Runtime-Extensible SQL Parsers Using Peg

The article discusses the limitations of traditional SQL parsers in data management systems, which often rely on outdated technologies like YACC. These parsers are inflexible and hinder innovation due to their monolithic design. The authors advocate for a shift towards modern Parsing Expression Grammars (PEG), which allow for dynamic changes to query syntax and improved error recovery. PEG parsers can be reconfigured at runtime, enabling the integration of new syntax and features without the need for extensive recompilation. This flexibility is particularly beneficial as SQL specifications evolve and new query languages emerge. The authors present a prototype PEG parser that successfully parses SQL queries and demonstrate its extensibility through experiments. Although the PEG parser shows a performance slowdown compared to traditional methods, the absolute parsing times remain efficient for analytical queries. The article emphasizes the need for modernizing parser infrastructure to enhance user experience and support the growing complexity of SQL dialects. The findings will be presented at the 2025 Conference on Innovative Data Systems Research (CIDR).

- Traditional SQL parsers are outdated and inflexible, limiting innovation.

- Modern PEG parsers allow for dynamic syntax changes and better error handling.

- A prototype PEG parser has been developed, demonstrating extensibility and efficiency.

- Despite some performance trade-offs, PEG parsers maintain acceptable parsing times for analytical queries.

- The research highlights the importance of updating parser technology in data management systems.