Horton

Re: cataloguing NOT NULL constraints

Alexander Lakhin <exclusion@gmail.com>

View thread
From: Alexander Lakhin <exclusion@gmail.com>
To: Alvaro Herrera <alvherre@alvh.no-ip.org>
Cc: Dean Rasheed <dean.a.rasheed@gmail.com>, Andrew Bille <andrewbille@gmail.com>, Peter Eisentraut <peter@eisentraut.org>, Robert Haas <robertmhaas@gmail.com>, pgsql-hackers@lists.postgresql.org
Date: 2024-04-25T05:00:00Z
Lists: pgsql-hackers

Commits

Same data as JSON: GET /api/v1/messages/:b64id/commits the thread's linked commits as JSON, with link sources. API reference →

  1. Revert structural changes to not-null constraints

  2. Fix inconsistencies in error messages

  3. Disallow direct change of NO INHERIT of not-null constraints

  4. Disallow NO INHERIT not-null constraints on partitioned tables

  5. Better handle indirect constraint drops

  6. Don't try to assign smart names to constraints

  7. Fix restore of not-null constraints with inheritance

  8. ATTACH PARTITION: Don't match a PK with a UNIQUE constraint

  9. Fix propagating attnotnull in multiple inheritance

  10. Check stack depth in new recursive functions

  11. Move privilege check to the right place

  12. Update information_schema definition for not-null constraints

  13. Fix not-null constraint test

  14. Disallow changing NO INHERIT status of a not-null constraint

  15. Catalog not-null constraints

  16. parallel_schedule: add comment on event_trigger test dependency

  17. Revert "Catalog NOT NULL constraints" and fallout

  18. Adjust contrib/sepgsql regression test expected outputs.

  19. Fix table name clash in recently introduced test

  20. Catalog NOT NULL constraints

  21. Change the rules for inherited CHECK constraints to be essentially the same

24.04.2024 20:36, Alvaro Herrera wrote:
> So I added a restriction that we only accept such a change when
> recursively adding a constraint, or during binary upgrade.  This should
> limit the damage: you're no longer able to change an existing constraint
> from NO INHERIT to YES INHERIT merely by doing another ALTER TABLE ADD
> CONSTRAINT.
>
> One thing that has me a little nervous about this whole business is
> whether we're set up to error out where some child table down the
> hierarchy has nulls, and we add a not-null constraint to it but fail to
> do a verification scan.  I tried a couple of cases and AFAICS it works
> correctly, but maybe there are other cases I haven't thought about where
> it doesn't.
>

Thank you for the fix!

While studying the NO INHERIT option, I've noticed that the documentation
probably misses it's specification for NOT NULL:
https://www.postgresql.org/docs/devel/sql-createtable.html

where column_constraint is:
...
[ CONSTRAINT constraint_name ]
{ NOT NULL |
   NULL |
   CHECK ( expression ) [ NO INHERIT ] |

Also, I've found a weird behaviour with a non-inherited NOT NULL
constraint for a partitioned table:
CREATE TABLE pt(a int NOT NULL NO INHERIT) PARTITION BY LIST (a);
CREATE TABLE dp(a int NOT NULL);
ALTER TABLE pt ATTACH PARTITION dp DEFAULT;
ALTER TABLE pt DETACH PARTITION dp;
fails with:
ERROR:  relation 16389 has non-inherited constraint "dp_a_not_null"

Though with an analogous check constraint, I get:
CREATE TABLE pt(a int, CONSTRAINT nna CHECK (a IS NOT NULL) NO INHERIT) PARTITION BY LIST (a);
ERROR:  cannot add NO INHERIT constraint to partitioned table "pt"

Best regards,
Alexander