JACWF (Just Another Calibre Web Frontend)

JACWF is a lightweight, unpretentious web frontend for your Calibre library. It's designed to be fast, simple, and easy to deploy on NixOS or via Docker.

Features

• Calibre Integration: Reads directly from your metadata.db and uses calibredb for uploads.
• In-Browser Reader:
  • EPUB: Custom "Coffee & Parchment" theme with synced reading progress.
  • PDF: Native browser integration.
  • Text/Markdown: Simple iframe-based viewer.
• NixOS Ready: Includes a Flake and a NixOS module for easy service management.
• Toggleable Auth: Simple password-based login that can be disabled for local-only use.
• Responsive Design: A bold, "neo-brutalist" aesthetic that works on mobile and desktop.

AI Disclaimer

This project was developed with significant assistance from GitHub Copilot (using Gemini 3 Flash, et al.). The architecture, styling, and Nix integration were co-authored by AI to ensure rapid development and adherence to modern SvelteKit 5 patterns.

Installation

NixOS (Recommended)

Add JACWF to your flake inputs:

inputs.jacwf.url = "github:youruser/jacwf";

Then enable the service in your configuration:

services.jacwf = {
  enable = true;
  booksPath = "/path/to/your/calibre/library";
  # Optional: enable authentication
  # auth.enable = true;
  # auth.passwordFile = "/run/secrets/jacwf-password";
};

Docker

docker build -t jacwf .
docker run -p 3000:3000 -v /path/to/books:/app/books jacwf

Manual Development

pnpm install
pnpm run dev

Configuration

JACWF uses environment variables for configuration:

• JACWF_BOOKS_PATH: Path to your books or Calibre library (default: ./books).
• JACWF_DB_PATH: Path to the internal SQLite database for settings/progress (default: ./data/library.sqlite3).
• JACWF_AUTH_ENABLED: Set to true to enable the login page.
• JACWF_PASSWORD: The password for authentication (if enabled).

License

This project is licensed under the GNU General Public License v3.0. See the LICENSE file for details.

• On first visit you will be prompted to pick or create a profile.
• EPUB progress is saved/restored per profile.
• PDF progress is not synced (PDFs open in a new tab, so the app does not control the viewer).

Build & preview

pnpm run build
pnpm run preview

Container (nerdctl / containerd)

This app is a server (SvelteKit routes + SQLite), so it must run as a Node process.

Build

nerdctl build -t library:dev .

Run (bind-mount books + data)

mkdir -p data

nerdctl run --rm -p 3000:3000 \
   -v "$PWD/books:/app/books:ro" \
   -v "$PWD/data:/app/data" \
   -e LIBRARY_DB_PATH=/app/data/library.sqlite3 \
   library:dev

Or use the helper script:

chmod +x scripts/nerdctl-build-run.sh
./scripts/nerdctl-build-run.sh

Then open http://localhost:3000.

Notes:

• books/ is mounted read-only by default. To enable uploads in the container, mount it writable (or run the helper script with BOOKS_RW=1).
• data/ holds the SQLite DB (profiles + progress).

Configuration

• LIBRARY_DB_PATH (default: ./data/library.sqlite3)
  • Path to the SQLite database file.
• BOOK_SCAN_CACHE_TTL_MS (default: 5000)
  • In-memory scan cache TTL. Set to 0 to disable caching.
• UPLOAD_MAX_BYTES (default: 536870912)
  • Max upload size in bytes (default is 512MB).

Notes / limitations

• Scanning happens on request (cached with a TTL, but no persistent index yet), so very large libraries may feel slower.
• Metadata is intentionally simple right now (filename/folder based).
• EPUB progress sync requires selecting a profile.
• PDF progress is not synced.
• Deployment needs filesystem access to the books/ directory; depending on where you deploy, you may want a dedicated SvelteKit adapter and/or a storage strategy.

Logical next steps

1. Indexing & caching

   • Persist a book index in SQLite (upsert on scan) to avoid filesystem walks for every request.
   • Add incremental updates by comparing mtime and removing missing paths.

2. Better metadata

   • EPUB: parse title/author/cover from metadata.
   • PDF: extract title/author when available.
   • Markdown/TXT: derive title from first heading / first line.

3. Security hardening for file serving

   • Add extension allow-listing in GET /api/book/[...path] (defense-in-depth).

4. UX improvements that stay simple

   • Display “pretty” title in the reader header (currently shows the path).
   • Add a clear “back to library” and show author/title consistently.

5. Progress & “currently reading”

   • Show the last 5 in-progress books for the active profile (by last updated progress).
   • Consider moving TXT/MD rendering into the app (instead of iframe) so progress can be tracked.

6. PDF progress sync (optional)

   • Add an in-app PDF viewer (e.g. pdf.js) if you want cross-device PDF progress.

7. Quality & maintainability

   • Add a couple of tests for the scanner (filename parsing, extension filtering, directory walking).
   • Add a small API contract test for /api/books.