vvysotskyi commented on a change in pull request #1844: DRILL-7326: Support
repeated lists for CTAS parquet format
URL: https://github.com/apache/drill/pull/1844#discussion_r315675367
##########
File path:
exec/java-exec/src/main/java/org/apache/drill/exec/store/parquet/ParquetRecordWriter.java
##########
@@ -302,6 +308,40 @@ private Type getType(MaterializedField field) {
}
}
+ /**
+ * Adds element type to {@code listBuilder} based on Drill's
+ * {@code elementField}.
+ *
+ * @param listBuilder list schema builder
+ * @param elementField Drill's type of list elements
+ */
+ private void addElementType(ListBuilder<GroupType> listBuilder,
MaterializedField elementField) {
+ if (elementField.getDataMode() == DataMode.REPEATED) {
+ MaterializedField child2 = elementField.getChildren().iterator().next();
+ ListBuilder<GroupType> inner =
org.apache.parquet.schema.Types.requiredList();
+ addElementType(inner, child2);
+ listBuilder.setElementType(inner.named("element"));
+ } else {
+ Type element = getType(elementField);
+ // element may have internal name '$data$',
+ // rename it to 'element' according to Parquet list schema
+ if (element.isPrimitive()) {
Review comment:
Is it possible when the `element` is not primitive? Or if we currently don't
assume such cases, should we throw an exception for non-primitive elements?
----------------------------------------------------------------
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