matthewgapp commented on issue #7636: URL: https://github.com/apache/arrow-datafusion/issues/7636#issuecomment-1739864817
Thanks @xhwhis, this seems like a separate bug (one whose root cause is because the values exec sets the schema for [all of its columns as nullable here](https://github.com/matthewgapp/arrow-datafusion/blob/3325ab09c74391583ef3b435c9eef99c9068c11f/datafusion/expr/src/logical_plan/builder.rs#L185). This causes the record batch to inherit this nullable schema. The record batch is then inserted into the memory table without complaining (we do check and complain for the inverse - if nullable columns inserted into a nullable table [here](https://github.com/matthewgapp/arrow-datafusion/blob/3325ab09c74391583ef3b435c9eef99c9068c11f/datafusion/physical-plan/src/insert.rs#L137)). The fix is might be passing down the expected table schema to the values logical constructor if there is one. I would open a separate ticket. -- 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]
