From: Daniels Umanovskis <daniels@umanovskis.se>
Date: Thu, 11 Oct 2018 17:29:58 GMT
Subject: Re: [PATCH v2 1/1] branch: introduce --show-current display option
Message-ID: <0f401a50-3c4c-5d09-a8b1-58ae80f5a210@umanovskis.se>
In-Reply-To: <xmqq4ldtgehs.fsf@gitster-ct.c.googlers.com>

On 10/11/18 8:54 AM, Junio C Hamano wrote:
> Is it a normal situation to have refname==NULL, or is it something
> worth reporting as an error?

Looks like that would be in the case of looping symrefs or file backend
failure, so seems a good idea to die() in that case.

> Without passing the &flag argument, I do not think there is a
> reliable way to ask resolve_ref_unsafe() if "HEAD" is a symbolic
> ref.

If I'm reading the code correctly, resolve_ref_unsafe() will return
"HEAD" or NULL if there's no symbolic reference, so anything else would
indicate a symref, but even in that case checking the flag explicitly is
definitely better to clearly show intent.

Will soon reply with v3 cleaning up the suggested patch accordingly.