fix(highlight): highlight in the world the selector resolves in

[Yury Semikhatsky (yury-s)]

Summary

• Custom selector engines registered without contentScript run in the main world, but HighlightController only rendered highlights in the utility world, so locator.highlight() matched nothing for them
• Resolve each highlight in the world its selector needs and clear both worlds in hideHighlights()

Fixes #42849

[github-actions]

Test results for "tests 1"

6 flaky

⚠️ [chromium-library] › library/video.spec.ts:725 › screencast › should capture full viewport on hidpi `@chromium-ubuntu-22.04-arm-node20`
⚠️ [chromium-library] › library/video.spec.ts:690 › screencast › should capture full viewport `@realtime-time-library-chromium-linux`
⚠️ [chromium-library] › library/beforeunload.spec.ts:130 › should support dismissing the dialog multiple times `@chromium-ubuntu-22.04-node22`
⚠️ [chromium-library] › library/video.spec.ts:762 › screencast › should work with video+trace `@chromium-ubuntu-22.04-node22`
⚠️ [firefox-library] › library/browsercontext-cookies-third-party.spec.ts:257 › third party 'Partitioned;' cookies `@firefox-ubuntu-22.04-node20`
⚠️ [firefox-library] › library/browsercontext-cookies-third-party.spec.ts:470 › top level 'Partitioned;' cookie and same origin iframe `@firefox-ubuntu-22.04-node20`

52087 passed, 1252 skipped

---

Merge workflow run.

[github-actions]

Test results for "MCP"

4 failed
❌ [chrome] › mcp/cli-devtools.spec.ts:285 › video-show-actions and video-hide-actions @mcp-macos-latest-chrome
❌ [firefox] › mcp/cli-drag.spec.ts:19 › drag between elements @mcp-windows-latest-firefox
❌ [firefox] › mcp/cli-session.spec.ts:54 › idle timeout shuts the session down @mcp-windows-latest-firefox
❌ [msedge] › mcp/http.spec.ts:145 › http transport browser sigint @mcp-windows-latest-msedge

8688 passed, 1474 skipped

---

Merge workflow run.

[github-actions]

🟡 Nothing here points at this PR, but one failure I can't fully clear

Hi, I'm the Playwright bot and I triaged the failing CI runs.

Three of the four MCP failures are known flakes with history on unrelated SHAs. The fourth — mcp/cli-devtools.spec.ts:285 on Chrome — is a bare 30s timeout I can't tie to anything, and it has no prior history, so I'm leaving it as uncertain rather than calling it clean. The tests 1 report has no failures, only 6 flaky.

Details

This PR touches highlightController.ts (resolving highlights in the selector's world) plus one new test in tests/library/locator-highlight.spec.ts. None of the four failing tests exercise highlighting.

Pre-existing flake / infra

• [firefox] › mcp/cli-session.spec.ts:54 › idle timeout shuts the session down @mcp-windows-latest-firefox — a well-established flake: failed 25 of 173 runs (14%), passed the other 148, across many unrelated PRs (fix(fetch): apply proxy.bypass to redirect targets #42848, feat(content): add includeShadow option to page.content() #42845, fix(firefox): preserve response header field boundaries #42832, fix(mcp): rebind page-registered WebMCP tools when the tab or frame changes #42821, …) and pushes to main (52a0e72, 9ae5377, b8627f9, d1ead3e, …). Windows/Firefox only.
• [firefox] › mcp/cli-drag.spec.ts:19 › drag between elements @mcp-windows-latest-firefox — failed 7 of 723 runs, on unrelated PRs (feat(trace-viewer): show screencast frames during playback #42751, fix(reporter): fail perfetto writes when the output stream errors #42715, feat(mcp): add media emulation #42668, fix(trace-viewer): label filter inputs and keyboard-accessible error source #42469, fix(debug-controller): drop recorded actions when recording session ends #42458) and pushes to main (b856094, 0a2a33c). Windows/Firefox only.
• [msedge] › mcp/http.spec.ts:145 › http transport browser sigint @mcp-windows-latest-msedge — failed 5 of 714 runs, including pushes to main (b8627f9, 5967263, 52a0e72) and PR fix(net): apply socketTimeout from socket creation in httpRequest #42660. Signal-handling on Windows; this PR can't reach it.

Uncertain

• [chrome] › mcp/cli-devtools.spec.ts:285 › video-show-actions and video-hide-actions @mcp-macos-latest-chrome — Test timeout of 30000ms exceeded., no assertion error. The test only checks CLI output strings for video-show-actions / video-hide-actions; it never highlights anything, so the diff has no path to it and the signature reads like a stuck CLI/browser launch. But this is the only failure of this test in 732 runs in the aggregated results, and all of them are on this PR — so I can't cite it failing anywhere the PR isn't responsible for, and the flake rule says that's not proof. A re-run on the same sha, or one green run of this test elsewhere, would settle it.

_{Triaged by the Playwright bot - agent run}