yaooqinn commented on code in PR #45000:
URL: https://github.com/apache/spark/pull/45000#discussion_r1479174067
##########
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/util/ResolveDefaultColumnsUtil.scala:
##########
@@ -311,36 +313,35 @@ object ResolveDefaultColumns extends QueryErrorsBase
defaultSQL: String): Expression = {
val supplanted = CharVarcharUtils.replaceCharVarcharWithString(dataType)
// Perform implicit coercion from the provided expression type to the
required column type.
- val ret = if (supplanted == analyzed.dataType) {
- analyzed
- } else if (Cast.canUpCast(analyzed.dataType, supplanted)) {
- Cast(analyzed, supplanted)
- } else {
- // If the provided default value is a literal of a wider type than the
target column, but the
- // literal value fits within the narrower type, just coerce it for
convenience. Exclude
- // boolean/array/struct/map types from consideration for this type
coercion to avoid
- // surprising behavior like interpreting "false" as integer zero.
- val result = if (analyzed.isInstanceOf[Literal] &&
- !Seq(supplanted, analyzed.dataType).exists(_ match {
+ val ret = analyzed match {
+ case equivalent if equivalent.dataType == supplanted =>
+ equivalent
+ case canUpCast if Cast.canUpCast(canUpCast.dataType, supplanted) =>
+ Cast(analyzed, supplanted, Some(conf.sessionLocalTimeZone))
+ case l: Literal
+ if !Seq(supplanted, l.dataType).exists(_ match {
case _: BooleanType | _: ArrayType | _: StructType | _: MapType =>
true
case _ => false
- })) {
- try {
- val casted = Cast(analyzed, supplanted, evalMode =
EvalMode.TRY).eval()
Review Comment:
The syntax using match-case here is more elegant than if-else. Although it
increases the LOC, it's limited to a single method, so it's safe from the
perspective of `git diff`. However, if you insist, I can change it.
--
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]
