Rust implementation of dtn bundle protocol 7 draft https://tools.ietf.org/html/draft-ietf-dtn-bpbis-20
This library only handles encoding and decoding of bundles, not transmission or other processing of the data. A full daemon using this library can be found here: https://github.com/dtn7/dtn7-rs
This code is (probably) not production ready!
A simple benchmark is shipped with the library. It (de)serializes Bundles with a primary block, bundle age block and a payload block with the contents (b"ABC"). This benchmark can be used to compare the rust implementation to the golang, python or java implementations.
cargo run --release --example benchmark
Finished release [optimized] target(s) in 0.29s
Running `target/release/examples/benchmark`
Creating 100000 bundles with CRC_NO: 510059 bundles/second
Creating 100000 bundles with CRC_16: 293399 bundles/second
Creating 100000 bundles with CRC_32: 291399 bundles/second
Encoding 100000 bundles with CRC_NO: 1090996 bundles/second
Encoding 100000 bundles with CRC_16: 436836 bundles/second
Encoding 100000 bundles with CRC_32: 432774 bundles/second
Loading 100000 bundles with CRC_NO: 564817 bundles/second
Loading 100000 bundles with CRC_16: 473768 bundles/second
Loading 100000 bundles with CRC_32: 462013 bundles/second
These numbers were generated on a MBP 13" 2018 with i5 CPU and 16GB of ram.
For debugging a small helper tool is shipped:
```
$ cargo install bp7
[...]
$ usage "bp7"
$ bp7 decode 9f8907000182016b6e6f6465332f696e626f7882016b6e6f6465332f696e626f7882016b6e6f6465332f696e626f78821a247966ba001ad693a4004225b686010000014341424342237186080100010042dbccff
[src/main.rs:17] &bndl = Bundle { primary: PrimaryBlock { version: 7, bundlecontrolflags: 0, crctype: 1, destination: Dtn( 1, "node3/inbox" ), source: Dtn( 1, "node3/inbox" ), reportto: Dtn( 1, "node3/inbox" ), creationtimestamp: CreationTimestamp( 611935930, 0 ), lifetime: 3600000000, fragmentationoffset: 0, totaldatalength: 0, crc: [ 37, 182 ] }, canonicals: [ CanonicalBlock { blocktype: 1, blocknumber: 0, blockcontrolflags: 0, crctype: 1, data: Data( [ 65, 66, 67 ] ), crc: [ 35, 113 ] }, CanonicalBlock { blocktype: 8, blocknumber: 1, blockcontrolflags: 0, crctype: 1, data: BundleAge( 0 ), crc: [ 219, 204 ] } ] } ```
The generated hex string can also be directly discplayed as raw cbor on the awesome cbor.me website, e.g. http://cbor.me/?bytes=9f8907000182016b6e6f6465332f696e626f7882016b6e6f6465332f696e626f7882016b6e6f6465332f696e626f78821a247966ba001ad693a4004225b686010000014341424342237186080100010042dbccff
The library should build for wasm even though only very few functions get exported. The example benchmark can also be used in the browser through the cargo-web crate:
cargo web start --target wasm32-unknown-unknown --example benchmark --release
Results should be shown in the javascript console on http://127.0.0.1:8000.
The performance is quite similar to the native performance:
Creating 100000 bundles with CRC_NO: 441696 bundles/second
Creating 100000 bundles with CRC_16: 416484 bundles/second
Creating 100000 bundles with CRC_32: 405022 bundles/second
Encoding 100000 bundles with CRC_NO: 1647039 bundles/second
Encoding 100000 bundles with CRC_16: 908059 bundles/second
Encoding 100000 bundles with CRC_32: 867603 bundles/second
Loading 100000 bundles with CRC_NO: 401727 bundles/second
Loading 100000 bundles with CRC_16: 388394 bundles/second
Loading 100000 bundles with CRC_32: 384186 bundles/second
Some functions can easily be used from javascript (cargo web deploy --release):
javascript
Rust.bp7.then(function(bp7) {
var b = bp7.rnd_bundle_now();
var enc = bp7.encode_to_cbor(b);
var payload = bp7.payload_from_bundle(b)
console.log(payload);
console.log(String.fromCharCode.apply(null, payload));
console.log(bp7.cbor_is_administrative_record(enc));
console.log(bp7.sender_from_cbor(enc));
console.log(bp7.recipient_from_bundle(b));
console.log(bp7.valid_bundle(b));
});
Note that at the moment all functions have a variant working on the binary bundle and one working on the decoded bundle struct.
Licensed under either of Apache License, Version 2.0 or MIT license at your option.
Unless you explicitly state otherwise, any contribution intentionally submitted for inclusion in bp7-rs by you, as defined in the Apache-2.0 license, shall be dual licensed as above, without any additional terms or conditions.