chore(recorder): move fill coalescing back to collapseActions

[Dmitry Gozman (dgozman)]

Summary

Partially reverts #41958:

• RecorderSignalProcessor no longer buffers fill actions; click and navigate coalescing stays there.
• Brings back shouldMergeAction()/collapseActions() (fill-only now) and the actionUpdated protocol event.

#41958 forced quick typing into the same input - a pause over 500ms would generate two separate fill() actions, which is a behavior change we don't necessarily need to make yet.

[github-actions]

Test results for "tests 1"

10 flaky

⚠️ [chromium-library] › library/chromium/chromium.spec.ts:179 › serviceWorker(), and fromServiceWorker() work `@frozen-time-library-chromium-linux`
⚠️ [chromium-library] › library/browsertype-connect.spec.ts:714 › run-server › should record trace with sources `@chromium-ubuntu-22.04-arm-node20`
⚠️ [chromium-library] › library/beforeunload.spec.ts:130 › should support dismissing the dialog multiple times `@chromium-ubuntu-22.04-node24`
⚠️ [chromium-library] › library/popup.spec.ts:260 › should not throw when click closes popup `@chromium-ubuntu-22.04-node24`
⚠️ [chromium-library] › library/video.spec.ts:736 › screencast › should work with video+trace `@chromium-ubuntu-22.04-node24`
⚠️ [chromium-library] › library/video.spec.ts:664 › screencast › should capture full viewport `@chromium-ubuntu-22.04-node20`
⚠️ [chromium-library] › library/video.spec.ts:699 › screencast › should capture full viewport on hidpi `@chromium-ubuntu-22.04-node22`
⚠️ [firefox-library] › library/heap.spec.ts:223 › should not leak workers `@firefox-ubuntu-22.04-node20`
⚠️ [firefox-page] › page/page-event-request.spec.ts:181 › should return response body when Cross-Origin-Opener-Policy is set `@firefox-ubuntu-22.04-node20`
⚠️ [playwright-test] › ui-mode-test-update.spec.ts:202 › should update test locations `@ubuntu-latest-node22`

51243 passed, 1240 skipped

---

Merge workflow run.

[github-actions]

Test results for "MCP"

3 failed
❌ [chromium] › mcp/clipboard.spec.ts:19 › clipboard write without permission dialog @mcp-macos-latest-chromium
❌ [chromium] › mcp/clipboard.spec.ts:19 › clipboard write without permission dialog @mcp-windows-latest-chromium
❌ [chromium] › mcp/clipboard.spec.ts:19 › clipboard write without permission dialog @mcp-ubuntu-latest-chromium

8256 passed, 1361 skipped

---

Merge workflow run.

[github-actions]

Hi, I'm the Playwright bot and I took a look at the failing CI.

🟢 CI is clear — the one failure is pre-existing, not from this PR

The only real failures are mcp/clipboard.spec.ts:19 › clipboard write without permission dialog on Chromium (ubuntu/macos/windows). The same test fails identically on this PR's base commit 6748802c (a push to main) and on a string of unrelated PRs, so it's already broken on trunk. This PR only reworks recorder fill coalescing and doesn't touch the MCP clipboard/permissions path. The tests 1 report had no failures — just flakes.

Details

Overall: nothing in this diff explains the failures. It moves fill coalescing out of RecorderSignalProcessor back into collapseActions()/shouldMergeAction() and re-adds the actionUpdated recorder event — all recorder codegen. The failing test exercises browser_navigate + browser_evaluate clipboard permissions, which this PR never reaches.

Pre-existing flake / infra

• [chromium] › mcp/clipboard.spec.ts:19 › clipboard write without permission dialog (macos / windows / ubuntu) — pre-existing failure, not caused by this PR. Same expect(...).toEqual(...) deep-equality mismatch on all three OSes across every recent run, including this PR's own base sha 6748802c (push to main) and PRs feat(firefox): roll to r1541 #42441, chore(chromium): switch linux arm64 to cft builds #42440, chore(trace): remove snapshot pointers from action events #42431, feat(chromium): roll to r1243 #42434, feat(chromium): roll to r1242 #42356. In the test-results DB it's failed the last ~27 runs consistently on those three bots regardless of PR — a broken-on-main test, not something this branch introduced.

No real failures elsewhere

• The tests 1 report shows 10 flaky, 0 failed (51243 passed). Nothing to triage there.

_{Triaged by the Playwright bot - agent run}