[
https://issues.apache.org/jira/browse/SPARK-60637?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
ASF GitHub Bot updated SPARK-60637:
-----------------------------------
Labels: pull-request-available (was: )
> Account for decimal coercion when deriving VALUES nullability
> -------------------------------------------------------------
>
> Key: SPARK-60637
> URL: https://issues.apache.org/jira/browse/SPARK-60637
> Project: Spark
> Issue Type: Bug
> Components: SQL
> Affects Versions: 5.0.0
> Reporter: Linhong Liu
> Priority: Major
> Labels: pull-request-available
>
> VALUES currently derives output nullability from the original expressions
> rather than their coerced forms. With ANSI mode disabled and
> spark.sql.legacy.decimal.retainFractionDigitsOnTruncate=true, combining a
> 38-digit integer with 0.1 can overflow the integer to NULL but leave the
> output marked nonnullable. IS NULL and COUNT(column) can then return
> incorrect results. Derive metadata from coerced expressions. For eagerly
> materialized tables, retain nonnullable metadata for originally nonnullable
> columns when every evaluated value is non-NULL, without evaluating
> expressions again. Cover both eager evaluation settings and the
> runtime-expression Expand path. CURRENT_LIKE-deferred tables may retain
> conservative nullability.
--
This message was sent by Atlassian Jira
(v10.3.26#10030026)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
