montana-zstd

Zstandard (zstd) compression implementation for the Montana batch submission pipeline.

Overview

This crate provides a ZstdCompressor that implements the Compressor trait from montana-pipeline. Zstandard offers an excellent balance of compression ratio and speed, making it ideal for L2 batch compression.

Usage

use montana_zstd::ZstdCompressor;
use montana_pipeline::Compressor;

// Create a compressor with default settings
let compressor = ZstdCompressor::default();

// Or use presets
let fast = ZstdCompressor::fast();           // Level 1 - fastest
let balanced = ZstdCompressor::balanced();   // Level 3 - good balance
let best = ZstdCompressor::best();           // Level 19 - best compression

// Compress data
let data = b"Hello, World!";
let compressed = compressor.compress(data).unwrap();

// Decompress data
let decompressed = compressor.decompress(&compressed).unwrap();
assert_eq!(decompressed, data);

Compression Levels

Zstandard supports compression levels 1-22:

• 1: Fastest compression
• 3: Default/balanced compression
• 19: Best compression (zstd --ultra not enabled)
• 22: Maximum compression (very slow)

Why Zstandard?

Zstandard was developed by Facebook and offers:

• Fast decompression: Critical for L2 derivation
• Good compression ratios: Competitive with Brotli
• Dictionary support: Can achieve even better ratios with pre-trained dictionaries
• Widely adopted: Used by many blockchain projects