Jobalots Watcher

Objective

Create a local Windows GUI that watches Jobalots for user-defined search terms every six hours and notifies when new matching auction listings appear.

Outcome

A working Python/Tkinter desktop application was built and verified against live Jobalots data.

Features

• Add and save multiple search terms.
• Run a manual "check now" on demand.
• Start a six-hour automated watch cycle.
• Open Jobalots searches and individual results directly in the browser.
• Remember previously seen listing URLs to avoid repeat alerts.
• Notify only when genuinely new matches appear.
• Display listing titles, prices and product links.
• Launch through a Windows batch file.

Technical approach

The first HTML-parsing approach was rejected because Jobalots renders product data client-side. The watcher instead calls the same JSON backend the site itself uses:

This avoids running a heavyweight embedded browser and keeps the app lightweight enough for everyday local use.

Verification

• Python syntax check passed.
• A live search for "lego" returned 24 current listings.
• Product titles, prices and URLs were returned correctly.
• Price display was switched to plain "GBP" text to avoid Windows console encoding problems with the £ symbol.

Deliverables

• jobalots_watcher.py — the application itself
• run_jobalots_watcher.bat — one-click launcher
• README_Jobalots_Watcher.md — usage documentation

Known limitation

The desktop application has to stay running for its internal six-hour schedule to execute — there's no background service yet.

Possible improvements

• Run automatically at Windows sign-in.
• Move scheduling to Windows Task Scheduler so the GUI doesn't need to stay open.
• Add email, Telegram or mobile push notifications.
• Add maximum-price and exact-match filters.
• Add a history/export view for discovered listings.
• Add retry/backoff logic and endpoint-change diagnostics.
• Package as a signed standalone .exe.

Related