Обсуждение: Cleanup explain_memoize function after test

Hi!

I was doing some hacking today, when I noticed that the memoize
regression test does not clean up the explain_memoize function after
its end. I checked other regression tests and they tend to do it. So,
I am not sure if there is a rule about `you should drop all objects
that you create in regression test`, but if there is such a rule, PFA.



-- 
Best regards,
Kirill Reshke

> On Apr 22, 2026, at 14:49, Kirill Reshke <reshkekirill@gmail.com> wrote:
>
> Hi!
>
> I was doing some hacking today, when I noticed that the memoize
> regression test does not clean up the explain_memoize function after
> its end. I checked other regression tests and they tend to do it. So,
> I am not sure if there is a rule about `you should drop all objects
> that you create in regression test`, but if there is such a rule, PFA.
>
>
>
> --
> Best regards,
> Kirill Reshke
> <v1-0001-Cleanup-explain_memoize-after-resgression-test.patch>

Good catch.

You also need to update the corresponding out file, otherwise the test will fail. I have helped update the out file in
v2,now “make check” passes with v2 applied. 

Best regards,
--
Chao Li (Evan)
HighGo Software Co., Ltd.
https://www.highgo.com/

Hi,

On Wed, Apr 22, 2026 at 05:24:33PM +0800, Chao Li wrote:
> 
> 
> > On Apr 22, 2026, at 14:49, Kirill Reshke <reshkekirill@gmail.com> wrote:
> > 
> > Hi!
> > 
> > I was doing some hacking today, when I noticed that the memoize
> > regression test does not clean up the explain_memoize function after
> > its end. I checked other regression tests and they tend to do it. So,
> > I am not sure if there is a rule about `you should drop all objects
> > that you create in regression test`, but if there is such a rule, PFA.
> > 
> > 
> > 
> > -- 
> > Best regards,
> > Kirill Reshke
> > <v1-0001-Cleanup-explain_memoize-after-resgression-test.patch>
> 
> Good catch.
> 
> You also need to update the corresponding out file, otherwise the test will fail. I have helped update the out file
inv2, now “make check” passes with v2 applied.
 


I'm not sure fixing "only" this single missing clean up makes sense. I think,
that the main purpose of those would be to be able to run "make installcheck EXTRA_REGRESS_OPTS="--use-existing""
multiple times on the same running instance without any errors. If that's the
case, then from what I can see after a quick test, the task would be way larger
that this patch.

Regards,

-- 
Bertrand Drouvot
PostgreSQL Contributors Team
RDS Open Source Databases
Amazon Web Services: https://aws.amazon.com