Woo-dashboard — tests¶
Frontend: 15 files · 77 tests. E2E: none yet.
The Woo dashboard is the most thinly tested of the four boards by test count, while reporting high coverage percentages. Both things are true at once, and the combination is worth understanding before reading the numbers as reassurance.
Frontend¶
Re-derived on 9 October 2026 at ebec288 (v2026.10.1), where every count
and coverage figure below is the same as on 30 September and 3 October —
neither v2026.10.0 nor v2026.10.1 touched this board. The frontend package was
measured the same day — 130 files, 1540 tests, all passing — but Vitest's
console reports only that total, so the counts below are taken from the
source, each test file parsed with its .each tables expanded; summed over
the whole package the method gives exactly the runner's 1540.
| Area | Files | Tests |
|---|---|---|
components/WooDashboard |
12 | 43 |
pages/woo (data and config) |
2 | 22 |
pages/WooDashboard.test.tsx |
1 | 12 |
| File | Tests | Covers |
|---|---|---|
pages/woo/woo.data.test.ts |
19 | The board's data module |
pages/WooDashboard.test.tsx |
12 | The page container |
components/WooDashboard/Register.test.tsx |
8 | The register section |
components/WooDashboard/WooCommandPalette.test.tsx |
8 | Command palette |
components/WooDashboard/charts.test.tsx |
7 | Chart rendering |
components/WooDashboard/WooSectionRouter.test.tsx |
7 | Section routing |
pages/woo/modes.config.test.ts |
3 | Mode configuration |
components/WooDashboard/Bezwaar.test.tsx |
2 | Objections |
components/WooDashboard/Proces.test.tsx |
2 | Process view |
components/WooDashboard/Publicatie.test.tsx |
2 | Publication |
components/WooDashboard/Verzoeken.test.tsx |
2 | Requests |
components/WooDashboard/WooDock.test.tsx |
2 | The dock |
components/WooDashboard/Overzicht.test.tsx |
1 | Overview |
components/WooDashboard/Tijdigheid.test.tsx |
1 | Timeliness |
components/WooDashboard/WooNoAccessPanel.test.tsx |
1 | The no-access panel |
woo.data.test.ts (+4) and WooDashboard.test.tsx (+3) grew in v2026.09.15's
branch-margin work; the other counts are as they were on 28 September.
Coverage on 9 October, as on 30 September and 3 October: components/WooDashboard
98.26 / 94.78 / 98.52 / 98.57 and pages/woo 99.13 / 100 / 100 / 99 —
the latter up from 96.55 / 84.12 / 94.44 / 98.01.
High coverage, few assertions
Eight of the twelve component files carry one or two tests each. Those render the section and assert it renders — which executes nearly every line in a presentational component and so scores 98%, without asserting much about what it renders.
That is a legitimate scoping choice for presentational sections, and it is
the same "critical interactions only" approach used across the frontend.
But it means this board's coverage percentage is a weaker signal than the
identical percentage on, say, pages/infra-board, where 108 tests across
six files are asserting real behaviour. Read the test count alongside the
percentage.
E2E¶
None. There is no Playwright spec that drives this board — verified against
packages/frontend/e2e/ on 30 August 2026, at ae06c9e on 30 September, in
the full frontend run of 3 October and in that of 9 October, whose 47 tests
include none that drives this board — login-redirect only checks that
test-woo-flevoland lands on it, and v2026.10.1's landing-access-denied
opens the Woo card as a caseworker to assert the landing page's
Geen toegang tot Woo-dashboard dialog, which is the landing page's
behaviour, not the board's — not inferred from a changelog. It is still the
only board in that position; see
Coverage per board.
As with Infra-board, this is a gap rather than a decision, and the case for closing it is a little stronger here: the combination of few assertions and no end-to-end coverage means a broken write path or an empty panel would pass everything currently in place.
The PA cockpit suite is the model — mock mode driven against the real store, no mocking, deterministic fixtures. See PA cockpit → Why mock mode is worth an E2E suite.