Introduction

compressionz is a fast, ergonomic compression library for Zig that provides a unified API across multiple compression algorithms.

The Problem

When working with compression in most languages, you face several challenges:

Fragmented APIs — Each codec has its own interface, options, and error handling
System dependencies — Libraries like zstd or brotli require system installation
Performance trade-offs — Choosing between speed, ratio, and memory usage
Missing features — Many bindings lack streaming, dictionaries, or proper error handling

The Solution

compressionz solves all of these with:

Unified API

// Same interface for every codec
const compressed = try cz.compress(.zstd, data, allocator);
const compressed = try cz.compress(.lz4, data, allocator);
const compressed = try cz.compress(.gzip, data, allocator);
const compressed = try cz.compress(.snappy, data, allocator);
const compressed = try cz.compress(.brotli, data, allocator);

Switch codecs by changing a single enum value. No API changes, no refactoring.

Zero System Dependencies

All C libraries are vendored directly in the repository:

zstd 1.5.7 — BSD licensed
zlib 1.3.1 — zlib licensed
brotli — MIT licensed

No brew install, no apt-get, no pkg-config. Just zig build.

Pure Zig Performance

LZ4 and Snappy are implemented in pure Zig with explicit SIMD optimizations:

// 16-byte vectorized comparison
const v1: @Vector(16, u8) = src[pos..][0..16].*;
const v2: @Vector(16, u8) = src[match_pos..][0..16].*;
const mask = @as(u16, @bitCast(v1 == v2));

This achieves 36+ GB/s compression throughput — competitive with hand-optimized C.

Complete Feature Set

Streaming — Process large files without loading into memory
Dictionaries — Better ratios for small data with known patterns
Zero-copy — Compress into pre-allocated buffers
Auto-detection — Identify format from magic bytes
Archives — ZIP and TAR support built-in
Safety — Decompression bomb protection with max_output_size

When to Use compressionz

Use compressionz when you need:

Multiple compression algorithms in one project
Predictable, cross-platform builds
High performance without sacrificing ergonomics
Streaming or dictionary compression
ZIP/TAR archive handling

Consider alternatives when you need:

Only one specific codec (use specialized bindings)
Async I/O integration (compressionz is synchronous)
Formats not yet supported (xz, lzma, etc.)

Codec Overview

Codec	Best For	Speed	Ratio
Zstd	General purpose	Fast	Excellent
LZ4	Maximum speed	Fastest	Good
Snappy	Real-time applications	Very fast	Moderate
Gzip	HTTP compatibility	Moderate	Good
Brotli	Static web assets	Slow	Best

See Choosing a Codec for detailed guidance.

Next Steps

Ready to get started?