Github user nongli commented on a diff in the pull request:
https://github.com/apache/spark/pull/10333#discussion_r47833170
--- Diff:
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/expressions/Expression.scala
---
@@ -424,19 +431,30 @@ abstract class BinaryExpression extends Expression {
val eval1 = left.gen(ctx)
val eval2 = right.gen(ctx)
val resultCode = f(eval1.value, eval2.value)
- s"""
- ${eval1.code}
- boolean ${ev.isNull} = ${eval1.isNull};
- ${ctx.javaType(dataType)} ${ev.value} =
${ctx.defaultValue(dataType)};
- if (!${ev.isNull}) {
- ${eval2.code}
- if (!${eval2.isNull}) {
- $resultCode
- } else {
- ${ev.isNull} = true;
+ if (nullable) {
+ s"""
+ ${eval1.code}
+ boolean ${ev.isNull} = ${eval1.isNull};
+ ${ctx.javaType(dataType)} ${ev.value} =
${ctx.defaultValue(dataType)};
+ if (!${ev.isNull}) {
+ ${eval2.code}
+ if (!${eval2.isNull}) {
+ $resultCode
+ } else {
+ ${ev.isNull} = true;
+ }
}
- }
- """
+ """
+
+ } else {
+ ev.isNull = "false"
--- End diff --
Is this branch necessary? (not suggesting you change it) but does the
nullable path collapse correctly if left and right are non nullable? What I
mean is:
if eval1.isNull and eval2.isNull is always just false, do we get the same
behavior as this special casing from the compiler optimizations?
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
