Fix DataflowOutputCounter calculation for ValueInEmptyWindows

[shunping]

When processing shuffle or streaming data in Dataflow Legacy Runner (e.g., from GroupingShuffleReader or WindowingWindmillReader), KeyedWorkItems are wrapped inside a ValueInEmptyWindows (windows.size() == 0).

Previously, DataflowOutputCounter.update() counted these as 1 element. This caused inaccurate element counts because:

1. A KeyedWorkItem can contain multiple elements.
2. Elements may belong to multiple windows and need to be fanned out accordingly.
3. KeyedWorkItems containing only timers were incorrectly incrementing element counters.

Internal bug id: 509652238

[gemini-code-assist]

Caution

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

[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 added as fallback since no labels match configuration

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).

[kennknowles]

Based on the description of the problem, would it make sense to find a way to factor the code so we can have separate streaming/batch implementation of the output counter? The ways they pass elements is just pretty fundamentally different.

[shunping]

Based on the description of the problem, would it make sense to find a way to factor the code so we can have separate streaming/batch implementation of the output counter? The ways they pass elements is just pretty fundamentally different.

The difference between the streaming and batch output counters really only matters during the GBK step, specifically regarding what gets wrapped inside ValueInEmptyWindows. Per your comment, I updated the PR to separate the implementations. Particularly I subclassed DataflowOutputCounter and extracted the specific routine that handles the window_size == 0 case.

PTAL! Thanks!

[codecov]

Codecov Report

❌ Patch coverage is 62.90323% with 23 lines in your changes missing coverage. Please review.
✅ Project coverage is 55.81%. Comparing base (dca1253) to head (c1fb2d6).
⚠️ Report is 100 commits behind head on master.

Files with missing lines	Patch %	Lines
...runners/dataflow/worker/WindmillKeyedWorkItem.java	41.66%	13 Missing and 1 partial ⚠️
...ataflow/worker/StreamingDataflowOutputCounter.java	68.75%	3 Missing and 2 partials ⚠️
...rs/dataflow/worker/BatchDataflowOutputCounter.java	66.66%	2 Missing ⚠️
...runners/dataflow/worker/DataflowOutputCounter.java	71.42%	2 Missing ⚠️

Additional details and impacted files

@@             Coverage Diff              @@
##             master   #39487      +/-   ##
============================================
- Coverage     58.05%   55.81%   -2.24%     
+ Complexity    13057     2272   -10785     
============================================
  Files          2518     1115    -1403     
  Lines        263812   174219   -89593     
  Branches      10765     1460    -9305     
============================================
- Hits         153164    97246   -55918     
+ Misses       104899    74518   -30381     
+ Partials       5749     2455    -3294     

Flag	Coverage Δ
java	75.54% <62.90%> (+11.31%)	⬆️

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.

[scwhittle]

LGTM with small comment
Thanks!