Re: Collation versioning - Mailing list pgsql-hackers

From Julien Rouhaud
Subject Re: Collation versioning
Date
Msg-id 20200227074535.4qmrehctqi6nrqei@nol
Whole thread Raw
In response to Re: Collation versioning  (Thomas Munro <thomas.munro@gmail.com>)
Responses Re: Collation versioning
List pgsql-hackers
On Thu, Feb 27, 2020 at 04:10:14PM +1300, Thomas Munro wrote:
> On Thu, Feb 27, 2020 at 3:29 AM Julien Rouhaud <rjuju123@gmail.com> wrote:
> > [v10]
>
> Thanks.  I'll do some more testing and review soon.  It'd be really
> cool to get this into PG13.

Thanks!

>
> FYI cfbot said:
>
> +++ /home/travis/build/postgresql-cfbot/postgresql/src/test/regress/results/collate.icu.utf8.out
> 2020-02-26 14:45:52.114401999 +0000
> ...
> - icuidx06_d_en_fr_ga | "default" | up to date
> + icuidx06_d_en_fr_ga       | "default"  | out of date

Oh.  It turns out that pg_collation_actual_version() isn't handling the
default collation:

# SELECT pg_collation_actual_version(100), pg_collation_actual_version(c.oid)
  FROM pg_database d
  JOIN pg_collation c on c.collname = d.datcollate
  WHERE datname = current_database();
 pg_collation_actual_version | pg_collation_actual_version
-----------------------------+-----------------------------
 <NULL>                      | 2.30
(1 row)


Fixed in v11 by changing pg_collation_actual_version() to handle default
collation too, as it seems a better behavior.

Attachment

pgsql-hackers by date:

Previous
From: Michael Paquier
Date:
Subject: Re: reindex concurrently and two toast indexes
Next
From: Kyotaro Horiguchi
Date:
Subject: Re: Crash by targetted recovery