Skip to content

[SPARK-30844][SQL]Static partition should also follow StoreAssignmentPolicy when insert into table #27597

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
Ngone51 wants to merge 5 commits into from
Closed

[SPARK-30844][SQL]Static partition should also follow StoreAssignmentPolicy when insert into table #27597

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
9ca4a77
fix
Ngone51 Feb 16, 2020
1abe474
remove conf
Ngone51 Feb 17, 2020
f38e056
test legacy and strict
Ngone51 Feb 17, 2020
a70c89c
add comments
Ngone51 Feb 21, 2020
3d10ca7
address nit
Ngone51 Feb 21, 2020
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
15 changes: 12 additions & 3 deletions sql/core/src/main/scala/org/apache/spark/sql/execution/datasources/DataSourceStrategy.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -39,6 +39,7 @@ import org.apache.spark.sql.catalyst.rules.Rule
import org.apache.spark.sql.execution.{RowDataSourceScanExec, SparkPlan}
import org.apache.spark.sql.execution.command._
import org.apache.spark.sql.internal.SQLConf
import org.apache.spark.sql.internal.SQLConf.StoreAssignmentPolicy
import org.apache.spark.sql.sources._
import org.apache.spark.sql.types._
import org.apache.spark.unsafe.types.UTF8String
Expand All @@ -59,7 +60,8 @@ case class DataSourceAnalysis(conf: SQLConf) extends Rule[LogicalPlan] with Cast
sourceAttributes: Seq[Attribute],
providedPartitions: Map[String, Option[String]],
targetAttributes: Seq[Attribute],
targetPartitionSchema: StructType): Seq[NamedExpression] = {
targetPartitionSchema: StructType,
conf: SQLConf): Seq[NamedExpression] = {
Copy link
Member

@maropu maropu Feb 16, 2020

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Why do we need conf here, the constructor has conf though?

Copy link
Member Author

@Ngone51 Ngone51 Feb 17, 2020

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Removed, thanks!


assert(providedPartitions.exists(_._2.isDefined))

Expand Down Expand Up @@ -104,7 +106,13 @@ case class DataSourceAnalysis(conf: SQLConf) extends Rule[LogicalPlan] with Cast
None
} else if (potentialSpecs.size == 1) {
val partValue = potentialSpecs.head._2
Some(Alias(cast(Literal(partValue), field.dataType), field.name)())
conf.storeAssignmentPolicy match {
case StoreAssignmentPolicy.ANSI | StoreAssignmentPolicy.STRICT =>
Some(Alias(AnsiCast(Literal(partValue), field.dataType,
Option(conf.sessionLocalTimeZone)), field.name)())
case _ =>
Some(Alias(cast(Literal(partValue), field.dataType), field.name)())
}
} else {
throw new AnalysisException(
s"Partition column ${field.name} have multiple values specified, " +
Expand Down Expand Up @@ -176,7 +184,8 @@ case class DataSourceAnalysis(conf: SQLConf) extends Rule[LogicalPlan] with Cast
sourceAttributes = query.output,
providedPartitions = parts,
targetAttributes = l.output,
targetPartitionSchema = t.partitionSchema)
targetPartitionSchema = t.partitionSchema,
conf)
Project(projectList, query)
} else {
query
Expand Down
49 changes: 34 additions & 15 deletions sql/core/src/test/scala/org/apache/spark/sql/sources/DataSourceAnalysisSuite.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -22,9 +22,10 @@ import org.scalatest.BeforeAndAfterAll
import org.apache.spark.SparkFunSuite
import org.apache.spark.sql.AnalysisException
import org.apache.spark.sql.catalyst.dsl.expressions._
import org.apache.spark.sql.catalyst.expressions.{Alias, Attribute, Cast, Expression, Literal}
import org.apache.spark.sql.catalyst.expressions.{Alias, AnsiCast, Attribute, Cast, Expression, Literal}
import org.apache.spark.sql.execution.datasources.DataSourceAnalysis
import org.apache.spark.sql.internal.SQLConf
import org.apache.spark.sql.internal.SQLConf.StoreAssignmentPolicy
import org.apache.spark.sql.types.{DataType, IntegerType, StructType}

class DataSourceAnalysisSuite extends SparkFunSuite with BeforeAndAfterAll {
Expand Down Expand Up @@ -52,7 +53,12 @@ class DataSourceAnalysisSuite extends SparkFunSuite with BeforeAndAfterAll {
Seq(true, false).foreach { caseSensitive =>
val conf = new SQLConf().copy(SQLConf.CASE_SENSITIVE -> caseSensitive)
def cast(e: Expression, dt: DataType): Expression = {
Cast(e, dt, Option(conf.sessionLocalTimeZone))
conf.storeAssignmentPolicy match {
case StoreAssignmentPolicy.ANSI | StoreAssignmentPolicy.STRICT =>
AnsiCast(e, dt, Option(conf.sessionLocalTimeZone))
case _ =>
Cast(e, dt, Option(conf.sessionLocalTimeZone))
}
}
val rule = DataSourceAnalysis(conf)
test(
Expand All @@ -63,7 +69,8 @@ class DataSourceAnalysisSuite extends SparkFunSuite with BeforeAndAfterAll {
sourceAttributes = Seq('e.int, 'f.int),
providedPartitions = Map("b" -> None, "c" -> None),
targetAttributes = targetAttributes,
targetPartitionSchema = targetPartitionSchema)
targetPartitionSchema = targetPartitionSchema,
conf)
}
}

Expand All @@ -74,7 +81,8 @@ class DataSourceAnalysisSuite extends SparkFunSuite with BeforeAndAfterAll {
sourceAttributes = Seq('e.int),
providedPartitions = Map("b" -> Some("1"), "c" -> None),
targetAttributes = targetAttributes,
targetPartitionSchema = targetPartitionSchema)
targetPartitionSchema = targetPartitionSchema,
conf)
}
}

Expand All @@ -85,7 +93,8 @@ class DataSourceAnalysisSuite extends SparkFunSuite with BeforeAndAfterAll {
sourceAttributes = Seq('e.int, 'f.int),
providedPartitions = Map("b" -> Some("1")),
targetAttributes = targetAttributes,
targetPartitionSchema = targetPartitionSchema)
targetPartitionSchema = targetPartitionSchema,
conf)
}

// Missing partitioning columns.
Expand All @@ -94,7 +103,8 @@ class DataSourceAnalysisSuite extends SparkFunSuite with BeforeAndAfterAll {
sourceAttributes = Seq('e.int, 'f.int, 'g.int),
providedPartitions = Map("b" -> Some("1")),
targetAttributes = targetAttributes,
targetPartitionSchema = targetPartitionSchema)
targetPartitionSchema = targetPartitionSchema,
conf)
}

// Wrong partitioning columns.
Expand All @@ -103,7 +113,8 @@ class DataSourceAnalysisSuite extends SparkFunSuite with BeforeAndAfterAll {
sourceAttributes = Seq('e.int, 'f.int),
providedPartitions = Map("b" -> Some("1"), "d" -> None),
targetAttributes = targetAttributes,
targetPartitionSchema = targetPartitionSchema)
targetPartitionSchema = targetPartitionSchema,
conf)
}
}

Expand All @@ -114,7 +125,8 @@ class DataSourceAnalysisSuite extends SparkFunSuite with BeforeAndAfterAll {
sourceAttributes = Seq('e.int, 'f.int),
providedPartitions = Map("b" -> Some("1"), "d" -> Some("2")),
targetAttributes = targetAttributes,
targetPartitionSchema = targetPartitionSchema)
targetPartitionSchema = targetPartitionSchema,
conf)
}

// Wrong partitioning columns.
Expand All @@ -123,7 +135,8 @@ class DataSourceAnalysisSuite extends SparkFunSuite with BeforeAndAfterAll {
sourceAttributes = Seq('e.int),
providedPartitions = Map("b" -> Some("1"), "c" -> Some("3"), "d" -> Some("2")),
targetAttributes = targetAttributes,
targetPartitionSchema = targetPartitionSchema)
targetPartitionSchema = targetPartitionSchema,
conf)
}

if (caseSensitive) {
Expand All @@ -133,7 +146,8 @@ class DataSourceAnalysisSuite extends SparkFunSuite with BeforeAndAfterAll {
sourceAttributes = Seq('e.int, 'f.int),
providedPartitions = Map("b" -> Some("1"), "C" -> Some("3")),
targetAttributes = targetAttributes,
targetPartitionSchema = targetPartitionSchema)
targetPartitionSchema = targetPartitionSchema,
conf)
}
}
}
Expand All @@ -147,7 +161,8 @@ class DataSourceAnalysisSuite extends SparkFunSuite with BeforeAndAfterAll {
sourceAttributes = Seq('e.int, 'f.int),
providedPartitions = Map("b" -> None, "c" -> Some("3")),
targetAttributes = targetAttributes,
targetPartitionSchema = targetPartitionSchema)
targetPartitionSchema = targetPartitionSchema,
conf)
}
}

Expand All @@ -160,7 +175,8 @@ class DataSourceAnalysisSuite extends SparkFunSuite with BeforeAndAfterAll {
sourceAttributes = nonPartitionedAttributes,
providedPartitions = Map("b" -> Some("1"), "C" -> Some("3")),
targetAttributes = targetAttributes,
targetPartitionSchema = targetPartitionSchema)
targetPartitionSchema = targetPartitionSchema,
conf)
checkProjectList(actual, expected)
}

Expand All @@ -172,7 +188,8 @@ class DataSourceAnalysisSuite extends SparkFunSuite with BeforeAndAfterAll {
sourceAttributes = nonPartitionedAttributes,
providedPartitions = Map("b" -> Some("1"), "c" -> Some("3")),
targetAttributes = targetAttributes,
targetPartitionSchema = targetPartitionSchema)
targetPartitionSchema = targetPartitionSchema,
conf)
checkProjectList(actual, expected)
}

Expand All @@ -184,7 +201,8 @@ class DataSourceAnalysisSuite extends SparkFunSuite with BeforeAndAfterAll {
sourceAttributes = nonPartitionedAttributes,
providedPartitions = Map("b" -> Some("1")),
targetAttributes = Seq('a.int, 'd.int, 'b.int),
targetPartitionSchema = new StructType().add("b", IntegerType))
targetPartitionSchema = new StructType().add("b", IntegerType),
conf)
checkProjectList(actual, expected)
}
}
Expand All @@ -200,7 +218,8 @@ class DataSourceAnalysisSuite extends SparkFunSuite with BeforeAndAfterAll {
sourceAttributes = nonPartitionedAttributes ++ dynamicPartitionAttributes,
providedPartitions = Map("b" -> Some("1"), "c" -> None),
targetAttributes = targetAttributes,
targetPartitionSchema = targetPartitionSchema)
targetPartitionSchema = targetPartitionSchema,
conf)
checkProjectList(actual, expected)
}
}
Expand Down
13 changes: 13 additions & 0 deletions sql/core/src/test/scala/org/apache/spark/sql/sources/InsertSuite.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -753,6 +753,19 @@ class InsertSuite extends DataSourceTest with SharedSparkSession {
}
}

test("SPARK-30844: static partition should also follow StoreAssignmentPolicy") {
withSQLConf(
SQLConf.STORE_ASSIGNMENT_POLICY.key -> SQLConf.StoreAssignmentPolicy.ANSI.toString) {
Copy link
Member

@maropu maropu Feb 16, 2020

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

How about testing the other modes here, e.g., LEGACY and STRICT?

Copy link
Member Author

@Ngone51 Ngone51 Feb 17, 2020

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

updated!

withTable("t") {
sql("create table t(a int, b string) using parquet partitioned by (a)")
val errorMsg = intercept[NumberFormatException] {
sql("insert into t partition(a='ansi') values('ansi')")
}.getMessage
assert(errorMsg.contains("invalid input syntax for type numeric: ansi"))
}
}
}

test("SPARK-24860: dynamic partition overwrite specified per source without catalog table") {
withTempPath { path =>
Seq((1, 1), (2, 2)).toDF("i", "part")
Expand Down