RE: [Question] worktree vs. gitdir in [includeIf]

From: Randall S. Becker <rsbecker@nexbridge.com>
Date: Jul 9, 2021, 22:37 UTC
Message-ID: <009d01d77513$07b638d0$1722aa70$@nexbridge.com>
In-Reply-To: <xmqqk0m0brq1.fsf@gitster.g>

On July 8, 2021 2:12 PM, Junio C Hamano wrote:

Show 11 quoted lines

>To: Randall S. Becker <rsbecker@nexbridge.com>
>Cc: git@vger.kernel.org
>Subject: Re: [Question] worktree vs. gitdir in [includeIf]
>
>"Randall S. Becker" <rsbecker@nexbridge.com> writes:
>
>> I am wondering whether there is, or are plans for, something analogous
>> to [includeIf "worktree:path"] with a similar semantic to [includeIf
>> "gitdir:path"].
>
>I do not think there currently is an implementation, nor a plan, but I offhand do not see downsides in a feature to let you

conditionally

>enable some configuration based on the location of the worktree.

I'm scratching my head on this a bit. The [includeIf "onbranch:foo"] form is a bit confusing when dealing with a worktree. It looks like the code either does not work if in a worktree or will get the branch of the main repo instead of the worktree ref. It looks like include_by_branch() always returns 0 if in a worktree. Is there something being done in skip_prefix() or resolve_ref_unsafe() that would deal with this?

Thanks, Randall