zhengruifeng commented on code in PR #37728:
URL: https://github.com/apache/spark/pull/37728#discussion_r959257404


##########
sql/core/src/test/resources/tpcds-query-results/v1_4/q77.sql.out:
##########
@@ -4,8 +4,8 @@
 
struct<channel:string,id:int,sales:decimal(27,2),returns:decimal(27,2),profit:decimal(28,2)>
 -- !query output
 NULL   NULL    238379361.39    11949589.80     -69066318.65
-catalog channel        NULL    116209.49       1989207.49      -1103184.43

Review Comment:
   `RDD.takeOrdered` and `RDD.top` are used in many place.
   In SQL, it's used in 
[TakeOrderedAndProjectExec](https://github.com/apache/spark/blob/d1ca0ea4ae55bf19c7569eacae5b377c3340dde9/sql/core/src/main/scala/org/apache/spark/sql/execution/limit.scala#L285-L300),
 to pass the `TPCDS querys`, I had to swap two line in the results of `q77` and 
`q77a`. the 
[`q77`](https://github.com/apache/spark/blob/a2d8d767d933321426a4eb9df1583e017722d7d6/sql/core/src/test/resources/tpcds/q77.sql#L99-L100)
 and 
[`q77a`](https://github.com/apache/spark/blob/5f653d4f7c84e6147cd323cd650da65e0381ebe8/sql/core/src/test/resources/tpcds-v2.7.0/q77a.sql#L119-L121)
 both end with `order by channel, id limit 100`, and the swapped lines have the 
same `(channel, id)`, so I guess this change is fine.
   
   But I think it still need you to take a look @cloud-fan @HyukjinKwon 



-- 
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]

Reply via email to