "License"); you may not use this file except in compliance with | ||
* the License. You may obtain a copy of the License at | ||
* | ||
* http://www.apache.org/licenses/LICENSE-2.0 | ||
* | ||
* Unless required by applicable law or agreed to in writing, software | ||
* distributed under the License is distributed on an "AS IS" BASIS, | ||
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. | ||
* See the License for the specific language governing permissions and | ||
* limitations under the License. | ||
*/ | ||
#include <DataTypes/DataTypeArray.h> | ||
#include <Functions/FunctionsMiscellaneous.h> | ||
#include <Parser/FunctionParser.h> | ||
#include <Common/Exception.h> | ||
#include <Common/assert_cast.h> | ||
|
||
namespace DB | ||
{ | ||
namespace ErrorCodes | ||
{ | ||
extern const int SIZES_OF_COLUMNS_DOESNT_MATCH; | ||
}; | ||
}; | ||
|
||
namespace local_engine | ||
{ | ||
class FunctionParserArrayRemove : public FunctionParser | ||
{ | ||
public: | ||
FunctionParserArrayRemove(SerializedPlanParser * plan_parser_) : FunctionParser(plan_parser_) { } | ||
~FunctionParserArrayRemove() override = default; | ||
|
||
static constexpr auto name = "array_remove"; | ||
String getName() const override { return name; } | ||
|
||
const DB::ActionsDAG::Node * | ||
parse(const substrait::Expression_ScalarFunction & substrait_func, DB::ActionsDAG & actions_dag) const override | ||
{ | ||
auto parsed_args = parseFunctionArguments(substrait_func, actions_dag); | ||
if (parsed_args.size() != 2) | ||
throw Exception(DB::ErrorCodes::SIZES_OF_COLUMNS_DOESNT_MATCH, "Function {} requires exactly two arguments", getName()); | ||
|
||
/// Parse spark array_remove(arr, elem) | ||
/// if (arr == null || elem == null) return null | ||
/// else return arrayFilter(x -> x != assumeNotNull(elem), assumeNotNull(arr)) | ||
const auto * arr_arg = parsed_args[0]; | ||
const auto * elem_arg = parsed_args[1]; | ||
const auto * arr_not_null = toFunctionNode(actions_dag, "assumeNotNull", {arr_arg}); | ||
const auto * elem_not_null = toFunctionNode(actions_dag, "assumeNotNull", {elem_arg}); | ||
const auto & arr_not_null_type = assert_cast<const DataTypeArray &>(*arr_not_null->result_type); | ||
|
||
/// Create lambda function x -> ifNull(x != assumeNotNull(elem), 1) | ||
/// Note that notEquals in CH is not null safe, so we need to wrap it with ifNull | ||
ActionsDAG lambda_actions_dag; | ||
const auto * x_in_lambda = &lambda_actions_dag.addInput("x", arr_not_null_type.getNestedType()); | ||
const auto * elem_in_lambda = &lambda_actions_dag.addInput(elem_not_null->result_name, elem_not_null->result_type); | ||
const auto * not_equals_in_lambda = toFunctionNode(lambda_actions_dag, "notEquals", {x_in_lambda, elem_in_lambda}); | ||
const auto * const_one_in_lambda = addColumnToActionsDAG(lambda_actions_dag, std::make_shared<DataTypeUInt8>(), {1}); | ||
const auto * if_null_in_lambda = toFunctionNode(lambda_actions_dag, "ifNull", {not_equals_in_lambda, const_one_in_lambda}); | ||
const auto * lambda_output = if_null_in_lambda; | ||
lambda_actions_dag.getOutputs().push_back(lambda_output); | ||
lambda_actions_dag.removeUnusedActions(Names(1, lambda_output->result_name)); | ||
|
||
auto expression_actions_settings = DB::ExpressionActionsSettings::fromContext(getContext(), DB::CompileExpressions::yes); | ||
auto lambda_actions = std::make_shared<DB::ExpressionActions>(std::move(lambda_actions_dag), expression_actions_settings); | ||
|
||
DB::Names captured_column_names{elem_in_lambda->result_name}; | ||
NamesAndTypesList lambda_arguments_names_and_types; | ||
lambda_arguments_names_and_types.emplace_back(x_in_lambda->result_name, x_in_lambda->result_type); | ||
DB::Names required_column_names = lambda_actions->getRequiredColumns(); | ||
auto function_capture = std::make_shared<FunctionCaptureOverloadResolver>( | ||
lambda_actions, | ||
captured_column_names, | ||
lambda_arguments_names_and_types, | ||
lambda_output->result_type, | ||
lambda_output->result_name); | ||
const auto * lambda_function = &actions_dag.addFunction(function_capture, {elem_not_null}, lambda_output->result_name); | ||
|
||
/// Apply arrayFilter with the lambda function | ||
const auto * array_filter = toFunctionNode(actions_dag, "arrayFilter", {lambda_function, arr_not_null}); | ||
|
||
/// Return null if arr or elem is null | ||
const auto * arr_is_null = toFunctionNode(actions_dag, "isNull", {arr_arg}); | ||
const auto * elem_is_null = toFunctionNode(actions_dag, "isNull", {elem_arg}); | ||
const auto * arr_or_elem_is_null = toFunctionNode(actions_dag, "or", {arr_is_null, elem_is_null}); | ||
const auto * null_array_node | ||
= addColumnToActionsDAG(actions_dag, std::make_shared<DataTypeNullable>(arr_not_null->result_type), {}); | ||
const auto * if_node = toFunctionNode(actions_dag, "if", {arr_or_elem_is_null, null_array_node, array_filter}); | ||
return convertNodeTypeIfNeeded(substrait_func, if_node, actions_dag); | ||
} | ||
}; | ||
|
||
static FunctionParserRegister<FunctionParserArrayRemove> register_array_remove; | ||
} |
97 changes: 97 additions & 0 deletions
97
cpp-ch/local-engine/Parser/scalar_function_parser/arrayRepeat.cpp
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,97 @@ | ||
|
||
/* | ||
* Licensed to the Apache Software Foundation (ASF) under one or more | ||
* contributor license agreements. See the NOTICE file distributed with | ||
* this work for additional information regarding copyright ownership. | ||
* The ASF licenses this file to You under the Apache License, Version 2.0 | ||
* (the "License"); you may not use this file except in compliance with | ||
* the License. You may obtain a copy of the License at | ||
* | ||
* http://www.apache.org/licenses/LICENSE-2.0 | ||
* | ||
* Unless required by applicable law or agreed to in writing, software | ||
* distributed under the License is distributed on an "AS IS" BASIS, | ||
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. | ||
* See the License for the specific language governing permissions and | ||
* limitations under the License. | ||
*/ | ||
#include <DataTypes/DataTypeArray.h> | ||
#include <Functions/FunctionsMiscellaneous.h> | ||
#include <Parser/FunctionParser.h> | ||
#include <Common/Exception.h> | ||
#include <Common/assert_cast.h> | ||
|
||
namespace DB | ||
{ | ||
namespace ErrorCodes | ||
{ | ||
extern const int SIZES_OF_COLUMNS_DOESNT_MATCH; | ||
}; | ||
}; | ||
|
||
namespace local_engine | ||
{ | ||
class FunctionParserArrayRepeat : public FunctionParser | ||
{ | ||
public: | ||
FunctionParserArrayRepeat(SerializedPlanParser * plan_parser_) : FunctionParser(plan_parser_) { } | ||
~FunctionParserArrayRepeat() override = default; | ||
|
||
static constexpr auto name = "array_repeat"; | ||
String getName() const override { return name; } | ||
|
||
const DB::ActionsDAG::Node * | ||
parse(const substrait::Expression_ScalarFunction & substrait_func, DB::ActionsDAG & actions_dag) const override | ||
{ | ||
auto parsed_args = parseFunctionArguments(substrait_func, actions_dag); | ||
if (parsed_args.size() != 2) | ||
throw Exception(DB::ErrorCodes::SIZES_OF_COLUMNS_DOESNT_MATCH, "Function {} requires exactly two arguments", getName()); | ||
|
||
/// Parse spark array_repeat(elem, n) | ||
/// if (n == null) return null | ||
/// else return arrayMap(x -> elem, range(greatest(assumeNotNull(n)))) | ||
const auto * elem_arg = parsed_args[0]; | ||
const auto * n_arg = parsed_args[1]; | ||
const auto * n_not_null_arg = toFunctionNode(actions_dag, "assumeNotNull", {n_arg}); | ||
const auto * const_zero_node = addColumnToActionsDAG(actions_dag, n_not_null_arg->result_type, {0}); | ||
const auto * greatest_node = toFunctionNode(actions_dag, "greatest", {n_not_null_arg, const_zero_node}); | ||
const auto * range_node = toFunctionNode(actions_dag, "range", {greatest_node}); | ||
const auto & range_type = assert_cast<const DataTypeArray & >(*removeNullable(range_node->result_type)); | ||
|
||
// Create lambda function x -> elem | ||
ActionsDAG lambda_actions_dag; | ||
const auto * x_in_lambda = &lambda_actions_dag.addInput("x", range_type.getNestedType()); | ||
const auto * elem_in_lambda = &lambda_actions_dag.addInput(elem_arg->result_name, elem_arg->result_type); | ||
const auto * lambda_output = elem_in_lambda; | ||
lambda_actions_dag.getOutputs().push_back(lambda_output); | ||
lambda_actions_dag.removeUnusedActions(Names(1, lambda_output->result_name)); | ||
|
||
auto expression_actions_settings = DB::ExpressionActionsSettings::fromContext(getContext(), DB::CompileExpressions::yes); | ||
auto lambda_actions = std::make_shared<DB::ExpressionActions>(std::move(lambda_actions_dag), expression_actions_settings); | ||
|
||
DB::Names captured_column_names{elem_in_lambda->result_name}; | ||
NamesAndTypesList lambda_arguments_names_and_types; | ||
lambda_arguments_names_and_types.emplace_back(x_in_lambda->result_name, x_in_lambda->result_type); | ||
DB::Names required_column_names = lambda_actions->getRequiredColumns(); | ||
auto function_capture = std::make_shared<FunctionCaptureOverloadResolver>( | ||
lambda_actions, | ||
captured_column_names, | ||
lambda_arguments_names_and_types, | ||
lambda_output->result_type, | ||
lambda_output->result_name); | ||
const auto * lambda_function = &actions_dag.addFunction(function_capture, {elem_arg}, lambda_output->result_name); | ||
|
||
/// Apply arrayMap with the lambda function | ||
const auto * array_map_node = toFunctionNode(actions_dag, "arrayMap", {lambda_function, range_node}); | ||
|
||
/// Return null if n is null | ||
const auto * n_is_null_node = toFunctionNode(actions_dag, "isNull", {n_arg}); | ||
const auto * null_array_node | ||
= addColumnToActionsDAG(actions_dag, std::make_shared<DataTypeNullable>(array_map_node->result_type), {}); | ||
const auto * if_node = toFunctionNode(actions_dag, "if", {n_is_null_node, null_array_node, array_map_node}); | ||
return convertNodeTypeIfNeeded(substrait_func, if_node, actions_dag); | ||
} | ||
}; | ||
|
||
static FunctionParserRegister<FunctionParserArrayRepeat> register_array_repeat; | ||
} |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters