ODIn/spark-instrumented-optimizer
1
0
Fork 0
Code Issues Pull requests Projects Releases Wiki Activity

[SPARK-31166][SQL] UNION map<null, null> and other maps should not fail

Browse source 
### What changes were proposed in this pull request?

After https://github.com/apache/spark/pull/27542, `map()` returns `map<null, null>` instead of `map<string, string>`. However, this breaks queries which union `map()` and other maps.

The reason is, `TypeCoercion` rules and `Cast` think it's illegal to cast null type map key to other types, as it makes the key nullable, but it's actually legal. This PR fixes it.

### Why are the changes needed?

To avoid breaking queries.

### Does this PR introduce any user-facing change?

Yes, now some queries that work in 2.x can work in 3.0 as well.

### How was this patch tested?

new test

Closes #27926 from cloud-fan/bug.

Authored-by: Wenchen Fan <wenchen@databricks.com>
Signed-off-by: Wenchen Fan <wenchen@databricks.com>
This commit is contained in:
Wenchen Fan 2020-03-17 12:01:29 +08:00
parent 93088f79cc
commit d7b97a1d0d
3 changed files with 13 additions and 3 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/TypeCoercion.scala
View file

@ -158,7 +158,7 @@ object TypeCoercion {
}
case (MapType(kt1, vt1, valueContainsNull1), MapType(kt2, vt2, valueContainsNull2)) =>
findTypeFunc(kt1, kt2)
.filter { kt => !Cast.forceNullable(kt1, kt) && !Cast.forceNullable(kt2, kt) }
.filter { kt => Cast.canCastMapKeyNullSafe(kt1, kt) && Cast.canCastMapKeyNullSafe(kt2, kt) }
.flatMap { kt =>
findTypeFunc(vt1, vt2).map { vt =>
MapType(kt, vt, valueContainsNull1 || valueContainsNull2 ||

8
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/expressions/Cast.scala
View file

@ -77,8 +77,7 @@ object Cast {
resolvableNullability(fn || forceNullable(fromType, toType), tn)
case (MapType(fromKey, fromValue, fn), MapType(toKey, toValue, tn)) =>
canCast(fromKey, toKey) &&
(!forceNullable(fromKey, toKey)) &&
canCast(fromKey, toKey) && canCastMapKeyNullSafe(fromKey, toKey) &&
canCast(fromValue, toValue) &&
resolvableNullability(fn || forceNullable(fromValue, toValue), tn)
@ -98,6 +97,11 @@ object Cast {
case _ => false
}
def canCastMapKeyNullSafe(fromType: DataType, toType: DataType): Boolean = {
// If the original map key type is NullType, it's OK as the map must be empty.
fromType == NullType || !forceNullable(fromType, toType)
}
/**
* Return true if we need to use the `timeZone` information casting `from` type to `to` type.
* The patterns matched reflect the current implementation in the Cast node.

6
sql/core/src/test/scala/org/apache/spark/sql/SQLQuerySuite.scala
View file

@ -3487,6 +3487,12 @@ class SQLQuerySuite extends QueryTest with SharedSparkSession with AdaptiveSpark
}
}
}
test("SPARK-31166: UNION map<null, null> and other maps should not fail") {
checkAnswer(
sql("(SELECT map()) UNION ALL (SELECT map(1, 2))"),
Seq(Row(Map[Int, Int]()), Row(Map(1 -> 2))))
}
}
case class Foo(bar: Option[String])