Re: not null constraints, again

jian he <jian.universality@gmail.com>

From: jian he <jian.universality@gmail.com>
To: Alvaro Herrera <alvherre@alvh.no-ip.org>
Cc: Tender Wang <tndrwang@gmail.com>, Pg Hackers <pgsql-hackers@lists.postgresql.org>
Date: 2024-09-19T03:30:00Z
Lists: pgsql-hackers
On Tue, Sep 17, 2024 at 1:47 AM Alvaro Herrera <alvherre@alvh.no-ip.org> wrote:
>


still digging inheritance related issues.

drop table if exists pp1,cc1, cc2;
create table pp1 (f1 int, constraint nn check (f1 > 1));
create table cc1 (f2 text, f3 int ) inherits (pp1);
create table cc2(f4 float, constraint nn check (f1 > 1)) inherits(pp1,cc1);
execute constr_meta('{pp1,cc1, cc2}');
alter table only cc2 drop constraint nn;
ERROR:  cannot drop inherited constraint "nn" of relation "cc2"

So:

drop table if exists pp1,cc1, cc2;
create table pp1 (f1 int not null);
create table cc1 (f2 text, f3 int not null no inherit) inherits (pp1);
create table cc2(f4 float, f1 int not null) inherits(pp1,cc1);
execute constr_meta('{pp1,cc1, cc2}');
alter table only cc2 drop constraint cc2_f1_not_null;

Last "alter table only cc2" should fail?
because it violates catalog-pg-constraint coninhcount description:
"The number of direct inheritance ancestors this constraint has. A
constraint with a nonzero number of ancestors cannot be dropped nor
renamed."

also
alter table only cc2 drop constraint cc2_f1_not_null;
success executed.
some pg_constraint attribute info may change.
but constraint cc2_f1_not_null still exists.



Commits

  1. Suppress "may be used uninitialized" warnings from older compilers.

  2. Elide not-null constraint checks on child tables during PK creation

  3. Remove unnecessary code to handle CONSTR_NOTNULL

  4. Silence compilers about extractNotNullColumn()

  5. Add pg_constraint rows for not-null constraints