[R6RS] Unicode scalar value escape sequences

Michael Sperber sperber
Tue Mar 22 07:37:38 EST 2005


>>>>> "Manuel" == Manuel Serrano <Manuel.Serrano at sophia.inria.fr> writes:

Manuel> Won't it be acceptable to use a representation with a fixed
Manuel> number of digits? For instance #uXXXX or #uXXXXXX. This
Manuel> eliminates the need for an ending delimiter.  I agree that it
Manuel> imposes extra 0 at the beginning of the token but is it really
Manuel> important?

In principle, that's correct.  However, the specific representation
that was picked for the string escape sequences is the way it is (if I
understand Matthew correctly) because of compatibility with C and/or
Java.  Going to a fixed number of digits would be an improvement,
IMHO, but it would break that compatibility.

If we're going to break compatibility, we might as well pick some
other representation, was my thinking, especially as I like Marc's
method of using the regular numeric literals.

I'd be having trouble remembering which of x, u, U is 2, 4, 6 digits
respectively.

-- 
Cheers =8-} Mike
Friede, V?lkerverst?ndigung und ?berhaupt blabla


More information about the R6RS mailing list