Skip to content

[SPARK-59678][SQL] Fix pivot failing on array and struct columns with non-nullable nested fields #58934

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

Open
jiwen624 wants to merge 2 commits into
base: master
Choose a base branch
from
Open

[SPARK-59678][SQL] Fix pivot failing on array and struct columns with non-nullable nested fields #58934

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
1424696
[SPARK-59678][SQL] Ignore nullability when matching pivot values to t…
jiwen624 Sep 20, 2026
0ef0869
[SPARK-59678][SQL] Keep the pivot column type when the value already …
jiwen624 Sep 22, 2026
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
12 changes: 9 additions & 3 deletions sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/PivotTransformer.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -97,15 +97,16 @@ object PivotTransformer extends AliasHelper with SQLConfHelper {
aggregates: Seq[Expression],
childOutput: Seq[Attribute],
newAlias: (Expression, Option[String]) => Alias): LogicalPlan = {
val pivotColumnType = pivotColumn.dataType.asNullable
val evalPivotValues = pivotValues.map { value =>
val foldable = trimAliases(value).foldable
if (!foldable) {
throw QueryCompilationErrors.nonLiteralPivotValError(value)
}
if (!Cast.canCast(value.dataType, pivotColumn.dataType)) {
if (!Cast.canCast(value.dataType, pivotColumnType)) {
throw QueryCompilationErrors.pivotValDataTypeMismatchError(value, pivotColumn)
}
Cast(value, pivotColumn.dataType, Some(conf.sessionLocalTimeZone)).eval(EmptyRow)
Cast(value, pivotColumnType, Some(conf.sessionLocalTimeZone)).eval(EmptyRow)
}
val groupByExpressions = groupByExpressionsOpt.getOrElse {
val pivotColumnAndAggregatesRefs = pivotColumn.references ++ AttributeSet(aggregates)
Expand Down Expand Up @@ -211,10 +212,15 @@ object PivotTransformer extends AliasHelper with SQLConfHelper {
expression: Expression,
pivotColumn: Expression,
value: Expression) = {
val valueType = if (Cast.canCast(value.dataType, pivotColumn.dataType)) {
pivotColumn.dataType
} else {
pivotColumn.dataType.asNullable
}
If(
EqualNullSafe(
pivotColumn,
Cast(value, pivotColumn.dataType, Some(conf.sessionLocalTimeZone))
Cast(value, valueType, Some(conf.sessionLocalTimeZone))
),
expression,
Literal(null)
Expand Down
22 changes: 22 additions & 0 deletions sql/core/src/test/scala/org/apache/spark/sql/DataFramePivotSuite.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -342,6 +342,28 @@ class DataFramePivotSuite extends SharedSparkSession {
checkAnswer(actual, expected)
}

test("SPARK-59678: pivot with explicit values on an array column with non-nullable elements") {
val df = Seq((1.0d, "x"), (2.0d, "y")).toDF("v", "s").selectExpr("v", "s", "array(v) AS a")
assert(!df.schema("a").dataType.asInstanceOf[ArrayType].containsNull)
checkAnswer(
df.groupBy("v").pivot($"a", Seq(Array(1.0d))).count(),
Row(1.0d, 1L) :: Row(2.0d, null) :: Nil)
checkAnswer(
df.groupBy("v").pivot($"a", Seq(Array(1.0d))).agg(first("s")),
Row(1.0d, "x") :: Row(2.0d, null) :: Nil)
}

test("SPARK-59678: SQL PIVOT on a struct column with a non-nullable field") {
val source = "SELECT named_struct('x', id) AS c, id AS v, CAST(id AS STRING) AS s FROM range(2)"
val value = "named_struct('x', if(true, 1L, null))"
checkAnswer(
sql(s"SELECT * FROM (SELECT c, v FROM ($source)) PIVOT (sum(v) FOR c IN ($value))"),
Row(1L))
checkAnswer(
sql(s"SELECT * FROM (SELECT c, s FROM ($source)) PIVOT (first(s) FOR c IN ($value))"),
Row("1"))
}

test("SPARK-35480: percentile_approx should work with pivot") {
val actual = Seq(
("a", -1.0), ("a", 5.5), ("a", 2.5), ("b", 3.0), ("b", 5.2)).toDF("type", "value")
Expand Down