From: Eric Sunshine <sunshine@sunshineco.com>
Date: Sun, 07 Mar 2021 20:52:24 GMT
Subject: Re: [PATCH] mktag tests: fix broken "&&" chain
Message-ID: <CAPig+cRnAFaHWGFqNes9SVXDeW=zgV4ONvY9=RTKcyJ+CnZO8Q@mail.gmail.com>
In-Reply-To: <xmqqy2ey7kzx.fsf@gitster.c.googlers.com>

On Sun, Mar 7, 2021 at 3:43 PM Junio C Hamano <gitster@pobox.com> wrote:
> Ævar Arnfjörð Bjarmason  <avarab@gmail.com> writes:
> > -                     test_must_fail git mktag --no-strict <tag.sig 2>message.no-strict &&xb
> > +                     test_must_fail git mktag --no-strict <tag.sig 2>message.no-strict &&
> >                       grep '$2' message.no-strict
>
> I do not think the chainlint check is designed to deal with helper
> functions, but I wonder why nobody noticed a runtime failure.  Is
> this an unused/dead codepath?

The `git mktag` invocation would still have produced the expected
stderr output regardless of what happens later (by "later", I mean the
failing invocation of the non-existent `xb` command), which means that
the subsequent `grep` will find what it's looking for, making the
function succeed as expected.

Someone looking at the verbose output might have had a chance of
spotting the "xp: not found" error message, but there likely wasn't
any reason to scan the verbose output since the test succeeded.