Re: [GSoC][RFC PATCH v4 3/5] builtin/refs: add list subcommand

From: Phillip Wood <phillip.wood123@gmail.com>
Date: Aug 4, 2025, 09:28 UTC
Message-ID: <9544a669-3e3a-492f-a312-748b5d49bec3@gmail.com>
In-Reply-To: <xmqqwm7nklqf.fsf@gitster.g>

On 01/08/2025 18:14, Junio C Hamano wrote:

Show 9 quoted lines

> Phillip Wood <phillip.wood123@gmail.com> writes:
> 
>> It is confusing that refs/heads does a prefix match but refs/heads/m
>> does not (unless there is a hierarchy boundary after the m).
> 
> And refs/hea would not show branches, for the same reason.  It is
> what "anchored at hierarchy boundary" in the message you are
> responding to means.  In other words, it is not a simple textual
> prefix match.

The last sentence sums up why I find it confusing - it is not possible to tell from the pattern what sort of match will be returned. If a script wants to query some property of a ref it either has to know it exists (which is racy but probably not a practical problem most of the time) or it has to include %(refname) in the format argument and check that it matches the ref in the query string. A script can force a prefix match by adding a trailing slash but there is no way to force an exact match.

Thanks

Phillip