Re: [PATCH v3 2/4] refs: introduce get_worktree_ref_store()

From: Duy Nguyen <pclouds@gmail.com>
Date: Mar 26, 2017, 08:26 UTC
Message-ID: <CACsJy8BRUXx7WcCTrgDEv++Lj+d=fWbLyysMCW3Rw_Z+QRcS_w@mail.gmail.com>
In-Reply-To: <0d437f43-30d6-3920-5298-0c426b098c1a@alum.mit.edu>

On Mon, Mar 20, 2017 at 9:25 PM, Michael Haggerty <mhagger@alum.mit.edu> wrote:

Show 5 quoted lines

> Instead of moving all of the `for_each_*_submodule()` functions over, I
> encourage you to consider getting rid of them entirely and let the
> end-users call the `refs_for_each_*()` versions of the functions. Again,
> I'm not sure that there won't be friction in doing so, but it seems like
> it's worth a try.

They are getting rid of. If you look at pu (or nd/prune-in-worktree actually) there's only head_ref_submodule() and for_each_remote_ref_submodule() left. head_ref_submodule() has no caller but is still there, I'll need to kill it. Killing the latter can be done separately since the callers in submodule.c need some update.

-- 
Duy