MaxGekk commented on a change in pull request #34596:
URL: https://github.com/apache/spark/pull/34596#discussion_r756272464
##########
File path:
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/csv/CSVInferSchema.scala
##########
@@ -160,6 +169,17 @@ class CSVInferSchema(val options: CSVOptions) extends
Serializable {
private def tryParseDouble(field: String): DataType = {
if ((allCatch opt field.toDouble).isDefined || isInfOrNan(field)) {
DoubleType
+ } else {
+ tryParseTimestampNTZ(field)
Review comment:
When I supported timestamp inference in JSON
https://github.com/apache/spark/pull/23201, I had to introduce new option in
https://github.com/apache/spark/pull/23455, and disable the inference by
default in https://github.com/apache/spark/pull/28966 because such inference
slowed down user queries. If you are sure that is not the case, please, extend
CSV benchmarks to proof the performance doesn't degrade.
--
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]
