[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]


    Date: Thu, 23 Mar 89 23:11:13 GMT
    From: Jeff Dalton <jeff%aiai.edinburgh.ac.uk@NSS.Cs.Ucl.AC.UK>

    It's very late, but here it is.


The write-up looks quite good to me.  Even though it's late, I plan to
support it.

I prefer option READTABLE-KEYWORDS for two reasons:
(1) your recent arguments about printer issues being simpler that way
(2) the lack of need to add CHAR-INVERT-CASE (which I don't think is very 
    useful outside of this context).

I do want to mention that I think having an arbitrary function is not as
hard on the printer as you might expect, so I'm not sure READTABLE-FUNCTION
really needs to restrict its inputs.  When the system sees an unknown
function, it could iterate over both-case-p characters, calling the function
and figuring out the mappings. Since we're talking about the system, it will
know which chars those are even in the face of international char sets.
It can determine from such a table whether the mapping is one-to-one for
a particular character (so it will know if escaping is needed) as well as
whether the mapping is case-preserving for any given character.  However,
even though I think this is possible, I admit it is a pain and probably
just plain not worth the effort so I don't think it's the way to go. Based
on this, I don't really oppose READTABLE-FUNCTION, but I'd much prefer the
simpler READTABLE-KEYWORDS proposal anyway.

Btw, there may be some overlap between this and 
PRINT-CASE-PRINT-ESCAPE-INTERACTION. I don't have time to check this to be
sure, but you might want to double-check to avoid last-minute snags.