At 02:24 AM 17-11-00 +0100, Luc Franzen wrote in the mers list:

>In german we've got the 'Umlaute' (äöü) and I haven't had a need to save
>field-names including them so far (okay, I can bypass it by ae, oe, ue
>and it wouldn't lead to misunderstandings, but quite often I use english
>names anyway - sql is in english, too...) - but there are more annoying
>things about that issue: like NOT BEING ABLE to save Unicode (or at
>least CharSets) in exceptions.
>
>Hope, this will be changed.

Can we do this in some sensible way, even perhaps moving the RDB$MESSAGE
and RDB$DESCRIPTION into a separate relation in which the charset can be
specified

create exception TOO_DAMP charset FSS_UNICODE 'Es gibt Sprühregen in
Osnabrück', charset NONE 'It is drizzling in Osnabruck'

would store

in RDB$EXCEPTIONS:

RDB$EXCEPTION_NAME 'TOO_DAMP'
RDB$EXCEPTION_NUMBER 99
RDB$DESCRIPTION <BLOB> (doesn't seem to be accessible, anyway)

RDB$EXCEPTION_NUMBER 99
RDB$CHARSET FSS_UNICODE
RDB$MESSAGE 'Es gibt Sprühregen in Osnabrück'
RDB$SYSTEM_FLAG 0

RDB$EXCEPTION_NUMBER 99
RDB$CHARSET NONE
RDB$MESSAGE It is drizzling in Osnabruck'
RDB$SYSTEM_FLAG 0

and the database would be smart about binding the message choice to the
charset specified in SET NAMES.

One more thing for gbak to know about for upward conversions, I guess.

Cheers,
Helen

All for Open and Open for All
InterBase Developer Initiative · http://www.interbase2000.org
___________________________________________________