gengliangwang commented on PR #38379: URL: https://github.com/apache/spark/pull/38379#issuecomment-1289658114
I confirmed that the regression is caused by the refactoring PR https://github.com/apache/spark/pull/36698/. Before the refactor, the query will look like ``` coalesce(CheckOverflow((promote_precision(cast(CheckOverflow((promote_precision(cast(a#0 as decimal(14,6))) + promote_precision(cast(b#1 as decimal(14,6)))), DecimalType(14,6), true) as decimal(15,6))) + promote_precision(cast(1.75 as decimal(15,6)))), DecimalType(15,6), true), a#0) ``` All the children of Add are cast as the final data type. Thus reordering Add for canonicalization won’t matter. -- 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]
