On Wed, Nov 30, 2011 at 10:53 AM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> On the whole, it might not be a bad idea to have two allowed signatures
> for the validator function, rather than inventing an additional column
> in pg_language. But the fundamental point IMHO is that there needs to
> be a provision to pass language-dependent validation options to the
> function, whether it's the existing validator or a separate checker
> entry point.
Something like:
CHECK FUNCTION proname(proargs) WITH (...fdw-style elastic options...)
?
--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company