test(integration): stabilize watch helper using run-completion sync

[mostafaNazari702]

PR Checklist

• Addresses an existing open issue: fixes 🛠️ Repo: CI run failures (test failures) #5361
• That issue was marked as status: accepting prs
• Steps in CONTRIBUTING.md were taken

Overview

Replace fixed-sleep timing in runMochaWatchAsync with stdout synchronisation to remove flaky watch integration tests.
old approach relied on sleeps around the "change" callback which could race with incomplete runs on CI causing inconsistent test results.

we buffer instead stdout to reliably detect "stats" run-completion markers across chunk boundaries and replace the multi-waiter loop with a single progress waite, and snapshots the run baseline before triggering "change" to prevent race conditions.
an expectedRuns option is added to avoid unnecessary rerun waits in negative tests and tests are updated to use it where appropriate.

Now we have faster and more stable watch tests (from 120 seconds down to, from my test, 51) with all integration tests still passing.

EDIT:

from 120 seconds down to, from my test, 51
I actually ran a focused test rather than the full integration suite. Time still same. I apologize for the mistake.

[mostafaNazari702]

Wrong branch.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 81.00%. Comparing base (6695fba) to head (716b6ec).

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #5879      +/-   ##
==========================================
+ Coverage   80.89%   81.00%   +0.10%     
==========================================
  Files          64       64              
  Lines        4602     4602              
  Branches      976      976              
==========================================
+ Hits         3723     3728       +5     
+ Misses        879      874       -5     

☔ View full report in Codecov by Sentry.
📢 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.

[mostafaNazari702]

I was actually right, thought that i pushed code to the previous active PR that i have. Sorry.
I would also like to mention that the PR body was rewritten from my notes, which normally sound like gibberish. using AI.

[mark-wiemer]

Nice, this could finally close that issue :) updated the branch to get another run to see if there is flakiness left, thank you :)

[JoshuaKGoldberg]

@mostafaNazari702 the most recent CI run has integration failures ☹️ can you investigate? It looks like this might not have actually fixed the flake?

[mostafaNazari702]

Accidental push without committing closed the PR, excuse me.
Anyway, looked into and rewrote the helper synchronisation.