21 DI tokens selected their implementation like this:
factory: () => (environment.useMockData ? inject(XLocal) : inject(XApi))
That reads as a toggle and is not one. Naming both classes in the factory
keeps both reachable, so every mock shipped regardless of the flag - and
`useMockData` is false in both environment files, so none of them were
ever the selected implementation in the first place. Verified: a fixture
string from partner-hierarchy-local.gateway.ts was present in a
production bundle.
Token factories now inject the API gateway unconditionally. Mock
overrides move to src/app/mock-gateway.providers.ts, swapped for a
production copy that imports nothing, via the same fileReplacements
mechanism mock-data.interceptor.production.ts already uses. Dev behaviour
is unchanged - flip useMockData in environment.ts exactly as before.
useExisting rather than useClass: the local gateways are already
providedIn: 'root' singletons, and an app-level provider for the token
wins over its tree-shakable default.
scan-bundle.sh gains two patterns so this cannot come back: any
*LocalGateway class name, and known fixture literals. Verified in both
directions - clean against the real dist, exit 1 against a planted
OfferLocalGateway.
Result: zero LocalGateway classes and zero fixtures in the production
bundle, down from 21 classes and 75 kB of source. Initial bundle is
unchanged at 1.55 MB because these all sat in lazy chunks; the win is
that production can no longer serve seeded fixtures as real data, not
bytes off the critical path.
Not addressed here: MediaRepository is still bound to MockMediaRepository
unconditionally in app.config.ts. That one cannot be deleted - no real
implementation exists yet - so it is a missing API gateway, not dead
weight. Tracked separately.
256 tests pass. Build green, boundaries and cycles green.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
None of these 5 facades had a single test before this - the highest-risk
gap flagged in Q7 of the delivery plan: real business logic (dashboard stat
computation, health scoring, bulk actions, draft save routing) with zero
coverage on domains that just got real backends behind them (Block 3).
Each spec stubs the gateway TOKEN via jasmine.createSpyObj, not a concrete
class - the correct pattern per this session's earlier fix to
admin-analytics.facade.spec.ts / admin-order-watcher.service.spec.ts, so
these don't repeat that same latent bug.
66 new tests, several written specifically to catch a real regression
class rather than pad a count:
- averageRating/averageOrder must be null/0 for an empty set, never NaN
or a divide-by-zero artifact
- a zero rating must not drag down an average across other real ratings
- a customer with 2+ orders counts as returning; 1 order does not
- moderationHealthPercent is 100 for an empty queue by convention, not 0
- toggleSelection/toggleAll must not produce duplicate ids
- saveDraft must route to createProduct vs updateProduct correctly and
must do nothing when no draft is loaded - a real prior bug class
(silently creating a duplicate on a no-op save)
- an events-load failure in AdminMonitoringFacade must not block queues/
webhooks from loading independently
One own mistake caught before commit, not after: the first updateDraft
test asserted the updatedAt timestamp changed after two synchronous calls
- both landed in the same millisecond, so the ISO string was correctly
identical and the assertion was the bug, not the facade. Replaced with an
assertion that actually matches what's worth testing (a valid timestamp
gets stamped, not that two back-to-back calls differ).
Verified: 205/205 unit tests (66 new), arch:check clean, production build
succeeds.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Block 3 of the frontend backlog. 18 gateways were hardcoded to their local
(mock/localStorage) implementation with no seam to a real backend at all -
this closes that gap for everything with a documented contract to build
against.
10 core gateways, token now resolves environment.useMockData ? local : api,
same pattern already proven on fx-quote earlier this session:
permission, analytics, cart (server-cart), finance, vk-id (identity),
connector (integrations), marketplace (registry), offer, seller,
mall-content
8 admin gateways, same pattern:
orders, products, users, transactions, monitoring, moderation,
notifications, dashboard-metrics
Endpoints came from the matching contract doc where one exists (Phase 1-10,
Track A, Track S - cited per file). Three domains have no dedicated contract
doc yet (transactions, monitoring, moderation) - those gateways call the
established /api/admin/v2/{resource} convention used throughout the rest of
docs/backend/, flagged in each file's own comment as inferred rather than
specified, for whoever writes that contract to confirm or correct.
One real fix along the way: offer-api.gateway.ts's publish() translates a
422 + details[] response (contract §7's actual failure mode) into the
interface's { ok: false, errors } shape, rather than letting an HTTP error
leak past a caller that expects a value back.
Media repository (mock-media-repository.service.ts) deliberately NOT
swapped - no backend contract exists for it anywhere in docs/backend/, and
inventing endpoint shapes with zero grounding is worse than leaving it mock.
Regression this surfaced, fixed as part of the same change: three spec files
stubbed a gateway's concrete Local class directly via useValue. That worked
by accident while the token unconditionally resolved to the local class; once
the token became conditional on useMockData, those specs silently injected
the real (unmocked) API gateway instead and failed. Fixed by providing the
token instead of the class - the pattern the ADMIN_CATEGORIES_GATEWAY entry
in the same spec file already used correctly, because categories was
already token-swapped before this session:
- admin-analytics.facade.spec.ts (orders/products/moderation gateways)
- admin-order-watcher.service.spec.ts (orders gateway, 3 call sites)
Verified: 115/115 unit tests, 5/5 E2E, arch:check clean, production build
succeeds.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Adds InjectionToken + factory for Orders, Products, Users, Transactions,
Monitoring, Moderation (mirrors existing Categories/Dashboard pattern) and
repoints their facades plus the derived Analytics/Customers facades and
admin-order-watcher off the mock LocalGateway class directly. No behavior
change today - still resolves to the mock - but a real backend can now be
bound per domain with zero facade edits.
Docs: full gap analysis of Product Plan v3.1 against current repo state,
and a phased delivery plan (10 phases, 34 sprints, 5 tracks) breaking every
identified gap into scoped, sequenced work.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
The bulk button read 'Archive' but called applyBulkDelete() ->
gateway.deleteReview() - a real hard delete, and reviews have no
archived status in the model at all, so 'archive' was never a real
concept here. Relabeled to 'Delete' and added an app-confirm-dialog
gate, same pattern as products/categories/orders.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Phase 8 (RC-01): these 4 list/dashboard pages had no error-state handling
on their primary data-load subscriptions — on a gateway error, `loading`
was either never reset (Users, Monitoring, Analytics: genuine infinite-
spinner risk, nested subscribe chain in Analytics never resolved on
failure) or there was no loading/empty/error handling at all (Reports
queue: raw table with zero skeleton or fallback).
- admin-users.facade.ts, admin-monitoring.facade.ts: add `error` signal,
error callback on the primary load subscribe so `loading` always
resolves.
- admin-analytics.facade.ts: add `error` signal; every level of the
4-deep nested gateway subscribe chain (orders -> products ->
categories -> reviews) now has an error handler that resolves loading
instead of leaving it stuck true.
- admin-moderation.facade.ts: add `reportsLoading`/`reportsError` signals
(reports list had none previously).
- Templates: reuse existing `app-skeleton`/`app-empty-state`/`app-button`
primitives for the new error branch, `common.retry` label, two new
generic `common.errorTitle`/`common.errorDescription` i18n keys added
to en/ru/hy (reused across all 4 fixes instead of one-off per-page
copy).
Verified: tsc --noEmit clean, `npm run build` green (pre-existing bundle-
budget warning only, unrelated). Live-checked Home (375px) and Backoffice
Products (1024px) — no console errors, tables/cards render without
overflow.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
- Replaced hardcoded English audit/timeline text (order status changes,
review moderation events, user role/status changes) with proper
adminXxx.timelineEvent.*/adminUsers.audit.* i18n keys, so Recent
Activity/Timeline/Audit panels no longer mix English into ru/hy UI.
- Translated raw internal codes rendered directly to users: transaction
payment method ('card'/'qr'/'cash_on_delivery' -> adminTransactions.methodValue.*)
and user roles/permissions ('products.manage' etc -> adminUsers.roleValue.*/
adminUsers.permission.*), replacing developer-facing enum leakage with
real copy.
- Fixed wrong-noun list-footer counts: Orders/Transactions/Moderation
list pages all reused adminProducts.items ("N товаров"/"N products")
regardless of what was actually listed; each now has its own itemsCount
key ("N заказов", "N транзакций", "N отзывов").
- Fixed customer detail page's "Back" button reusing adminOrders.back
("Back to orders") instead of a customers-specific label.
- Added translation keys to en/ru/hy + translations.ts interface for all
of the above.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
RC A11Y-01 audit pass, Backoffice (admin/*) area. Builds on 712a7b4/63c9cee's
composition pass (scope="col", skeletons, empty-states) without redoing that
work.
- Sidebar nav landmark: admin-layout's <nav aria-label> reused the
"Dashboard" nav-item translation key as its landmark label, misleadingly
announcing the whole sidebar as "Dashboard" - added a dedicated
adminShell.sidebarLabel key ("Admin sidebar navigation") in en/ru/hy.
Skip link, #admin-content main landmark, Escape handling, and mobile-drawer
focus management were already correct - verified, not touched.
- Categories tree drag-and-drop keyboard fallback (WCAG 2.1.1): the category
tree's native HTML5 DnD (draggable/dragstart/drop) reorders siblings with
no keyboard equivalent - existing arrow-key tree navigation only expands/
collapses/selects, never reorders. Added per-row move-up/move-down icon
buttons (disabled at sibling boundaries), reusing the existing `reorder`
output so the facade's reorder logic is untouched; new
adminCategories.moveUp/moveDown keys in en/ru/hy.
- Screen-reader loading announcements: skeleton-row loading states across
Products, Categories, Customers, Orders, Transactions, Users, Reviews,
Monitoring (webhooks/events), and Analytics (summary cards + top products)
were purely visual (app-skeleton is aria-hidden by design) with no
accessible "loading" text, unlike the storefront/product-details pattern -
added role="status"/aria-live="polite"/aria-busy + sr-only text using the
existing common.loading key.
- Table row headers: added scope="row" to the primary identifying cell
(product/category/customer name, order number, transaction order number,
user name, review customer, report target, top-products/low-stock product
name, webhook endpoint) on 9 tables that only had scope="col". Added
matching `tbody th[scope='row'] { font-weight/color/text-align/
vertical-align }` + last-row border resets in each component's own scss so
the semantic change doesn't alter visuals (the shared app-table stylesheet
styles all <th> as bold/muted by default).
Verified via `git show --stat` of fb1afb7/a03260e and `docs/UI-COMPOSITION-
REVIEW.md`'s Backoffice sections first, per instructions - confirmed
scope="col" coverage already complete, all admin modals already route
through the shared app-dialog (focus-trap/Escape/return-focus already
correct, nothing to fix), and the bare-<select> filters still carry
aria-label per the accepted Sprint 28 decision (not re-migrated to
app-select).
Flagged, not fixed:
- No toast/notification system exists anywhere in this codebase (product/
category save and delete call the gateway with no success/error UI at
all, not even a subscribe error handler) - there is nothing to wire
aria-live onto without adding a new UI mechanism, which is out of scope
for an a11y-only pass. A prerequisite feature-level fix, not an a11y
regression.
- Dashboard's per-card metric/status-row/timeline skeletons (dashboard-
metric, dashboard-status-row, dashboard-timeline) were left without
aria-live wiring - wrapping each of the ~10 simultaneous mini-widgets in
its own live region would fire a burst of redundant announcements; needs
a single page-level "loading dashboard" region instead, a larger change
than this surgical pass.
- Monitoring's events table and the notifications dropdown (role="menu"
with a static empty-state message, aria-haspopup="true") were left as-is -
matches the same partial-widget-pattern precedent already accepted for
locale-tabs/product-tabs in the storefront and builder passes.
- Analytics `lowStockProducts` table's missing loading-skeleton branch
(already flagged, not fixed, in the RC-Visual-02 pass) - untouched again
here for the same reason.
Verified: npx tsc --noEmit clean; npm run build green (only the pre-existing
768.57 kB vs 700 kB initial-bundle budget warning, unrelated to this pass).
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
- Transactions list: hardcoded #fff background replaced with var(--bg-primary); missing th scope=col added on all 7 headers
- Customers list: hardcoded #fff background replaced with var(--bg-primary); missing th scope=col added on all 7 headers
- Customer detail: fake CSS vars --border-subtle/--brand-primary (never defined, silently falling back to hex) replaced with real --border-color/--primary-color
- Reviews list: hardcoded #fff card background and sticky-header background replaced with var(--bg-primary)/var(--bg-secondary); missing th scope=col added across dynamic column table header
- Reports list: hardcoded #fff background replaced with var(--bg-primary); fake --brand-primary var replaced with --primary-color; missing th scope=col added
- Review health widget: fake --surface-muted/--brand-primary vars replaced with real --bg-tertiary/--primary-color (matches product-health-widget precedent)
- Users page: hardcoded #fff background replaced with var(--bg-primary); missing th scope=col added on both tables
- Monitoring page: hardcoded #fff background replaced with var(--bg-primary); missing th scope=col added on webhooks and events tables
- Analytics page: fake --color-primary var (undefined anywhere in codebase) replaced with real --primary-color across tabs/chart/focus rings; fake --surface-muted/--brand-primary on health bar replaced with --bg-tertiary/--primary-color; hardcoded #fff card/summary-card backgrounds replaced with var(--bg-primary); missing th scope=col added across 4 tables
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Mechanical, value-preserving: font-weight: 400/500/600/700 -> var(--font-weight-normal/medium/semibold/bold, <same value>) across src/app.
Note: cart.component.scss now sits ~771 bytes over its per-file budget
in angular.json due to longer var() strings; non-fatal build warning,
noted in docs/UI-DESIGN-REVIEW.md as a follow-up (either bump the
component style budget slightly or accept the warning).
Mechanical sweep for border-radius: 4px/8px/12px/13px/999px replaced
with var(--radius-xs/sm/md/lg/full, <same value>) across src/app.
Only exact matches to existing token values were touched (20px, 16px,
10px, 6px, 3px, 2px etc. were left as-is since no token maps to them
without a visible size change on at least one tenant theme — see
docs/UI-DESIGN-REVIEW.md).
Mechanical, value-preserving substitution: every literal font-size
declaration across src/app (89 files) that matched one of the 9
typography scale steps introduced earlier (--font-size-xs..4xl) was
replaced with var(--font-size-STEP, <same-or-nearest-step-value>).
Values within ~0.03rem/1px of a scale step were snapped to that step
(e.g. 0.85rem and 0.8rem both -> --font-size-sm/0.8125rem; 0.9rem and
0.875rem -> --font-size-base/0.875rem) to consolidate roughly 15
near-duplicate sizes down to the 9-step scale, per the RC design-system
finalization brief. This eliminates most of the font-size fragmentation
found across the app (previously: 0.7/0.72/0.75/0.78/0.8/0.8125/0.85/
0.875/0.9/0.9375/0.95/1/1.05/1.1/1.125/1.15/1.2/1.25/1.3/1.35/1.4/1.5/
1.75/2rem all in live use simultaneously).
Not touched (deliberately, see docs/UI-DESIGN-REVIEW.md): 3rem+ display
sizes (too large a jump to any existing step, would need a --font-size-5xl
addition), font-size values expressed via clamp()/calc(), and any
component listed as intentionally distinct (code-editor syntax tokens,
theme brand colors).
Same select-all/row-checkbox accessible-name gap in table view; grid
view already had aria-label on its row checkbox (customer name),
table view was the gap. New adminModeration.selectAllRows key added
across en/ru/hy.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
- Fix AdminLayoutComponent.readRouteData crash: leaf route snapshot/data is now optional with dashboard-title fallback, so admin deep links never crash the shell when route metadata is missing (root cause of every blocked browser test since Sprint 12)
- Never-settling promises fixed: all gateway ensureData() bridges (products, categories, moderation) and the catalog category resolver now resolve with an empty list on transport failure instead of hanging forever, so API outages surface as empty states with guidance rather than permanent skeletons plus global console errors
- Request de-duplication: BackofficeDataService caches products/categories with shareReplay - one in-flight request per endpoint shared by all consuming gateways (was 5+ duplicate requests per admin page load); failures clear the cache so the next call retries
- Gateway contract audit: all 8 admin gateways (products, categories, orders, customers/moderation, transactions, users, dashboard metrics, monitoring) now implement an explicit *Gateway interface - added the missing AdminMonitoringGateway; media already swaps via the abstract MediaRepository DI class
- Mock mode untouched: provider selection still flows through RuntimeProviderStrategyService/BACKOFFICE_DATA_PROVIDER
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
- Tokenized every admin card/container radius to the design-system scale: raw 16px/12px/8px replaced with var(--radius-md)/var(--radius-sm) per DESIGN.md (cards 12px, fields keep 10px rounded.field)
- All 16 admin surfaces now share one card language: 1px --border-color border, --radius-md, white fill, 16px padding
- Normalized type-ramp outliers: 0.88rem -> 0.85rem (order timeline), 0.9em -> 0.9rem (category form breadcrumb)
- Visual-only pass: no markup, logic, or layout changes
New Reviews & Moderation feature (frontend only): moderation dashboard (real pending/approved/rejected/reported/spam counts, average rating, recent activity, moderation-health%, computed from the full review queue); reviews list with table/cards, density, saved column visibility, search/status/rating filters, bulk approve/reject/spam/hide/export; review detail shows customer/product/rating/text/photos/timeline/moderator-notes with a real moderation workflow (approve/reject/spam/hide/restore/pin/feature, feature disabled with an explanation unless the review is approved); reusable ReviewHealthWidget (rating/text/media/moderated/report-status/visible + completion%); reports queue lists reports against products/reviews/customers/categories with resolve/dismiss, and honestly renders 'Not available yet' rather than fabricating a value wherever a target can't be resolved (e.g. photos, customer-target reports). Backed by a new in-memory AdminModerationLocalGateway seeded from real product data - the same mock-gateway pattern already used by every other admin feature in this app (orders/products/categories), since no review/report backend exists to reuse. Replaced the 'Reviews' comingSoon nav placeholder with a working link; added full adminModeration i18n coverage.
