Test-writing discipline

Correct behavior is the goal, not green. A test that asserts nothing, or asserts the buggy output, certifies the bug. Fix a red test by correcting the code or tightening the test. These rules apply on top of handsontable-playwright-e2e, handsontable-e2e-testing, handsontable-unit-testing.

Intent, not implementation

Write the test from the requirement (what the user or API is supposed to get), ideally before the implementation. For E2E, state the user-observable expectation before wiring a selector.

When a test is red

The code is the prime suspect. Check whether the expectation matches the intended behavior:

• Expectation correct, code wrong: fix the code, leave the test.
• Expectation mis-encoded the intent: tighten the test toward the real behavior.
• Cannot tell: re-read the requirement.

Banned ways of faking green

Reach green only by fixing code or tightening the test. Reject:

• Deleting or loosening an assertion, or widening a tolerance, to match the output.
• .skip / xit / xdescribe, or .only / test.only / describe.only / fit / fdescribe (focusing silently drops the rest of the suite).
• try/catch swallowing a failure.
• Asserting whatever the code produced (a "snapshot of the bug").
• it.flaky / retries over a real intermittent failure.

The Stop hook and pre-push run the test you touched and block on red; that forces reconciliation, it does not authorize weakening.

Weakening over an admitted race needs a ticket in the same PR

An assertion weakened or deleted with a race as the reason ("held 15/15 locally", "flakes on classic", "racy on develop") ships only with a filed ticket in the same PR, named beside the change. "Still open, no ticket" is a review blocker (#13332).

"Passes on retry" and "passes in isolation" are not determinism evidence

• Before calling a test deterministic, run it focused across every leg: cd tests && HOT_E2E_FULL_MATRIX=1 npx playwright test e2e/<spec>.spec.ts --repeat-each 20 (no --project filter, and the variable makes all six theme × bundle legs run it whatever the spec declares; without it only the declared legs run). For a test tagged @cross-browser, repeat it on the engine legs too: npx playwright test --config playwright-engines.config.ts e2e/<spec>.spec.ts --repeat-each 20. The hidden-init migration needed ~700 such runs to expose three rAF-starvation timeouts.
• A test failing about half the time under that load on unchanged develop reports a product race: file a product ticket and keep the assertion.

Bug fixes: failing test first

1. Reproduce the bug as a test and watch it fail for the right reason (the missing behavior, not a typo or bad selector).
2. Apply the fix and watch the same test pass.
3. On a bugfix PR, name the spec that fails without the fix.

Cover what the change adds

• Verify a timing-semantics claim against the primitive. setTimeout(0) is the next macrotask, not end-of-task, so a JSDoc saying "closes when the task ends" above it is wrong; that mismatch shipped a data-corruption bug (an editor stranded by alter() committed through stale coordinates, #13332). Read the primitive (setTimeout, queueMicrotask, requestAnimationFrame, a hook's call site) and test what it does.
• A dedicated X-during-Y path needs a test that drives X during Y. flush, drain, cancel, pending, suspend, a depth counter are the tell. Fire X from inside Y: a hook callback, a nested alter(), an edit inside a batch.
• Execute every documented form of a new option at least once, the guide's lead form first.

Before saying "done"

• Run the exact test command fresh, read the full output and exit code, state the result with that evidence. Say so when you did not run it.
• Banned phrasings: "should work", "this fixes it" without a run, "tested manually, looks fine".
• After editing source, run the impacted test and confirm green.

No hollow assertions

• Assert the behavior: expect(getDataAtCell(0, 0)).toBe('x'), not expect(true).toBe(true) and not an it() with no expect at all.
• Coverage-on-new-code can be satisfied by a line executed without a checked result, so judge quality by the assertion.

Mocking

• Mock at the real boundary (network, timers, ResizeObserver; see test/__mocks__/), never the unit under test, and mock the complete real data shape.
• Where a unit test would need to mock a module, write an E2E test.

Broken legacy tests

The Jasmine suite is frozen. Rewrite a broken or flaky legacy *.spec.js as a Playwright test in tests/e2e/ and delete the Jasmine one (no sleep() patches). Routine, low-risk edits to an existing Jasmine spec are fine.