Zouxxyy commented on code in PR #11076: URL: https://github.com/apache/incubator-gluten/pull/11076#discussion_r2529689419
########## gluten-substrait/src/main/scala/org/apache/gluten/extension/columnar/RemovePushedDownFilter.scala: ########## @@ -0,0 +1,49 @@ +/* + * 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. + */ +package org.apache.gluten.extension.columnar + +import org.apache.gluten.backendsapi.BackendsApiManager +import org.apache.gluten.execution.{BasicScanExecTransformer, FilterExecTransformerBase, FilterHandler} + +import org.apache.spark.sql.catalyst.expressions.{And, Expression, PredicateHelper} +import org.apache.spark.sql.catalyst.rules.Rule +import org.apache.spark.sql.execution.SparkPlan + +/** Remove the filter that have been pushed down to native scan. */ +object RemovePushedDownFilter extends Rule[SparkPlan] with PredicateHelper { + + override def apply(plan: SparkPlan): SparkPlan = plan.transform { + case filter: FilterExecTransformerBase if filter.cond != null => + val scanFilters = filter.child match { + // Get the filters including the manually pushed down ones. + case basicScanExecTransformer: BasicScanExecTransformer => + basicScanExecTransformer.filterExprs() + // For fallback scan, we need to keep original filter. + case _ => + Seq.empty[Expression] + } + if (scanFilters.isEmpty) { + filter + } else { + val remainingFilters = + FilterHandler.getRemainingFilters(scanFilters, splitConjunctivePredicates(filter.cond)) + val newCond = remainingFilters.reduceLeftOption(And).orNull Review Comment: > facing some attribute mismatch errors from Catalyst. @zhztheplayer I checked the logic of FilterExec and confirmed that its output indeed changes—it equals the child's output plus some NOT NULL attributes added by the condition. Therefore, I made the following modifications: - Extracted a helper method `buildNewOutput(output, conds)` from FilterExec. - If pushedFilter nonEmpty: - Set ScanNode's newOutput = buildNewOutput(output, pushedFilter). - Set FilterNode's newCond = cond -- pushedFilter. (If newCond == null, remove the filter node.) I believe this better reflects Node's correct behavior, although perhaps all golden files should be updated. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
