Tom Lane writes:
> In any case we need to distinguish implicit coercion for a store from
> explicit coercion. Do your recent pg_cast changes make that any easier?
Possibly we could allow for registering both implicit and explicit cast
functions for the same combination. Or we could make cast functions take
an optional second argument that tells whether the cast is implicit or
explicit. We probably also need to take into account the typemod coercion
functions that are currently handled in a semi-internal way. (Any
formalization of those should probably consider the possibility of
allowing typemods on arbitrary types.) Not sure which way to go.
--
Peter Eisentraut peter_e@gmx.net