Fastest LZ4 implementation in Rust. Originally based on redox-os' lz4 compression, but now a complete rewrite. The results in the table are from a benchmark in this project (66Kb JSON).
| Compressor | Compression | Decompression | |------------------|-------------|---------------| | lz4flex unsafe | 711 MiB/s | 3733 MiB/s | | lz4flex safe | 628 MiB/s | 1433 MiB/s | | lz4cpp | 914 MiB/s | 3793 MiB/s | | lz4fear | 443 MiB/s | 836 MiB/s |
Compression and decompression uses no usafe via the default feature flags "safe-encode" and "safe-decode". If you need more performance you can disable them (e.g. with no-default-features).
Safe:
lz4_flex = { version = "0.7.0" }
Performance:
lz4_flex = { version = "0.7.0", default-features = false }
```rust use lz4flex::{compressprependsize, decompresssize_prepended};
fn main(){ let input: &[u8] = b"Hello people, what's up?"; let compressed = compressprependsize(input); let uncompressed = decompresssizeprepended(&compressed).unwrap(); assert_eq!(input, uncompressed); } ```
The benchmark is run with criterion, the test files are in the benches folder.
Currently 3 implementations are compared, this one, the redox-version, lz-fear and the c++ version via rust bindings. The lz4-flex version is tested with the feature flags safe-decode and safe-encode switched on and off.
cargo bench --no-default-features
Executed on Core i7-6700 Win10 WSL.
cargo bench
Executed on Core i7-6700 Win10 WSL.
Miri can be used to find issues related to incorrect unsafe usage:
MIRIFLAGS="-Zmiri-disable-isolation -Zmiri-disable-stacked-borrows" cargo miri test --no-default-features
This fuzz target fuzzes, and asserts compression and decompression returns the original input.
cargo fuzz run fuzz_roundtrip
This fuzz target fuzzes, and asserts compression with cpp and decompression returns the original input.
cargo fuzz run fuzz_roundtrip_cpp_compress