You are viewing a plain text version of this content. The canonical link for it is here.
Posted to reviews@spark.apache.org by "HyukjinKwon (via GitHub)" <gi...@apache.org> on 2023/03/16 09:19:50 UTC
[GitHub] [spark] HyukjinKwon opened a new pull request, #40456: [SPARK-42720][PS][SQL] Uses expression for distributed-sequence default index instead of plan
HyukjinKwon opened a new pull request, #40456:
URL: https://github.com/apache/spark/pull/40456
### What changes were proposed in this pull request?
This PR replaces `DataFrame.withSequenceColumn` to `DataFrame.select(distributed_sequence_column, col("*")` internally because this essentially attaches a column and it should be treated as a scalar expression at the logical level.
This is used to generate the unique index only for pandas API on Spark.
### Why are the changes needed?
For better readability of codes, and for cleaner definition of Spark Connect protobuf message, see also https://github.com/apache/spark/pull/40270.
### Does this PR introduce _any_ user-facing change?
No, it's internal change only.
### How was this patch tested?
Existing test cases in pandas API on Spark verify this 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.
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org
[GitHub] [spark] cloud-fan commented on a diff in pull request #40456: [SPARK-42720][PS][SQL] Uses expression for distributed-sequence default index instead of plan
Posted by "cloud-fan (via GitHub)" <gi...@apache.org>.
cloud-fan commented on code in PR #40456:
URL: https://github.com/apache/spark/pull/40456#discussion_r1139667580
##########
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/ExtractDistributedSequenceID.scala:
##########
@@ -0,0 +1,42 @@
+/*
+ * 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.expressions.{AttributeReference, DistributedSequenceID}
+import org.apache.spark.sql.catalyst.plans.logical.{AttachDistributedSequence, LogicalPlan}
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.trees.TreePattern.DISTRIBUTED_SEQUENCE_ID
+import org.apache.spark.sql.types.LongType
+
+/**
+ * Extracts [[DistributedSequenceID]] in logical plans, and replace it to
+ * [[AttachDistributedSequence]] because this expressions requires a shuffle
+ * to generate a sequence that needs the context of the whole data, e.g.,
+ * [[org.apache.spark.rdd.RDD.zipWithIndex]].
+ */
+object ExtractDistributedSequenceID extends Rule[LogicalPlan] {
+ def apply(plan: LogicalPlan): LogicalPlan = {
+ plan.resolveOperatorsUpWithPruning(_.containsPattern(DISTRIBUTED_SEQUENCE_ID)) {
+ case plan: LogicalPlan if
+ plan.expressions.exists(_.exists(_.isInstanceOf[DistributedSequenceID])) =>
Review Comment:
Does it mean we can put `DistributedSequenceID` in any plan now? Then we should be more defensive here. We need to add a new project to project away the extra id column if it was not included in `plan.output`
think about `Filter(id > 5, child)`, the final plan should be `Project(child.output, Filter(id > 5, AttachDistributedSequence(child)))`
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org
[GitHub] [spark] HyukjinKwon commented on a diff in pull request #40456: [SPARK-42720][PS][SQL] Uses expression for distributed-sequence default index instead of plan
Posted by "HyukjinKwon (via GitHub)" <gi...@apache.org>.
HyukjinKwon commented on code in PR #40456:
URL: https://github.com/apache/spark/pull/40456#discussion_r1140080892
##########
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/ExtractDistributedSequenceID.scala:
##########
@@ -0,0 +1,42 @@
+/*
+ * 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.expressions.{AttributeReference, DistributedSequenceID}
+import org.apache.spark.sql.catalyst.plans.logical.{AttachDistributedSequence, LogicalPlan}
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.trees.TreePattern.DISTRIBUTED_SEQUENCE_ID
+import org.apache.spark.sql.types.LongType
+
+/**
+ * Extracts [[DistributedSequenceID]] in logical plans, and replace it to
+ * [[AttachDistributedSequence]] because this expressions requires a shuffle
+ * to generate a sequence that needs the context of the whole data, e.g.,
+ * [[org.apache.spark.rdd.RDD.zipWithIndex]].
+ */
+object ExtractDistributedSequenceID extends Rule[LogicalPlan] {
+ def apply(plan: LogicalPlan): LogicalPlan = {
+ plan.resolveOperatorsUpWithPruning(_.containsPattern(DISTRIBUTED_SEQUENCE_ID)) {
+ case plan: LogicalPlan if
+ plan.expressions.exists(_.exists(_.isInstanceOf[DistributedSequenceID])) =>
Review Comment:
This actually is only used with projection always .. but let me fix it like to make it future prooven.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org
[GitHub] [spark] HyukjinKwon commented on pull request #40456: [SPARK-42720][PS][SQL] Uses expression for distributed-sequence default index instead of plan
Posted by "HyukjinKwon (via GitHub)" <gi...@apache.org>.
HyukjinKwon commented on PR #40456:
URL: https://github.com/apache/spark/pull/40456#issuecomment-1471582823
cc @cloud-fan @hvanhovell FYI
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org
[GitHub] [spark] HyukjinKwon closed pull request #40456: [SPARK-42720][PS][SQL] Uses expression for distributed-sequence default index instead of plan
Posted by "HyukjinKwon (via GitHub)" <gi...@apache.org>.
HyukjinKwon closed pull request #40456: [SPARK-42720][PS][SQL] Uses expression for distributed-sequence default index instead of plan
URL: https://github.com/apache/spark/pull/40456
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org
[GitHub] [spark] cloud-fan commented on a diff in pull request #40456: [SPARK-42720][PS][SQL] Uses expression for distributed-sequence default index instead of plan
Posted by "cloud-fan (via GitHub)" <gi...@apache.org>.
cloud-fan commented on code in PR #40456:
URL: https://github.com/apache/spark/pull/40456#discussion_r1140294460
##########
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/ExtractDistributedSequenceID.scala:
##########
@@ -0,0 +1,43 @@
+/*
+ * 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.expressions.{AttributeReference, DistributedSequenceID}
+import org.apache.spark.sql.catalyst.plans.logical.{AttachDistributedSequence, LogicalPlan, Project}
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.trees.TreePattern.DISTRIBUTED_SEQUENCE_ID
+import org.apache.spark.sql.types.LongType
+
+/**
+ * Extracts [[DistributedSequenceID]] in logical plans, and replace it to
+ * [[AttachDistributedSequence]] because this expressions requires a shuffle
+ * to generate a sequence that needs the context of the whole data, e.g.,
+ * [[org.apache.spark.rdd.RDD.zipWithIndex]].
+ */
+object ExtractDistributedSequenceID extends Rule[LogicalPlan] {
+ def apply(plan: LogicalPlan): LogicalPlan = {
+ plan.resolveOperatorsUpWithPruning(_.containsPattern(DISTRIBUTED_SEQUENCE_ID)) {
+ case plan: LogicalPlan if
+ plan.expressions.exists(_.exists(_.isInstanceOf[DistributedSequenceID])) =>
+ val attr = AttributeReference("distributed_sequence_id", LongType, nullable = false)()
+ val newPlan = plan.withNewChildren(plan.children.map(AttachDistributedSequence(attr, _)))
+ .transformExpressions { case _: DistributedSequenceID => attr }
+ Project(plan.output, newPlan)
Review Comment:
We can only call `plan.outout` if `plan` is resolved.
##########
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/ExtractDistributedSequenceID.scala:
##########
@@ -0,0 +1,43 @@
+/*
+ * 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.expressions.{AttributeReference, DistributedSequenceID}
+import org.apache.spark.sql.catalyst.plans.logical.{AttachDistributedSequence, LogicalPlan, Project}
+import org.apache.spark.sql.catalyst.rules.Rule
+import org.apache.spark.sql.catalyst.trees.TreePattern.DISTRIBUTED_SEQUENCE_ID
+import org.apache.spark.sql.types.LongType
+
+/**
+ * Extracts [[DistributedSequenceID]] in logical plans, and replace it to
+ * [[AttachDistributedSequence]] because this expressions requires a shuffle
+ * to generate a sequence that needs the context of the whole data, e.g.,
+ * [[org.apache.spark.rdd.RDD.zipWithIndex]].
+ */
+object ExtractDistributedSequenceID extends Rule[LogicalPlan] {
+ def apply(plan: LogicalPlan): LogicalPlan = {
+ plan.resolveOperatorsUpWithPruning(_.containsPattern(DISTRIBUTED_SEQUENCE_ID)) {
+ case plan: LogicalPlan if
Review Comment:
```suggestion
case plan: LogicalPlan if plan.resolved &&
```
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org
[GitHub] [spark] zhengruifeng commented on a diff in pull request #40456: [SPARK-42720][PS][SQL] Uses expression for distributed-sequence default index instead of plan
Posted by "zhengruifeng (via GitHub)" <gi...@apache.org>.
zhengruifeng commented on code in PR #40456:
URL: https://github.com/apache/spark/pull/40456#discussion_r1138370203
##########
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/expressions/DistributedSequenceID.scala:
##########
@@ -0,0 +1,42 @@
+/*
+ * 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.expressions
+
+import org.apache.spark.sql.catalyst.trees.TreePattern.{DISTRIBUTED_SEQUENCE_ID, TreePattern}
+import org.apache.spark.sql.types.{DataType, LongType}
+
+/**
+ * Returns increasing 64-bit integers consecutive from 0.
+ * The generated ID is guaranteed to be increasing consecutive started from 0.
+ *
+ * @note this expression is dedicated for Pandas API on Spark to use.
+ */
+case class DistributedSequenceID() extends LeafExpression with Unevaluable with NonSQLExpression {
Review Comment:
~~will it be `Nondeterministic` like `SparkPartitionID` and `MonotonicallyIncreasingID`?~~
the underlying `AttachDistributedSequenceExec` directly relies on `RDD.zipWithIndex`, so no need to be `MonotonicallyIncreasingID`
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org
[GitHub] [spark] HyukjinKwon commented on pull request #40456: [SPARK-42720][PS][SQL] Uses expression for distributed-sequence default index instead of plan
Posted by "HyukjinKwon (via GitHub)" <gi...@apache.org>.
HyukjinKwon commented on PR #40456:
URL: https://github.com/apache/spark/pull/40456#issuecomment-1476121551
Merged to master.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org
[GitHub] [spark] zhengruifeng commented on a diff in pull request #40456: [SPARK-42720][PS][SQL] Uses expression for distributed-sequence default index instead of plan
Posted by "zhengruifeng (via GitHub)" <gi...@apache.org>.
zhengruifeng commented on code in PR #40456:
URL: https://github.com/apache/spark/pull/40456#discussion_r1138370203
##########
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/expressions/DistributedSequenceID.scala:
##########
@@ -0,0 +1,42 @@
+/*
+ * 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.expressions
+
+import org.apache.spark.sql.catalyst.trees.TreePattern.{DISTRIBUTED_SEQUENCE_ID, TreePattern}
+import org.apache.spark.sql.types.{DataType, LongType}
+
+/**
+ * Returns increasing 64-bit integers consecutive from 0.
+ * The generated ID is guaranteed to be increasing consecutive started from 0.
+ *
+ * @note this expression is dedicated for Pandas API on Spark to use.
+ */
+case class DistributedSequenceID() extends LeafExpression with Unevaluable with NonSQLExpression {
Review Comment:
will it be `Nondeterministic` like `SparkPartitionID` and `MonotonicallyIncreasingID`?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org
[GitHub] [spark] dongjoon-hyun commented on pull request #40456: [SPARK-42720][PS][SQL] Uses expression for distributed-sequence default index instead of plan
Posted by "dongjoon-hyun (via GitHub)" <gi...@apache.org>.
dongjoon-hyun commented on PR #40456:
URL: https://github.com/apache/spark/pull/40456#issuecomment-1473076150
BTW, according to the JIRA, this is only for Apache Spark 3.5?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org