Re: strange IS NULL behaviour

Bruce Momjian <bruce@momjian.us>

From: Bruce Momjian <bruce@momjian.us>

To: Kevin Grittner <kgrittn@ymail.com>

Cc: Robert Haas <robertmhaas@gmail.com>, Andres Freund <andres@2ndquadrant.com>, Tom Lane <tgl@sss.pgh.pa.us>, Hannu Krosing <hannu@2ndquadrant.com>, Peter Eisentraut <peter_e@gmx.net>, Andrew Dunstan <andrew@dunslane.net>, PostgreSQL-development <pgsql-hackers@postgresql.org>

Date: 2013-09-10T19:52:00Z

Lists: pgsql-hackers

On Tue, Sep 10, 2013 at 12:48:08PM -0700, Kevin Grittner wrote:
> Bruce Momjian <bruce@momjian.us> wrote:
> 
> > FYI, I think these queries below prove that NOT NULL constraints do not
> > follow the single-depth ROW NULL inspection rule that PL/pgSQL follows,
> > and that my patch was trying to promote for queries:
> >
> >     CREATE TABLE test2(x test NOT NULL);
> >     CREATE TABLE
> >     INSERT INTO test2 VALUES (null);
> >     ERROR:  null value in column "x" violates not-null constraint
> >     DETAIL:  Failing row contains (null).
> > -->    INSERT INTO test2 VALUES (row(null));
> >     INSERT 0 1
> 
> If I remember correctly, the standard wants a NOT NULL constraint
> on a column with a composite type to behave the same as
> 
>   CHECK (col IS DISTINCT FROM NULL)
> 
> ... which is consistent with the behavior you show.

Is IS DISTINCT FROM correct though?

	SELECT ROW(NULL) IS DISTINCT FROM NULL;
	 ?column?
	----------
	 t
	(1 row)

-- 
  Bruce Momjian  <bruce@momjian.us>        http://momjian.us
  EnterpriseDB                             http://enterprisedb.com

  + It's impossible for everything to be true. +