SquashA2

SquashA2 is a lossless data compression algorithm and reference implementation designed for use in embedded systems and resource-constrained environments. The name conveys the idea of “squashing” data size while the A2 designation signals its second-order adaptive context modeling approach. The project aims to provide a simple, fast, and portable compressor suitable for small payloads common in sensors, logs, and short messages.

The algorithm employs a two-pass encoding process with adaptive context modeling. In the first pass, symbol frequencies and short-range patterns are collected to inform a context-aware coding model. In the second pass, the data is encoded using a lightweight entropy coder that combines context-derived coding decisions with optional run-length encoding for sequences of repeated values. The design emphasizes low memory use and fast decoding, making it practical for microcontrollers and other environments with limited processing power and RAM.

SquashA2 originated in a collaborative effort among researchers and open-source contributors to address the needs of compact, real-time data workflows. It has been described in project documentation and online repositories as a reference implementation intended for educational use and for evaluating light-weight compression strategies in constrained devices.

Usage and availability: SquashA2 provides a C reference implementation with a permissive software license and a streaming API suitable for block-based or continuous data streams. It is intended for developers seeking modest compression without the complexity or resource demands of larger general-purpose algorithms. Limitations include less competitive performance on highly repetitive text or large multi-megabyte files compared with state-of-the-art general compressors.