Skip to content
patterueldevPublic

About

No description, website, or topics provided.

Resources

Stars

0 stars

Watchers

0 watching

Forks

Repository files navigation

Droparr

Drop a folder. It's in your *arr.

Droparr is a self-hosted web app that takes an existing movie or series folder — messy names, mixed seasons, whatever — and handles the tedious part of getting it into Sonarr and Radarr:

  • detects series vs movie, and routes anime vs standard drops to the right instance
  • matches against TVDB/TMDB through your *arr instances' own lookup APIs (no extra API keys)
  • gives you a review screen to fix the match, pick seasons and monitoring
  • stages files and lets Sonarr/Radarr do a native manual import — renaming, quality detection and moving stay inside the *arr, exactly like the built-in UI

Optional extras: LLM-assisted matching for ugly filenames (off by default), Jellyfin login with an admin approval queue for submissions from family and friends, and ntfy/Discord webhook notifications when a submission needs approval or an import finishes.

Status: v1.0.0 release candidate — core pipeline, Jellyfin auth, and shared mode (uploads, approval queue, submission status) are built; now refining live on a home server. See docs/PLAN.md for milestones.

Why not the built-in Library Import?

Sonarr and Radarr both ship a "Library Import" feature, but it is per-instance only, expects cleanly organized files inside a root folder, and cannot decide which of your instances (anime Sonarr? TV Sonarr? Radarr?) a drop belongs to. Droparr is the orchestration layer above that machinery. See Prior art.

How it works

  1. Add — drop folders (browser, chunked/resumable) or point at a server path / watch folder
  2. Analyze — heuristics parse names (S01E01, absolute anime numbering, year), then query your *arr lookups for matches
  3. Review — fix the match, choose a category, seasons and monitoring; duplicates are handled ("import only" mode)
  4. Import — files are staged, the title is added to Sonarr/Radarr, and a native manual import runs with live progress
  5. Play — optional Jellyfin library refresh once it's done

Roles

  • Admin — full configuration, approval queue, instant imports
  • Submitter — signs in with their Jellyfin account, drops files, tracks status; submissions wait in an approval queue unless an admin marks the account trusted

Documentation

  • Product plan — problem, scope, milestones, decision log
  • Architecture — stack, pipeline, verified *arr/Jellyfin API mechanics, Cloudflare Tunnel constraints
  • Deployment — Docker next to the *arrs, Cloudflare Tunnel, hardening checklist

Development

pnpm monorepo — React 19 + Vite (web), Node + Fastify (server), TypeScript throughout, SQLite + JSON config, Docker for deployment.

Prerequisites: Node ≥ 22, pnpm ≥ 10.

pnpm install
pnpm dev          # API on :3100, web on :5173 (proxies /api)

Other commands:

pnpm -r typecheck  # typecheck all packages
pnpm -r test       # unit tests (analyzer, path mapping, staging)
pnpm --filter @droparr/web build   # build the web UI

Deploy with Docker — prebuilt multi-arch images (linux/amd64 + linux/arm64) are published to GHCR, and latest follows the newest release candidate:

docker compose pull && docker compose up -d   # ghcr.io/patterueldev/droparr:latest

Pin an exact candidate (ghcr.io/patterueldev/droparr:v1.0.0-rcN) for reproducible installs and rollbacks. Candidates are cut by tagging main — CI builds and publishes both the pinned tag and latest.

The server serves the built UI and the API on a single port (3100). Exposing it at a subdomain with Cloudflare Tunnel — plus the hardening checklist for cookies, rate limits and upload body limits — is covered in the deployment guide.

Set PUID/PGID (default 1000:1000) to the user that owns ./config and that your *arr stack runs as, so quarantine uploads, config writes and SQLite files stay readable by everyone involved; UMASK (default 022) controls the mode of newly created files. Settings → Uploads & disk covers the quarantine directory, per-file and per-submission caps, the minimum free space kept on the volume (uploads are refused/aborted below it), and the retention window for abandoned and finished drops (default 7 days) with a manual "Run cleanup now" action. Settings → Notifications optionally posts to an ntfy topic or Discord webhook when a new submission is waiting for approval, and when an import finishes or is rejected (off by default; delivery failures are only logged).

First run: open the UI — a setup wizard asks for your Jellyfin URL, then a Jellyfin admin signs in and confirms; that account becomes the Droparr admin and the wizard locks permanently (the lock lives in SQLite, so deleting config.json cannot reopen it). While setup is incomplete, every API route except health/auth/setup is blocked.

After setup: sign in with your Jellyfin account and configure instances, categories and the staging directory in Settings. Jellyfin admins become Droparr admins automatically; the Users tab lets an admin promote or demote accounts, mark submitters trusted (skips the approval queue) and block them (signs the account out everywhere and refuses new logins). Other accounts can sign in and manage their own sessions; the submitter drop flow arrives in M3. Droparr never stores Jellyfin passwords.

Multi-arch note: images build on both linux/arm64 (dev Mac) and linux/amd64 (home server). Use docker buildx build --platform linux/amd64,linux/arm64.

License

MIT

About

No description, website, or topics provided.

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages