clean up test_rls_hooks module
Over at [1]/messages/by-id/47F87A0E-C0E5-43A6-89F6-D403F2B45175@enterprisedb.com Mark Dilger got led astray because he used
src/test/modules/test_rls_hooks as a template for his new test module.
It looks like that module was created with some over eager copying of
another test module, but it has a couple of things wrong with it.
. it isn't an extension, so the Makefile shouldn't have an EXTENSION
entry, and there shouldn't be a .control file
. it doesn't need to be preloaded so there is no requirement for a
special config, nor for corresponding REGRESS_OPTS and NO_INSTALLCHECK
lines in the Makefile.
Here's a patch to fix those things.
cheers
andrew
[1]: /messages/by-id/47F87A0E-C0E5-43A6-89F6-D403F2B45175@enterprisedb.com
/messages/by-id/47F87A0E-C0E5-43A6-89F6-D403F2B45175@enterprisedb.com
--
Andrew Dunstan
EDB: https://www.enterprisedb.com
Attachments:
0001-fix-test_rls_hooks.patchtext/x-patch; charset=UTF-8; name=0001-fix-test_rls_hooks.patchDownload+2-16
Andrew Dunstan <andrew@dunslane.net> writes:
Here's a patch to fix those things.
Seems reasonable to me, except that I think I'd leave the formatting
of the OBJS macro alone. IIRC, Andres or someone went around and
made all of those follow this one-file-per-line style some time ago.
regards, tom lane
Hi,
On 2022-03-23 10:07:06 -0400, Tom Lane wrote:
Andrew Dunstan <andrew@dunslane.net> writes:
Here's a patch to fix those things.
Seems reasonable to me, except that I think I'd leave the formatting
of the OBJS macro alone. IIRC, Andres or someone went around and
made all of those follow this one-file-per-line style some time ago.
Yea. merge conflicts are considerably rarer that way, and a heck of a lot
easier to deal with when they occur.
Greetings,
Andres Freund