Re: TAP tests and symlinks on Windows - Mailing list pgsql-hackers

From Michael Paquier
Subject Re: TAP tests and symlinks on Windows
Date
Msg-id 20200617074112.GA118592@paquier.xyz
Whole thread Raw
In response to Re: TAP tests and symlinks on Windows  (Andrew Dunstan <andrew.dunstan@2ndquadrant.com>)
List pgsql-hackers
On Tue, Jun 16, 2020 at 08:32:03AM -0400, Andrew Dunstan wrote:
> On 6/16/20 8:24 AM, Peter Eisentraut wrote:
>> MSYS2, which is basically Cygwin, emulates symlinks with junction
>> points, so this happens to work for our purpose.  We could therefore
>> enable these tests in that environment, if we could come up with a
>> reliable way to detect it.

Hmm.  In this case does perl's -l think that a junction point is
corrently a soft link or not?  We have a check based on that in
pg_basebackup's test and -l fails when it sees to a junction point,
forcing us to skip this test.

> From src/bin/pg_dump/t/010_dump_connstr.pl:
>
>     if ($^O eq 'msys' && `uname -or` =~ /^[2-9].*Msys/)

Smart.  This could become a central variable in TestLib.pm.
--
Michael

Attachment

pgsql-hackers by date:

Previous
From: Michael Paquier
Date:
Subject: Re: pg_regress cleans up tablespace twice.
Next
From: Michael Paquier
Date:
Subject: Re: TAP tests and symlinks on Windows