HyukjinKwon commented on a change in pull request #27728:
[SPARK-25556][SPARK-17636][SPARK-31026][SPARK-31060][SQL][test-hive1.2] Nested
Column Predicate Pushdown for Parquet
URL: https://github.com/apache/spark/pull/27728#discussion_r394853243
##########
File path:
sql/catalyst/src/main/scala/org/apache/spark/sql/sources/filters.scala
##########
@@ -32,6 +33,7 @@ import org.apache.spark.annotation.{Evolving, Stable}
sealed abstract class Filter {
/**
* List of columns that are referenced by this filter.
+ * Note that, if a column contains `dots` in name, it will be quoted to
avoid confusion.
Review comment:
@cloud-fan, https://github.com/apache/spark/pull/27728/files#r390853911, I
think it shouldn't be a legacy configuration but a proper configuration might
list which source will take the nested filter-push down.
There is no workaround for the behaviour change except this legacy
configuration; but legacy configurations are supposed to be removed. If we're
going to add this as a legacy configuration, we should have a way to don't
unquote this.
Quoting itself might not exist in some downstream datasource
implementations. Dots can be used for different meanings such as namespaces.
Some source don't have nested structures at all and presumably they won't also
have such quotes at all.
----------------------------------------------------------------
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
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
