rdblue commented on a change in pull request #1852: URL: https://github.com/apache/iceberg/pull/1852#discussion_r532855059
########## File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/OptimizeConditionsInRowLevelOperations.scala ########## @@ -0,0 +1,46 @@ +/* + * 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.spark.sql.catalyst.optimizer + +import org.apache.spark.sql.SparkSession +import org.apache.spark.sql.catalyst.expressions.{Expression, Literal, SubqueryExpression} +import org.apache.spark.sql.catalyst.plans.logical.{DeleteFromTable, Filter, LocalRelation, LogicalPlan} +import org.apache.spark.sql.catalyst.rules.Rule +import org.apache.spark.sql.execution.datasources.v2.DataSourceV2ScanRelation + +// we have to optimize expressions used in delete/update before we can rewrite row-level operations +// otherwise, we will have to deal with redundant casts and will not detect noop deletes +// it is a temp solution since we cannot inject rewrite of row-level ops after operator optimizations +object OptimizeConditionsInRowLevelOperations extends Rule[LogicalPlan] { + override def apply(plan: LogicalPlan): LogicalPlan = plan transform { + case d @ DeleteFromTable(table, cond) if !SubqueryExpression.hasSubquery(cond.getOrElse(Literal.TrueLiteral)) => + val optimizedCond = optimizeCondition(cond.getOrElse(Literal.TrueLiteral), table) + d.copy(condition = Some(optimizedCond)) + } + + private def optimizeCondition(condition: Expression, targetTable: LogicalPlan): Expression = { + val optimizer = SparkSession.active.sessionState.optimizer + optimizer.execute(Filter(condition, targetTable)) match { + case Filter(optimizedCondition, _) => optimizedCondition Review comment: I think the `LocalRelation` would happen if 0 records matched. Spark would replace the scan + filter with an empty `LocalRelation`. That means that 0 records match the delete filter, so we can replace the whole delete filter with `false` to match 0 rows. Similarly, if the filter is removed then it must have matched everything, so the filter should be `true`. ---------------------------------------------------------------- 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. For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
