Checks are failing. Will not request review until checks are succeeding. If you'd like to override that behavior, comment assign set of reviewers

Run Java PreCommit

I think task fails is not related to this PR.

Run Java PreCommit

Assigning reviewers. If you would like to opt out of this review, comment assign to next reviewer:

R: @damccorm for label build.

Available commands:

• stop reviewer notifications - opt out of the automated review tooling
• remind me after tests pass - tag the comment author after tests pass
• waiting on author - shift the attention set back to the author (any comment or push by the author will return the attention set to the reviewers)

The PR bot will only process comments in the main thread (not review comments).

LGTM overall, just one question from higher-level perspective. Given that the only partitioner that is ever passed to this method is returned by

  private static @Nullable Partitioner getPartitioner(SparkTranslationContext context) {
    Long bundleSize =
        context.serializablePipelineOptions.get().as(SparkPipelineOptions.class).getBundleSize();
    return (bundleSize > 0)
        ? null
        : new HashPartitioner(context.getSparkContext().defaultParallelism());
  }

does this change have any practical impact? I would expect that the HashPartitioner with default parallelism is what would be used when the partitioner is unused, is this assumption wrong?

Yes, as you mentioned, following the code ultimately leads to using the HashPartitioner.
However, I assumed that the partitioner would be passed as a parameter to the GroupNonMergingWindowsFunctions#groupByKeyInGlobalWindow method, so that it could be used if necessary.
While the current changes may not have a significant practical impact, this decision was made with future extensibility in mind, allowing the use of custom partitioners for specific Apache Beam PTransforms.
Of course, this may not be necessary in the end, but I wanted to leave room for this kind of flexibility.
Thank you.

Thanks @twosom!