Github user HyukjinKwon commented on a diff in the pull request:
https://github.com/apache/spark/pull/22527#discussion_r219686429
--- Diff: sql/core/src/main/scala/org/apache/spark/sql/SQLContext.scala ---
@@ -1100,13 +1101,24 @@ object SQLContext {
attrs: Seq[AttributeReference]): Iterator[InternalRow] = {
val extractors =
JavaTypeInference.getJavaBeanReadableProperties(beanClass).map(_.getReadMethod)
- val methodsToConverts = extractors.zip(attrs).map { case (e, attr) =>
- (e, CatalystTypeConverters.createToCatalystConverter(attr.dataType))
+ val methodsToTypes = extractors.zip(attrs).map { case (e, attr) =>
+ (e, attr.dataType)
+ }
+ def invoke(element: Any)(tuple: (Method, DataType)): Any = tuple match
{
+ case (e, structType: StructType) =>
+ val value = e.invoke(element)
+ val nestedExtractors =
JavaTypeInference.getJavaBeanReadableProperties(value.getClass)
+ .map(desc => desc.getName -> desc.getReadMethod)
+ .toMap
+ new GenericInternalRow(structType.map(nestedProperty =>
+ invoke(value)(nestedExtractors(nestedProperty.name) ->
nestedProperty.dataType)
+ ).toArray)
--- End diff --
Why should we use a map here while we don't need it for the root bean?
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
