PartsHub ⚡

A single place to search electronics parts across multiple Indian vendors, compare prices, and build a per-project bill of materials (BOM) — instead of hopping between vendor websites.

Built with FastAPI + SQLite + Jinja2 (HTML/CSS/vanilla JS) — a simple, beginner- friendly stack. See ARCHITECTURE.md for the full design and the roadmap toward unified checkout.

Features (MVP)

• 🔎 Unified search across all configured vendors (FTS5 full-text search)
• ⚖️ Price comparison — same part across vendors, cheapest in-stock first
• 📋 Project BOMs — group parts into projects with quantities and per-vendor subtotals
• 🛒 Deep-link checkout — at checkout each vendor shows a sign-in link (you log in on the vendor's own site; we never see or store it) plus pre-filled cart links: Shopify vendors get a one-click permalink that drops every item into the cart (/cart/<variant>:<qty>,…); WooCommerce vendors get per-item ?add-to-cart= links; others fall back to product-page links. You then pay on the vendor's site.

After updating, re-run python -m scripts.ingest once so the new cart_ref field (Shopify variant ids / Woo product ids) is captured — that's what powers the pre-filled cart links. The DB self-migrates to add the column.

Vendors included

Robocraze, ThinkRobotics, QuartzComponents (Shopify), Sharvi Electronics and FlyRobo (type: "auto", platform detected at ingestion time), Sunrom Electronics (custom platform, dedicated HTML scraper), Sun Electronics (sunelectronics.co.in, WooCommerce Store API), Ktron (www.ktron.in, type: "auto"), and Robu.in (headless Next.js + custom GraphQL API). Add more in app/config.py — for any Shopify/WooCommerce store it's a single entry.

Robu.in is a special case: it's a headless storefront with no public product feed, so app/scrapers/robu.py queries its internal GraphQL API (/api/proxy/graphql). That API is private/undocumented and Cloudflare-protected, so that adapter is more fragile than the public-feed vendors — run it from a normal network and re-capture the query if robu changes its schema.

Onboarding a new store: if you're unsure of its platform, run python scripts/check_vendor.py <store_url> from your machine. It reports whether the store is Shopify, WooCommerce, or HTML-only and which type to use. Some stores block datacenter IPs, so always run ingestion/diagnostics from your own network.

Quick start

# 1. (optional) create a virtual environment
python -m venv venv
venv\Scripts\activate        # Windows
# source venv/bin/activate   # macOS/Linux

# 2. install dependencies
pip install -r requirements.txt

# 3a. load REAL product data from the live vendor catalogs
python -m scripts.ingest

#     ...or 3b. just load a small demo dataset to try the UI instantly
python -m scripts.seed_demo

# 4. start the web app
uvicorn app.main:app --reload

# 5. open http://127.0.0.1:8000

Run python -m scripts.ingest from the project root (the folder with this README), not from inside scripts/.

How it works

1. Ingestion (scripts/ingest.py → app/scrapers/) fetches each vendor's public product catalog and stores a normalized snapshot in partshub.db (SQLite).
2. The web app (app/main.py) only reads from SQLite, so the site stays fast and keeps working even if a vendor is down. Re-run ingestion to refresh prices/stock.

Refreshing the catalog

Re-run python -m scripts.ingest periodically (e.g. a daily Windows Task Scheduler job or cron) to keep prices and stock current.

Project layout

app/
  main.py            FastAPI routes + page rendering
  config.py          vendor list + scraping settings
  database.py        SQLite schema, connection, title normalization
  search.py          full-text search + price-comparison grouping
  projects.py        project / BOM operations
  scrapers/
    base.py          shared helpers + normalized product shape
    shopify.py       Shopify /products.json adapter (works for any Shopify store)
    woocommerce.py   WooCommerce Store-API + HTML-fallback adapter
    runner.py        ingest all vendors → upsert into SQLite
  templates/         Jinja2 HTML
  static/            style.css + app.js
scripts/
  ingest.py          run live ingestion
  seed_demo.py       load a tiny demo dataset

Notes & limitations

• This MVP does not store vendor logins or place orders automatically — see the roadmap in ARCHITECTURE.md for why and the realistic path forward.
• Only public catalog data is ingested and cached, with links back to each vendor. Be polite (the scraper rate-limits and identifies itself) and review each vendor's Terms before scaling up.