mihailom-db commented on code in PR #48585:
URL: https://github.com/apache/spark/pull/48585#discussion_r1815655287
##########
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/TypeCoercion.scala:
##########
@@ -1014,7 +1016,9 @@ object TypeCoercion extends TypeCoercionBase {
case (_: StringType, datetime: DatetimeType) => datetime
case (_: StringType, AnyTimestampType) =>
AnyTimestampType.defaultConcreteType
case (_: StringType, BinaryType) => BinaryType
- // Cast any atomic type to string.
+ // Cast any atomic type to string except if there are strings with
different collations. In
+ // that case we skip the cast.
+ case (st1: StringType, st2: StringType) if st1.collationId !=
st2.collationId => null
Review Comment:
This seems like we are adding something unnecessary. The comment said we
cast any atomic type and that rule is checking for whether any is instance of
StringType, so we cannot hit this by any rule below actually, so we should be
fine with getting the default which is null? I am good with adding additional
comment as to why we do not cast StringType in the existing rule, but this is
just moving the default branch earlier.
--
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]
