Allow sequences (tuples and lists) as pivot values argument in PySpark.

[wrobell]

Both tuples and lists are accepted by PySpark on runtime.

[holdenk]

Jenkins ok to test

[holdenk]

This looks reasonable to me, I'm not very familiar with the typing code for Python yet so cc @zero323

[SparkQA]

Test build #140338 has finished for PR 33083 at commit 8187108.

• This patch passes all tests.
• This patch merges cleanly.
• This patch adds no public classes.

[SparkQA]

Kubernetes integration test unable to build dist.

exiting with code: 1
URL: https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder-K8s/44869/

[zero323]

Technically speaking, Sequence is more than Tuple or List (I mention that, because we have quite a few cases where we explicitly restrict inputs to these two), but Py4J should be able map any Sequence in the same way (I've done some rough testing on custom Sequence implementation to be sure, and it seems to work fine).

[zero323]

LGTM, subject to passing tests.

[AmplabJenkins]

Can one of the admins verify this patch?

[HyukjinKwon]

@wrobell, can you file a JIRA (see https://spark.apache.org/contributing.html), and enable GitHub Actions in your fork repo (see https://github.com/apache/spark/pull/33083/checks?check_run_id=2913538608)?

Also please keep the GIthub PR template (https://github.com/apache/spark/blob/master/.github/PULL_REQUEST_TEMPLATE) and format PR title properly.

[HyukjinKwon]

Otherwise, looks fine to me too. I'll leave it to him.

[zero323]

Otherwise, looks fine to me too. I'll leave it to him.

Sure, I'll handle this once pending comments are addressed.

[zero323]

Gentle ping @wrobell

[wrobell]

Sorry, but due to personal circumstances I will not be able to help with this for next couple of weeks.

[dchvn]

any update here? python/pyspark/sql/group.pyi has been removed by #34197, so can I create a JIRA ticket and a PR for this issue? @HyukjinKwon @zero323 @wrobell
Thanks!