“Elevating Parsing in Rust : A Peek into pest3"

On the surface, pest appears as just another parser generator within the Rust ecosystem, yet it stands out by being an external domain specific language with dedicated tooling. Developed with a commitment to accessibility, correctness, and performance, pest utilizes parsing expression grammars (PEG). These grammars, akin to regular expressions but more expressive, enable the parsing of intricate languages that handwritten parsers could easily make mistakes in.

As we delve into the current state of pest, it becomes clear that its journey is not without challenges. While it serves as a robust general-purpose parser generator, there are inherent limitations that users encounter. This talk aims to provide a reflective examination of these shortcomings, not merely as critiques, but as opportunities for growth and innovation.

Looking towards the horizon, we introduce "pest3" — a forthcoming iteration that promises to address these limitations head-on. With a working prototype already in development, this new version is set to improve the way we handle parsing in Rust. Central to this transformation are the newly introduced trivia handling sequence operators and a typed tree API, which collectively enhance the parser's capability to manage nuanced syntactical structures.

Tomas Tauber is a software engineer and researcher based in Hong Kong. He completed his PhD in Computer Science at The University of Hong Kong, where his research focused on programming language implementation and design.

Since 2018, Tomas has been working professionally with the Rust programming language. Prior to his current role, he was a Vice President of Blockchain Engineering at Crypto.com, building decentralized applications and infrastructure. Tomas is now self-employed, working primarily as a consultant and developer for IDVerse, a startup building novel digital identity verification solutions. He is passionate about exploring the capabilities of Rust for building secure, reliable, and high-performance systems.