On Sun, Jul 07, 2013 at 08:55:01PM -0400, Peter Eisentraut wrote:
> On Sun, 2013-07-07 at 20:15 -0400, Noah Misch wrote:
> > I mildly recommend we reject this patch as such, remove the TODO item,
> > remove
> > the XXX comments this patch removes, and plan not to add more trivial
> > SPI
> > wrappers. If consensus goes otherwise, I think we should at least
> > introduce
> > SPI_getcollation() at the same time. Code that needs to transfer one
> > of them
> > very often needs to transfer the other. Having API coverage for just
> > one
> > makes it easier for hackers to miss that.
>
> The question is, what would one do with those values? It's hard to see
> when you would need the typmod and the collation of a result set. There
> might be cases, but enough to provide a special API for it?
Good point. One of the ways PL/pgSQL uses it is to feed a result datum back
into a future query as a Param node.
--
Noah Misch
EnterpriseDB http://www.enterprisedb.com