Fixes #32968: focus the entity-type trigger before opening its listbox

[ShaileshParmar11]

Describe your changes:

Fixes #32968

TestLibrary.spec.ts has been ejecting PRs from the merge queue via playwright-summary — most recently #32524 (run, shard chromium-07), a PR that touches nothing in this path. Two tests fail on the same line, clicking the TABLE option in the entity-type dropdown.

Root cause. The listbox dismisses itself ~160 ms after it opens, deterministically — no CI load required:

295ms  focusin   button#root/entityType   ← the click
306ms  listbox+                           ← options appear
421ms  focusout  div{listbox}
426ms  listbox-                           ← destroyed, and nothing reopens it

React Aria's listbox is a non-modal popover. A press on a trigger that does not already hold focus both opens it and produces a focus transition, and that transition dismisses it a frame later. The trigger and form nodes are not remounted, so this is not an RJSF re-render; it is also unrelated to the doc panel (forcing every programmatic scroll instant changes nothing, and the teardown is identical with no fields filled).

Measured, 3/3 runs each:

Gesture	Listbox
locator.click() on an unfocused trigger	closes after ~160 ms
focus() + assert focused + click()	stays open
raw mouse.click() on an unfocused trigger	closes at ~0 ms
focus() + Enter	stays open

So the four cold-click sites have always had a ~160 ms window. A dev machine wins that race; a loaded 2-core runner does not — and since nothing reopens the listbox, option.click() retries until the test timeout.

The change. Four sites opened the Select on an unfocused trigger (:156, :375, :876, :1219); a fifth (:635-650) already focused first and passed in the same CI run that failed the other two. All five collapse into one selectEntityType helper built on that working sequence — net +27 / −42.

Type of change:

• Bug fix

Checklist:

• I have read the CONTRIBUTING document.
• My PR is adequately covered with tests

Manual test steps / verification:

Page-side CPU throttling does not reproduce this, because CI's lag is on the driver side. Injecting a delay between the trigger press and the option click reproduces it exactly:

Injected driver lag	Before	Signature
100 ms	2/2 fail	element is not stable + element was detached from the DOM — CI's exact first-attempt text
140 / 170 / 220 / 400 ms	fail	waiting for getByRole('option', …)

With the trigger focused first, the same scenario passes at 400 ms and at 1500 ms of injected lag — 10× the window.

Running the spec at 4 workers against a local stack:

Run	Result
origin/main unmodified	2 failed, 13 passed — incl. should maintain page on edit and reset to first page on delete, one of the two CI failures
this branch	1 failed, 14 passed — only a platform badges failure that also fails on unmodified origin/main and is unrelated

The two CI-affected tests, --repeat-each=3 --workers=4 on this branch: 6/6 passed.

Lint and types: the CI ui-checkstyle playwright sequence (organize-imports → eslint --fix → prettier) is clean and idempotent on the changed file, and tsc --project playwright/tsconfig.json reports 0 errors in it (165 pre-existing elsewhere in the project, untouched).

Note for reviewers — possible product bug

With a human-speed press (60 ms hold) rather than Playwright's instantaneous down/up, the listbox closed at ~0 ms on both the first and second click, 3/3. If that holds for real hardware input, the entity-type dropdown does not open on a normal mouse click and the real fix belongs in the ui-core-components Select. I did not confirm this by hand and CDP-injected press-and-hold may not match real input, so it is captured as a note in #32968 rather than a claim — but there are 118 getByRole('option') call sites across 38 spec files sitting on the same behaviour, so it is worth someone clicking it.

🤖 Generated with Claude Code

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

[github-actions]

✅ PR checks passed

The linked issue has a description and all required Shipping project fields set. Thanks!

[gitar-bot]

Code Review ✅ Approved

Fixes a race condition in TestLibrary.spec.ts where the entity-type dropdown listbox dismisses itself ~160 ms after opening when clicked on an unfocused trigger. Consolidates four separate click sites into a single selectEntityType helper that focuses the trigger before opening the listbox, resolving the flaky test failures (−42 lines, +27 lines net). Verified to pass the two previously failing tests at 6/6 with --repeat-each=3 --workers=4.

Options

Display: compact → Showing less information.

Comment with these commands to change the behavior for this request:

Compact

gitar display:verbose

_{Was this helpful? React with 👍 / 👎 | Powered by Gitar — free for open source}

[github-actions]

✅ Playwright Results — workflow succeeded

Validated commit 2194667887e2c57a4df1e503b61a7cd09a89a15f in Playwright run 34244477822, attempt 1.

✅ 61 passed · ❌ 0 failed · 🟡 0 flaky · ⏭️ 0 skipped · 🧰 0 lifecycle flaky

Performance

Blocking targets: ✅ met · Optimization targets: 🟡 in progress

Shard-job maxima below are not the full workflow wall time; the linked run includes build, fixture, planning, and reporting.

🕒 Full workflow signal wall (to summary) 35m 51s

⏱️ Max setup 3m 19s · max shard execution 6m 49s · max shard-job elapsed before upload 11m 1s · reporting 2s

🌐 140.15 requests/attempt · 1.46 app boots/UI scenario · 0.00% common-shard skew

Optimization targets still in progress:

• Application boot ratio was 1.46 per UI scenario (92 boots / 63 scenarios; convergence target: at most 1).

Shard	Passed	Failed	Flaky	Skipped	Lifecycle failed	Lifecycle flaky
✅ Shard chromium-01	61	0	0	0	0	0

📦 Download artifacts

How to debug locally

# Download playwright-test-results-<shard> artifact and unzip
npx playwright show-trace path/to/trace.zip    # view trace