Slug is an (incomplete) scriptable BitTorrent daemon for Unix-like systems.

I decided to create Slug because I'm fed up with the lackluster automation tools offered by other BitTorrent clients. I wanted a client that exposed a simple API combined with pattern matching which would allow an individual to write powerful pre- and post-hooks to be executed when a torrent that matches a pattern is started or finished.

Example usages/features:

• Downloading every torrent from an RSS feed and then moving it to another folder automatically.
• Moving a torrent to a specific folder based on a pattern, like moving torrents that match the regex "[Ll]ost" to "/tv/lost".
• Automatically unraring a torrent and moving the unrared file(s) to a folder.
• Contacting a metadata server and renaming a torrent based on the metadata recieved from the server.
• Scheduling specific hours during which Slug should run.
• Stopping a torrent after a certain amount of time or once it has seeded to a certain ratio.

All of these things will be possible with Slug and, most importantly, they won't stop the torrents from seeding -- even after being renamed, etc. This will be accomplished by embedding Lua for use as a scripting language inside of Slug and allowing users to write "rule files" to complete tasks. Slug's configuration will be exposed via a default rule file.

Slug, by itself, will not have a GUI but -- once I finish Slug -- I intend to write a user-friendly WebUI that can optionally be used alongside Slug. This will expose actions such as pausing and removing torrents and make torrent management a breeze, even when dealing with thousands of torrents. Ideally, you will be able to think of Slug's webUI as an even better version of Rutorrent.

Contributing is easy. Just fork the project, make your changes, and then submit a pull request.

Some stuff that needs done:

• Clean-up the codebase to match OpenBSD's style guidelines. All of the linked lists need be reimplemented to use the macros from <queue.h>.
• Lots of error checking has been ignored and needs added.
• Read through the code and find areas that ought to be explained by a comment but aren't. Comment them or tell me to comment them.
• Implement support for multi-file torrents. For more information, you can look at how unworkable does it.
• Improve our queueing strategy. If you know C++, you could go dig through rtorrent's code, figure out their download strategy, and explain it to me so I can implement it.
• Implement a variable length queue that's based on some network speed heuristic.
• Flesh out announce support.