yaooqinn commented on a change in pull request #28197: [SPARK-31431][SQL] Add
CalendarInterval encoder support
URL: https://github.com/apache/spark/pull/28197#discussion_r409957915
##########
File path:
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/ScalaReflection.scala
##########
@@ -235,6 +235,7 @@ object ScalaReflection extends ScalaReflection {
case t if isSubtype(t, localTypeOf[java.lang.String]) =>
createDeserializerForString(path, returnNullable = false)
+ case t if isSubtype(t, localTypeOf[CalendarInterval]) => path
Review comment:
Do you mean we work for scala Products? `Seq(Array(something)).toDF` seems
need that `something` to be java primitives or `scala.Product`s. Not yet
support `CalendarInterval` or e.g....
```scala
scala> case class A(a: Int)
defined class A
scala> Seq(Array(A(1))).toDF
res8: org.apache.spark.sql.DataFrame = [value: array<struct<a:int>>]
scala> class B(a: Int)
defined class B
scala> Seq(Array(new B(1))).toDF
<console>:29: error: value toDF is not a member of Seq[Array[B]]
Seq(Array(new B(1))).toDF
^
scala> Seq(new java.math.BigDecimal(1)).toDF
res10: org.apache.spark.sql.DataFrame = [value: decimal(38,18)]
scala> Seq(Array(new java.math.BigDecimal(1))).toDF
<console>:28: error: value toDF is not a member of
Seq[Array[java.math.BigDecimal]]
Seq(Array(new java.math.BigDecimal(1))).toDF
```
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
