[stepvhen]

I would have to check, but the resources on lex/yacc that I remember would make a point of their usability at scale. Generally it would be unlikely to wrote a faster lexer than what lex could do, but very likely to write a better parser than yacc. However writing a small lexer and/or parser _quickly_ would be easier with those tools than hand rolling one almost every time. Its great parser generators are still being developed and improved, and its fine to point out issues, but this article misses the point by a wide margin.