You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@iceberg.apache.org by GitBox <gi...@apache.org> on 2021/02/02 04:09:00 UTC
[GitHub] [iceberg] aokolnychyi opened a new pull request #2193: Spark: Support UPDATE statements without subqueries
aokolnychyi opened a new pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193
This PR introduces support for UPDATE statements in Spark. Support for subqueries will come in a follow-up PR.
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] mehtaashish23 commented on a change in pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
mehtaashish23 commented on a change in pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#discussion_r568382200
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/analysis/AlignUpdateTable.scala
##########
@@ -0,0 +1,32 @@
+/*
+ * 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.analysis
+
+import org.apache.spark.sql.catalyst.plans.logical.{LogicalPlan, UpdateTable}
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.internal.SQLConf
+
+case class AlignUpdateTable(conf: SQLConf) extends Rule[LogicalPlan] with AssignmentAlignmentSupport {
+
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case u: UpdateTable if u.resolved =>
Review comment:
You want to put the existing check i.e. to not Align, if targetTable is not an Iceberg table?
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteUpdate.scala
##########
@@ -0,0 +1,87 @@
+/*
+ * 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.AnalysisException
+import org.apache.spark.sql.SparkSession
+import org.apache.spark.sql.catalyst.expressions.Alias
+import org.apache.spark.sql.catalyst.expressions.Expression
+import org.apache.spark.sql.catalyst.expressions.If
+import org.apache.spark.sql.catalyst.expressions.SubqueryExpression
+import org.apache.spark.sql.catalyst.plans.logical.Assignment
+import org.apache.spark.sql.catalyst.plans.logical.Filter
+import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
+import org.apache.spark.sql.catalyst.plans.logical.Project
+import org.apache.spark.sql.catalyst.plans.logical.ReplaceData
+import org.apache.spark.sql.catalyst.plans.logical.UpdateTable
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.utils.PlanUtils.isIcebergRelation
+import org.apache.spark.sql.catalyst.utils.RewriteRowLevelOperationHelper
+import org.apache.spark.sql.execution.datasources.v2.DataSourceV2Relation
+import org.apache.spark.sql.execution.datasources.v2.ExtendedDataSourceV2Implicits
+
+// TODO: should be part of early scan push down after the delete condition is optimized
+case class RewriteUpdate(spark: SparkSession) extends Rule[LogicalPlan] with RewriteRowLevelOperationHelper {
+
+ import ExtendedDataSourceV2Implicits._
+
+ // TODO: can we do any better for no-op updates? when conditions evaluate to false/true?
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond))
+ if isIcebergRelation(r) && SubqueryExpression.hasSubquery(cond) =>
+ throw new AnalysisException("UPDATE statements with subqueries are not currently supported")
+
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond)) if isIcebergRelation(r) =>
+ val writeInfo = newWriteInfo(r.schema)
+ val mergeBuilder = r.table.asMergeable.newMergeBuilder("update", writeInfo)
+
+ val matchingRowsPlanBuilder = scanRelation => Filter(cond, scanRelation)
+ val scanPlan = buildDynamicFilterScanPlan(spark, r.table, r.output, mergeBuilder, cond, matchingRowsPlanBuilder)
+
+ val updateProjection = buildUpdateProjection(r, scanPlan, assignments, cond)
+
+ val mergeWrite = mergeBuilder.asWriteBuilder.buildForBatch()
+ val writePlan = buildWritePlan(updateProjection, r.table)
+ ReplaceData(r, mergeWrite, writePlan)
+ }
+
+ private def buildUpdateProjection(
+ relation: DataSourceV2Relation,
+ scanPlan: LogicalPlan,
+ assignments: Seq[Assignment],
+ cond: Expression): LogicalPlan = {
+
+ // this method relies on the fact that the assignments have been aligned before
+ require(relation.output.size == assignments.size, "assignments must be aligned")
+
+ // Spark is going to execute the condition for each column but it seems we cannot avoid this
Review comment:
@aokolnychyi You mean running the same set of data with
1. All updates merged via "Merge Into"
2. All updates updated via "Update table"
And see the benchmarks? I wonder that considering MergeInto will do a join, irrespective of input, so I am not sure whether the Filter expression can be compared with a shuffle of data. I hope I am getting the expectation right.
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteUpdate.scala
##########
@@ -0,0 +1,87 @@
+/*
+ * 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.AnalysisException
+import org.apache.spark.sql.SparkSession
+import org.apache.spark.sql.catalyst.expressions.Alias
+import org.apache.spark.sql.catalyst.expressions.Expression
+import org.apache.spark.sql.catalyst.expressions.If
+import org.apache.spark.sql.catalyst.expressions.SubqueryExpression
+import org.apache.spark.sql.catalyst.plans.logical.Assignment
+import org.apache.spark.sql.catalyst.plans.logical.Filter
+import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
+import org.apache.spark.sql.catalyst.plans.logical.Project
+import org.apache.spark.sql.catalyst.plans.logical.ReplaceData
+import org.apache.spark.sql.catalyst.plans.logical.UpdateTable
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.utils.PlanUtils.isIcebergRelation
+import org.apache.spark.sql.catalyst.utils.RewriteRowLevelOperationHelper
+import org.apache.spark.sql.execution.datasources.v2.DataSourceV2Relation
+import org.apache.spark.sql.execution.datasources.v2.ExtendedDataSourceV2Implicits
+
+// TODO: should be part of early scan push down after the delete condition is optimized
+case class RewriteUpdate(spark: SparkSession) extends Rule[LogicalPlan] with RewriteRowLevelOperationHelper {
+
+ import ExtendedDataSourceV2Implicits._
+
+ // TODO: can we do any better for no-op updates? when conditions evaluate to false/true?
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond))
+ if isIcebergRelation(r) && SubqueryExpression.hasSubquery(cond) =>
+ throw new AnalysisException("UPDATE statements with subqueries are not currently supported")
+
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond)) if isIcebergRelation(r) =>
+ val writeInfo = newWriteInfo(r.schema)
+ val mergeBuilder = r.table.asMergeable.newMergeBuilder("update", writeInfo)
+
+ val matchingRowsPlanBuilder = scanRelation => Filter(cond, scanRelation)
+ val scanPlan = buildDynamicFilterScanPlan(spark, r.table, r.output, mergeBuilder, cond, matchingRowsPlanBuilder)
+
+ val updateProjection = buildUpdateProjection(r, scanPlan, assignments, cond)
+
+ val mergeWrite = mergeBuilder.asWriteBuilder.buildForBatch()
+ val writePlan = buildWritePlan(updateProjection, r.table)
+ ReplaceData(r, mergeWrite, writePlan)
+ }
+
+ private def buildUpdateProjection(
+ relation: DataSourceV2Relation,
+ scanPlan: LogicalPlan,
+ assignments: Seq[Assignment],
+ cond: Expression): LogicalPlan = {
+
+ // this method relies on the fact that the assignments have been aligned before
+ require(relation.output.size == assignments.size, "assignments must be aligned")
+
+ // Spark is going to execute the condition for each column but it seems we cannot avoid this
Review comment:
@aokolnychyi You mean running the same set of data with
1. All updates merged via "Merge Into"
2. All updates updated via "Update table"
And see the benchmarks? I wonder that considering MergeInto will do a join, irrespective of input, so I am not sure whether the Filter expression can be compared with a shuffle of data. I hope I am getting the expectation right.
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteUpdate.scala
##########
@@ -0,0 +1,87 @@
+/*
+ * 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.AnalysisException
+import org.apache.spark.sql.SparkSession
+import org.apache.spark.sql.catalyst.expressions.Alias
+import org.apache.spark.sql.catalyst.expressions.Expression
+import org.apache.spark.sql.catalyst.expressions.If
+import org.apache.spark.sql.catalyst.expressions.SubqueryExpression
+import org.apache.spark.sql.catalyst.plans.logical.Assignment
+import org.apache.spark.sql.catalyst.plans.logical.Filter
+import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
+import org.apache.spark.sql.catalyst.plans.logical.Project
+import org.apache.spark.sql.catalyst.plans.logical.ReplaceData
+import org.apache.spark.sql.catalyst.plans.logical.UpdateTable
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.utils.PlanUtils.isIcebergRelation
+import org.apache.spark.sql.catalyst.utils.RewriteRowLevelOperationHelper
+import org.apache.spark.sql.execution.datasources.v2.DataSourceV2Relation
+import org.apache.spark.sql.execution.datasources.v2.ExtendedDataSourceV2Implicits
+
+// TODO: should be part of early scan push down after the delete condition is optimized
+case class RewriteUpdate(spark: SparkSession) extends Rule[LogicalPlan] with RewriteRowLevelOperationHelper {
+
+ import ExtendedDataSourceV2Implicits._
+
+ // TODO: can we do any better for no-op updates? when conditions evaluate to false/true?
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond))
+ if isIcebergRelation(r) && SubqueryExpression.hasSubquery(cond) =>
+ throw new AnalysisException("UPDATE statements with subqueries are not currently supported")
+
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond)) if isIcebergRelation(r) =>
+ val writeInfo = newWriteInfo(r.schema)
+ val mergeBuilder = r.table.asMergeable.newMergeBuilder("update", writeInfo)
+
+ val matchingRowsPlanBuilder = scanRelation => Filter(cond, scanRelation)
+ val scanPlan = buildDynamicFilterScanPlan(spark, r.table, r.output, mergeBuilder, cond, matchingRowsPlanBuilder)
+
+ val updateProjection = buildUpdateProjection(r, scanPlan, assignments, cond)
+
+ val mergeWrite = mergeBuilder.asWriteBuilder.buildForBatch()
+ val writePlan = buildWritePlan(updateProjection, r.table)
+ ReplaceData(r, mergeWrite, writePlan)
+ }
+
+ private def buildUpdateProjection(
+ relation: DataSourceV2Relation,
+ scanPlan: LogicalPlan,
+ assignments: Seq[Assignment],
+ cond: Expression): LogicalPlan = {
+
+ // this method relies on the fact that the assignments have been aligned before
+ require(relation.output.size == assignments.size, "assignments must be aligned")
+
+ // Spark is going to execute the condition for each column but it seems we cannot avoid this
Review comment:
@aokolnychyi You mean running the same set of data with
1. All updates merged via "Merge Into"
2. All updates updated via "Update table"
And see the benchmarks? I wonder that considering MergeInto will do a join, irrespective of input, so I am not sure whether the If expression can be compared with a shuffle of data. I hope I am getting the expectation right.
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteUpdate.scala
##########
@@ -0,0 +1,87 @@
+/*
+ * 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.AnalysisException
+import org.apache.spark.sql.SparkSession
+import org.apache.spark.sql.catalyst.expressions.Alias
+import org.apache.spark.sql.catalyst.expressions.Expression
+import org.apache.spark.sql.catalyst.expressions.If
+import org.apache.spark.sql.catalyst.expressions.SubqueryExpression
+import org.apache.spark.sql.catalyst.plans.logical.Assignment
+import org.apache.spark.sql.catalyst.plans.logical.Filter
+import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
+import org.apache.spark.sql.catalyst.plans.logical.Project
+import org.apache.spark.sql.catalyst.plans.logical.ReplaceData
+import org.apache.spark.sql.catalyst.plans.logical.UpdateTable
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.utils.PlanUtils.isIcebergRelation
+import org.apache.spark.sql.catalyst.utils.RewriteRowLevelOperationHelper
+import org.apache.spark.sql.execution.datasources.v2.DataSourceV2Relation
+import org.apache.spark.sql.execution.datasources.v2.ExtendedDataSourceV2Implicits
+
+// TODO: should be part of early scan push down after the delete condition is optimized
+case class RewriteUpdate(spark: SparkSession) extends Rule[LogicalPlan] with RewriteRowLevelOperationHelper {
+
+ import ExtendedDataSourceV2Implicits._
+
+ // TODO: can we do any better for no-op updates? when conditions evaluate to false/true?
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond))
+ if isIcebergRelation(r) && SubqueryExpression.hasSubquery(cond) =>
+ throw new AnalysisException("UPDATE statements with subqueries are not currently supported")
+
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond)) if isIcebergRelation(r) =>
+ val writeInfo = newWriteInfo(r.schema)
+ val mergeBuilder = r.table.asMergeable.newMergeBuilder("update", writeInfo)
+
+ val matchingRowsPlanBuilder = scanRelation => Filter(cond, scanRelation)
+ val scanPlan = buildDynamicFilterScanPlan(spark, r.table, r.output, mergeBuilder, cond, matchingRowsPlanBuilder)
+
+ val updateProjection = buildUpdateProjection(r, scanPlan, assignments, cond)
+
+ val mergeWrite = mergeBuilder.asWriteBuilder.buildForBatch()
+ val writePlan = buildWritePlan(updateProjection, r.table)
+ ReplaceData(r, mergeWrite, writePlan)
+ }
+
+ private def buildUpdateProjection(
+ relation: DataSourceV2Relation,
+ scanPlan: LogicalPlan,
+ assignments: Seq[Assignment],
+ cond: Expression): LogicalPlan = {
+
+ // this method relies on the fact that the assignments have been aligned before
+ require(relation.output.size == assignments.size, "assignments must be aligned")
+
+ // Spark is going to execute the condition for each column but it seems we cannot avoid this
Review comment:
Roger, we can help with that but would recommend merging any of the approach to be functionally ready. And as we get bandwidth we can run the benchmark between the two approaches.
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] aokolnychyi commented on pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
aokolnychyi commented on pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#issuecomment-771342449
cc @dilipbiswal @rdblue @RussellSpitzer
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] aokolnychyi commented on pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
aokolnychyi commented on pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#issuecomment-772215412
Thanks for reviewing, @dilipbiswal @mehtaashish23 @rdblue!
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] aokolnychyi commented on a change in pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
aokolnychyi commented on a change in pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#discussion_r568890500
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/utils/RewriteRowLevelOperationHelper.scala
##########
@@ -197,6 +203,22 @@ trait RewriteRowLevelOperationHelper extends PredicateHelper with Logging {
}
}
}
+
+ protected def buildWritePlan(childPlan: LogicalPlan, table: Table): LogicalPlan = {
Review comment:
I think we may consider doing that too.
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] aokolnychyi merged pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
aokolnychyi merged pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] mehtaashish23 commented on a change in pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
mehtaashish23 commented on a change in pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#discussion_r568924173
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteUpdate.scala
##########
@@ -0,0 +1,87 @@
+/*
+ * 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.AnalysisException
+import org.apache.spark.sql.SparkSession
+import org.apache.spark.sql.catalyst.expressions.Alias
+import org.apache.spark.sql.catalyst.expressions.Expression
+import org.apache.spark.sql.catalyst.expressions.If
+import org.apache.spark.sql.catalyst.expressions.SubqueryExpression
+import org.apache.spark.sql.catalyst.plans.logical.Assignment
+import org.apache.spark.sql.catalyst.plans.logical.Filter
+import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
+import org.apache.spark.sql.catalyst.plans.logical.Project
+import org.apache.spark.sql.catalyst.plans.logical.ReplaceData
+import org.apache.spark.sql.catalyst.plans.logical.UpdateTable
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.utils.PlanUtils.isIcebergRelation
+import org.apache.spark.sql.catalyst.utils.RewriteRowLevelOperationHelper
+import org.apache.spark.sql.execution.datasources.v2.DataSourceV2Relation
+import org.apache.spark.sql.execution.datasources.v2.ExtendedDataSourceV2Implicits
+
+// TODO: should be part of early scan push down after the delete condition is optimized
+case class RewriteUpdate(spark: SparkSession) extends Rule[LogicalPlan] with RewriteRowLevelOperationHelper {
+
+ import ExtendedDataSourceV2Implicits._
+
+ // TODO: can we do any better for no-op updates? when conditions evaluate to false/true?
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond))
+ if isIcebergRelation(r) && SubqueryExpression.hasSubquery(cond) =>
+ throw new AnalysisException("UPDATE statements with subqueries are not currently supported")
+
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond)) if isIcebergRelation(r) =>
+ val writeInfo = newWriteInfo(r.schema)
+ val mergeBuilder = r.table.asMergeable.newMergeBuilder("update", writeInfo)
+
+ val matchingRowsPlanBuilder = scanRelation => Filter(cond, scanRelation)
+ val scanPlan = buildDynamicFilterScanPlan(spark, r.table, r.output, mergeBuilder, cond, matchingRowsPlanBuilder)
+
+ val updateProjection = buildUpdateProjection(r, scanPlan, assignments, cond)
+
+ val mergeWrite = mergeBuilder.asWriteBuilder.buildForBatch()
+ val writePlan = buildWritePlan(updateProjection, r.table)
+ ReplaceData(r, mergeWrite, writePlan)
+ }
+
+ private def buildUpdateProjection(
+ relation: DataSourceV2Relation,
+ scanPlan: LogicalPlan,
+ assignments: Seq[Assignment],
+ cond: Expression): LogicalPlan = {
+
+ // this method relies on the fact that the assignments have been aligned before
+ require(relation.output.size == assignments.size, "assignments must be aligned")
+
+ // Spark is going to execute the condition for each column but it seems we cannot avoid this
Review comment:
@aokolnychyi You mean running the same set of data with
1. All updates merged via "Merge Into"
2. All updates updated via "Update table"
And see the benchmarks? I wonder that considering MergeInto will do a join, irrespective of input, so I am not sure whether the Filter expression can be compared with a shuffle of data. I hope I am getting the expectation right.
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] aokolnychyi commented on a change in pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
aokolnychyi commented on a change in pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#discussion_r568841003
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/analysis/AlignUpdateTable.scala
##########
@@ -0,0 +1,32 @@
+/*
+ * 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.analysis
+
+import org.apache.spark.sql.catalyst.plans.logical.{LogicalPlan, UpdateTable}
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.internal.SQLConf
+
+case class AlignUpdateTable(conf: SQLConf) extends Rule[LogicalPlan] with AssignmentAlignmentSupport {
+
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case u: UpdateTable if u.resolved =>
Review comment:
Good catch, I forgot to update it.
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] aokolnychyi commented on a change in pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
aokolnychyi commented on a change in pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#discussion_r568928996
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteUpdate.scala
##########
@@ -0,0 +1,87 @@
+/*
+ * 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.AnalysisException
+import org.apache.spark.sql.SparkSession
+import org.apache.spark.sql.catalyst.expressions.Alias
+import org.apache.spark.sql.catalyst.expressions.Expression
+import org.apache.spark.sql.catalyst.expressions.If
+import org.apache.spark.sql.catalyst.expressions.SubqueryExpression
+import org.apache.spark.sql.catalyst.plans.logical.Assignment
+import org.apache.spark.sql.catalyst.plans.logical.Filter
+import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
+import org.apache.spark.sql.catalyst.plans.logical.Project
+import org.apache.spark.sql.catalyst.plans.logical.ReplaceData
+import org.apache.spark.sql.catalyst.plans.logical.UpdateTable
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.utils.PlanUtils.isIcebergRelation
+import org.apache.spark.sql.catalyst.utils.RewriteRowLevelOperationHelper
+import org.apache.spark.sql.execution.datasources.v2.DataSourceV2Relation
+import org.apache.spark.sql.execution.datasources.v2.ExtendedDataSourceV2Implicits
+
+// TODO: should be part of early scan push down after the delete condition is optimized
+case class RewriteUpdate(spark: SparkSession) extends Rule[LogicalPlan] with RewriteRowLevelOperationHelper {
+
+ import ExtendedDataSourceV2Implicits._
+
+ // TODO: can we do any better for no-op updates? when conditions evaluate to false/true?
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond))
+ if isIcebergRelation(r) && SubqueryExpression.hasSubquery(cond) =>
+ throw new AnalysisException("UPDATE statements with subqueries are not currently supported")
+
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond)) if isIcebergRelation(r) =>
+ val writeInfo = newWriteInfo(r.schema)
+ val mergeBuilder = r.table.asMergeable.newMergeBuilder("update", writeInfo)
+
+ val matchingRowsPlanBuilder = scanRelation => Filter(cond, scanRelation)
+ val scanPlan = buildDynamicFilterScanPlan(spark, r.table, r.output, mergeBuilder, cond, matchingRowsPlanBuilder)
+
+ val updateProjection = buildUpdateProjection(r, scanPlan, assignments, cond)
+
+ val mergeWrite = mergeBuilder.asWriteBuilder.buildForBatch()
+ val writePlan = buildWritePlan(updateProjection, r.table)
+ ReplaceData(r, mergeWrite, writePlan)
+ }
+
+ private def buildUpdateProjection(
+ relation: DataSourceV2Relation,
+ scanPlan: LogicalPlan,
+ assignments: Seq[Assignment],
+ cond: Expression): LogicalPlan = {
+
+ // this method relies on the fact that the assignments have been aligned before
+ require(relation.output.size == assignments.size, "assignments must be aligned")
+
+ // Spark is going to execute the condition for each column but it seems we cannot avoid this
Review comment:
I meant if we have an alternate implementation to what is currently proposed, will it be possible to perform tests at some reasonable scale to see if that alternative solution works better?
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] aokolnychyi commented on a change in pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
aokolnychyi commented on a change in pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#discussion_r568890120
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteUpdate.scala
##########
@@ -0,0 +1,87 @@
+/*
+ * 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.AnalysisException
+import org.apache.spark.sql.SparkSession
+import org.apache.spark.sql.catalyst.expressions.Alias
+import org.apache.spark.sql.catalyst.expressions.Expression
+import org.apache.spark.sql.catalyst.expressions.If
+import org.apache.spark.sql.catalyst.expressions.SubqueryExpression
+import org.apache.spark.sql.catalyst.plans.logical.Assignment
+import org.apache.spark.sql.catalyst.plans.logical.Filter
+import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
+import org.apache.spark.sql.catalyst.plans.logical.Project
+import org.apache.spark.sql.catalyst.plans.logical.ReplaceData
+import org.apache.spark.sql.catalyst.plans.logical.UpdateTable
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.utils.PlanUtils.isIcebergRelation
+import org.apache.spark.sql.catalyst.utils.RewriteRowLevelOperationHelper
+import org.apache.spark.sql.execution.datasources.v2.DataSourceV2Relation
+import org.apache.spark.sql.execution.datasources.v2.ExtendedDataSourceV2Implicits
+
+// TODO: should be part of early scan push down after the delete condition is optimized
Review comment:
Copy and paste from the delete rule, will remove.
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] aokolnychyi commented on a change in pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
aokolnychyi commented on a change in pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#discussion_r568310632
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteUpdate.scala
##########
@@ -0,0 +1,87 @@
+/*
+ * 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.AnalysisException
+import org.apache.spark.sql.SparkSession
+import org.apache.spark.sql.catalyst.expressions.Alias
+import org.apache.spark.sql.catalyst.expressions.Expression
+import org.apache.spark.sql.catalyst.expressions.If
+import org.apache.spark.sql.catalyst.expressions.SubqueryExpression
+import org.apache.spark.sql.catalyst.plans.logical.Assignment
+import org.apache.spark.sql.catalyst.plans.logical.Filter
+import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
+import org.apache.spark.sql.catalyst.plans.logical.Project
+import org.apache.spark.sql.catalyst.plans.logical.ReplaceData
+import org.apache.spark.sql.catalyst.plans.logical.UpdateTable
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.utils.PlanUtils.isIcebergRelation
+import org.apache.spark.sql.catalyst.utils.RewriteRowLevelOperationHelper
+import org.apache.spark.sql.execution.datasources.v2.DataSourceV2Relation
+import org.apache.spark.sql.execution.datasources.v2.ExtendedDataSourceV2Implicits
+
+// TODO: should be part of early scan push down after the delete condition is optimized
+case class RewriteUpdate(spark: SparkSession) extends Rule[LogicalPlan] with RewriteRowLevelOperationHelper {
+
+ import ExtendedDataSourceV2Implicits._
+
+ // TODO: can we do any better for no-op updates? when conditions evaluate to false/true?
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond))
+ if isIcebergRelation(r) && SubqueryExpression.hasSubquery(cond) =>
+ throw new AnalysisException("UPDATE statements with subqueries are not currently supported")
+
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond)) if isIcebergRelation(r) =>
+ val writeInfo = newWriteInfo(r.schema)
+ val mergeBuilder = r.table.asMergeable.newMergeBuilder("update", writeInfo)
+
+ val matchingRowsPlanBuilder = scanRelation => Filter(cond, scanRelation)
+ val scanPlan = buildDynamicFilterScanPlan(spark, r.table, r.output, mergeBuilder, cond, matchingRowsPlanBuilder)
+
+ val updateProjection = buildUpdateProjection(r, scanPlan, assignments, cond)
+
+ val mergeWrite = mergeBuilder.asWriteBuilder.buildForBatch()
+ val writePlan = buildWritePlan(updateProjection, r.table)
+ ReplaceData(r, mergeWrite, writePlan)
+ }
+
+ private def buildUpdateProjection(
+ relation: DataSourceV2Relation,
+ scanPlan: LogicalPlan,
+ assignments: Seq[Assignment],
+ cond: Expression): LogicalPlan = {
+
+ // this method relies on the fact that the assignments have been aligned before
+ require(relation.output.size == assignments.size, "assignments must be aligned")
+
+ // Spark is going to execute the condition for each column but it seems we cannot avoid this
Review comment:
Any ideas whether we can avoid this are welcome. I did check the generated code and the condition was evaluated for every column.
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] rdblue commented on a change in pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
rdblue commented on a change in pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#discussion_r568885574
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/utils/RewriteRowLevelOperationHelper.scala
##########
@@ -197,6 +203,22 @@ trait RewriteRowLevelOperationHelper extends PredicateHelper with Logging {
}
}
}
+
+ protected def buildWritePlan(childPlan: LogicalPlan, table: Table): LogicalPlan = {
Review comment:
For a follow up, should we add a rule that does this for Append and Overwrite plans as well? That would be nice so that we don't have to wait until 3.2.0 to get write distribution and ordering for normal writes.
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] rdblue commented on a change in pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
rdblue commented on a change in pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#discussion_r568879510
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteUpdate.scala
##########
@@ -0,0 +1,87 @@
+/*
+ * 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.AnalysisException
+import org.apache.spark.sql.SparkSession
+import org.apache.spark.sql.catalyst.expressions.Alias
+import org.apache.spark.sql.catalyst.expressions.Expression
+import org.apache.spark.sql.catalyst.expressions.If
+import org.apache.spark.sql.catalyst.expressions.SubqueryExpression
+import org.apache.spark.sql.catalyst.plans.logical.Assignment
+import org.apache.spark.sql.catalyst.plans.logical.Filter
+import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
+import org.apache.spark.sql.catalyst.plans.logical.Project
+import org.apache.spark.sql.catalyst.plans.logical.ReplaceData
+import org.apache.spark.sql.catalyst.plans.logical.UpdateTable
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.utils.PlanUtils.isIcebergRelation
+import org.apache.spark.sql.catalyst.utils.RewriteRowLevelOperationHelper
+import org.apache.spark.sql.execution.datasources.v2.DataSourceV2Relation
+import org.apache.spark.sql.execution.datasources.v2.ExtendedDataSourceV2Implicits
+
+// TODO: should be part of early scan push down after the delete condition is optimized
Review comment:
I'm not sure about this. I thought we wanted to explore moving these into the analyzer?
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] aokolnychyi commented on a change in pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
aokolnychyi commented on a change in pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#discussion_r568889377
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/analysis/AlignUpdateTable.scala
##########
@@ -0,0 +1,32 @@
+/*
+ * 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.analysis
+
+import org.apache.spark.sql.catalyst.plans.logical.{LogicalPlan, UpdateTable}
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.internal.SQLConf
+
+case class AlignUpdateTable(conf: SQLConf) extends Rule[LogicalPlan] with AssignmentAlignmentSupport {
+
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case u: UpdateTable if u.resolved =>
+ u.copy(assignments = alignAssignments(u.table, u.assignments))
Review comment:
I have no preference. `AlignMergeIntoTable` is way more involved but adding updates there would be alright.
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] aokolnychyi commented on a change in pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
aokolnychyi commented on a change in pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#discussion_r568310415
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteUpdate.scala
##########
@@ -0,0 +1,87 @@
+/*
+ * 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.AnalysisException
+import org.apache.spark.sql.SparkSession
+import org.apache.spark.sql.catalyst.expressions.Alias
+import org.apache.spark.sql.catalyst.expressions.Expression
+import org.apache.spark.sql.catalyst.expressions.If
+import org.apache.spark.sql.catalyst.expressions.SubqueryExpression
+import org.apache.spark.sql.catalyst.plans.logical.Assignment
+import org.apache.spark.sql.catalyst.plans.logical.Filter
+import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
+import org.apache.spark.sql.catalyst.plans.logical.Project
+import org.apache.spark.sql.catalyst.plans.logical.ReplaceData
+import org.apache.spark.sql.catalyst.plans.logical.UpdateTable
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.utils.PlanUtils.isIcebergRelation
+import org.apache.spark.sql.catalyst.utils.RewriteRowLevelOperationHelper
+import org.apache.spark.sql.execution.datasources.v2.DataSourceV2Relation
+import org.apache.spark.sql.execution.datasources.v2.ExtendedDataSourceV2Implicits
+
+// TODO: should be part of early scan push down after the delete condition is optimized
+case class RewriteUpdate(spark: SparkSession) extends Rule[LogicalPlan] with RewriteRowLevelOperationHelper {
+
+ import ExtendedDataSourceV2Implicits._
+
+ // TODO: can we do any better for no-op updates? when conditions evaluate to false/true?
Review comment:
Right now, `UPDATE t SET ... WHERE false` with result in a job and a commit with no changes. We may want to handle such cases differently but I am not sure it is a big deal.
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] rdblue commented on pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
rdblue commented on pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#issuecomment-771946663
I see no major blockers with this. There are a few things to consider, but overall I think this can go in and we can improve it from there.
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] aokolnychyi commented on pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
aokolnychyi commented on pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#issuecomment-771883039
The PR is ready for another round.
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] aokolnychyi commented on a change in pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
aokolnychyi commented on a change in pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#discussion_r568844834
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/analysis/AlignUpdateTable.scala
##########
@@ -0,0 +1,32 @@
+/*
+ * 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.analysis
+
+import org.apache.spark.sql.catalyst.plans.logical.{LogicalPlan, UpdateTable}
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.internal.SQLConf
+
+case class AlignUpdateTable(conf: SQLConf) extends Rule[LogicalPlan] with AssignmentAlignmentSupport {
+
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case u: UpdateTable if u.resolved =>
Review comment:
Fixed.
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/OptimizeConditionsInRowLevelOperations.scala
##########
@@ -21,7 +21,7 @@ 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.plans.logical.{DeleteFromTable, Filter, LocalRelation, LogicalPlan, UpdateTable}
Review comment:
Fixed.
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/analysis/AlignUpdateTable.scala
##########
@@ -0,0 +1,32 @@
+/*
+ * 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.analysis
+
+import org.apache.spark.sql.catalyst.plans.logical.{LogicalPlan, UpdateTable}
Review comment:
Fixed.
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] aokolnychyi edited a comment on pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
aokolnychyi edited a comment on pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#issuecomment-771342449
cc @dilipbiswal @rdblue @RussellSpitzer @mehtaashish23
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] aokolnychyi commented on a change in pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
aokolnychyi commented on a change in pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#discussion_r568309937
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/analysis/AlignUpdateTable.scala
##########
@@ -0,0 +1,32 @@
+/*
+ * 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.analysis
+
+import org.apache.spark.sql.catalyst.plans.logical.{LogicalPlan, UpdateTable}
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.internal.SQLConf
+
+case class AlignUpdateTable(conf: SQLConf) extends Rule[LogicalPlan] with AssignmentAlignmentSupport {
+
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case u: UpdateTable if u.resolved =>
+ u.copy(assignments = alignAssignments(u.table, u.assignments))
Review comment:
We apply the same logic as for MERGE.
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteDelete.scala
##########
@@ -55,8 +54,6 @@ case class RewriteDelete(spark: SparkSession) extends Rule[LogicalPlan] with Rew
import ExtendedDataSourceV2Implicits._
import RewriteRowLevelOperationHelper._
- override def resolver: Resolver = spark.sessionState.conf.resolver
Review comment:
Moved this to parent as I need both `conf` and `resolver` now.
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteMergeInto.scala
##########
@@ -224,22 +216,6 @@ case class RewriteMergeInto(spark: SparkSession) extends Rule[LogicalPlan] with
}
!(actions.size == 1 && hasUnconditionalDelete(actions.headOption))
}
-
- private def buildWritePlan(childPlan: LogicalPlan, table: Table): LogicalPlan = {
Review comment:
Moved to parent.
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteUpdate.scala
##########
@@ -0,0 +1,87 @@
+/*
+ * 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.AnalysisException
+import org.apache.spark.sql.SparkSession
+import org.apache.spark.sql.catalyst.expressions.Alias
+import org.apache.spark.sql.catalyst.expressions.Expression
+import org.apache.spark.sql.catalyst.expressions.If
+import org.apache.spark.sql.catalyst.expressions.SubqueryExpression
+import org.apache.spark.sql.catalyst.plans.logical.Assignment
+import org.apache.spark.sql.catalyst.plans.logical.Filter
+import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
+import org.apache.spark.sql.catalyst.plans.logical.Project
+import org.apache.spark.sql.catalyst.plans.logical.ReplaceData
+import org.apache.spark.sql.catalyst.plans.logical.UpdateTable
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.utils.PlanUtils.isIcebergRelation
+import org.apache.spark.sql.catalyst.utils.RewriteRowLevelOperationHelper
+import org.apache.spark.sql.execution.datasources.v2.DataSourceV2Relation
+import org.apache.spark.sql.execution.datasources.v2.ExtendedDataSourceV2Implicits
+
+// TODO: should be part of early scan push down after the delete condition is optimized
+case class RewriteUpdate(spark: SparkSession) extends Rule[LogicalPlan] with RewriteRowLevelOperationHelper {
+
+ import ExtendedDataSourceV2Implicits._
+
+ // TODO: can we do any better for no-op updates? when conditions evaluate to false/true?
Review comment:
Right now, `UPDATE t SET ... WHERE false` with result in a job and a commit with no changes. We may want to handle such cases differently but I am not sure it is a big deal.
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteUpdate.scala
##########
@@ -0,0 +1,87 @@
+/*
+ * 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.AnalysisException
+import org.apache.spark.sql.SparkSession
+import org.apache.spark.sql.catalyst.expressions.Alias
+import org.apache.spark.sql.catalyst.expressions.Expression
+import org.apache.spark.sql.catalyst.expressions.If
+import org.apache.spark.sql.catalyst.expressions.SubqueryExpression
+import org.apache.spark.sql.catalyst.plans.logical.Assignment
+import org.apache.spark.sql.catalyst.plans.logical.Filter
+import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
+import org.apache.spark.sql.catalyst.plans.logical.Project
+import org.apache.spark.sql.catalyst.plans.logical.ReplaceData
+import org.apache.spark.sql.catalyst.plans.logical.UpdateTable
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.utils.PlanUtils.isIcebergRelation
+import org.apache.spark.sql.catalyst.utils.RewriteRowLevelOperationHelper
+import org.apache.spark.sql.execution.datasources.v2.DataSourceV2Relation
+import org.apache.spark.sql.execution.datasources.v2.ExtendedDataSourceV2Implicits
+
+// TODO: should be part of early scan push down after the delete condition is optimized
+case class RewriteUpdate(spark: SparkSession) extends Rule[LogicalPlan] with RewriteRowLevelOperationHelper {
+
+ import ExtendedDataSourceV2Implicits._
+
+ // TODO: can we do any better for no-op updates? when conditions evaluate to false/true?
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond))
+ if isIcebergRelation(r) && SubqueryExpression.hasSubquery(cond) =>
+ throw new AnalysisException("UPDATE statements with subqueries are not currently supported")
+
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond)) if isIcebergRelation(r) =>
+ val writeInfo = newWriteInfo(r.schema)
+ val mergeBuilder = r.table.asMergeable.newMergeBuilder("update", writeInfo)
+
+ val matchingRowsPlanBuilder = scanRelation => Filter(cond, scanRelation)
+ val scanPlan = buildDynamicFilterScanPlan(spark, r.table, r.output, mergeBuilder, cond, matchingRowsPlanBuilder)
+
+ val updateProjection = buildUpdateProjection(r, scanPlan, assignments, cond)
+
+ val mergeWrite = mergeBuilder.asWriteBuilder.buildForBatch()
+ val writePlan = buildWritePlan(updateProjection, r.table)
+ ReplaceData(r, mergeWrite, writePlan)
+ }
+
+ private def buildUpdateProjection(
+ relation: DataSourceV2Relation,
+ scanPlan: LogicalPlan,
+ assignments: Seq[Assignment],
+ cond: Expression): LogicalPlan = {
+
+ // this method relies on the fact that the assignments have been aligned before
+ require(relation.output.size == assignments.size, "assignments must be aligned")
+
+ // Spark is going to execute the condition for each column but it seems we cannot avoid this
Review comment:
Any ideas whether we can avoid this are welcome. I did check the generated code and the condition was evaluated for every column.
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteUpdate.scala
##########
@@ -0,0 +1,87 @@
+/*
+ * 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.AnalysisException
+import org.apache.spark.sql.SparkSession
+import org.apache.spark.sql.catalyst.expressions.Alias
+import org.apache.spark.sql.catalyst.expressions.Expression
+import org.apache.spark.sql.catalyst.expressions.If
+import org.apache.spark.sql.catalyst.expressions.SubqueryExpression
+import org.apache.spark.sql.catalyst.plans.logical.Assignment
+import org.apache.spark.sql.catalyst.plans.logical.Filter
+import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
+import org.apache.spark.sql.catalyst.plans.logical.Project
+import org.apache.spark.sql.catalyst.plans.logical.ReplaceData
+import org.apache.spark.sql.catalyst.plans.logical.UpdateTable
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.utils.PlanUtils.isIcebergRelation
+import org.apache.spark.sql.catalyst.utils.RewriteRowLevelOperationHelper
+import org.apache.spark.sql.execution.datasources.v2.DataSourceV2Relation
+import org.apache.spark.sql.execution.datasources.v2.ExtendedDataSourceV2Implicits
+
+// TODO: should be part of early scan push down after the delete condition is optimized
+case class RewriteUpdate(spark: SparkSession) extends Rule[LogicalPlan] with RewriteRowLevelOperationHelper {
+
+ import ExtendedDataSourceV2Implicits._
+
+ // TODO: can we do any better for no-op updates? when conditions evaluate to false/true?
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond))
+ if isIcebergRelation(r) && SubqueryExpression.hasSubquery(cond) =>
+ throw new AnalysisException("UPDATE statements with subqueries are not currently supported")
+
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond)) if isIcebergRelation(r) =>
+ val writeInfo = newWriteInfo(r.schema)
+ val mergeBuilder = r.table.asMergeable.newMergeBuilder("update", writeInfo)
+
+ val matchingRowsPlanBuilder = scanRelation => Filter(cond, scanRelation)
+ val scanPlan = buildDynamicFilterScanPlan(spark, r.table, r.output, mergeBuilder, cond, matchingRowsPlanBuilder)
+
+ val updateProjection = buildUpdateProjection(r, scanPlan, assignments, cond)
+
+ val mergeWrite = mergeBuilder.asWriteBuilder.buildForBatch()
+ val writePlan = buildWritePlan(updateProjection, r.table)
+ ReplaceData(r, mergeWrite, writePlan)
+ }
+
+ private def buildUpdateProjection(
+ relation: DataSourceV2Relation,
+ scanPlan: LogicalPlan,
+ assignments: Seq[Assignment],
+ cond: Expression): LogicalPlan = {
+
+ // this method relies on the fact that the assignments have been aligned before
+ require(relation.output.size == assignments.size, "assignments must be aligned")
+
+ // Spark is going to execute the condition for each column but it seems we cannot avoid this
Review comment:
Any ideas whether we can avoid this are welcome. I did check the generated code and the condition was evaluated for every column that is subject to change.
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/utils/RewriteRowLevelOperationHelper.scala
##########
@@ -197,6 +203,22 @@ trait RewriteRowLevelOperationHelper extends PredicateHelper with Logging {
}
}
}
+
+ protected def buildWritePlan(childPlan: LogicalPlan, table: Table): LogicalPlan = {
Review comment:
Copied as it was.
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteUpdate.scala
##########
@@ -0,0 +1,87 @@
+/*
+ * 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.AnalysisException
+import org.apache.spark.sql.SparkSession
+import org.apache.spark.sql.catalyst.expressions.Alias
+import org.apache.spark.sql.catalyst.expressions.Expression
+import org.apache.spark.sql.catalyst.expressions.If
+import org.apache.spark.sql.catalyst.expressions.SubqueryExpression
+import org.apache.spark.sql.catalyst.plans.logical.Assignment
+import org.apache.spark.sql.catalyst.plans.logical.Filter
+import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
+import org.apache.spark.sql.catalyst.plans.logical.Project
+import org.apache.spark.sql.catalyst.plans.logical.ReplaceData
+import org.apache.spark.sql.catalyst.plans.logical.UpdateTable
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.utils.PlanUtils.isIcebergRelation
+import org.apache.spark.sql.catalyst.utils.RewriteRowLevelOperationHelper
+import org.apache.spark.sql.execution.datasources.v2.DataSourceV2Relation
+import org.apache.spark.sql.execution.datasources.v2.ExtendedDataSourceV2Implicits
+
+// TODO: should be part of early scan push down after the delete condition is optimized
+case class RewriteUpdate(spark: SparkSession) extends Rule[LogicalPlan] with RewriteRowLevelOperationHelper {
+
+ import ExtendedDataSourceV2Implicits._
+
+ // TODO: can we do any better for no-op updates? when conditions evaluate to false/true?
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond))
+ if isIcebergRelation(r) && SubqueryExpression.hasSubquery(cond) =>
+ throw new AnalysisException("UPDATE statements with subqueries are not currently supported")
Review comment:
Yeah, I am already looking into this. I did not want to add a condition to the branch below so that it can stay as is.
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteUpdate.scala
##########
@@ -0,0 +1,87 @@
+/*
+ * 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.AnalysisException
+import org.apache.spark.sql.SparkSession
+import org.apache.spark.sql.catalyst.expressions.Alias
+import org.apache.spark.sql.catalyst.expressions.Expression
+import org.apache.spark.sql.catalyst.expressions.If
+import org.apache.spark.sql.catalyst.expressions.SubqueryExpression
+import org.apache.spark.sql.catalyst.plans.logical.Assignment
+import org.apache.spark.sql.catalyst.plans.logical.Filter
+import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
+import org.apache.spark.sql.catalyst.plans.logical.Project
+import org.apache.spark.sql.catalyst.plans.logical.ReplaceData
+import org.apache.spark.sql.catalyst.plans.logical.UpdateTable
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.utils.PlanUtils.isIcebergRelation
+import org.apache.spark.sql.catalyst.utils.RewriteRowLevelOperationHelper
+import org.apache.spark.sql.execution.datasources.v2.DataSourceV2Relation
+import org.apache.spark.sql.execution.datasources.v2.ExtendedDataSourceV2Implicits
+
+// TODO: should be part of early scan push down after the delete condition is optimized
+case class RewriteUpdate(spark: SparkSession) extends Rule[LogicalPlan] with RewriteRowLevelOperationHelper {
+
+ import ExtendedDataSourceV2Implicits._
+
+ // TODO: can we do any better for no-op updates? when conditions evaluate to false/true?
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond))
+ if isIcebergRelation(r) && SubqueryExpression.hasSubquery(cond) =>
+ throw new AnalysisException("UPDATE statements with subqueries are not currently supported")
+
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond)) if isIcebergRelation(r) =>
+ val writeInfo = newWriteInfo(r.schema)
+ val mergeBuilder = r.table.asMergeable.newMergeBuilder("update", writeInfo)
+
+ val matchingRowsPlanBuilder = scanRelation => Filter(cond, scanRelation)
+ val scanPlan = buildDynamicFilterScanPlan(spark, r.table, r.output, mergeBuilder, cond, matchingRowsPlanBuilder)
+
+ val updateProjection = buildUpdateProjection(r, scanPlan, assignments, cond)
+
+ val mergeWrite = mergeBuilder.asWriteBuilder.buildForBatch()
+ val writePlan = buildWritePlan(updateProjection, r.table)
+ ReplaceData(r, mergeWrite, writePlan)
+ }
+
+ private def buildUpdateProjection(
+ relation: DataSourceV2Relation,
+ scanPlan: LogicalPlan,
+ assignments: Seq[Assignment],
+ cond: Expression): LogicalPlan = {
+
+ // this method relies on the fact that the assignments have been aligned before
+ require(relation.output.size == assignments.size, "assignments must be aligned")
+
+ // Spark is going to execute the condition for each column but it seems we cannot avoid this
Review comment:
You mean using `MergeInto` node? Let me add a TODO item here. I think we would need to test this out at some reasonable scale.
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/OptimizeConditionsInRowLevelOperations.scala
##########
@@ -35,6 +35,10 @@ object OptimizeConditionsInRowLevelOperations extends Rule[LogicalPlan] {
if !SubqueryExpression.hasSubquery(cond.getOrElse(Literal.TrueLiteral)) && isIcebergRelation(table) =>
val optimizedCond = optimizeCondition(cond.getOrElse(Literal.TrueLiteral), table)
d.copy(condition = Some(optimizedCond))
+ case u @ UpdateTable(table, _, cond)
Review comment:
Took another look. Not sure, actually. We need both branches to get `cond`, the main shared logic is in a separate method but I'd welcome any ideas I could overlook.
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/analysis/AlignUpdateTable.scala
##########
@@ -0,0 +1,32 @@
+/*
+ * 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.analysis
+
+import org.apache.spark.sql.catalyst.plans.logical.{LogicalPlan, UpdateTable}
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.internal.SQLConf
+
+case class AlignUpdateTable(conf: SQLConf) extends Rule[LogicalPlan] with AssignmentAlignmentSupport {
+
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case u: UpdateTable if u.resolved =>
Review comment:
Good catch, I forgot to update it.
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteUpdate.scala
##########
@@ -0,0 +1,87 @@
+/*
+ * 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.AnalysisException
+import org.apache.spark.sql.SparkSession
+import org.apache.spark.sql.catalyst.expressions.Alias
+import org.apache.spark.sql.catalyst.expressions.Expression
+import org.apache.spark.sql.catalyst.expressions.If
+import org.apache.spark.sql.catalyst.expressions.SubqueryExpression
+import org.apache.spark.sql.catalyst.plans.logical.Assignment
+import org.apache.spark.sql.catalyst.plans.logical.Filter
+import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
+import org.apache.spark.sql.catalyst.plans.logical.Project
+import org.apache.spark.sql.catalyst.plans.logical.ReplaceData
+import org.apache.spark.sql.catalyst.plans.logical.UpdateTable
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.utils.PlanUtils.isIcebergRelation
+import org.apache.spark.sql.catalyst.utils.RewriteRowLevelOperationHelper
+import org.apache.spark.sql.execution.datasources.v2.DataSourceV2Relation
+import org.apache.spark.sql.execution.datasources.v2.ExtendedDataSourceV2Implicits
+
+// TODO: should be part of early scan push down after the delete condition is optimized
+case class RewriteUpdate(spark: SparkSession) extends Rule[LogicalPlan] with RewriteRowLevelOperationHelper {
+
+ import ExtendedDataSourceV2Implicits._
+
+ // TODO: can we do any better for no-op updates? when conditions evaluate to false/true?
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond))
+ if isIcebergRelation(r) && SubqueryExpression.hasSubquery(cond) =>
+ throw new AnalysisException("UPDATE statements with subqueries are not currently supported")
+
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond)) if isIcebergRelation(r) =>
+ val writeInfo = newWriteInfo(r.schema)
+ val mergeBuilder = r.table.asMergeable.newMergeBuilder("update", writeInfo)
+
+ val matchingRowsPlanBuilder = scanRelation => Filter(cond, scanRelation)
+ val scanPlan = buildDynamicFilterScanPlan(spark, r.table, r.output, mergeBuilder, cond, matchingRowsPlanBuilder)
+
+ val updateProjection = buildUpdateProjection(r, scanPlan, assignments, cond)
+
+ val mergeWrite = mergeBuilder.asWriteBuilder.buildForBatch()
+ val writePlan = buildWritePlan(updateProjection, r.table)
+ ReplaceData(r, mergeWrite, writePlan)
+ }
+
+ private def buildUpdateProjection(
+ relation: DataSourceV2Relation,
+ scanPlan: LogicalPlan,
+ assignments: Seq[Assignment],
+ cond: Expression): LogicalPlan = {
+
+ // this method relies on the fact that the assignments have been aligned before
+ require(relation.output.size == assignments.size, "assignments must be aligned")
+
+ // Spark is going to execute the condition for each column but it seems we cannot avoid this
Review comment:
@dilipbiswal @mehtaashish23 Do you have an existing infra for benchmarks? Will it be easy for you to try this?
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/analysis/AlignUpdateTable.scala
##########
@@ -0,0 +1,32 @@
+/*
+ * 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.analysis
+
+import org.apache.spark.sql.catalyst.plans.logical.{LogicalPlan, UpdateTable}
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.internal.SQLConf
+
+case class AlignUpdateTable(conf: SQLConf) extends Rule[LogicalPlan] with AssignmentAlignmentSupport {
+
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case u: UpdateTable if u.resolved =>
Review comment:
Fixed.
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/OptimizeConditionsInRowLevelOperations.scala
##########
@@ -21,7 +21,7 @@ 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.plans.logical.{DeleteFromTable, Filter, LocalRelation, LogicalPlan, UpdateTable}
Review comment:
Fixed.
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/analysis/AlignUpdateTable.scala
##########
@@ -0,0 +1,32 @@
+/*
+ * 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.analysis
+
+import org.apache.spark.sql.catalyst.plans.logical.{LogicalPlan, UpdateTable}
Review comment:
Fixed.
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteUpdate.scala
##########
@@ -0,0 +1,87 @@
+/*
+ * 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.AnalysisException
+import org.apache.spark.sql.SparkSession
+import org.apache.spark.sql.catalyst.expressions.Alias
+import org.apache.spark.sql.catalyst.expressions.Expression
+import org.apache.spark.sql.catalyst.expressions.If
+import org.apache.spark.sql.catalyst.expressions.SubqueryExpression
+import org.apache.spark.sql.catalyst.plans.logical.Assignment
+import org.apache.spark.sql.catalyst.plans.logical.Filter
+import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
+import org.apache.spark.sql.catalyst.plans.logical.Project
+import org.apache.spark.sql.catalyst.plans.logical.ReplaceData
+import org.apache.spark.sql.catalyst.plans.logical.UpdateTable
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.utils.PlanUtils.isIcebergRelation
+import org.apache.spark.sql.catalyst.utils.RewriteRowLevelOperationHelper
+import org.apache.spark.sql.execution.datasources.v2.DataSourceV2Relation
+import org.apache.spark.sql.execution.datasources.v2.ExtendedDataSourceV2Implicits
+
+// TODO: should be part of early scan push down after the delete condition is optimized
+case class RewriteUpdate(spark: SparkSession) extends Rule[LogicalPlan] with RewriteRowLevelOperationHelper {
+
+ import ExtendedDataSourceV2Implicits._
+
+ // TODO: can we do any better for no-op updates? when conditions evaluate to false/true?
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond))
+ if isIcebergRelation(r) && SubqueryExpression.hasSubquery(cond) =>
+ throw new AnalysisException("UPDATE statements with subqueries are not currently supported")
+
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond)) if isIcebergRelation(r) =>
+ val writeInfo = newWriteInfo(r.schema)
+ val mergeBuilder = r.table.asMergeable.newMergeBuilder("update", writeInfo)
+
+ val matchingRowsPlanBuilder = scanRelation => Filter(cond, scanRelation)
+ val scanPlan = buildDynamicFilterScanPlan(spark, r.table, r.output, mergeBuilder, cond, matchingRowsPlanBuilder)
+
+ val updateProjection = buildUpdateProjection(r, scanPlan, assignments, cond)
+
+ val mergeWrite = mergeBuilder.asWriteBuilder.buildForBatch()
+ val writePlan = buildWritePlan(updateProjection, r.table)
+ ReplaceData(r, mergeWrite, writePlan)
+ }
+
+ private def buildUpdateProjection(
+ relation: DataSourceV2Relation,
+ scanPlan: LogicalPlan,
+ assignments: Seq[Assignment],
+ cond: Expression): LogicalPlan = {
+
+ // this method relies on the fact that the assignments have been aligned before
+ require(relation.output.size == assignments.size, "assignments must be aligned")
+
+ // Spark is going to execute the condition for each column but it seems we cannot avoid this
Review comment:
@rdblue, yes, I confirm your thoughts as I tried that.
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/analysis/AlignUpdateTable.scala
##########
@@ -0,0 +1,32 @@
+/*
+ * 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.analysis
+
+import org.apache.spark.sql.catalyst.plans.logical.{LogicalPlan, UpdateTable}
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.internal.SQLConf
+
+case class AlignUpdateTable(conf: SQLConf) extends Rule[LogicalPlan] with AssignmentAlignmentSupport {
+
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case u: UpdateTable if u.resolved =>
+ u.copy(assignments = alignAssignments(u.table, u.assignments))
Review comment:
I have no preference. `AlignMergeIntoTable` is way more involved but adding updates there would be alright.
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteUpdate.scala
##########
@@ -0,0 +1,87 @@
+/*
+ * 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.AnalysisException
+import org.apache.spark.sql.SparkSession
+import org.apache.spark.sql.catalyst.expressions.Alias
+import org.apache.spark.sql.catalyst.expressions.Expression
+import org.apache.spark.sql.catalyst.expressions.If
+import org.apache.spark.sql.catalyst.expressions.SubqueryExpression
+import org.apache.spark.sql.catalyst.plans.logical.Assignment
+import org.apache.spark.sql.catalyst.plans.logical.Filter
+import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
+import org.apache.spark.sql.catalyst.plans.logical.Project
+import org.apache.spark.sql.catalyst.plans.logical.ReplaceData
+import org.apache.spark.sql.catalyst.plans.logical.UpdateTable
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.utils.PlanUtils.isIcebergRelation
+import org.apache.spark.sql.catalyst.utils.RewriteRowLevelOperationHelper
+import org.apache.spark.sql.execution.datasources.v2.DataSourceV2Relation
+import org.apache.spark.sql.execution.datasources.v2.ExtendedDataSourceV2Implicits
+
+// TODO: should be part of early scan push down after the delete condition is optimized
Review comment:
Copy and paste from the delete rule, will remove.
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/utils/RewriteRowLevelOperationHelper.scala
##########
@@ -197,6 +203,22 @@ trait RewriteRowLevelOperationHelper extends PredicateHelper with Logging {
}
}
}
+
+ protected def buildWritePlan(childPlan: LogicalPlan, table: Table): LogicalPlan = {
Review comment:
I think we may consider doing that too.
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteUpdate.scala
##########
@@ -0,0 +1,87 @@
+/*
+ * 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.AnalysisException
+import org.apache.spark.sql.SparkSession
+import org.apache.spark.sql.catalyst.expressions.Alias
+import org.apache.spark.sql.catalyst.expressions.Expression
+import org.apache.spark.sql.catalyst.expressions.If
+import org.apache.spark.sql.catalyst.expressions.SubqueryExpression
+import org.apache.spark.sql.catalyst.plans.logical.Assignment
+import org.apache.spark.sql.catalyst.plans.logical.Filter
+import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
+import org.apache.spark.sql.catalyst.plans.logical.Project
+import org.apache.spark.sql.catalyst.plans.logical.ReplaceData
+import org.apache.spark.sql.catalyst.plans.logical.UpdateTable
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.utils.PlanUtils.isIcebergRelation
+import org.apache.spark.sql.catalyst.utils.RewriteRowLevelOperationHelper
+import org.apache.spark.sql.execution.datasources.v2.DataSourceV2Relation
+import org.apache.spark.sql.execution.datasources.v2.ExtendedDataSourceV2Implicits
+
+// TODO: should be part of early scan push down after the delete condition is optimized
+case class RewriteUpdate(spark: SparkSession) extends Rule[LogicalPlan] with RewriteRowLevelOperationHelper {
+
+ import ExtendedDataSourceV2Implicits._
+
+ // TODO: can we do any better for no-op updates? when conditions evaluate to false/true?
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond))
+ if isIcebergRelation(r) && SubqueryExpression.hasSubquery(cond) =>
+ throw new AnalysisException("UPDATE statements with subqueries are not currently supported")
+
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond)) if isIcebergRelation(r) =>
+ val writeInfo = newWriteInfo(r.schema)
+ val mergeBuilder = r.table.asMergeable.newMergeBuilder("update", writeInfo)
+
+ val matchingRowsPlanBuilder = scanRelation => Filter(cond, scanRelation)
+ val scanPlan = buildDynamicFilterScanPlan(spark, r.table, r.output, mergeBuilder, cond, matchingRowsPlanBuilder)
+
+ val updateProjection = buildUpdateProjection(r, scanPlan, assignments, cond)
+
+ val mergeWrite = mergeBuilder.asWriteBuilder.buildForBatch()
+ val writePlan = buildWritePlan(updateProjection, r.table)
+ ReplaceData(r, mergeWrite, writePlan)
+ }
+
+ private def buildUpdateProjection(
+ relation: DataSourceV2Relation,
+ scanPlan: LogicalPlan,
+ assignments: Seq[Assignment],
+ cond: Expression): LogicalPlan = {
+
+ // this method relies on the fact that the assignments have been aligned before
+ require(relation.output.size == assignments.size, "assignments must be aligned")
+
+ // Spark is going to execute the condition for each column but it seems we cannot avoid this
Review comment:
I meant if we have an alternate implementation to what is currently proposed, will it be possible to perform tests at some reasonable scale to see if that alternative solution works better?
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] dilipbiswal commented on pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
dilipbiswal commented on pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#issuecomment-771495594
@aokolnychyi Looks good to me.
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] mehtaashish23 commented on a change in pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
mehtaashish23 commented on a change in pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#discussion_r568382200
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/analysis/AlignUpdateTable.scala
##########
@@ -0,0 +1,32 @@
+/*
+ * 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.analysis
+
+import org.apache.spark.sql.catalyst.plans.logical.{LogicalPlan, UpdateTable}
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.internal.SQLConf
+
+case class AlignUpdateTable(conf: SQLConf) extends Rule[LogicalPlan] with AssignmentAlignmentSupport {
+
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case u: UpdateTable if u.resolved =>
Review comment:
You want to put the existing check i.e. to not Align, if targetTable is not an Iceberg table?
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] aokolnychyi commented on a change in pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
aokolnychyi commented on a change in pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#discussion_r568310060
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteDelete.scala
##########
@@ -55,8 +54,6 @@ case class RewriteDelete(spark: SparkSession) extends Rule[LogicalPlan] with Rew
import ExtendedDataSourceV2Implicits._
import RewriteRowLevelOperationHelper._
- override def resolver: Resolver = spark.sessionState.conf.resolver
Review comment:
Moved this to parent as I need both `conf` and `resolver` now.
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteMergeInto.scala
##########
@@ -224,22 +216,6 @@ case class RewriteMergeInto(spark: SparkSession) extends Rule[LogicalPlan] with
}
!(actions.size == 1 && hasUnconditionalDelete(actions.headOption))
}
-
- private def buildWritePlan(childPlan: LogicalPlan, table: Table): LogicalPlan = {
Review comment:
Moved to parent.
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] aokolnychyi commented on pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
aokolnychyi commented on pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#issuecomment-771342449
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] aokolnychyi commented on a change in pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
aokolnychyi commented on a change in pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#discussion_r568310632
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteUpdate.scala
##########
@@ -0,0 +1,87 @@
+/*
+ * 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.AnalysisException
+import org.apache.spark.sql.SparkSession
+import org.apache.spark.sql.catalyst.expressions.Alias
+import org.apache.spark.sql.catalyst.expressions.Expression
+import org.apache.spark.sql.catalyst.expressions.If
+import org.apache.spark.sql.catalyst.expressions.SubqueryExpression
+import org.apache.spark.sql.catalyst.plans.logical.Assignment
+import org.apache.spark.sql.catalyst.plans.logical.Filter
+import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
+import org.apache.spark.sql.catalyst.plans.logical.Project
+import org.apache.spark.sql.catalyst.plans.logical.ReplaceData
+import org.apache.spark.sql.catalyst.plans.logical.UpdateTable
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.utils.PlanUtils.isIcebergRelation
+import org.apache.spark.sql.catalyst.utils.RewriteRowLevelOperationHelper
+import org.apache.spark.sql.execution.datasources.v2.DataSourceV2Relation
+import org.apache.spark.sql.execution.datasources.v2.ExtendedDataSourceV2Implicits
+
+// TODO: should be part of early scan push down after the delete condition is optimized
+case class RewriteUpdate(spark: SparkSession) extends Rule[LogicalPlan] with RewriteRowLevelOperationHelper {
+
+ import ExtendedDataSourceV2Implicits._
+
+ // TODO: can we do any better for no-op updates? when conditions evaluate to false/true?
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond))
+ if isIcebergRelation(r) && SubqueryExpression.hasSubquery(cond) =>
+ throw new AnalysisException("UPDATE statements with subqueries are not currently supported")
+
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond)) if isIcebergRelation(r) =>
+ val writeInfo = newWriteInfo(r.schema)
+ val mergeBuilder = r.table.asMergeable.newMergeBuilder("update", writeInfo)
+
+ val matchingRowsPlanBuilder = scanRelation => Filter(cond, scanRelation)
+ val scanPlan = buildDynamicFilterScanPlan(spark, r.table, r.output, mergeBuilder, cond, matchingRowsPlanBuilder)
+
+ val updateProjection = buildUpdateProjection(r, scanPlan, assignments, cond)
+
+ val mergeWrite = mergeBuilder.asWriteBuilder.buildForBatch()
+ val writePlan = buildWritePlan(updateProjection, r.table)
+ ReplaceData(r, mergeWrite, writePlan)
+ }
+
+ private def buildUpdateProjection(
+ relation: DataSourceV2Relation,
+ scanPlan: LogicalPlan,
+ assignments: Seq[Assignment],
+ cond: Expression): LogicalPlan = {
+
+ // this method relies on the fact that the assignments have been aligned before
+ require(relation.output.size == assignments.size, "assignments must be aligned")
+
+ // Spark is going to execute the condition for each column but it seems we cannot avoid this
Review comment:
Any ideas whether we can avoid this are welcome. I did check the generated code and the condition was evaluated for every column that is subject to change.
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] rdblue commented on pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
rdblue commented on pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#issuecomment-771946663
I see no major blockers with this. There are a few things to consider, but overall I think this can go in and we can improve it from there.
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] aokolnychyi commented on a change in pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
aokolnychyi commented on a change in pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#discussion_r568888315
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteUpdate.scala
##########
@@ -0,0 +1,87 @@
+/*
+ * 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.AnalysisException
+import org.apache.spark.sql.SparkSession
+import org.apache.spark.sql.catalyst.expressions.Alias
+import org.apache.spark.sql.catalyst.expressions.Expression
+import org.apache.spark.sql.catalyst.expressions.If
+import org.apache.spark.sql.catalyst.expressions.SubqueryExpression
+import org.apache.spark.sql.catalyst.plans.logical.Assignment
+import org.apache.spark.sql.catalyst.plans.logical.Filter
+import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
+import org.apache.spark.sql.catalyst.plans.logical.Project
+import org.apache.spark.sql.catalyst.plans.logical.ReplaceData
+import org.apache.spark.sql.catalyst.plans.logical.UpdateTable
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.utils.PlanUtils.isIcebergRelation
+import org.apache.spark.sql.catalyst.utils.RewriteRowLevelOperationHelper
+import org.apache.spark.sql.execution.datasources.v2.DataSourceV2Relation
+import org.apache.spark.sql.execution.datasources.v2.ExtendedDataSourceV2Implicits
+
+// TODO: should be part of early scan push down after the delete condition is optimized
+case class RewriteUpdate(spark: SparkSession) extends Rule[LogicalPlan] with RewriteRowLevelOperationHelper {
+
+ import ExtendedDataSourceV2Implicits._
+
+ // TODO: can we do any better for no-op updates? when conditions evaluate to false/true?
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond))
+ if isIcebergRelation(r) && SubqueryExpression.hasSubquery(cond) =>
+ throw new AnalysisException("UPDATE statements with subqueries are not currently supported")
+
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond)) if isIcebergRelation(r) =>
+ val writeInfo = newWriteInfo(r.schema)
+ val mergeBuilder = r.table.asMergeable.newMergeBuilder("update", writeInfo)
+
+ val matchingRowsPlanBuilder = scanRelation => Filter(cond, scanRelation)
+ val scanPlan = buildDynamicFilterScanPlan(spark, r.table, r.output, mergeBuilder, cond, matchingRowsPlanBuilder)
+
+ val updateProjection = buildUpdateProjection(r, scanPlan, assignments, cond)
+
+ val mergeWrite = mergeBuilder.asWriteBuilder.buildForBatch()
+ val writePlan = buildWritePlan(updateProjection, r.table)
+ ReplaceData(r, mergeWrite, writePlan)
+ }
+
+ private def buildUpdateProjection(
+ relation: DataSourceV2Relation,
+ scanPlan: LogicalPlan,
+ assignments: Seq[Assignment],
+ cond: Expression): LogicalPlan = {
+
+ // this method relies on the fact that the assignments have been aligned before
+ require(relation.output.size == assignments.size, "assignments must be aligned")
+
+ // Spark is going to execute the condition for each column but it seems we cannot avoid this
Review comment:
@rdblue, yes, I confirm your thoughts as I tried that.
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] aokolnychyi commented on a change in pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
aokolnychyi commented on a change in pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#discussion_r569102648
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/analysis/AlignUpdateTable.scala
##########
@@ -0,0 +1,32 @@
+/*
+ * 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.analysis
+
+import org.apache.spark.sql.catalyst.plans.logical.{LogicalPlan, UpdateTable}
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.internal.SQLConf
+
+case class AlignUpdateTable(conf: SQLConf) extends Rule[LogicalPlan] with AssignmentAlignmentSupport {
+
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case u: UpdateTable if u.resolved =>
+ u.copy(assignments = alignAssignments(u.table, u.assignments))
Review comment:
Done.
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] aokolnychyi commented on a change in pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
aokolnychyi commented on a change in pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#discussion_r569102983
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteUpdate.scala
##########
@@ -0,0 +1,87 @@
+/*
+ * 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.AnalysisException
+import org.apache.spark.sql.SparkSession
+import org.apache.spark.sql.catalyst.expressions.Alias
+import org.apache.spark.sql.catalyst.expressions.Expression
+import org.apache.spark.sql.catalyst.expressions.If
+import org.apache.spark.sql.catalyst.expressions.SubqueryExpression
+import org.apache.spark.sql.catalyst.plans.logical.Assignment
+import org.apache.spark.sql.catalyst.plans.logical.Filter
+import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
+import org.apache.spark.sql.catalyst.plans.logical.Project
+import org.apache.spark.sql.catalyst.plans.logical.ReplaceData
+import org.apache.spark.sql.catalyst.plans.logical.UpdateTable
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.utils.PlanUtils.isIcebergRelation
+import org.apache.spark.sql.catalyst.utils.RewriteRowLevelOperationHelper
+import org.apache.spark.sql.execution.datasources.v2.DataSourceV2Relation
+import org.apache.spark.sql.execution.datasources.v2.ExtendedDataSourceV2Implicits
+
+// TODO: should be part of early scan push down after the delete condition is optimized
+case class RewriteUpdate(spark: SparkSession) extends Rule[LogicalPlan] with RewriteRowLevelOperationHelper {
+
+ import ExtendedDataSourceV2Implicits._
+
+ // TODO: can we do any better for no-op updates? when conditions evaluate to false/true?
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond))
+ if isIcebergRelation(r) && SubqueryExpression.hasSubquery(cond) =>
+ throw new AnalysisException("UPDATE statements with subqueries are not currently supported")
+
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond)) if isIcebergRelation(r) =>
+ val writeInfo = newWriteInfo(r.schema)
+ val mergeBuilder = r.table.asMergeable.newMergeBuilder("update", writeInfo)
+
+ val matchingRowsPlanBuilder = scanRelation => Filter(cond, scanRelation)
+ val scanPlan = buildDynamicFilterScanPlan(spark, r.table, r.output, mergeBuilder, cond, matchingRowsPlanBuilder)
+
+ val updateProjection = buildUpdateProjection(r, scanPlan, assignments, cond)
+
+ val mergeWrite = mergeBuilder.asWriteBuilder.buildForBatch()
+ val writePlan = buildWritePlan(updateProjection, r.table)
+ ReplaceData(r, mergeWrite, writePlan)
+ }
+
+ private def buildUpdateProjection(
+ relation: DataSourceV2Relation,
+ scanPlan: LogicalPlan,
+ assignments: Seq[Assignment],
+ cond: Expression): LogicalPlan = {
+
+ // this method relies on the fact that the assignments have been aligned before
+ require(relation.output.size == assignments.size, "assignments must be aligned")
+
+ // Spark is going to execute the condition for each column but it seems we cannot avoid this
+ val assignedExprs = assignments.map(_.value)
+ val updatedExprs = assignedExprs.zip(relation.output).map { case (assignedExpr, attr) =>
+ // use semanticEquals to avoid unnecessary if expressions as we may run after operator optimization
+ val updatedExpr = if (attr.semanticEquals(assignedExpr)) {
+ attr
+ } else {
+ If(cond, assignedExpr, attr)
+ }
+ Alias(updatedExpr, attr.name)()
+ }
+ Project(updatedExprs, scanPlan)
Review comment:
Done.
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] rdblue commented on a change in pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
rdblue commented on a change in pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#discussion_r568878203
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/OptimizeConditionsInRowLevelOperations.scala
##########
@@ -35,6 +35,10 @@ object OptimizeConditionsInRowLevelOperations extends Rule[LogicalPlan] {
if !SubqueryExpression.hasSubquery(cond.getOrElse(Literal.TrueLiteral)) && isIcebergRelation(table) =>
val optimizedCond = optimizeCondition(cond.getOrElse(Literal.TrueLiteral), table)
d.copy(condition = Some(optimizedCond))
+ case u @ UpdateTable(table, _, cond)
Review comment:
This is short enough that I think it's okay.
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] aokolnychyi commented on a change in pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
aokolnychyi commented on a change in pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#discussion_r568310756
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/utils/RewriteRowLevelOperationHelper.scala
##########
@@ -197,6 +203,22 @@ trait RewriteRowLevelOperationHelper extends PredicateHelper with Logging {
}
}
}
+
+ protected def buildWritePlan(childPlan: LogicalPlan, table: Table): LogicalPlan = {
Review comment:
Copied as it was.
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] aokolnychyi commented on a change in pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
aokolnychyi commented on a change in pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#discussion_r569102838
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteUpdate.scala
##########
@@ -0,0 +1,87 @@
+/*
+ * 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.AnalysisException
+import org.apache.spark.sql.SparkSession
+import org.apache.spark.sql.catalyst.expressions.Alias
+import org.apache.spark.sql.catalyst.expressions.Expression
+import org.apache.spark.sql.catalyst.expressions.If
+import org.apache.spark.sql.catalyst.expressions.SubqueryExpression
+import org.apache.spark.sql.catalyst.plans.logical.Assignment
+import org.apache.spark.sql.catalyst.plans.logical.Filter
+import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
+import org.apache.spark.sql.catalyst.plans.logical.Project
+import org.apache.spark.sql.catalyst.plans.logical.ReplaceData
+import org.apache.spark.sql.catalyst.plans.logical.UpdateTable
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.utils.PlanUtils.isIcebergRelation
+import org.apache.spark.sql.catalyst.utils.RewriteRowLevelOperationHelper
+import org.apache.spark.sql.execution.datasources.v2.DataSourceV2Relation
+import org.apache.spark.sql.execution.datasources.v2.ExtendedDataSourceV2Implicits
+
+// TODO: should be part of early scan push down after the delete condition is optimized
Review comment:
Fixed.
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] aokolnychyi commented on a change in pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
aokolnychyi commented on a change in pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#discussion_r568840738
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/OptimizeConditionsInRowLevelOperations.scala
##########
@@ -35,6 +35,10 @@ object OptimizeConditionsInRowLevelOperations extends Rule[LogicalPlan] {
if !SubqueryExpression.hasSubquery(cond.getOrElse(Literal.TrueLiteral)) && isIcebergRelation(table) =>
val optimizedCond = optimizeCondition(cond.getOrElse(Literal.TrueLiteral), table)
d.copy(condition = Some(optimizedCond))
+ case u @ UpdateTable(table, _, cond)
Review comment:
Took another look. Not sure, actually. We need both branches to get `cond`, the main shared logic is in a separate method but I'd welcome any ideas I could overlook.
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] aokolnychyi commented on a change in pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
aokolnychyi commented on a change in pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#discussion_r568828696
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteUpdate.scala
##########
@@ -0,0 +1,87 @@
+/*
+ * 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.AnalysisException
+import org.apache.spark.sql.SparkSession
+import org.apache.spark.sql.catalyst.expressions.Alias
+import org.apache.spark.sql.catalyst.expressions.Expression
+import org.apache.spark.sql.catalyst.expressions.If
+import org.apache.spark.sql.catalyst.expressions.SubqueryExpression
+import org.apache.spark.sql.catalyst.plans.logical.Assignment
+import org.apache.spark.sql.catalyst.plans.logical.Filter
+import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
+import org.apache.spark.sql.catalyst.plans.logical.Project
+import org.apache.spark.sql.catalyst.plans.logical.ReplaceData
+import org.apache.spark.sql.catalyst.plans.logical.UpdateTable
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.utils.PlanUtils.isIcebergRelation
+import org.apache.spark.sql.catalyst.utils.RewriteRowLevelOperationHelper
+import org.apache.spark.sql.execution.datasources.v2.DataSourceV2Relation
+import org.apache.spark.sql.execution.datasources.v2.ExtendedDataSourceV2Implicits
+
+// TODO: should be part of early scan push down after the delete condition is optimized
+case class RewriteUpdate(spark: SparkSession) extends Rule[LogicalPlan] with RewriteRowLevelOperationHelper {
+
+ import ExtendedDataSourceV2Implicits._
+
+ // TODO: can we do any better for no-op updates? when conditions evaluate to false/true?
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond))
+ if isIcebergRelation(r) && SubqueryExpression.hasSubquery(cond) =>
+ throw new AnalysisException("UPDATE statements with subqueries are not currently supported")
+
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond)) if isIcebergRelation(r) =>
+ val writeInfo = newWriteInfo(r.schema)
+ val mergeBuilder = r.table.asMergeable.newMergeBuilder("update", writeInfo)
+
+ val matchingRowsPlanBuilder = scanRelation => Filter(cond, scanRelation)
+ val scanPlan = buildDynamicFilterScanPlan(spark, r.table, r.output, mergeBuilder, cond, matchingRowsPlanBuilder)
+
+ val updateProjection = buildUpdateProjection(r, scanPlan, assignments, cond)
+
+ val mergeWrite = mergeBuilder.asWriteBuilder.buildForBatch()
+ val writePlan = buildWritePlan(updateProjection, r.table)
+ ReplaceData(r, mergeWrite, writePlan)
+ }
+
+ private def buildUpdateProjection(
+ relation: DataSourceV2Relation,
+ scanPlan: LogicalPlan,
+ assignments: Seq[Assignment],
+ cond: Expression): LogicalPlan = {
+
+ // this method relies on the fact that the assignments have been aligned before
+ require(relation.output.size == assignments.size, "assignments must be aligned")
+
+ // Spark is going to execute the condition for each column but it seems we cannot avoid this
Review comment:
You mean using `MergeInto` node? Let me add a TODO item here. I think we would need to test this out at some reasonable scale.
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] rdblue commented on a change in pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
rdblue commented on a change in pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#discussion_r568897443
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/analysis/AlignUpdateTable.scala
##########
@@ -0,0 +1,32 @@
+/*
+ * 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.analysis
+
+import org.apache.spark.sql.catalyst.plans.logical.{LogicalPlan, UpdateTable}
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.internal.SQLConf
+
+case class AlignUpdateTable(conf: SQLConf) extends Rule[LogicalPlan] with AssignmentAlignmentSupport {
+
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case u: UpdateTable if u.resolved =>
+ u.copy(assignments = alignAssignments(u.table, u.assignments))
Review comment:
One fewer class/file seems like a good thing.
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] aokolnychyi commented on a change in pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
aokolnychyi commented on a change in pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#discussion_r568843955
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteUpdate.scala
##########
@@ -0,0 +1,87 @@
+/*
+ * 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.AnalysisException
+import org.apache.spark.sql.SparkSession
+import org.apache.spark.sql.catalyst.expressions.Alias
+import org.apache.spark.sql.catalyst.expressions.Expression
+import org.apache.spark.sql.catalyst.expressions.If
+import org.apache.spark.sql.catalyst.expressions.SubqueryExpression
+import org.apache.spark.sql.catalyst.plans.logical.Assignment
+import org.apache.spark.sql.catalyst.plans.logical.Filter
+import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
+import org.apache.spark.sql.catalyst.plans.logical.Project
+import org.apache.spark.sql.catalyst.plans.logical.ReplaceData
+import org.apache.spark.sql.catalyst.plans.logical.UpdateTable
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.utils.PlanUtils.isIcebergRelation
+import org.apache.spark.sql.catalyst.utils.RewriteRowLevelOperationHelper
+import org.apache.spark.sql.execution.datasources.v2.DataSourceV2Relation
+import org.apache.spark.sql.execution.datasources.v2.ExtendedDataSourceV2Implicits
+
+// TODO: should be part of early scan push down after the delete condition is optimized
+case class RewriteUpdate(spark: SparkSession) extends Rule[LogicalPlan] with RewriteRowLevelOperationHelper {
+
+ import ExtendedDataSourceV2Implicits._
+
+ // TODO: can we do any better for no-op updates? when conditions evaluate to false/true?
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond))
+ if isIcebergRelation(r) && SubqueryExpression.hasSubquery(cond) =>
+ throw new AnalysisException("UPDATE statements with subqueries are not currently supported")
+
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond)) if isIcebergRelation(r) =>
+ val writeInfo = newWriteInfo(r.schema)
+ val mergeBuilder = r.table.asMergeable.newMergeBuilder("update", writeInfo)
+
+ val matchingRowsPlanBuilder = scanRelation => Filter(cond, scanRelation)
+ val scanPlan = buildDynamicFilterScanPlan(spark, r.table, r.output, mergeBuilder, cond, matchingRowsPlanBuilder)
+
+ val updateProjection = buildUpdateProjection(r, scanPlan, assignments, cond)
+
+ val mergeWrite = mergeBuilder.asWriteBuilder.buildForBatch()
+ val writePlan = buildWritePlan(updateProjection, r.table)
+ ReplaceData(r, mergeWrite, writePlan)
+ }
+
+ private def buildUpdateProjection(
+ relation: DataSourceV2Relation,
+ scanPlan: LogicalPlan,
+ assignments: Seq[Assignment],
+ cond: Expression): LogicalPlan = {
+
+ // this method relies on the fact that the assignments have been aligned before
+ require(relation.output.size == assignments.size, "assignments must be aligned")
+
+ // Spark is going to execute the condition for each column but it seems we cannot avoid this
Review comment:
@dilipbiswal @mehtaashish23 Do you have an existing infra for benchmarks? Will it be easy for you to try this?
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] mehtaashish23 commented on a change in pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
mehtaashish23 commented on a change in pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#discussion_r568924173
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteUpdate.scala
##########
@@ -0,0 +1,87 @@
+/*
+ * 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.AnalysisException
+import org.apache.spark.sql.SparkSession
+import org.apache.spark.sql.catalyst.expressions.Alias
+import org.apache.spark.sql.catalyst.expressions.Expression
+import org.apache.spark.sql.catalyst.expressions.If
+import org.apache.spark.sql.catalyst.expressions.SubqueryExpression
+import org.apache.spark.sql.catalyst.plans.logical.Assignment
+import org.apache.spark.sql.catalyst.plans.logical.Filter
+import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
+import org.apache.spark.sql.catalyst.plans.logical.Project
+import org.apache.spark.sql.catalyst.plans.logical.ReplaceData
+import org.apache.spark.sql.catalyst.plans.logical.UpdateTable
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.utils.PlanUtils.isIcebergRelation
+import org.apache.spark.sql.catalyst.utils.RewriteRowLevelOperationHelper
+import org.apache.spark.sql.execution.datasources.v2.DataSourceV2Relation
+import org.apache.spark.sql.execution.datasources.v2.ExtendedDataSourceV2Implicits
+
+// TODO: should be part of early scan push down after the delete condition is optimized
+case class RewriteUpdate(spark: SparkSession) extends Rule[LogicalPlan] with RewriteRowLevelOperationHelper {
+
+ import ExtendedDataSourceV2Implicits._
+
+ // TODO: can we do any better for no-op updates? when conditions evaluate to false/true?
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond))
+ if isIcebergRelation(r) && SubqueryExpression.hasSubquery(cond) =>
+ throw new AnalysisException("UPDATE statements with subqueries are not currently supported")
+
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond)) if isIcebergRelation(r) =>
+ val writeInfo = newWriteInfo(r.schema)
+ val mergeBuilder = r.table.asMergeable.newMergeBuilder("update", writeInfo)
+
+ val matchingRowsPlanBuilder = scanRelation => Filter(cond, scanRelation)
+ val scanPlan = buildDynamicFilterScanPlan(spark, r.table, r.output, mergeBuilder, cond, matchingRowsPlanBuilder)
+
+ val updateProjection = buildUpdateProjection(r, scanPlan, assignments, cond)
+
+ val mergeWrite = mergeBuilder.asWriteBuilder.buildForBatch()
+ val writePlan = buildWritePlan(updateProjection, r.table)
+ ReplaceData(r, mergeWrite, writePlan)
+ }
+
+ private def buildUpdateProjection(
+ relation: DataSourceV2Relation,
+ scanPlan: LogicalPlan,
+ assignments: Seq[Assignment],
+ cond: Expression): LogicalPlan = {
+
+ // this method relies on the fact that the assignments have been aligned before
+ require(relation.output.size == assignments.size, "assignments must be aligned")
+
+ // Spark is going to execute the condition for each column but it seems we cannot avoid this
Review comment:
@aokolnychyi You mean running the same set of data with
1. All updates merged via "Merge Into"
2. All updates updated via "Update table"
And see the benchmarks? I wonder that considering MergeInto will do a join, irrespective of input, so I am not sure whether the Filter expression can be compared with a shuffle of data. I hope I am getting the expectation right.
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] aokolnychyi commented on a change in pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
aokolnychyi commented on a change in pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#discussion_r568309937
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/analysis/AlignUpdateTable.scala
##########
@@ -0,0 +1,32 @@
+/*
+ * 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.analysis
+
+import org.apache.spark.sql.catalyst.plans.logical.{LogicalPlan, UpdateTable}
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.internal.SQLConf
+
+case class AlignUpdateTable(conf: SQLConf) extends Rule[LogicalPlan] with AssignmentAlignmentSupport {
+
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case u: UpdateTable if u.resolved =>
+ u.copy(assignments = alignAssignments(u.table, u.assignments))
Review comment:
We apply the same logic as for MERGE.
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] dilipbiswal commented on a change in pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
dilipbiswal commented on a change in pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#discussion_r568374762
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/analysis/AlignUpdateTable.scala
##########
@@ -0,0 +1,32 @@
+/*
+ * 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.analysis
+
+import org.apache.spark.sql.catalyst.plans.logical.{LogicalPlan, UpdateTable}
Review comment:
import separate lines ?
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/OptimizeConditionsInRowLevelOperations.scala
##########
@@ -35,6 +35,10 @@ object OptimizeConditionsInRowLevelOperations extends Rule[LogicalPlan] {
if !SubqueryExpression.hasSubquery(cond.getOrElse(Literal.TrueLiteral)) && isIcebergRelation(table) =>
val optimizedCond = optimizeCondition(cond.getOrElse(Literal.TrueLiteral), table)
d.copy(condition = Some(optimizedCond))
+ case u @ UpdateTable(table, _, cond)
Review comment:
@aokolnychyi Both the cases have same actions. Is there a way we can keep it in once place ?
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/OptimizeConditionsInRowLevelOperations.scala
##########
@@ -21,7 +21,7 @@ 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.plans.logical.{DeleteFromTable, Filter, LocalRelation, LogicalPlan, UpdateTable}
Review comment:
imports in separate lines ?
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteUpdate.scala
##########
@@ -0,0 +1,87 @@
+/*
+ * 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.AnalysisException
+import org.apache.spark.sql.SparkSession
+import org.apache.spark.sql.catalyst.expressions.Alias
+import org.apache.spark.sql.catalyst.expressions.Expression
+import org.apache.spark.sql.catalyst.expressions.If
+import org.apache.spark.sql.catalyst.expressions.SubqueryExpression
+import org.apache.spark.sql.catalyst.plans.logical.Assignment
+import org.apache.spark.sql.catalyst.plans.logical.Filter
+import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
+import org.apache.spark.sql.catalyst.plans.logical.Project
+import org.apache.spark.sql.catalyst.plans.logical.ReplaceData
+import org.apache.spark.sql.catalyst.plans.logical.UpdateTable
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.utils.PlanUtils.isIcebergRelation
+import org.apache.spark.sql.catalyst.utils.RewriteRowLevelOperationHelper
+import org.apache.spark.sql.execution.datasources.v2.DataSourceV2Relation
+import org.apache.spark.sql.execution.datasources.v2.ExtendedDataSourceV2Implicits
+
+// TODO: should be part of early scan push down after the delete condition is optimized
+case class RewriteUpdate(spark: SparkSession) extends Rule[LogicalPlan] with RewriteRowLevelOperationHelper {
+
+ import ExtendedDataSourceV2Implicits._
+
+ // TODO: can we do any better for no-op updates? when conditions evaluate to false/true?
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond))
+ if isIcebergRelation(r) && SubqueryExpression.hasSubquery(cond) =>
+ throw new AnalysisException("UPDATE statements with subqueries are not currently supported")
Review comment:
@aokolnychyi We are keeping it here and not in checkAnalysis as we are going to open it up in follow-up ?
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteUpdate.scala
##########
@@ -0,0 +1,87 @@
+/*
+ * 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.AnalysisException
+import org.apache.spark.sql.SparkSession
+import org.apache.spark.sql.catalyst.expressions.Alias
+import org.apache.spark.sql.catalyst.expressions.Expression
+import org.apache.spark.sql.catalyst.expressions.If
+import org.apache.spark.sql.catalyst.expressions.SubqueryExpression
+import org.apache.spark.sql.catalyst.plans.logical.Assignment
+import org.apache.spark.sql.catalyst.plans.logical.Filter
+import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
+import org.apache.spark.sql.catalyst.plans.logical.Project
+import org.apache.spark.sql.catalyst.plans.logical.ReplaceData
+import org.apache.spark.sql.catalyst.plans.logical.UpdateTable
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.utils.PlanUtils.isIcebergRelation
+import org.apache.spark.sql.catalyst.utils.RewriteRowLevelOperationHelper
+import org.apache.spark.sql.execution.datasources.v2.DataSourceV2Relation
+import org.apache.spark.sql.execution.datasources.v2.ExtendedDataSourceV2Implicits
+
+// TODO: should be part of early scan push down after the delete condition is optimized
+case class RewriteUpdate(spark: SparkSession) extends Rule[LogicalPlan] with RewriteRowLevelOperationHelper {
+
+ import ExtendedDataSourceV2Implicits._
+
+ // TODO: can we do any better for no-op updates? when conditions evaluate to false/true?
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond))
+ if isIcebergRelation(r) && SubqueryExpression.hasSubquery(cond) =>
+ throw new AnalysisException("UPDATE statements with subqueries are not currently supported")
+
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond)) if isIcebergRelation(r) =>
+ val writeInfo = newWriteInfo(r.schema)
+ val mergeBuilder = r.table.asMergeable.newMergeBuilder("update", writeInfo)
+
+ val matchingRowsPlanBuilder = scanRelation => Filter(cond, scanRelation)
+ val scanPlan = buildDynamicFilterScanPlan(spark, r.table, r.output, mergeBuilder, cond, matchingRowsPlanBuilder)
+
+ val updateProjection = buildUpdateProjection(r, scanPlan, assignments, cond)
+
+ val mergeWrite = mergeBuilder.asWriteBuilder.buildForBatch()
+ val writePlan = buildWritePlan(updateProjection, r.table)
+ ReplaceData(r, mergeWrite, writePlan)
+ }
+
+ private def buildUpdateProjection(
+ relation: DataSourceV2Relation,
+ scanPlan: LogicalPlan,
+ assignments: Seq[Assignment],
+ cond: Expression): LogicalPlan = {
+
+ // this method relies on the fact that the assignments have been aligned before
+ require(relation.output.size == assignments.size, "assignments must be aligned")
+
+ // Spark is going to execute the condition for each column but it seems we cannot avoid this
Review comment:
@aokolnychyi I guess, if we can try to look at this in the future. I think there is not a good way to know the cost of the condition expression at the time of rewrite. So not sure if an alternate plan would do better in all condition, wdyt ?
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] rdblue commented on a change in pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
rdblue commented on a change in pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#discussion_r568884683
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteUpdate.scala
##########
@@ -0,0 +1,87 @@
+/*
+ * 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.AnalysisException
+import org.apache.spark.sql.SparkSession
+import org.apache.spark.sql.catalyst.expressions.Alias
+import org.apache.spark.sql.catalyst.expressions.Expression
+import org.apache.spark.sql.catalyst.expressions.If
+import org.apache.spark.sql.catalyst.expressions.SubqueryExpression
+import org.apache.spark.sql.catalyst.plans.logical.Assignment
+import org.apache.spark.sql.catalyst.plans.logical.Filter
+import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
+import org.apache.spark.sql.catalyst.plans.logical.Project
+import org.apache.spark.sql.catalyst.plans.logical.ReplaceData
+import org.apache.spark.sql.catalyst.plans.logical.UpdateTable
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.utils.PlanUtils.isIcebergRelation
+import org.apache.spark.sql.catalyst.utils.RewriteRowLevelOperationHelper
+import org.apache.spark.sql.execution.datasources.v2.DataSourceV2Relation
+import org.apache.spark.sql.execution.datasources.v2.ExtendedDataSourceV2Implicits
+
+// TODO: should be part of early scan push down after the delete condition is optimized
+case class RewriteUpdate(spark: SparkSession) extends Rule[LogicalPlan] with RewriteRowLevelOperationHelper {
+
+ import ExtendedDataSourceV2Implicits._
+
+ // TODO: can we do any better for no-op updates? when conditions evaluate to false/true?
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond))
+ if isIcebergRelation(r) && SubqueryExpression.hasSubquery(cond) =>
+ throw new AnalysisException("UPDATE statements with subqueries are not currently supported")
+
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond)) if isIcebergRelation(r) =>
+ val writeInfo = newWriteInfo(r.schema)
+ val mergeBuilder = r.table.asMergeable.newMergeBuilder("update", writeInfo)
+
+ val matchingRowsPlanBuilder = scanRelation => Filter(cond, scanRelation)
+ val scanPlan = buildDynamicFilterScanPlan(spark, r.table, r.output, mergeBuilder, cond, matchingRowsPlanBuilder)
+
+ val updateProjection = buildUpdateProjection(r, scanPlan, assignments, cond)
+
+ val mergeWrite = mergeBuilder.asWriteBuilder.buildForBatch()
+ val writePlan = buildWritePlan(updateProjection, r.table)
+ ReplaceData(r, mergeWrite, writePlan)
+ }
+
+ private def buildUpdateProjection(
+ relation: DataSourceV2Relation,
+ scanPlan: LogicalPlan,
+ assignments: Seq[Assignment],
+ cond: Expression): LogicalPlan = {
+
+ // this method relies on the fact that the assignments have been aligned before
+ require(relation.output.size == assignments.size, "assignments must be aligned")
+
+ // Spark is going to execute the condition for each column but it seems we cannot avoid this
+ val assignedExprs = assignments.map(_.value)
+ val updatedExprs = assignedExprs.zip(relation.output).map { case (assignedExpr, attr) =>
+ // use semanticEquals to avoid unnecessary if expressions as we may run after operator optimization
+ val updatedExpr = if (attr.semanticEquals(assignedExpr)) {
+ attr
+ } else {
+ If(cond, assignedExpr, attr)
+ }
+ Alias(updatedExpr, attr.name)()
+ }
+ Project(updatedExprs, scanPlan)
Review comment:
Nit: white space after control flow blocks.
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] aokolnychyi commented on a change in pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
aokolnychyi commented on a change in pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#discussion_r568827563
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteUpdate.scala
##########
@@ -0,0 +1,87 @@
+/*
+ * 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.AnalysisException
+import org.apache.spark.sql.SparkSession
+import org.apache.spark.sql.catalyst.expressions.Alias
+import org.apache.spark.sql.catalyst.expressions.Expression
+import org.apache.spark.sql.catalyst.expressions.If
+import org.apache.spark.sql.catalyst.expressions.SubqueryExpression
+import org.apache.spark.sql.catalyst.plans.logical.Assignment
+import org.apache.spark.sql.catalyst.plans.logical.Filter
+import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
+import org.apache.spark.sql.catalyst.plans.logical.Project
+import org.apache.spark.sql.catalyst.plans.logical.ReplaceData
+import org.apache.spark.sql.catalyst.plans.logical.UpdateTable
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.utils.PlanUtils.isIcebergRelation
+import org.apache.spark.sql.catalyst.utils.RewriteRowLevelOperationHelper
+import org.apache.spark.sql.execution.datasources.v2.DataSourceV2Relation
+import org.apache.spark.sql.execution.datasources.v2.ExtendedDataSourceV2Implicits
+
+// TODO: should be part of early scan push down after the delete condition is optimized
+case class RewriteUpdate(spark: SparkSession) extends Rule[LogicalPlan] with RewriteRowLevelOperationHelper {
+
+ import ExtendedDataSourceV2Implicits._
+
+ // TODO: can we do any better for no-op updates? when conditions evaluate to false/true?
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond))
+ if isIcebergRelation(r) && SubqueryExpression.hasSubquery(cond) =>
+ throw new AnalysisException("UPDATE statements with subqueries are not currently supported")
Review comment:
Yeah, I am already looking into this. I did not want to add a condition to the branch below so that it can stay as is.
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] dilipbiswal commented on pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
dilipbiswal commented on pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#issuecomment-771495594
@aokolnychyi Looks good to me.
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] aokolnychyi edited a comment on pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
aokolnychyi edited a comment on pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#issuecomment-771342449
cc @dilipbiswal @rdblue @RussellSpitzer @mehtaashish23
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] rdblue commented on a change in pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
rdblue commented on a change in pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#discussion_r568877523
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/analysis/AlignUpdateTable.scala
##########
@@ -0,0 +1,32 @@
+/*
+ * 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.analysis
+
+import org.apache.spark.sql.catalyst.plans.logical.{LogicalPlan, UpdateTable}
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.internal.SQLConf
+
+case class AlignUpdateTable(conf: SQLConf) extends Rule[LogicalPlan] with AssignmentAlignmentSupport {
+
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case u: UpdateTable if u.resolved =>
+ u.copy(assignments = alignAssignments(u.table, u.assignments))
Review comment:
Should we merge the rules into one then? What is the value of separate rules for Update and Merge?
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] rdblue commented on a change in pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
rdblue commented on a change in pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#discussion_r568884496
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteUpdate.scala
##########
@@ -0,0 +1,87 @@
+/*
+ * 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.AnalysisException
+import org.apache.spark.sql.SparkSession
+import org.apache.spark.sql.catalyst.expressions.Alias
+import org.apache.spark.sql.catalyst.expressions.Expression
+import org.apache.spark.sql.catalyst.expressions.If
+import org.apache.spark.sql.catalyst.expressions.SubqueryExpression
+import org.apache.spark.sql.catalyst.plans.logical.Assignment
+import org.apache.spark.sql.catalyst.plans.logical.Filter
+import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
+import org.apache.spark.sql.catalyst.plans.logical.Project
+import org.apache.spark.sql.catalyst.plans.logical.ReplaceData
+import org.apache.spark.sql.catalyst.plans.logical.UpdateTable
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.utils.PlanUtils.isIcebergRelation
+import org.apache.spark.sql.catalyst.utils.RewriteRowLevelOperationHelper
+import org.apache.spark.sql.execution.datasources.v2.DataSourceV2Relation
+import org.apache.spark.sql.execution.datasources.v2.ExtendedDataSourceV2Implicits
+
+// TODO: should be part of early scan push down after the delete condition is optimized
+case class RewriteUpdate(spark: SparkSession) extends Rule[LogicalPlan] with RewriteRowLevelOperationHelper {
+
+ import ExtendedDataSourceV2Implicits._
+
+ // TODO: can we do any better for no-op updates? when conditions evaluate to false/true?
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond))
+ if isIcebergRelation(r) && SubqueryExpression.hasSubquery(cond) =>
+ throw new AnalysisException("UPDATE statements with subqueries are not currently supported")
+
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond)) if isIcebergRelation(r) =>
+ val writeInfo = newWriteInfo(r.schema)
+ val mergeBuilder = r.table.asMergeable.newMergeBuilder("update", writeInfo)
+
+ val matchingRowsPlanBuilder = scanRelation => Filter(cond, scanRelation)
+ val scanPlan = buildDynamicFilterScanPlan(spark, r.table, r.output, mergeBuilder, cond, matchingRowsPlanBuilder)
+
+ val updateProjection = buildUpdateProjection(r, scanPlan, assignments, cond)
+
+ val mergeWrite = mergeBuilder.asWriteBuilder.buildForBatch()
+ val writePlan = buildWritePlan(updateProjection, r.table)
+ ReplaceData(r, mergeWrite, writePlan)
+ }
+
+ private def buildUpdateProjection(
+ relation: DataSourceV2Relation,
+ scanPlan: LogicalPlan,
+ assignments: Seq[Assignment],
+ cond: Expression): LogicalPlan = {
+
+ // this method relies on the fact that the assignments have been aligned before
+ require(relation.output.size == assignments.size, "assignments must be aligned")
+
+ // Spark is going to execute the condition for each column but it seems we cannot avoid this
Review comment:
Yeah, we could use a custom node like `MergeIntoExec` that runs the expression and then performs the right projection. I think if we tried to use two projections, the optimizer would collapse them and rewrite everything equivalent to this, where the condition runs for each column.
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] aokolnychyi commented on a change in pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
aokolnychyi commented on a change in pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#discussion_r569102769
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteUpdate.scala
##########
@@ -0,0 +1,87 @@
+/*
+ * 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.AnalysisException
+import org.apache.spark.sql.SparkSession
+import org.apache.spark.sql.catalyst.expressions.Alias
+import org.apache.spark.sql.catalyst.expressions.Expression
+import org.apache.spark.sql.catalyst.expressions.If
+import org.apache.spark.sql.catalyst.expressions.SubqueryExpression
+import org.apache.spark.sql.catalyst.plans.logical.Assignment
+import org.apache.spark.sql.catalyst.plans.logical.Filter
+import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
+import org.apache.spark.sql.catalyst.plans.logical.Project
+import org.apache.spark.sql.catalyst.plans.logical.ReplaceData
+import org.apache.spark.sql.catalyst.plans.logical.UpdateTable
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.utils.PlanUtils.isIcebergRelation
+import org.apache.spark.sql.catalyst.utils.RewriteRowLevelOperationHelper
+import org.apache.spark.sql.execution.datasources.v2.DataSourceV2Relation
+import org.apache.spark.sql.execution.datasources.v2.ExtendedDataSourceV2Implicits
+
+// TODO: should be part of early scan push down after the delete condition is optimized
+case class RewriteUpdate(spark: SparkSession) extends Rule[LogicalPlan] with RewriteRowLevelOperationHelper {
+
+ import ExtendedDataSourceV2Implicits._
+
+ // TODO: can we do any better for no-op updates? when conditions evaluate to false/true?
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond))
+ if isIcebergRelation(r) && SubqueryExpression.hasSubquery(cond) =>
+ throw new AnalysisException("UPDATE statements with subqueries are not currently supported")
+
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond)) if isIcebergRelation(r) =>
+ val writeInfo = newWriteInfo(r.schema)
+ val mergeBuilder = r.table.asMergeable.newMergeBuilder("update", writeInfo)
+
+ val matchingRowsPlanBuilder = scanRelation => Filter(cond, scanRelation)
+ val scanPlan = buildDynamicFilterScanPlan(spark, r.table, r.output, mergeBuilder, cond, matchingRowsPlanBuilder)
+
+ val updateProjection = buildUpdateProjection(r, scanPlan, assignments, cond)
+
+ val mergeWrite = mergeBuilder.asWriteBuilder.buildForBatch()
+ val writePlan = buildWritePlan(updateProjection, r.table)
+ ReplaceData(r, mergeWrite, writePlan)
+ }
+
+ private def buildUpdateProjection(
+ relation: DataSourceV2Relation,
+ scanPlan: LogicalPlan,
+ assignments: Seq[Assignment],
+ cond: Expression): LogicalPlan = {
+
+ // this method relies on the fact that the assignments have been aligned before
+ require(relation.output.size == assignments.size, "assignments must be aligned")
+
+ // Spark is going to execute the condition for each column but it seems we cannot avoid this
Review comment:
Sounds good. I'll cc once done.
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] rdblue commented on a change in pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
rdblue commented on a change in pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#discussion_r568877523
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/analysis/AlignUpdateTable.scala
##########
@@ -0,0 +1,32 @@
+/*
+ * 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.analysis
+
+import org.apache.spark.sql.catalyst.plans.logical.{LogicalPlan, UpdateTable}
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.internal.SQLConf
+
+case class AlignUpdateTable(conf: SQLConf) extends Rule[LogicalPlan] with AssignmentAlignmentSupport {
+
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case u: UpdateTable if u.resolved =>
+ u.copy(assignments = alignAssignments(u.table, u.assignments))
Review comment:
Should we merge the rules into one then? What is the value of separate rules for Update and Merge?
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/OptimizeConditionsInRowLevelOperations.scala
##########
@@ -35,6 +35,10 @@ object OptimizeConditionsInRowLevelOperations extends Rule[LogicalPlan] {
if !SubqueryExpression.hasSubquery(cond.getOrElse(Literal.TrueLiteral)) && isIcebergRelation(table) =>
val optimizedCond = optimizeCondition(cond.getOrElse(Literal.TrueLiteral), table)
d.copy(condition = Some(optimizedCond))
+ case u @ UpdateTable(table, _, cond)
Review comment:
This is short enough that I think it's okay.
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteUpdate.scala
##########
@@ -0,0 +1,87 @@
+/*
+ * 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.AnalysisException
+import org.apache.spark.sql.SparkSession
+import org.apache.spark.sql.catalyst.expressions.Alias
+import org.apache.spark.sql.catalyst.expressions.Expression
+import org.apache.spark.sql.catalyst.expressions.If
+import org.apache.spark.sql.catalyst.expressions.SubqueryExpression
+import org.apache.spark.sql.catalyst.plans.logical.Assignment
+import org.apache.spark.sql.catalyst.plans.logical.Filter
+import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
+import org.apache.spark.sql.catalyst.plans.logical.Project
+import org.apache.spark.sql.catalyst.plans.logical.ReplaceData
+import org.apache.spark.sql.catalyst.plans.logical.UpdateTable
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.utils.PlanUtils.isIcebergRelation
+import org.apache.spark.sql.catalyst.utils.RewriteRowLevelOperationHelper
+import org.apache.spark.sql.execution.datasources.v2.DataSourceV2Relation
+import org.apache.spark.sql.execution.datasources.v2.ExtendedDataSourceV2Implicits
+
+// TODO: should be part of early scan push down after the delete condition is optimized
Review comment:
I'm not sure about this. I thought we wanted to explore moving these into the analyzer?
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteUpdate.scala
##########
@@ -0,0 +1,87 @@
+/*
+ * 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.AnalysisException
+import org.apache.spark.sql.SparkSession
+import org.apache.spark.sql.catalyst.expressions.Alias
+import org.apache.spark.sql.catalyst.expressions.Expression
+import org.apache.spark.sql.catalyst.expressions.If
+import org.apache.spark.sql.catalyst.expressions.SubqueryExpression
+import org.apache.spark.sql.catalyst.plans.logical.Assignment
+import org.apache.spark.sql.catalyst.plans.logical.Filter
+import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
+import org.apache.spark.sql.catalyst.plans.logical.Project
+import org.apache.spark.sql.catalyst.plans.logical.ReplaceData
+import org.apache.spark.sql.catalyst.plans.logical.UpdateTable
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.utils.PlanUtils.isIcebergRelation
+import org.apache.spark.sql.catalyst.utils.RewriteRowLevelOperationHelper
+import org.apache.spark.sql.execution.datasources.v2.DataSourceV2Relation
+import org.apache.spark.sql.execution.datasources.v2.ExtendedDataSourceV2Implicits
+
+// TODO: should be part of early scan push down after the delete condition is optimized
+case class RewriteUpdate(spark: SparkSession) extends Rule[LogicalPlan] with RewriteRowLevelOperationHelper {
+
+ import ExtendedDataSourceV2Implicits._
+
+ // TODO: can we do any better for no-op updates? when conditions evaluate to false/true?
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond))
+ if isIcebergRelation(r) && SubqueryExpression.hasSubquery(cond) =>
+ throw new AnalysisException("UPDATE statements with subqueries are not currently supported")
+
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond)) if isIcebergRelation(r) =>
+ val writeInfo = newWriteInfo(r.schema)
+ val mergeBuilder = r.table.asMergeable.newMergeBuilder("update", writeInfo)
+
+ val matchingRowsPlanBuilder = scanRelation => Filter(cond, scanRelation)
+ val scanPlan = buildDynamicFilterScanPlan(spark, r.table, r.output, mergeBuilder, cond, matchingRowsPlanBuilder)
+
+ val updateProjection = buildUpdateProjection(r, scanPlan, assignments, cond)
+
+ val mergeWrite = mergeBuilder.asWriteBuilder.buildForBatch()
+ val writePlan = buildWritePlan(updateProjection, r.table)
+ ReplaceData(r, mergeWrite, writePlan)
+ }
+
+ private def buildUpdateProjection(
+ relation: DataSourceV2Relation,
+ scanPlan: LogicalPlan,
+ assignments: Seq[Assignment],
+ cond: Expression): LogicalPlan = {
+
+ // this method relies on the fact that the assignments have been aligned before
+ require(relation.output.size == assignments.size, "assignments must be aligned")
+
+ // Spark is going to execute the condition for each column but it seems we cannot avoid this
Review comment:
Yeah, we could use a custom node like `MergeIntoExec` that runs the expression and then performs the right projection. I think if we tried to use two projections, the optimizer would collapse them and rewrite everything equivalent to this, where the condition runs for each column.
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteUpdate.scala
##########
@@ -0,0 +1,87 @@
+/*
+ * 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.AnalysisException
+import org.apache.spark.sql.SparkSession
+import org.apache.spark.sql.catalyst.expressions.Alias
+import org.apache.spark.sql.catalyst.expressions.Expression
+import org.apache.spark.sql.catalyst.expressions.If
+import org.apache.spark.sql.catalyst.expressions.SubqueryExpression
+import org.apache.spark.sql.catalyst.plans.logical.Assignment
+import org.apache.spark.sql.catalyst.plans.logical.Filter
+import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
+import org.apache.spark.sql.catalyst.plans.logical.Project
+import org.apache.spark.sql.catalyst.plans.logical.ReplaceData
+import org.apache.spark.sql.catalyst.plans.logical.UpdateTable
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.utils.PlanUtils.isIcebergRelation
+import org.apache.spark.sql.catalyst.utils.RewriteRowLevelOperationHelper
+import org.apache.spark.sql.execution.datasources.v2.DataSourceV2Relation
+import org.apache.spark.sql.execution.datasources.v2.ExtendedDataSourceV2Implicits
+
+// TODO: should be part of early scan push down after the delete condition is optimized
+case class RewriteUpdate(spark: SparkSession) extends Rule[LogicalPlan] with RewriteRowLevelOperationHelper {
+
+ import ExtendedDataSourceV2Implicits._
+
+ // TODO: can we do any better for no-op updates? when conditions evaluate to false/true?
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond))
+ if isIcebergRelation(r) && SubqueryExpression.hasSubquery(cond) =>
+ throw new AnalysisException("UPDATE statements with subqueries are not currently supported")
+
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond)) if isIcebergRelation(r) =>
+ val writeInfo = newWriteInfo(r.schema)
+ val mergeBuilder = r.table.asMergeable.newMergeBuilder("update", writeInfo)
+
+ val matchingRowsPlanBuilder = scanRelation => Filter(cond, scanRelation)
+ val scanPlan = buildDynamicFilterScanPlan(spark, r.table, r.output, mergeBuilder, cond, matchingRowsPlanBuilder)
+
+ val updateProjection = buildUpdateProjection(r, scanPlan, assignments, cond)
+
+ val mergeWrite = mergeBuilder.asWriteBuilder.buildForBatch()
+ val writePlan = buildWritePlan(updateProjection, r.table)
+ ReplaceData(r, mergeWrite, writePlan)
+ }
+
+ private def buildUpdateProjection(
+ relation: DataSourceV2Relation,
+ scanPlan: LogicalPlan,
+ assignments: Seq[Assignment],
+ cond: Expression): LogicalPlan = {
+
+ // this method relies on the fact that the assignments have been aligned before
+ require(relation.output.size == assignments.size, "assignments must be aligned")
+
+ // Spark is going to execute the condition for each column but it seems we cannot avoid this
+ val assignedExprs = assignments.map(_.value)
+ val updatedExprs = assignedExprs.zip(relation.output).map { case (assignedExpr, attr) =>
+ // use semanticEquals to avoid unnecessary if expressions as we may run after operator optimization
+ val updatedExpr = if (attr.semanticEquals(assignedExpr)) {
+ attr
+ } else {
+ If(cond, assignedExpr, attr)
+ }
+ Alias(updatedExpr, attr.name)()
+ }
+ Project(updatedExprs, scanPlan)
Review comment:
Nit: white space after control flow blocks.
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/utils/RewriteRowLevelOperationHelper.scala
##########
@@ -197,6 +203,22 @@ trait RewriteRowLevelOperationHelper extends PredicateHelper with Logging {
}
}
}
+
+ protected def buildWritePlan(childPlan: LogicalPlan, table: Table): LogicalPlan = {
Review comment:
For a follow up, should we add a rule that does this for Append and Overwrite plans as well? That would be nice so that we don't have to wait until 3.2.0 to get write distribution and ordering for normal writes.
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/analysis/AlignUpdateTable.scala
##########
@@ -0,0 +1,32 @@
+/*
+ * 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.analysis
+
+import org.apache.spark.sql.catalyst.plans.logical.{LogicalPlan, UpdateTable}
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.internal.SQLConf
+
+case class AlignUpdateTable(conf: SQLConf) extends Rule[LogicalPlan] with AssignmentAlignmentSupport {
+
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case u: UpdateTable if u.resolved =>
+ u.copy(assignments = alignAssignments(u.table, u.assignments))
Review comment:
One fewer class/file seems like a good thing.
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] mehtaashish23 commented on a change in pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
mehtaashish23 commented on a change in pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#discussion_r568924173
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteUpdate.scala
##########
@@ -0,0 +1,87 @@
+/*
+ * 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.AnalysisException
+import org.apache.spark.sql.SparkSession
+import org.apache.spark.sql.catalyst.expressions.Alias
+import org.apache.spark.sql.catalyst.expressions.Expression
+import org.apache.spark.sql.catalyst.expressions.If
+import org.apache.spark.sql.catalyst.expressions.SubqueryExpression
+import org.apache.spark.sql.catalyst.plans.logical.Assignment
+import org.apache.spark.sql.catalyst.plans.logical.Filter
+import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
+import org.apache.spark.sql.catalyst.plans.logical.Project
+import org.apache.spark.sql.catalyst.plans.logical.ReplaceData
+import org.apache.spark.sql.catalyst.plans.logical.UpdateTable
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.utils.PlanUtils.isIcebergRelation
+import org.apache.spark.sql.catalyst.utils.RewriteRowLevelOperationHelper
+import org.apache.spark.sql.execution.datasources.v2.DataSourceV2Relation
+import org.apache.spark.sql.execution.datasources.v2.ExtendedDataSourceV2Implicits
+
+// TODO: should be part of early scan push down after the delete condition is optimized
+case class RewriteUpdate(spark: SparkSession) extends Rule[LogicalPlan] with RewriteRowLevelOperationHelper {
+
+ import ExtendedDataSourceV2Implicits._
+
+ // TODO: can we do any better for no-op updates? when conditions evaluate to false/true?
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond))
+ if isIcebergRelation(r) && SubqueryExpression.hasSubquery(cond) =>
+ throw new AnalysisException("UPDATE statements with subqueries are not currently supported")
+
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond)) if isIcebergRelation(r) =>
+ val writeInfo = newWriteInfo(r.schema)
+ val mergeBuilder = r.table.asMergeable.newMergeBuilder("update", writeInfo)
+
+ val matchingRowsPlanBuilder = scanRelation => Filter(cond, scanRelation)
+ val scanPlan = buildDynamicFilterScanPlan(spark, r.table, r.output, mergeBuilder, cond, matchingRowsPlanBuilder)
+
+ val updateProjection = buildUpdateProjection(r, scanPlan, assignments, cond)
+
+ val mergeWrite = mergeBuilder.asWriteBuilder.buildForBatch()
+ val writePlan = buildWritePlan(updateProjection, r.table)
+ ReplaceData(r, mergeWrite, writePlan)
+ }
+
+ private def buildUpdateProjection(
+ relation: DataSourceV2Relation,
+ scanPlan: LogicalPlan,
+ assignments: Seq[Assignment],
+ cond: Expression): LogicalPlan = {
+
+ // this method relies on the fact that the assignments have been aligned before
+ require(relation.output.size == assignments.size, "assignments must be aligned")
+
+ // Spark is going to execute the condition for each column but it seems we cannot avoid this
Review comment:
@aokolnychyi You mean running the same set of data with
1. All updates merged via "Merge Into"
2. All updates updated via "Update table"
And see the benchmarks? I wonder that considering MergeInto will do a join, irrespective of input, so I am not sure whether the If expression can be compared with a shuffle of data. I hope I am getting the expectation right.
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
[GitHub] [iceberg] dilipbiswal commented on a change in pull request #2193: Spark: Support UPDATE statements without subqueries
Posted by GitBox <gi...@apache.org>.
dilipbiswal commented on a change in pull request #2193:
URL: https://github.com/apache/iceberg/pull/2193#discussion_r568374762
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/analysis/AlignUpdateTable.scala
##########
@@ -0,0 +1,32 @@
+/*
+ * 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.analysis
+
+import org.apache.spark.sql.catalyst.plans.logical.{LogicalPlan, UpdateTable}
Review comment:
import separate lines ?
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/OptimizeConditionsInRowLevelOperations.scala
##########
@@ -35,6 +35,10 @@ object OptimizeConditionsInRowLevelOperations extends Rule[LogicalPlan] {
if !SubqueryExpression.hasSubquery(cond.getOrElse(Literal.TrueLiteral)) && isIcebergRelation(table) =>
val optimizedCond = optimizeCondition(cond.getOrElse(Literal.TrueLiteral), table)
d.copy(condition = Some(optimizedCond))
+ case u @ UpdateTable(table, _, cond)
Review comment:
@aokolnychyi Both the cases have same actions. Is there a way we can keep it in once place ?
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/OptimizeConditionsInRowLevelOperations.scala
##########
@@ -21,7 +21,7 @@ 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.plans.logical.{DeleteFromTable, Filter, LocalRelation, LogicalPlan, UpdateTable}
Review comment:
imports in separate lines ?
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteUpdate.scala
##########
@@ -0,0 +1,87 @@
+/*
+ * 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.AnalysisException
+import org.apache.spark.sql.SparkSession
+import org.apache.spark.sql.catalyst.expressions.Alias
+import org.apache.spark.sql.catalyst.expressions.Expression
+import org.apache.spark.sql.catalyst.expressions.If
+import org.apache.spark.sql.catalyst.expressions.SubqueryExpression
+import org.apache.spark.sql.catalyst.plans.logical.Assignment
+import org.apache.spark.sql.catalyst.plans.logical.Filter
+import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
+import org.apache.spark.sql.catalyst.plans.logical.Project
+import org.apache.spark.sql.catalyst.plans.logical.ReplaceData
+import org.apache.spark.sql.catalyst.plans.logical.UpdateTable
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.utils.PlanUtils.isIcebergRelation
+import org.apache.spark.sql.catalyst.utils.RewriteRowLevelOperationHelper
+import org.apache.spark.sql.execution.datasources.v2.DataSourceV2Relation
+import org.apache.spark.sql.execution.datasources.v2.ExtendedDataSourceV2Implicits
+
+// TODO: should be part of early scan push down after the delete condition is optimized
+case class RewriteUpdate(spark: SparkSession) extends Rule[LogicalPlan] with RewriteRowLevelOperationHelper {
+
+ import ExtendedDataSourceV2Implicits._
+
+ // TODO: can we do any better for no-op updates? when conditions evaluate to false/true?
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond))
+ if isIcebergRelation(r) && SubqueryExpression.hasSubquery(cond) =>
+ throw new AnalysisException("UPDATE statements with subqueries are not currently supported")
Review comment:
@aokolnychyi We are keeping it here and not in checkAnalysis as we are going to open it up in follow-up ?
##########
File path: spark3-extensions/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteUpdate.scala
##########
@@ -0,0 +1,87 @@
+/*
+ * 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.AnalysisException
+import org.apache.spark.sql.SparkSession
+import org.apache.spark.sql.catalyst.expressions.Alias
+import org.apache.spark.sql.catalyst.expressions.Expression
+import org.apache.spark.sql.catalyst.expressions.If
+import org.apache.spark.sql.catalyst.expressions.SubqueryExpression
+import org.apache.spark.sql.catalyst.plans.logical.Assignment
+import org.apache.spark.sql.catalyst.plans.logical.Filter
+import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
+import org.apache.spark.sql.catalyst.plans.logical.Project
+import org.apache.spark.sql.catalyst.plans.logical.ReplaceData
+import org.apache.spark.sql.catalyst.plans.logical.UpdateTable
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.utils.PlanUtils.isIcebergRelation
+import org.apache.spark.sql.catalyst.utils.RewriteRowLevelOperationHelper
+import org.apache.spark.sql.execution.datasources.v2.DataSourceV2Relation
+import org.apache.spark.sql.execution.datasources.v2.ExtendedDataSourceV2Implicits
+
+// TODO: should be part of early scan push down after the delete condition is optimized
+case class RewriteUpdate(spark: SparkSession) extends Rule[LogicalPlan] with RewriteRowLevelOperationHelper {
+
+ import ExtendedDataSourceV2Implicits._
+
+ // TODO: can we do any better for no-op updates? when conditions evaluate to false/true?
+ override def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond))
+ if isIcebergRelation(r) && SubqueryExpression.hasSubquery(cond) =>
+ throw new AnalysisException("UPDATE statements with subqueries are not currently supported")
+
+ case UpdateTable(r: DataSourceV2Relation, assignments, Some(cond)) if isIcebergRelation(r) =>
+ val writeInfo = newWriteInfo(r.schema)
+ val mergeBuilder = r.table.asMergeable.newMergeBuilder("update", writeInfo)
+
+ val matchingRowsPlanBuilder = scanRelation => Filter(cond, scanRelation)
+ val scanPlan = buildDynamicFilterScanPlan(spark, r.table, r.output, mergeBuilder, cond, matchingRowsPlanBuilder)
+
+ val updateProjection = buildUpdateProjection(r, scanPlan, assignments, cond)
+
+ val mergeWrite = mergeBuilder.asWriteBuilder.buildForBatch()
+ val writePlan = buildWritePlan(updateProjection, r.table)
+ ReplaceData(r, mergeWrite, writePlan)
+ }
+
+ private def buildUpdateProjection(
+ relation: DataSourceV2Relation,
+ scanPlan: LogicalPlan,
+ assignments: Seq[Assignment],
+ cond: Expression): LogicalPlan = {
+
+ // this method relies on the fact that the assignments have been aligned before
+ require(relation.output.size == assignments.size, "assignments must be aligned")
+
+ // Spark is going to execute the condition for each column but it seems we cannot avoid this
Review comment:
@aokolnychyi I guess, if we can try to look at this in the future. I think there is not a good way to know the cost of the condition expression at the time of rewrite. So not sure if an alternate plan would do better in all condition, wdyt ?
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org