TeslaPricing

Aggregated Tesla used-car listings scraped from multiple French marketplaces. React frontend backed directly by Supabase, with a Node.js scraper suite.

Features

• Listings page — card-style filter sidebar (Model tiles, Price, Year, Mileage with "new only" toggle, Drivetrain, Autopilot, Seats, Color, Source) plus a horizontal sort bar at the top (latest, price ↑/↓, mileage ↑/↓, year, biggest drop in € and %). Infinite scroll with debounced updates, 4-per-row grid.
• Click-to-filter source — clicking a listing card's source pill (e.g. "TESLA", "LACENTRALE") toggles the source filter for the whole grid.
• Hide & reset — eye icon on each card hides a listing; sidebar surfaces hidden count with both "Show hidden" and "Reset hidden" actions.
• Watchlist & Compare — bookmark listings or compare up to 3 side-by-side (price, mileage, options, history).
• Deals page — biggest active price drops, with top-3 podium.
• Auctions page — separate view for Alcopa auction listings.
• Map view — geolocated dealer markers filtered by the same sidebar.
• Trends page — average price evolution per model over time (Model 3, Y, S, X) with price-change tracking (new history entry on every price change).
• In-feed ad scaffolding — placeholder card every 20 listings, gated behind a SHOW_ADS flag and ready to wire to AdSense once a CMP is set up.
• i18n (EN/FR), light/dark theme, Supabase auth (Google sign-in syncs watchlist/hidden across devices).
• Deployed automatically to GitHub Pages on every push to main.

Sources

Source	Command	Method
Tesla	tesla	tesla-inventory npm package — new & used inventory across 25 European markets (see --markets flag)
CapCar	capcar	Algolia API (no browser needed)
Leboncoin	leboncoin	Playwright — intercepts internal JSON API (Datadome protected)
GMECars	gmecars	HTTP + HTML regex parsing
AramisAuto	aramisauto	Playwright — DOM extraction (Nuxt SSR)
Renew Auto	renew	HTTP — parses window.APP_STATE JSON blob
LB Automobiles	lbauto	HTTP — parses application/ld+json structured data
La Centrale	lacentrale	Playwright — DOM extraction from SSR HTML (supports account login)
Heycar	heycar	HTTP — parses Next.js RSC payload for listing_id JSON fragments
Alcopa Auction	alcopa	HTTP + HTML regex parsing (auction listings)

Note: tesla and alcopa block GitHub-hosted runner IPs, so they only run locally — not from the scheduled GitHub Action.

Stack

• Backend (scraper suite) — Node.js 26 (ESM)
  • commander — CLI argument parsing (scraper/cli.js)
  • playwright + playwright-extra + puppeteer-extra-plugin-stealth — headless browser scraping with bot-detection evasion (Leboncoin, AramisAuto, La Centrale)
  • tesla-inventory — Tesla new/used inventory API client
  • pg — direct PostgreSQL access to the Supabase database
  • p-retry — retry logic for flaky requests
  • cli-progress, debug / debug-logfmt, dotenv — progress bars, logging, env config
• Frontend (web app) — React 19 + React DOM 19, TypeScript 5.6
  • vite 8 + @vitejs/plugin-react — build tooling and dev server
  • @supabase/supabase-js — direct client-side reads, auth (Google sign-in), and watchlist/hidden sync
  • leaflet + react-leaflet + leaflet.markercluster — map view with clustered dealer markers
• Database — Supabase (PostgreSQL)
• CI/CD — GitHub Actions
  • ci.yml — lint/test on push
  • scrape.yml — scheduled scraping (runner-IP-blocked sources excluded)
  • deploy.yml — build and deploy the frontend to GitHub Pages on every push to main

Setup

Install dependencies

cd backend && npm install
cd frontend && npm install

Playwright browsers (required for leboncoin, aramisauto, and lacentrale)

cd backend && npx playwright install chromium

Environment variables

# backend/.env
DATABASE_URL=postgresql://...

# frontend/.env
VITE_SUPABASE_URL=https://xxxx.supabase.co
VITE_SUPABASE_ANON_KEY=...

Usage

Start the frontend dev server

./dev.sh

Frontend runs on http://localhost:5173.

Run scrapers

./scrape.sh <source> [options]

Command	Description
./scrape.sh tesla	Fetch all Tesla models (new + used) — France only by default
./scrape.sh tesla --models m3,my	Specific models only
./scrape.sh tesla --markets fr,es,be	Specific markets — supported keys: at, be, ch, cz, de, dk, es, fi, fr, gb, gr, hr, hu, ie, is, it, lu, nl, no, pl, pt, ro, se, si, tr. Tesla also serves EE, LT, SK but the tesla-inventory library has no entry for them, so they're not yet scrapable.
./scrape.sh capcar	Scrape CapCar via Algolia (up to 10 pages)
./scrape.sh leboncoin	Scrape Leboncoin (1 page, Playwright)
./scrape.sh leboncoin --headed	Open browser window — required on first run to solve captcha
./scrape.sh leboncoin --pages 3	Multiple pages
./scrape.sh gmecars	Scrape GMECars
./scrape.sh aramisauto	Scrape AramisAuto (Playwright)
./scrape.sh aramisauto --headed	Open browser window if blocked
./scrape.sh renew	Scrape Renew Auto
./scrape.sh lbauto	Scrape LB Automobiles
./scrape.sh heycar	Scrape Heycar
./scrape.sh alcopa	Scrape Alcopa Auction (local only — blocks GitHub IPs)
./scrape.sh lacentrale	Scrape La Centrale (1 page, headless)
./scrape.sh lacentrale --headed	Open browser window — required on first run to solve captcha
./scrape.sh lacentrale --pages 5	Multiple pages
./scrape.sh lacentrale --login	Log in to your account and save the session (reduces bot detection)
./scrape.sh all	Run all scrapers in sequence

Deployment

The frontend is deployed automatically to GitHub Pages via GitHub Actions on every push to main. The workflow lives in .github/workflows/deploy.yml.

Live: https://tolsadus.github.io/TeslaPricing/