From: Tom Lane <tgl@sss.pgh.pa.us>
To: Peter Praxmarer <ppraxmarer@gmail.com>
cc: pgsql-bugs@lists.postgresql.org
Subject: Re: BUG: IS NOT NULL on RECORD variable fails in 17.5-dev
In-reply-to: 
 <CABBx1iF9pQfB3pOP9EBBHe0uyQo2oPRR-iWv0d_OOB-vn3rD+A@mail.gmail.com>
References: 
 <CABBx1iF9pQfB3pOP9EBBHe0uyQo2oPRR-iWv0d_OOB-vn3rD+A@mail.gmail.com>
Comments: In-reply-to Peter Praxmarer <ppraxmarer@gmail.com>
	message dated "Thu, 07 Aug 2025 12:43:54 -0700"
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <1603233.1754596435.1@sss.pgh.pa.us>
Date: Thu, 07 Aug 2025 15:53:55 -0400
Message-ID: <1603234.1754596435@sss.pgh.pa.us>
Archived-At: 
 <https://www.postgresql.org/message-id/1603234.1754596435%40sss.pgh.pa.us>
Precedence: bulk

Peter Praxmarer <ppraxmarer@gmail.com> writes:
> I have discovered what appears to be a severe bug in a development version
> of PostgreSQL 17 where a populated RECORD variable incorrectly fails an IS
> NOT NULL check.

No, that's behaving per SQL standard, and the same as many PG versions
before it.  Per [1]:

    If the expression is row-valued, then IS NULL is true when the row
    expression itself is null or when all the row's fields are null,
    while IS NOT NULL is true when the row expression itself is
    non-null and all the row's fields are non-null. Because of this
    behavior, IS NULL and IS NOT NULL do not always return inverse
    results for row-valued expressions; in particular, a row-valued
    expression that contains both null and non-null fields will return
    false for both tests.

IMHO this wasn't one of the SQL committee's better ideas, but we're
stuck with it.

			regards, tom lane

[1] https://www.postgresql.org/docs/current/functions-comparison.html