My Timetable

A free, installable PWA that turns a public Google Sheets timetable into a personal agenda: pick your specialisms once, filter what you see, with records saved locally and optional online integrations. Built with Vite + React + TypeScript; hosted on GitHub Pages and Vercel. See PLAN.md for the full plan and architecture, and DEPLOYMENT.md for how every piece is deployed (automatic and manual).

Using it

1. Open the app and paste your timetable's Google Sheets URL (open the tab you want first so the URL contains #gid=…). The sheet must be shared as "anyone with the link can view".
2. Pick your specialism(s) when asked — every other specialism session is hidden from then on.
3. Day view opens on today and scrolls through the coming days; Week and Month views are in the switcher. Tap any session for details and its Moodle link. Notes and records are saved locally; optional encrypted sync transfers record data, not attachment files.
4. Install it from the browser menu ("Add to Home Screen" / "Install app") for an app-like experience with offline support (last-fetched data is kept).

Calendar export

• Download .ics (Settings → Calendar export): a one-off snapshot of your filtered timetable to import into Google/Apple/Outlook calendars.

• Live feed (Settings → Calendar feed): a URL your calendar app polls so it stays in sync with the sheet. Needs the tiny worker in workers/ics-feed deployed to Cloudflare's free plan:

  cd workers/ics-feed
  npx wrangler deploy   # first run opens a browser to log in / create a free account
  

  Copy the printed *.workers.dev URL into Settings → Calendar feed. The app builds the full feed URL (sheet + your specialism choices) and you paste that into your calendar app's "subscribe by URL" option.

Background push (optional)

workers/push is a Cloudflare Worker (it fits the free plan at cohort scale; this deployment runs on Workers Paid) that sends session and key-date reminders even when the app is closed. Deploy:

cd workers/push
npx wrangler kv namespace create PUSH   # paste the printed id into wrangler.toml
npx wrangler deploy

Then paste the printed *.workers.dev URL into Settings → Background push and enable. VAPID keys generate themselves into KV on first use; the cron runs every 10 minutes.

Support

If this app saves you time, you can buy the developer a coffee on Ko-fi. ☕

Development

npm install
npm run dev      # local dev server
npm run build    # production build (includes PWA service worker)

Deployment is automatic on both frontend hosts for pushes to main. Both use the same npm run validate build, unit and browser checks (validate:ci also installs Chromium). Workers still deploy separately. See DEPLOYMENT.md before releasing cross-cutting changes.

Supported sources and privacy

The parser recognises timetable columns such as Title, Date, Start, End and Room. It does not support arbitrary spreadsheets, a column-mapping wizard, automatic tab discovery or CSV fallback yet. Use a public sheet and select the tab in its URL. Cached data supports offline viewing; live refresh, maps, routing and sync require online services.

Records start on this device. Optional encrypted sync transfers settings and record data; photo and wallet/document blobs remain local and can be moved using backup. Push stores a subscription and reminder configuration on the worker; background leave alerts also store the last app-open location when enabled. TfL receives route coordinates, OpenStreetMap serves map tiles, and address lookup uses postcodes.io or Nominatim. Calendar subscriptions expose their configured public-sheet selection to the feed service. Study groups share availability. Device usage pings use a pseudonymous identifier and feature counts and can be disabled in Settings; they are enabled by default outside the demo.

Safe development checks

npm run validate runs the production build, isolated worker/release unit tests and browser regressions. npm run validate:ci also installs Chromium; install OS browser dependencies on the runner where necessary. Browser tests block external services and worker tests use in-memory KV and synthetic device subscriptions. No live push or analytics test is needed. See tests/README.md for coverage and remaining checks.