uros-b commented on code in PR #56422:
URL: https://github.com/apache/spark/pull/56422#discussion_r3936126872
##########
sql/core/src/main/scala/org/apache/spark/sql/execution/datasources/parquet/types/ops/TimeTypeParquetOps.scala:
##########
@@ -89,13 +89,19 @@ case class TimeTypeParquetOps(t: TimeType) extends
ParquetTypeOps {
// ==================== Row-Based Read ====================
override def newConverter(
+ parquetType: Type,
+ updater: ParentContainerUpdater): Converter with
HasParentContainerUpdater =
+ newConverterInternal(parquetType, updater)
+
+ private def newConverterInternal(
parquetType: Type,
updater: ParentContainerUpdater): Converter with
HasParentContainerUpdater = {
// Framework-first dispatch in ParquetRowConverter routes here whenever the
// requested Spark type is TimeType, regardless of the actual Parquet
encoding.
- // Without this guard, files whose column is raw INT64, INT64
TIMESTAMP(MICROS),
- // INT32 TIME(MILLIS), etc. would silently decode as microsToNanos(value)
and
- // produce wrong results.
+ // Without this guard, files whose column is raw INT64, INT64 TIME(NANOS),
Review Comment:
Please drop `INT64 TIME(NANOS)` from the "would misdecode as microsToNanos"
list; TIME(NANOS) decodes correctly via the identity branch and is accepted by
the guard.
--
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.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
