Re: NOT ENFORCED constraint feature

On 2025-Mar-12, Amul Sul wrote:

> On Tue, Mar 11, 2025 at 11:13 PM Peter Eisentraut <peter(at)eisentraut(dot)org> wrote:

> > I think the next step here is that you work to fix Álvaro's concerns
> > about the recursion structure.
>
> Yes, I worked on that in the attached version. I refactored
> ATExecAlterConstraintInternal() and moved the code that updates the
> pg_constraint entry into a separate function (see 0001), so it can be
> called from the places where the entry needs to be updated, rather
> than revisiting ATExecAlterConstraintInternal(). In 0002,
> ATExecAlterConstraintInternal() is split into two functions:
> ATExecAlterConstrDeferrability() and
> ATExecAlterConstrInheritability(), which handle altering deferrability
> and inheritability, respectively. These functions are expected to
> recurse on themselves, rather than revisiting
> ATExecAlterConstraintInternal() as before. This approach simplifies
> things. Similarly can add ATExecAlterConstrEnforceability() which
> recurses itself.

Yeah, I gave this a look and I think this code layout is good. There
are more functions now, but the code flow is simpler.

Thanks!

--
Álvaro Herrera 48°01'N 7°57'E — https://www.EnterpriseDB.com/