Bruce Momjian wrote:
> Gregory Stark wrote:
> >
> > Alvaro Herrera <alvherre@commandprompt.com> writes:
> >
> > > Gregory Stark wrote:
> > >>
> > >> Well "char" doesn't have quite the same semantics as CHAR(1). If that's the
> > >> consensus though then I can work on either fixing "char" semantics to match
> > >> CHAR(1) or adding a separate type instead.
> > >
> > > What semantics?
> >
> > The main bit that comes to mind is 32::CHAR(1) give you '3' but 32::"char"
> > gives you ' '.
> >
> > Really it makes more sense if you think of "char" is a 1 byte integer type
> > with some extra text casts and operators to make C programmers happy, not a 1
> > byte character type.
>
> One very nifty trick would be to fix "char" to act as CHAR(), and map
> CHAR(1) automatically to "char".
Sorry, probably a stupid idea considering multi-byte encodings. I
suppose it could be an optimization for single-byte encodings, but that
seems very limiting.
-- Bruce Momjian bruce@momjian.us EnterpriseDB http://www.enterprisedb.com
+ If your life is a hard drive, Christ can be your backup. +