Re: Query ID Calculation Fix for DISTINCT / ORDER BY and LIMIT / OFFSET

Michael Paquier <michael@paquier.xyz>

From: Michael Paquier <michael@paquier.xyz>
To: Sami Imseih <samimseih@gmail.com>
Cc: David Rowley <dgrowleyml@gmail.com>, Ivan Bykov <I.Bykov@modernsys.ru>, pgsql-hackers@postgresql.org
Date: 2025-03-14T01:50:55Z
Lists: pgsql-hackers
> On Mar 14, 2025, at 8:15, Sami Imseih <samimseih@gmail.com> wrote:
>> FWIW, another idea I have on top of my mind is the addition of a
>> counter in JumbleState that we increment each time we enter
>> _jumbleNode(), then simply call JUMBLE_FIELD_SINGLE() after the
>> incrementation.  And we can rely on this counter to be unique each
>> time we jumble a node..
> 
> With this approach, the author of the custom function will need
> to remember to increment the counter, right?

Actually, no. _jumbleNode() is the unique entry point we use when jumbling a sub-Node in a bigger Node structure, so custom functions don’t need to touch it.

> Also, I am wondering what is the difference between appending an incrementing
> counter in JState vs just appending a constant for every node ( NULL
> or NOT NULL ) ?

Well, the problem is not NULL vs non-NULL, so..
—
Michael



Commits

  1. Optimize Query jumble