From: Nikolay Shaplov <dhyan@nataraj.su>
To: pgsql-hackers@lists.postgresql.org
Cc: Alvaro Herrera <alvherre@2ndquadrant.com>, Tom Lane <tgl@sss.pgh.pa.us>,
 Robert Haas <robertmhaas@gmail.com>,
 Nikita Glukhov <n.gluhov@postgrespro.ru>,
 "pgsql-hackers@postgresql.org" <pgsql-hackers@postgresql.org>
Subject: Re: [PATCH][PROPOSAL] Add enum releation option type
Date: Sun, 06 Jan 2019 18:28:11 +0300
Message-ID: <1777585.CqHLqLDU7p@x200m>
User-Agent: KMail/5.2.3 (Linux/4.9.0-8-amd64; KDE/5.28.0; x86_64; ; )
In-Reply-To: <201901032112.xpqqwmumqml2@alvherre.pgsql>
References: <201901032112.xpqqwmumqml2@alvherre.pgsql>
MIME-Version: 1.0
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

=D0=92 =D0=BF=D0=B8=D1=81=D1=8C=D0=BC=D0=B5 =D0=BE=D1=82 =D1=87=D0=B5=D1=82=
=D0=B2=D0=B5=D1=80=D0=B3, 3 =D1=8F=D0=BD=D0=B2=D0=B0=D1=80=D1=8F 2019 =D0=
=B3. 18:12:05 MSK =D0=BF=D0=BE=D0=BB=D1=8C=D0=B7=D0=BE=D0=B2=D0=B0=D1=82=D0=
=B5=D0=BB=D1=8C Alvaro Herrera=20
=D0=BD=D0=B0=D0=BF=D0=B8=D1=81=D0=B0=D0=BB:
> Attached version 7, with some renaming and rewording of comments.
> (I also pgindented.  Some things are not pretty because of lack of
> typedefs.list patching ... a minor issue at worst.)
Thanks! Imported it into my code tree..

> I'm still not satisfied with the way the builtin enum tables are passed.
> Can we settle for using add_enum_reloption instead at initialization
> time?  Maybe that would be less ugly.  Alternatively, we can have
> another member in relopt_enum which is a function pointer that returns
> the array of relopt_enum_elt_def.  Not sure at which point to call that
> function, though.
Actually I am not satisfied with it too... That is why I started that bigge=
r=20
reloptions refactoring work. So for now I would offer to keep initializatio=
n=20
as it was for other types: in initialize_reloptions  using [type]RelOpts[]=
=20
arrays. And then I will offer patch that changes it for all types and remov=
e=20
difference between biult-in reloptions and reloptions in extensions.

> I think it would be great to have a new enum option in, say,
> contrib/bloom, to make sure the add_enum_reloption stuff works
> correctly.  If there's nothing obvious to add there, let's add something
> to src/test/modules.
As far as I can see there is no obvious place in bloom where we can add enu=
m=20
options.

So I see several options here:

1. I can try to create some dummy index in  src/test/modules. And it would =
be=20
very useful for many other tests. For example we will have no real string=20
reloption when this patch is applied. But it may take a reasonable time to =
do,=20
because I've never created indexes before, and I do not have many time-slot=
s=20
for postgres development in my schedule.

2. Actually I am going to offer patch that will remove difference between=20
build-in and extension reloptions, all reloptions will use same API. After=
=20
applying that patch we would not need to test add_[type]_reloption calls=20
separately. So may be it would be good enough to check that add_enum_relopt=
ion=20
works right now (add an enum option to bloom, check that it works, and do n=
ot=20
commit that code anywhere) and then wait until we get rid of=20
add_[type]_reloption API.=20
This will save us some time. I am a bit worrying about Nikita Glukhov patch=
 it=20
is better to have reloptions reworked before adding opclass options.

2.5  May be this src/test/modules dummy index is subject to another patch. =
So=20
I will start working on it right now, but we will do this work not dependen=
t=20
to any other patches. And just add there what we need when it is ready. I=20
would actually add there some code that checks all types of options, becaus=
e=20
we actually never check that option value from reloptons successfully reach=
es=20
extension code. I did this checks manually, when I wrote that bigger relopt=
ion=20
patch, but there is no facilities to do regularly check is via test suit.=20
Creating dummy index will create such facilities.=20

=46or me all options are good enough, so it is for you too choose, I will d=
o as=20
you advices.