Github user liancheng commented on a diff in the pull request:
https://github.com/apache/spark/pull/13769#discussion_r67637019
--- Diff:
sql/core/src/main/scala/org/apache/spark/sql/execution/datasources/DataSourceStrategy.scala
---
@@ -43,8 +43,128 @@ import org.apache.spark.unsafe.types.UTF8String
* Replaces generic operations with specific variants that are designed to
work with Spark
* SQL Data Sources.
*/
-private[sql] object DataSourceAnalysis extends Rule[LogicalPlan] {
+private[sql] case class DataSourceAnalysis(conf: CatalystConf) extends
Rule[LogicalPlan] {
+
+ def resolver: Resolver = {
+ if (conf.caseSensitiveAnalysis) {
+ caseSensitiveResolution
+ } else {
+ caseInsensitiveResolution
+ }
+ }
+
+ // The access modifier is used to expose this method to tests.
+ private[sql] def convertStaticPartitions(
+ sourceAttributes: Seq[Attribute],
+ providedPartitions: Map[String, Option[String]],
+ targetAttributes: Seq[Attribute],
+ targetPartitionSchema: StructType): Seq[NamedExpression] = {
+
+ assert(providedPartitions.exists(_._2.isDefined))
+
+ val staticPartitions = providedPartitions.flatMap {
+ case (partKey, Some(partValue)) => (partKey, partValue) :: Nil
+ case (_, None) => Nil
+ }
+
+ // The sum of the number of static partition columns and columns
provided in the SELECT
+ // clause needs to match the number of columns of the target table.
+ if (staticPartitions.size + sourceAttributes.size !=
targetAttributes.size) {
+ throw new AnalysisException(
+ s"The data to be inserted needs to have the same number of " +
+ s"columns as the target table: target table has
${targetAttributes.size} " +
+ s"column(s) but the inserted data has ${sourceAttributes.size +
staticPartitions.size} " +
+ s"column(s), which contain ${staticPartitions.size} partition
column(s) having " +
+ s"assigned constant values.")
+ }
+
+ if (providedPartitions.size != targetPartitionSchema.fields.size) {
+ throw new AnalysisException(
+ s"The data to be inserted needs to have the same number of " +
+ s"partition columns as the target table: target table " +
+ s"has ${targetPartitionSchema.fields.size} partition column(s)
but the inserted " +
+ s"data has ${providedPartitions.size} partition columns
specified.")
+ }
+
+ staticPartitions.foreach {
+ case (partKey, partValue) =>
+ if (!targetPartitionSchema.fields.exists(field =>
resolver(field.name, partKey))) {
+ throw new AnalysisException(
+ s"$partKey is not a partition column. Partition columns are " +
+ s"${targetPartitionSchema.fields.map(_.name).mkString("[",
",", "]")}")
+ }
+ }
+
+ val partitionList = targetPartitionSchema.fields.map { field =>
+ val potentialSpecs = staticPartitions.filter {
+ case (partKey, partValue) => resolver(field.name, partKey)
+ }
+ if (potentialSpecs.size == 0) {
+ None
+ } else if (potentialSpecs.size == 1) {
+ val partValue = potentialSpecs.head._2
+ Some(Alias(Cast(Literal(partValue), field.dataType),
"_staticPart")())
+ } else {
+ throw new AnalysisException(
+ s"Partition column ${field.name} have multiple values specified,
" +
+ s"${potentialSpecs.mkString("[", ", ", "]")}. Please only
specify a single value.")
+ }
+ }
+
+ partitionList.sliding(2).foreach { v =>
--- End diff --
`sliding(2)` can be dangerous for single-element collections:
```
scala> Seq(1).sliding(2).foreach(println)
List(1)
```
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
