Lexer generators typically allow describing rules with regular expressions, but through Thompson's construction, subset construction, and DFA minimization, you can end up with a blazing fast/optimized DFA that processes your input at maximum speed.