Re: Inconvenience of pg_read_binary_file()

Tom Lane <tgl@sss.pgh.pa.us>

From: Tom Lane <tgl@sss.pgh.pa.us>
To: Michael Paquier <michael@paquier.xyz>
Cc: Kyotaro Horiguchi <horikyota.ntt@gmail.com>, pgsql-hackers@postgresql.org
Date: 2022-07-30T03:35:36Z
Lists: pgsql-hackers
Michael Paquier <michael@paquier.xyz> writes:
> On Fri, Jul 29, 2022 at 03:44:25PM -0400, Tom Lane wrote:
>> Pushed after some fooling with the docs and test cases.  (Notably,
>> I do not think we can assume that pg_hba.conf exists in $PGDATA; some
>> installations keep it elsewhere.  I used postgresql.auto.conf instead.)

> Are you sure that this last part is a good idea?  We don't force the
> creation of postgresql.auto.conf when starting a server, so this
> impacts the portability of the tests with installcheck if one decides
> to remove it from the data folder, and it sounds plausible to me that
> some distributions do exactly that..

Hm.  I considered reading PG_VERSION instead, or postmaster.pid.
PG_VERSION would be a very boring test case, but it should certainly
be present in $PGDATA.

			regards, tom lane



Commits

  1. Adjust new pg_read_file() test cases for more portability.

  2. Support pg_read_[binary_]file (filename, missing_ok).