مین‌گنجور — proof of concept

▶ Live demo

A single static file, no build step, no server — proof that ganjoor-data on its own (fetched live from jsDelivr) is enough to run a small reading app for classical Persian poetry.

• Loads manifest.json and shows every poet.
• Click a poet → their bio (poet.json) + their root category's contents (_cat.json).
• Click a category → its children, recursively, until you reach a poem.
• Click a poem → renders it as traditional right/left hemistich pairs (مصرع اول/دوم), with metre and rhyme shown when present.
• A "jump to poem by numeric id" box, demonstrating the bucketed id index (index/poems-by-id/{id / 2000}.json) resolving a bare id to a path.
• Routing is hash-based (#/hafez/ghazal/sh1) — works on GitHub Pages with zero server configuration, since there's no server-side route to configure in the first place.

Everything is fetched live, in the browser, straight from https://cdn.jsdelivr.net/gh/ganjoor/ganjoor-data@main/ — there is no backend anywhere in this app. That URL is the only "API base" set at the top of the <script> block in index.html.

1. Create a new, empty GitHub repo (or use an existing one).
2. Copy index.html into the root of that repo. That's the entire app — one file.
3. Push it.
4. Repo → Settings → Pages → Source: Deploy from a branch → pick main / (root) → Save.
5. GitHub gives you a URL like https://<username>.github.io/<repo>/ within a minute or two.

No build step, no npm install, no config beyond that. Any static host works the same way (Netlify, Cloudflare Pages, a plain python -m http.server locally to preview first) — GitHub Pages is just the one named in the ask.

cd mini-ganjoor
python3 -m http.server 8000

then open http://localhost:8000. (A plain file:// open also mostly works, but some browsers restrict fetch() from file:// origins — a local server sidesteps that.)

• No search — ganjoor-data doesn't have a search index yet (documented in its own API.md).
• Poems with unusual structures (free verse, unusual band forms) fall back to one line per verse rather than fully reproducing every classical form's exact traditional layout.
• No caching/offline support, no pagination for very large categories — it's a proof of concept, not a production reading app.