Since the lexer is using a state machine, that par...
# klaxon
c
Since the lexer is using a state machine, that part is probably fast, but creating the token from characters seems to be taking too long