[Spark] Support splittable DoFn self-checkpointing in portable batch

[Eliaaazzz]

Addresses #19468 for batch.

The problem

The portable Spark runner never passed a BundleCheckpointHandler to StageBundleFactory.getBundle, so a splittable DoFn that returns ProcessContinuation.resume() failed on its first bundle. Several SDF tests were skipped for this runner as a result.

The change

A batch stage containing a splittable DoFn now holds each residual in memory under a processing time timer, using the existing BundleCheckpointHandlers.StateAndTimerBundleCheckpointHandler with InMemoryTimerInternals and InMemoryStateInternals. This is the same model the portable Flink batch runner uses in FlinkExecutableStageFunction. Once the stage has drained its inputs, processing time advances to infinity and the held residuals are replayed until the SDK stops asking to resume, so a bounded restriction always runs out.

In the branch that also has user timers, the two drains alternate, since a replayed residual could set a timer and a fired timer could produce a residual. The fuser does not appear to produce a stage holding both today, so that loop is defensive.

What is deliberately still unsupported

• Streaming self-checkpointing. A residual has nowhere to live across micro-batches, so the streaming path rejects it with a message naming Spark portable runner: support SDF #19468. This replaces today's failure with a clearer one rather than changing behaviour.
• Unbounded splittable DoFn, which needs the above. UsesUnboundedSplittableParDo stays excluded and test_unbounded_source_read stays skipped.
• Bundle finalization (Spark portable runner: support Bundle Finalization #19517). It is now rejected explicitly rather than run early: InMemoryFinalizer is documented to run after the runner has durably committed the bundle's output, which this runner cannot report. The portable Flink batch runner rejects it the same way. UsesBundleFinalizer stays excluded.

I had an earlier revision of this PR that also added a driver-side relay to carry residuals across micro-batches in streaming. I took it out: it depended on the staged RDD executing exactly once, which persist(MEMORY_ONLY) does not guarantee, and its watermark could stall for a stage that emits output without ever checkpointing. That is worth doing properly rather than approximately, so it is left for follow-up.

Testing

SparkExecutableStageFunctionTest.sdfResidualsAreReplayedUntilDrained drives a stage whose first bundle self-checkpoints with one residual and whose replay bundle returns none, and asserts that the original element and then the residual both reach the input receiver. It fails if the replay loop is short-circuited.

Unskips the bounded SDF tests in spark_runner_test.py and removes the UsesBoundedSplittableParDo exclusion from the portable ValidatesRunner task.

---

Thank you for your contribution! Follow this checklist to help us incorporate your contribution quickly and easily:

• Choose reviewer(s) and mention them in a comment (R: @username).
• Mention the appropriate issue in your description (for example: addresses #123), if applicable. This will automatically add a link to the pull request in the issue. If you would like the issue to automatically close on merging the pull request, comment fixes #<ISSUE NUMBER> instead.
• Update CHANGES.md with noteworthy changes.
• If this contribution is large, please file an Apache Individual Contributor License Agreement.

See the Contributor Guide for more tips on how to make review process smoother.

To check the build health, please visit https://github.com/apache/beam/blob/master/.test-infra/BUILD_STATUS.md

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 54.76%. Comparing base (de21b1c) to head (b1428e7).

Additional details and impacted files

@@            Coverage Diff            @@
##             master   #39331   +/-   ##
=========================================
  Coverage     54.76%   54.76%           
  Complexity     1716     1716           
=========================================
  Files          1066     1066           
  Lines        169075   169075           
  Branches       1255     1255           
=========================================
+ Hits          92589    92591    +2     
+ Misses        74269    74267    -2     
  Partials       2217     2217           

Flag	Coverage Δ
python	79.71% <ø> (+<0.01%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[gemini-code-assist]

Caution

The consumer version of Gemini Code Assist on GitHub has been sunset. All code review activity has officially ceased.

[github-actions]

Assigning reviewers:

R: @jrmccluskey for label python.

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

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).