Re: [clisp-list] CHARSET questions & early vs late error signaling

[John K. H. <hi...@al...>]

 > I believe you should try to make CLISP match the database's
 > encoding. If you second-guess or hard-code ASCII, such code will not
 > work e.g. in the win32 port where the default encoding could be
 > OLD-DOS (whatever its name), while Oracle would probably use
 > ISO-8859-1.

Actually after some more thought I was thinking of trapping just the
specific case of where (a) the current CLISP encoding was ASCII and
(b) the Oracle charset was something anything other than ASCII.  This
should make for a minimum of second guessing on my part.  This library
is mainly for moving back-and-forth and not manipulating it.  I'm also
going to just make it a warning and not an error so that things are
not aborted in the case where there is only ASCII data extant in the
database.

   >> I'm using the FFI::c-string type.

 > Then custom:*foreign-encoding* is the place to set.

OK, I'll mention that in my warning message.

later,

---
John Hinsdale, Alma Mater Software, Inc., Tarrytown, NY 10591-3710 USA
hi...@al... | http://www.alma.com/staff/hin | +1 914 631 4690