[jira] [Created] (ARROW-8418) [Python] partition_filename_cb in write_to_dataset should be passed additional keyword arguments rather than just keys

["Varun Patil (Jira)" <ji...@apache.org>]

Varun Patil created ARROW-8418:
----------------------------------

             Summary: [Python] partition_filename_cb in write_to_dataset should be passed additional keyword arguments rather than just keys
                 Key: ARROW-8418
                 URL: https://issues.apache.org/jira/browse/ARROW-8418
             Project: Apache Arrow
          Issue Type: Improvement
          Components: Python
            Reporter: Varun Patil


I recently had a requirement where I would have liked to construct a filename based on additional context from Apache Airflow (specifically execution_date).

It would be nice to pass the additional kwargs to *partition_filename_cb* so that the filename can be constructed using additional information rather than just the keys used for partitioning.

I believe the fix should be as simple as passing kwargs to the *partition_filename_cb* inside *write_to_dataset.*



--
This message was sent by Atlassian Jira
(v8.3.4#803005)