Re: Implementing JDBC3 methods (Was: JDBC and fetching

On Thu, 2002-09-19 at 04:17, Michael Stephenson wrote:
> Dave wrote:
> >>I think so. I just read over the JDBC specification. It is not very
> >>specific, about what is considered a generated key. I suppose it would be
> >>any unique key (or field value?) that is automatically generated by the data
> >>source / database. In postgresql, this could be a SERIAL, a field with a
> >>default value from a SEQUENCE, of even an OID. This *could* also be a value
> >>supplied by a user-defined trigger, depends on the point of view.
> >
> > I would think it would have to be a sequence, and an index This should
> > be possible to find out through a select.
> > ie find if the table has a default on an index column which selects
> > nextval.
>
> That isn't really sufficient. There doesn't seem to be anything to stop you
> inserting an explicit value into a column which is of type serial, so you need
> to know whether the sequence was actually used or not. Trying to find that the
> obvious way you hit the problem that a sequence might have been used earlier in
> this transaction in a different statement (or even elsewhere in this statement)..
That's the only way, and will have to do. One thing to realize is that
once a sequence has been used, it can't be rolled back, so it is unique.
>
> Michael
>
> --
> Web Applications Developer
> Open World Ltd, 11 Riverside Court, Riverside Road, Bath, BA2 3DZ.
> Tel: +44 1225 444950  Fax: +44 1225 336738  http://www.openworld.co.uk/
>
> CONFIDENTIALITY NOTICE
> The information contained in this message is confidential, intended only for
> the use of the individual or the entity named as recipient. If the reader of
> this message is not that recipient, you are notified that any dissemination,
> distribution or copy of this message is strictly prohibited. If you have
> received this message in error, please immediately notify us by telephone on
> the number above. Your co-operation is appreciated.
>
>