Re: [PATCH 5/9] t7450: test verify_path() handling of gitmodules

From: Jeff King <peff@peff.net>
Date: May 3, 2021, 19:39 UTC
Message-ID: <YJBRVQ7+CUkYf5Us@coredump.intra.peff.net>
In-Reply-To: <CAPig+cQFUUncb17MFAvJAh2KFxK_OCq_6AoOwPViYyQJUDVJNg@mail.gmail.com>

On Sat, May 01, 2021 at 03:03:56PM -0400, Eric Sunshine wrote:

Show 13 quoted lines

> On Sat, May 1, 2021 at 2:55 PM Eric Sunshine <sunshine@sunshineco.com> wrote:
> > On Sat, May 1, 2021 at 11:42 AM Jeff King <peff@peff.net> wrote:
> > > +       tree=$(git -C symlink mktree <symlink/bad-tree)
> >
> > `tree` is an unusually generic name for this now-global variable. One
> > can easily imagine it being re-used by some unrelated test arbitrarily
> > inserted into this script, thus potentially breaking the following
> > tests which depend upon it. I wonder if a name such as `BAD_TREE`
> > would be more appropriate.
> 
> I see that all `$tree` references get encapsulated into a shell
> function by the next patch, so perhaps the generic name `tree` isn't a
> big deal after all.

Yeah. I'd like to think it is not that big a deal between even just adjacent tests, because anybody adding tests in the middle of a script would take care not to split related ones. But that may be too optimistic. ;)

At any rate, it seems OK to assume the function will all be used together.

-Peff