Lexical analysis is equivalent to using regular expressions (and hence
insufficiently powerful) -- indeed many lexical analyzers are build by
constructing a finite automaton from a set of regular expressions.
well, for that matter, almost the entirety of rfc 822 can be parsed
with a finite automaton - the main exception is nested parentheses.
so you probably could write a regular expression that would recognize
rfc 822 in practice (e.g. just be prepared to handle at most 20
levels of nested parentheses).
But I think the point is that it's hard to find a _simple_ regex/FSM
that will do an adequate job of recognizing 2047 for all fields.
Keith