Is this project just a "learning experience" or something that is meant to be a "productive" library at some point?
The reason I am asking is because as a productive library I fear it may have great difficulty competing with hand crafted parsers (Acorn/Esprima/Cherow/...) in terms of performance.