This saves having to save and load the parser state.
This could give an incorrect token in some cases where the parser
communicates to the lexer. However this is not applicable in any of the
current usages and this would require one to parse the current token
as normal which is exactly what you don't want to do in that scenario.