rqbit - bittorrent client in Rust
rqbit is a bittorrent client written in Rust.
Motivation
First of all, I love Rust. The project was created purely for the fun of the process of writing code in Rust.
I was not satisfied with my regular bittorrent client, and was wondering how much work would it be to create a new one from scratch.
Got it to the point where it downloads torrents reliably and pretty fast, and I was using for a few months myself. It works good enough for me, and at the moment of writing this I'm not planning to extend it further.
So in short, it's not "feature complete", but rather "good enough for my use-cases".
Open sourced it just in case anyone might find it useful and/or wants to contribute.
Build
Just a regular Rust binary build process.
cargo build --release
Usage quick start
Assuming you are downloading to ~/Downloads
rqbit 'magnet:?....' ~/Downloads
or
rqbit /some/file.torrent ~/Downloads
Useful options
-v
Increase verbosity. Possible values: trace, debug, info, warn, error.
--list
Will print the contents of the torrent file or the magnet link.
--overwrite
If you want to resume downloading a file that already exists, you'll need to add this option.
--peer-connect-timeout=10s
This will increase the default peer connect timeout. The default one is 2 seconds, and it's sometimes not enough.
-r / --filename-re
Use a regex here to select files by their names.
Features and missing features
Some supported features
- Sequential downloading (the default and only option)
- Resume downloading file(s) if they already exist on disk
- Selective downloading using a regular expression for filename
- DHT support. Allows magnet links to work, and makes more peers available.
- HTTP API
Code features
- Serde-based bencode serializer/deserializer
- Custom code for binary protocol serialization/deserialization. And for everything else too :)
- Supports several SHA1 implementations, as this seems to be the biggest performance bottleneck. Default is openssl as it's the fastest in my benchmarks.
- In theory, the libraries that rqbit is made of are re-usable.
- No unsafe
Bugs, missing features and other caveats
Below points are all easily fixable, PRs welcome.
- The CLI support only one mode of operation: download one torrent to a given folder.
- If you try to run multiple instances, there's some port conflicts (already listening on port). This happens because DHT stores persistence information on disk, and that includes the port it last listened on. So all instances try to listen on the same port. Which means we need to add support for managing multiple torrents rather than trying to run the client multiple times. Or at least add a switch like --disable-dht-persistence.
- HTTP API is rudimentary, mostly for looking at stats. E.g. you can't add a torrent through it.
- Only supports BitTorrent V1 over TCP
- As this was created for personal needs, and for educational purposes, documentation, commit message quality etc. leave a lot to be desired.
- Doesn't survive switching networks, i.e. doesn't reconnect to a peer once the TCP connection is closed.
Code organization
- crates/rqbit - main binary
- crates/librqbit - main library
- crates/librqbit-core - torrent utils
- crates/bencode - bencode serializing/deserializing
- crates/buffers - wrappers around binary buffers
- crates/clone_to_owned - a trait to make something owned
- crates/sha1w - wrappers around sha1 libraries
- crates/peer_binary_protocol - the protocol to talk to peers
- crates/dht - Distributed Hash Table implementation
HTTP API
By default it listens on http://127.0.0.1:3030.
curl -s 'http://127.0.0.1:3030/'
{
"apis": {
"GET /": "list all available APIs",
"GET /dht/stats": "DHT stats",
"GET /dht/table": "DHT routing table",
"GET /torrents": "List torrents (default torrent is 0)",
"GET /torrents/{index}": "Torrent details",
"GET /torrents/{index}/haves": "The bitfield of have pieces",
"GET /torrents/{index}/stats": "Torrent stats"
}
}