Ramadhan Sjamsani
fbc94daac7
Brings the mitra app to figma-bestie parity for Home (§1), Undangan
inbox with Curhat Baru + Perpanjang tabs (§2), and the incoming-popup
+ active-chat flow (§3). Home now lives inside a StatefulShellRoute
with BestieTabBar so Profil + Undangan + Home share one shell.
- Shell: features/shell/ (StatefulShellRoute, BestieTabBar, 3 branches)
- Undangan: features/undangan/ — Curhat Baru reads
chatRequestProvider.pendingInvites; row Terima delegates accept to
the notifier and ChatRequestOverlay owns nav (no double-push).
Perpanjang tab stubbed (empty state) until backend exposes
pendingExtensionsProvider.
- Profil: features/profile/ — Bestie-styled stub
- Home: refactored to body-only (shell owns chrome)
- Popup: chat_request_overlay + chat_request_notifier updated to
serve the list rows, not just the modal
- Chat: mitra_chat_screen polish
- Theme: accentAmber tokens for the Perpanjang tab + halo_orb widget
(loading spinner used by undangan list states)
- Login: replace broken GoRouterState location guard with
_expectOtpPush flag — was stacking duplicate /otp pages on OTP
resend (see project-otp-nav-bug-fixed-2026-05-21)
Maestro:
- 17 new flows under .maestro/flows/ts-mitra-{1,2,3}-* covering home
online/offline variants, undangan empty/populated/tolak states,
popup curhat-baru → accept → chat → ended banner, plus popup
dismiss/expire/cancelled edge cases
- 4 new §A OTP flows (07/08/09/10) for invalid/mismatch/expired/cooldown
- Helper scripts: force_mitra_online/offline, force_pairing_timeout,
force_session_expires_at, delete_mitra_status_row,
customer_blast_now (js), customer_cancel_latest_blast
- Backend: POST /internal/_test/delete-mitra-status-row supports the
"fresh mitra with no status row" test setup
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
121 lines
5.7 KiB
Markdown
121 lines
5.7 KiB
Markdown
# §A — Mitra Pre-Home (auth) test plan
|
||
|
||
Spec: [requirement/flow_mitra.mermaid.md §A](../../../requirement/flow_mitra.mermaid.md).
|
||
|
||
Tests use the naming convention `ts-mitra-<section>-<sub>-<description>.yaml`:
|
||
- `<section>` — flow_mitra.mermaid section identifier (`A` for pre-home auth).
|
||
- `<sub>` — sub-flow index within the section, zero-padded.
|
||
- `<description>` — snake_case summary of the branch under test.
|
||
|
||
## Implemented
|
||
|
||
| File | Branch (spec ref) | Expected destination |
|
||
|---|---|---|
|
||
| `ts-mitra-A-01-phone_invalid_inline_error.yaml` | §A.1 local CTA gate for short phones | "kirim kode" disabled for <9 subscriber digits, enables + navigates on valid input |
|
||
| `ts-mitra-A-03-success_login_to_home.yaml` | §A.2 200 + `is_active=true` | `/home` (BestieHome online; asserts "Kamu lagi ONLINE" — Stage 2 removed the Sesi Aktif / Riwayat Chat tiles) |
|
||
| `ts-mitra-A-04-account_inactive_screen.yaml` | §A.2 200 + `is_active=false` → 403 `ACCOUNT_INACTIVE` | `/auth/inactive` (AppBar back arrow omitted; system-back interception deferred) |
|
||
| `ts-mitra-A-05-phone_format_variants.yaml` | §A.1 phone-format normalization | 5 variants (`8…`, `08…`, `628…`, `+628…`, `0628…`) all reach S3b with `+628200000501` shown |
|
||
| `ts-mitra-A-06-back_to_login_and_retry.yaml` | §A.1 regression — back from S3b + re-submit | Second "kirim kode" tap still navigates to S3b after returning to S3a |
|
||
|
||
## Test infrastructure
|
||
|
||
JS scripts (live in [`../scripts/`](../scripts/)) called via `runScript:`:
|
||
|
||
| Script | Purpose | Backend endpoint |
|
||
|---|---|---|
|
||
| `reset_phone.js` | Clear `otp_requests` rows for `TEST_PHONE` so cooldown / phone-rate-limit don't trip on re-runs. | `POST /internal/_test/reset-phone` |
|
||
| `seed_mitra.js` | Upsert a mitra row with given phone + `is_active`. Idempotent. | `POST /internal/_test/seed-mitra` |
|
||
| `peek_otp.js` | Read the latest stub-generated OTP code for `TEST_PHONE`. Writes to `output.OTP`. | `GET /internal/_test/peek-otp` |
|
||
|
||
All three only exist on the **internal** listener (port 3001), so they're
|
||
network-isolated from production traffic.
|
||
|
||
### Phone-number convention
|
||
|
||
Each test uses a unique `+628200000<NN><SS>` phone to avoid cross-flow
|
||
interference:
|
||
- A-02 (deferred) → `+628200000201`
|
||
- A-03 → `+628200000301`
|
||
- A-04 → `+628200000401`
|
||
- A-05 → `+628200000501` (one phone, 5 input formats)
|
||
- A-06 → `+628200000601`
|
||
- §1 Home (ts-mitra-1-*) → `+62820000070{1..3}`
|
||
- §2 Undangan (ts-mitra-2-*) → `+62820000080{1..2}` (2-03 piggybacks on a
|
||
pre-signed-in device, no fresh OTP)
|
||
- §3 Popup + Chat (ts-mitra-3-*) → `+62820000090{1..4}`
|
||
|
||
If the same phone gets used across multiple flows in one run, the per-IP
|
||
rate-limit (10 OTP requests / hour, default) can trip and break A-03 or A-04
|
||
mid-suite. A-05 mitigates this by calling `reset_phone` between its 5
|
||
variants (each variant is one OTP request).
|
||
|
||
## Deferred (not yet implemented — see reasons)
|
||
|
||
### `ts-mitra-A-02-wrong_code_attempts_then_blocked.yaml`
|
||
**Branch:** §A.2 5× CODE_MISMATCH → OTP_ATTEMPTS_EXCEEDED → blocked dialog.
|
||
|
||
**Why deferred:** the 6-separate-TextField OTP pattern with `maxLength: 1`
|
||
per box doesn't play well with maestro's `inputText` on Android. Maestro
|
||
uses uiautomator2's `setText` under the hood, which delivers chars
|
||
non-deterministically across the per-box focus chain — even matching the
|
||
customer-app's exact `inputText: "000000"` × 6 pattern. After the 5th
|
||
auto-submit succeeds and the boxes clear, focus state races with the
|
||
next `inputText` call and digits silently drop. Verified manually: the
|
||
"Tersisa N percobaan" hint and blocked dialog both render correctly with
|
||
real keyboard typing.
|
||
|
||
**Possible future approach:** refactor S3b to use a single hidden
|
||
`TextField` with custom box decorations (the `pin_code_fields` pattern).
|
||
One `inputText: "000000"` per attempt would land all 6 chars in one IME
|
||
commit, matching how real users paste OTPs from SMS. Worth doing for
|
||
SMS-paste UX anyway.
|
||
|
||
### `ts-mitra-A-05-otp_cooldown_snackbar.yaml`
|
||
**Branch:** §A.1 second OTP request within 60s → `OTP_COOLDOWN` 429 + snackbar.
|
||
|
||
**Why deferred:** Maestro's `extendedWaitUntil` can't reliably assert a
|
||
floating snackbar that auto-dismisses in ~4s — the visible window is too
|
||
short for the polling cadence. Possible workaround: drive two consecutive
|
||
requests and rely on the CTA label switching to "coba lagi dalam Ns" (which
|
||
is non-floating and stable). Worth adding once the resend-cooldown UX
|
||
stabilizes.
|
||
|
||
### `ts-mitra-A-06-rate_limit_phone_popup.yaml`
|
||
**Branch:** §A.1 4th OTP request in 1h for same phone → `OTP_RATE_LIMIT_PHONE`
|
||
429 popup with `retry_after_seconds`.
|
||
|
||
**Why deferred:** The popup is asserted in manual testing (screenshot in
|
||
phase4-mitra-prehome-plan.md). Driving 4 sequential requests within one
|
||
maestro run is brittle if any earlier test bumped the counter. A backend
|
||
`_test/reset-phone-rate-limit` helper would make this reliable; not added yet
|
||
to keep the test-surface minimal.
|
||
|
||
### `ts-mitra-A-07-resend_after_cooldown.yaml`
|
||
**Branch:** §A.4 resend after 60s cooldown → fresh OTP, attempts counter resets.
|
||
|
||
**Why deferred:** 60s wall-clock wait per pass is too slow for CI. Drive
|
||
this manually until we have a `_test/expire-cooldown` helper that fast-forwards
|
||
the cooldown clock.
|
||
|
||
### `ts-mitra-A-08-otp_expired.yaml`
|
||
**Branch:** §A.2 5-minute TTL elapses → `OTP_EXPIRED` 410 → dialog → S3a.
|
||
|
||
**Why deferred:** Same wall-clock problem. Need a `_test/force-expire-otp`
|
||
helper before this is automatable.
|
||
|
||
## Running
|
||
|
||
From `mitra_app/`:
|
||
|
||
```bash
|
||
# Single flow
|
||
maestro test .maestro/flows/ts-mitra-A-01-phone_invalid_inline_error.yaml
|
||
|
||
# Whole §A suite
|
||
maestro test .maestro/flows/ts-mitra-A-*.yaml
|
||
```
|
||
|
||
The backend must be running with `OTP_STATIC_CODE` **unset** — peek_otp.js
|
||
relies on the stub generator returning a fresh code per request, not a
|
||
static one.
|