EMEX leaf parts fetch was averaging 12.5s cold per category (Honda Civic
prod traces). Almost all of that lived in the scraper's two waterfall page
loads, each of which (a) waited for `networkidle` — which only fires after
analytics/ads stop chattering — then (b) slept an unconditional 2 seconds.
The actual extraction work runs in <100ms once the DOM is parsed.
Three focused changes in scripts/emex-vin-scraper.js:
1) getParts() resolves the Unit.aspx URL via a plain `fetch()` of
QuickDetails.aspx instead of opening it in the browser. Probe showed
the page is fully server-rendered and the anchor is present in the
initial HTML, so the browser stage is dead weight (~3s saved). Falls
back to a browser load if the fetch fails or the anchor isn't there,
so catalogs that gate the link behind JS still work.
2) Both remaining `page.goto()` calls in getParts/getCategories/
getCategoryTree drop `waitUntil:'networkidle'` for `'domcontentloaded'`
plus a targeted `waitForSelector('img.dragger' | 'a[href*=…]', {timeout:5000-8000})`.
`.catch(()=>null)` makes the wait advisory — the evaluate() below has
its own null-safe fallbacks — but in practice the selector is present
well before networkidle would have fired.
3) Removes the two unconditional `setTimeout(r,2000)` sleeps in getParts.
They predate the selector-wait pattern and were belt-and-braces.
Expected: ~12.5s → ~3-5s cold leaf fetch. Warm path (DB-cached) is
unchanged at ~45ms. Dev-only push for verification before main merge.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Redesign emex_* tables from vehicle-centric to catalog-centric structure with
junction tables (emex_vehicle_group_links, emex_vehicle_part_links). Migrate
92M+ rows from emex DB via postgres_fdw. Add EmexCatalogService for browse
and VIN pre-scraped matching, with Redis caching. Frontend catalog page now
has PL24/Emex tabs with full drill-down routes (brand → vehicle → group → parts).
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Replace Playwright browser with pure HTTP for EMEX VIN decode:
- Vehicles.aspx, QuickGroups.aspx, QuickDetails.aspx all work
without authentication — no session/browser needed
- VIN decode drops from ~10-20s (Playwright) to ~1-5s (HTTP)
- Playwright browser kept as fallback if HTTP fails
Add multi-candidate selection for EMEX (mirrors PartsCatalogs flow):
- EmexCandidate interface matches PcatCar shape so VehicleSelectModal
renders without changes
- decodeVinOrCandidates(): single HTTP fetch returns vehicle OR
candidates (avoids double Vehicles.aspx request)
- decodeVinByIndex(): resolves user-selected candidate by index
- vehicles.service: emexCandidates in VinResolveResult, emexCarIndex
param in decodeVin/resolveVin, resolveEmexCarByIndex() method
- vehicles.controller: @Body("emexCarIndex") accepted
- search.tsx: candidateSource state, handleCandidateSelect sends
emexCarIndex or pcatCarId based on source
Test VIN NM417800006410193 (Fiat SIENA) now shows 2-option modal.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Add unavailable/remark/modelCodes/presel fields to categories and parts DB schema
- Pass PL24 unavailable flag through pipeline instead of filtering out records
- Show unavailable categories/parts at reduced opacity in grid, tree, and parts panel
- Display part remark and model codes as secondary info in parts table
- Move Google sign-in button above email form on login/register pages with branded icon
- Add public email existence check endpoint for better login error messages
- Update INDEX.md documentation
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Add oem_code_copies table and analytics module for tracking part code copies
- Integrate PostHog for frontend product analytics (VIN decode, OEM copy events)
- Switch email service from stub to Postal API with proper error handling
- Add copy button to parts panel with clipboard + server-side logging
- Add admin copy-logs page with filtering and top-copied-codes view
- Add VIN report endpoint for users to flag unrecognized chassis numbers
- Add Postal email env vars to config schema
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Add OpenTelemetry SDK with tracing, metrics, and OTLP export for API and worker
- Integrate Grafana Faro for frontend real-user monitoring
- Instrument health checks, database, Bull queues, and HTTP exception filter
- Add Grafana dashboard JSON for service overview
- Remove deprecated apps/web-nj (Next.js) — fully replaced by Vite+React frontend
- Update nginx config with OTEL collector proxy
- Minor UI fixes in schema viewer, category components, and subscription flow
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Split-panel auth layout with dark promo panel, sectioned dashboard nav with theme toggle,
testimonials/social proof on landing page, e-commerce Remotion demo, and static pages
(about, blog, contact, terms, privacy, KVKK).
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Add PL24 Ford legacy service for fordt_parts architecture
- Refactor EMEX to use persistent browser pool instead of per-call instances
- Make vehicle decode resilient: fallback to PL24 when Corgi doesn't recognize VIN
- Add collapsible sidebar with persistent user preference
- Improve brand access guard and categories service
- Add debug/test scripts for VIN e2e testing
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Add POST /admin/users endpoint with password hashing and role support
- Add user creation dialog to admin users page
- Migrate web from Next.js to Vite + TanStack Router
- Fix Dialog component positioning for Tailwind CSS v4
- Add @source directive for @sase/ui package scanning
- Add pl24 integration parsers and vehicle decode flow
- Backup old Next.js app to apps/web-nj
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
