[R6RS] Procedures that depend on Unicode character classification

William D Clinger will at ccs.neu.edu
Thu Jun 15 10:20:09 EDT 2006

> Is it really necessary to duplicate all that effort at this time?

I didn't realize there was a reference implementation of
those procedures.  Neither SRFI 75 nor the R6RS Subversion
repository had one, and I thought I had asked about the
availability of one in our last conference call and thought
I heard that none was available outside of the code for
specific systems such as Scheme48 and MzScheme.

I expect to finish the reference implementation within the
next couple of days, working just a couple of hours per day.
The guiding philosophy of this implementation is speed for
ASCII characters (and maybe ISO 8859-1), and compactness
for others.  It will be good enough to go into the next
version of Larceny, and it will be portable enough to drop
into virtually any other implementation.

Finally, as Matthew's assistant editor on the Unicode
subsystem, I thought this would be a good way for me to
learn a little about Unicode.


More information about the R6RS mailing list