Re: Allowing NOT IN to use ANTI joins

From: Marti Raudsepp <marti(at)juffo(dot)org>
To: David Rowley <dgrowleyml(at)gmail(dot)com>
Cc: PostgreSQL-development <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: Allowing NOT IN to use ANTI joins
Date: 2014-06-11 09:32:48
Message-ID: CABRT9RDtE+E0=VWm_aFvhfaQ1R+g54OOxs85y+KeYEgQmuYgWQ@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On Sun, Jun 8, 2014 at 3:36 PM, David Rowley <dgrowleyml(at)gmail(dot)com> wrote:
> Currently pull_up_sublinks_qual_recurse only changes the plan for NOT EXISTS
> queries and leaves NOT IN alone. The reason for this is because the values
> returned by a subquery in the IN clause could have NULLs.

There's a bug in targetListIsGuaranteedNotToHaveNulls, you have to
drill deeper into the query to guarantee the nullability of a result
column. If a table is OUTER JOINed, it can return NULLs even if the
original column specification has NOT NULL.

This test case produces incorrect results with your patch:

create table a (x int not null);
create table b (x int not null, y int not null);
insert into a values(1);
select * from a where x not in (select y from a left join b using (x));

Unpatched version correctly returns 0 rows since "y" will be NULL.
Your patch returns the value 1 from a.

Regards,
Marti

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message David Rowley 2014-06-11 09:43:55 Re: Allowing NOT IN to use ANTI joins
Previous Message Marti Raudsepp 2014-06-11 09:17:03 Re: Allowing NOT IN to use ANTI joins