wangyum opened a new pull request #34627:
URL: https://github.com/apache/spark/pull/34627
Backport #34580
### What changes were proposed in this pull request?
This pr fix push foldable into CaseWhen branches if elseValue is empty. For
example:
```scala
spark.sql("CREATE TABLE t1(bool boolean, number int) using parquet")
spark.sql("INSERT INTO t1 VALUES(false, 1)")
spark.sql(
"""
|SELECT *
|FROM (SELECT *,
| CASE
| WHEN bool THEN 'I am not null'
| END AS conditions
| FROM t1) t
|WHERE conditions IS NULL
|""".stripMargin).show
```
How do we optimize the filter conditions before this pr:
```
Filter isnull(CASE WHEN bool#7 THEN I am not null END) ->
Filter CASE WHEN bool#7 THEN isnull(I am not null) END -> // After
PushFoldableIntoBranches
Filter (bool#7 AND isnull(I am not null)) ->
Filter (bool#7 AND false) ->
Filter false
```
How do we optimize the filter conditions after this pr:
```
Filter isnull(CASE WHEN bool#7 THEN I am not null END) ->
Filter CASE WHEN bool#7 THEN isnull(I am not null) ELSE isnull(null) END ->
// After PushFoldableIntoBranches
Filter CASE WHEN bool#7 THEN false ELSE isnull(null) END ->
Filter CASE WHEN bool#7 THEN false ELSE true END ->
Filter NOT (bool#7 <=> true)
```
### Why are the changes needed?
Fix correctness issue.
### Does this PR introduce _any_ user-facing change?
No.
### How was this patch tested?
Unit test.
--
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]
