On Thu, 2003-06-05 at 11:11, Rod Taylor wrote:
> On Wed, 2003-06-04 at 22:28, Christopher Kings-Lynne wrote:
> > > Which ones are missing, and should we really be looking at creating
> a
> > > pg_definition_schema instead?
> >
> > Missing:
> >
> > Database, schema, table, domain, cast, conversion, function...
> >
> > Maybe a definition schema might be better.....dunno...it would need to
> use
> > the pg_get_*def functions anyway methinks.
yeah, i would think it would, but isn't the point of the information
schema to help hide the back end tech?
>
> As an interface writer, do you prefer dealing with functions like
> pg_get_constraintdef() or a view like the information schema provides?
>
I would think it is easier to get the information from the information
schema. That's most like what we're doing now getting the information
from the pg_* tables and istm it's easier to browse the information
schema than dig through function definitions. To be fair Chris tends to
hack on pg_dump at a much deeper level than I on either pg_dump or psql,
so he might be more familiar with the functions and have a different
viewpoint.
> The function doesn't easily allow determination of items such as the ON
> UPDATE or ON DELETE type (statement parsing is required), but the
> information schema gives the information in a segregated manner.
>
Well, the biggest pain in the arse I had to deal with in phpPgAdmin was
the handling of permissions, specifically due to having to parse through
the relacl information. I was so glad when Chris fixed up my half
working implementation.
> The pg_get_*def() functions seem to have been created primarily for psql
> and pg_dump.
yeah. again Chris tends to hack on pg_dump so he might see it
differently than I (and I haven't looked at psql in months).
(He's on holiday for the next few days btw which is why I'm chiming in)
Robert Treat
--
Build A Brighter Lamp :: Linux Apache {middleware} PostgreSQL
