Decode .torrent files

Motivation
Yesterday I tried Tribler, which is a next generation torrent client. Its novelty is that it is decentralized, thus it doesn’t need trackers at all.

I noticed that Tribler contacts other clients to collect .torrent files that are stored locally. This way every client becomes a small “tracker”. I got curious to see what those .torrent files are. How to decode them?

Solution
.torrent files have a so-called bencode format. At effbot I found a simple decoder in Python that I extended a little bit. My version can be found here (and it’s part of my jabbapylib library now).

Usage

./bdecode.py ubuntu.torrent

Output:

{   'announce': 'http://torrent.ubuntu.com:6969/announce',
    'announce-list': [   ['http://torrent.ubuntu.com:6969/announce'],
                         ['http://ipv6.torrent.ubuntu.com:6969/announce']],
    'comment': 'Ubuntu CD releases.ubuntu.com',
    'creation date': 1318500053,
    'info': {   'length': 731164672,
                'name': 'ubuntu-11.10-desktop-amd64.iso',
                'piece length': 524288}}

ubuntu-11.10-desktop-amd64.iso
697.3 MB

_{As you can see, information about pieces is omitted. If you need that too, use the variable torrent.long_torrent.}

About these ads