On Fri, 15 Aug 2003 08:03:04 -0700 (PDT)
Stephan Szabo <[EMAIL PROTECTED]> wrote:
> On Fri, 15 Aug 2003, expect wrote:
>
> > On Fri, 15 Aug 2003 00:32:01 -0700 (PDT)
> > Stephan Szabo <[EMAIL PROTECTED]> wrote:
> >
> > >
> > > On Thu, 14 Aug 2003, expect wrote:
> > >
> > > > On Thu, 14 Aug 2003 12:46:07 -0500
> > > > Bruno Wolff III <[EMAIL PROTECTED]> wrote:
> > > > > Note that it isn't obvious what empty strings should map to for numbers.
> > > > > NULL and 0 make about as much sense as using the default value.
> > > >
> > > > Well I'm new here but it seems to me they should map to the default value
> > > > for that column. Why wouldn't they?
> > >
> > > One problem with doing that is that it's inconsistent.
> >
> > Please elaborate. How is it inconsistent, exactly?
>
> See my comments below.
>
> > >
> > > Given
> > > create table test(
> > > a text default 'abc',
> > > b int default 5
> > > );
> > >
> > > copy test from stdin with delimiter ',';
> > > ,
> > > \.
> > >
> > > What would you expect the values of the row in test
> > > to be?
> >
> >
> > Oh a test....
> >
> > Does the \. end the STDIN input?
> >
> > Where's the null option? Don't you mean:
> >
> > copy test from stdin with delimiter ',' null '';
>
> No, I'm saying without a null specifier.
>
> > In this case I would expect the row to have:
> >
> > a | b
> > ----------
> > |
> > abc | 5
> >
> >
> > Is this too much to expect?
>
> Without a null specifier of '', empty string is a valid value for a.
> Why would it get the default, and how would you insert an empty string?
> Should it treat b differently because '' isn't valid for that type, that'd
> be inconsistent.
>
> With a null specifier of '', empty string is valid for both and means
> NULL. Otherwise, how would you specify a null when you have that
> null specifier?
Don't want to specify a null.
>
> What you probably really want is another specifier that inserts the
> default for a column, so you could say something like:
> copy test from stdin with delimiter ',' default '';
No, I don't want anyone else to do anymore work. I'd much rather see
improvements than new features, unless the new feature is a big
improvement.
>
>
>
>
---------------------------(end of broadcast)---------------------------
TIP 6: Have you searched our list archives?
http://archives.postgresql.org
