Github user viirya commented on a diff in the pull request:
https://github.com/apache/spark/pull/20277#discussion_r162536698
--- Diff:
sql/core/src/main/scala/org/apache/spark/sql/execution/ColumnarBatchScan.scala
---
@@ -50,7 +50,14 @@ private[sql] trait ColumnarBatchScan extends
CodegenSupport {
dataType: DataType,
nullable: Boolean): ExprCode = {
val javaType = ctx.javaType(dataType)
- val value = ctx.getValue(columnVar, dataType, ordinal)
+ val value = if (dataType.isInstanceOf[StructType]) {
+ // `ColumnVector.getStruct` is different from
`InternalRow.getStruct`, it only takes an
+ // `ordinal` parameter.
+ s"$columnVar.getStruct($ordinal)"
+ } else {
+ ctx.getValue(columnVar, dataType, ordinal)
+ }
--- End diff --
Can't we use this API?
```scala
/**
* Returns the specialized code to access a value from a column vector
for a given `DataType`.
*/
def getValue(vector: String, rowId: String, dataType: DataType): String =
{
...
}
```
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
