compression-algorithms

Compression algorithms, readable in Rust. Learn how your files get smaller.

What's Inside

Module	Algorithm	One-line summary
huffman	Huffman coding	Variable-length prefix codes from a frequency table
rle	Run-Length Encoding (RLE)	Collapse consecutive repeated bytes
lzw	Lempel-Ziv-Welch (LZW)	Dictionary compression that builds its codebook on the fly
lz77	LZ77 sliding window	Back-reference compression with a sliding window
bwt	Burrows-Wheeler Transform	Reversible permutation that clusters similar characters
mtf	Move-to-Front Transform	Converts BWT output into a stream of small integers
arithmetic	Arithmetic coding	Encodes a message as a single fractional number in [0, 1)
delta	Delta + ZigZag + VarInt	Stores differences between consecutive values
entropy	Shannon entropy & KL divergence	Theoretical lower bounds for lossless compression

Quick Start

use compression_algorithms::*;

// Huffman coding
let data = b"abracadabra";
let coder = HuffmanCoder::from_data(data);
let (packed, padding) = coder.encode_packed(data);
let recovered = coder.decode_packed(&packed, padding);
assert_eq!(recovered, data);

// Entropy analysis
let entropy = Entropy::shannon_entropy(data);
println!("Shannon entropy: {entropy:.3} bits/symbol");

// BWT → MTF → Huffman pipeline
let (bwt, idx) = BwtTransform::forward(data);
let mtf = MtfTransform::forward(&bwt);
let coder = HuffmanCoder::from_data(&mtf);
let (compressed, pad) = coder.encode_packed(&mtf);

Install

cargo add compression-algorithms

License

MIT OR Apache-2.0