hhhizzz commented on issue #1316: URL: https://github.com/apache/auron/issues/1316#issuecomment-3310420081
Casting in Spark is very complex. Auron's current implementation doesn't fully cover it. I think we can start by implementing **a simple handling for trim** to solve this issue first. And later port Spark's expression tests over completely to gradually address the differences in Spark expressions. Ref: Spark's trim logic PR for int/long: https://github.com/apache/spark/commit/2dd6807e421c96d0aaafc57ceb48f50f66f9d2e7#diff-5db49949ea798670cfe18d6dfd43621eec9c2745164f1e9e58217b2725bebc92 And Spark's trim logic for float/double is depend on the JDK's implement: https://github.com/openjdk/jdk/blob/5855fd2f654175c05341cc03ebf188d4db3e407d/src/java.base/share/classes/jdk/internal/math/FloatingDecimal.java#L1789C9-L1789C63 -- 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]
