Thread: sequence aliases?
I'm checking out Ruby on Rails and there's a tutorial about using Postgresql views and making them updateable (http://wiki.rubyonrails.com/rails/pages/HowtoUsePostgresViewsAsTables). The tutorial suggests renaming the sequence for the table to coincide with the view so that Rails can automatically access the sequence. I don't like this as it may break other things. Is there a way to create an alias or something (e.g. quotes_seq_id points to quotes_table_seq_id), or is there a better way? Thanks, CSN __________________________________ Yahoo! Mail - PC Magazine Editors' Choice 2005 http://mail.yahoo.com
CSN <cool_screen_name90001@yahoo.com> writes: > I'm checking out Ruby on Rails and there's a tutorial > about using Postgresql views and making them > updateable > (http://wiki.rubyonrails.com/rails/pages/HowtoUsePostgresViewsAsTables). > The tutorial suggests renaming the sequence for the > table to coincide with the view so that Rails can > automatically access the sequence. I don't like this > as it may break other things. Not to mention that it won't actually work --- the proposed approach fails because PG will refuse to let you drop the original sequence. (It has failed in every release since 7.3, so I don't put a lot of stock in that tutorial having been tested much.) Furthermore, it's quite likely that future releases won't let you fool with the default expression of a serial column at all. My advice to the Rails people would be to fix whatever it is in their code that is assuming a particular sequence name, or indeed assuming a sequence at all... regards, tom lane
Tom Lane <tgl@sss.pgh.pa.us> writes: > My advice to the Rails people would be to fix whatever it is in their > code that is assuming a particular sequence name, or indeed assuming > a sequence at all... Well how else do you find the id of the last inserted record without assuming a sequence? -- greg
On Sun, Nov 06, 2005 at 12:15:45AM -0500, Greg Stark wrote: > > Tom Lane <tgl@sss.pgh.pa.us> writes: > > > My advice to the Rails people would be to fix whatever it is in their > > code that is assuming a particular sequence name, or indeed assuming > > a sequence at all... > > Well how else do you find the id of the last inserted record without assuming > a sequence? I suppose using something like: select currval( pg_get_serial_sequence( 'table','col' ) ); That avoids hardcoding the sequence id. Not assuming a sequences at all may be trickier. Have a nice day, -- Martijn van Oosterhout <kleptog@svana.org> http://svana.org/kleptog/ > Patent. n. Genius is 5% inspiration and 95% perspiration. A patent is a > tool for doing 5% of the work and then sitting around waiting for someone > else to do the other 95% so you can sue them.
Attachment
--- Martijn van Oosterhout <kleptog@svana.org> wrote:
> On Sun, Nov 06, 2005 at 12:15:45AM -0500, Greg Stark
> wrote:
> >
> > Tom Lane <tgl@sss.pgh.pa.us> writes:
> >
> > > My advice to the Rails people would be to fix
> whatever it is in their
> > > code that is assuming a particular sequence
> name, or indeed assuming
> > > a sequence at all...
> >
> > Well how else do you find the id of the last
> inserted record without assuming
> > a sequence?
>
> I suppose using something like:
>
> select currval( pg_get_serial_sequence(
> 'table','col' ) );
I like that better than the current 'table_col_seq'.
Perhaps just currval('table', 'col'). Or perhaps
tables could be made to have a default sequence (one
that's associated with the primary key).
CSN
>
> That avoids hardcoding the sequence id. Not assuming
> a sequences at all
> may be trickier.
>
> Have a nice day,
> --
> Martijn van Oosterhout <kleptog@svana.org>
> http://svana.org/kleptog/
> > Patent. n. Genius is 5% inspiration and 95%
> perspiration. A patent is a
> > tool for doing 5% of the work and then sitting
> around waiting for someone
> > else to do the other 95% so you can sue them.
>
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com