| From: | Michael Paquier <michael(at)paquier(dot)xyz> |
|---|---|
| To: | Dagfinn Ilmari Mannsåker <ilmari(at)ilmari(dot)org> |
| Cc: | Peter Smith <smithpb2250(at)gmail(dot)com>, Andrew Dunstan <andrew(at)dunslane(dot)net>, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>, PostgreSQL Hackers <pgsql-hackers(at)lists(dot)postgresql(dot)org> |
| Subject: | Re: pg_createsubscriber TAP test wrapping makes command options hard to read. |
| Date: | 2025-01-22 05:50:47 |
| Message-ID: | Z5CHN6WajoZP6Rqz@paquier.xyz |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
On Tue, Jan 21, 2025 at 02:17:03PM +0000, Dagfinn Ilmari Mannsåker wrote:
> Thanks again for reviewing this monster patch.
+$node->command_checks_all(
+ [ 'pg_amcheck', '--exclude-user' => 'no_such_user', 'postgres' ],
Extra error spotted here with s/--exclude-user/--username/.
The double --verbose for test 'run pg_createsubscriber on node S' is
intentional, as mentioned by Euler. Fixed this one by making the two
--verbose be side-by-side, and added a comment to document the
intention.
Perhaps pg_resetwal/t/001_basic.pl should also use more long option
names? Note that there are a few more under "Locale provider
tests" in the initdb tests. I've left that as a follow-up thing,
that was already a lot..
Found some more inconsistencies with the ordering of the options, like
a few start commands with pg_ctl in the recovery tests. I've
maintained some consistency here, even if your patch was OK.
--
Michael
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Japin Li | 2025-01-22 06:09:48 | Re: [RFC] Lock-free XLog Reservation from WAL |
| Previous Message | Nisha Moond | 2025-01-22 05:19:28 | Re: Introduce XID age and inactive timeout based replication slot invalidation |