iceberg-kitten opened a new issue, #2668:
URL: https://github.com/apache/iceberg-go/issues/2668

   ### Apache Iceberg version
   
   main (development)
   
   ### Please describe the bug 🐞
   
   #### Why
   
   `table.WriteRecords` can panic when writing a bucket-partitioned Iceberg 
`long` from an Arrow `Int32`, `Int16`, `Int8`, `Uint16`, or `Uint8` column. 
Schema validation accepts the promotion, but partition extraction passes an 
`int32` to the bucket transformer, which asserts `int64`. [Arrow value 
conversion](https://github.com/apache/iceberg-go/blob/74b684cf085355bfddba0d2c74743bacd5d2eae3/table/partitioned_fanout_writer.go#L1266-L1301)
 · [Bucket 
transformer](https://github.com/apache/iceberg-go/blob/74b684cf085355bfddba0d2c74743bacd5d2eae3/transforms.go#L664-L679)
 · [Review on 
#2658](https://github.com/apache/iceberg-go/pull/2658#pullrequestreview-5986244814)
   
   #### What
   
   Create a table with `id` of type `long` and partition it by `bucket(16, 
id)`. Run `table.WriteRecords` with the default writer and a one-row Arrow 
batch containing `id = -1` as `Int32`. The write should succeed and produce the 
same bucket as an Arrow `Int64` containing `-1`.
   
   #### Implementation
   
   In `getArrowValueAsIcebergValue`, widen those five Arrow types to `int64` 
when the Iceberg source is `long`, following the existing `Uint32` branch.
   


-- 
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