On 2026-02-18 We 7:10 PM, Tom Lane wrote:
Haritabh Gupta <[email protected]> writes:
Thanks for addressing the comments. I tested v7 and found that
type modifiers (typmod) are lost in the base type output.
This report crystallized something that's been bothering me
about not only pg_get_domain_ddl() but all the similar patches
that are in the queue. They are adding a large amount of new
code that will have to be kept in sync with behavior elsewhere,
and there is basically zero forcing function to ensure that
that happens. Even the rather-overly-voluminous test cases
proposed for the functions cannot catch errors of omission,
especially not future errors of omission.
I don't really know what to do about this, but I don't like the
implementation approach that's being proposed. I think it's
loading too much development effort and future maintenance effort
onto us in comparison to the expected benefit of having these
functions.
Do you have an alternative suggestion? We could create an extension, but
keeping that in sync might in fact be harder, and we know from
experience that extensions are not universally available. That would
make leveraging these functions for something like Matheus Alcantara's
schema cloning proposal (as I think Alvaro suggested) pretty much
impossible.
I'm not sure how much maintenance effort you think will be needed. We
don't change the shape of database objects all that often.
cheers
andrew
--
Andrew Dunstan
EDB: https://www.enterprisedb.com
