Re: [PATCH] mailmap: add an entry for Phillip Wood

From: Jeff King <peff@peff.net>
Date: Jan 20, 2026, 19:24 UTC
Message-ID: <20260120192424.GA3295894@coredump.intra.peff.net>
In-Reply-To: <xmqqtswgywb4.fsf@gitster.g>

On Tue, Jan 20, 2026 at 10:26:55AM -0800, Junio C Hamano wrote:

Show 32 quoted lines

> "Kristoffer Haugsbakk" <kristofferhaugsbakk@fastmail.com> writes:
> 
> > On Tue, Jan 20, 2026, at 12:01, Phillip Wood wrote:
> >> From: Phillip Wood <phillip.wood@dunelm.org.uk>
> >>
> >> While all my commits appear under the same address, other addresses
> >> appear in some commit trailers. Map those addresses to the canonical
> >> one.
> >>
> >> Signed-off-by: Phillip Wood <phillip.wood@dunelm.org.uk>
> >> ---
> >> I'm not sure if we actually remap the address in trailers but
> >> we've certainly talked about doing it in the past.
> >
> > Yeah it does.
> 
> Hmph, are you sure?  My quick experiment tells me that the trailers
> are not munged.
> 
>     $ git show --pretty=fuller -s | grep '@'
>     Author:     Patrick Steinhardt <ps@pks.im>
>     Commit:     Junio C Hamano <gitster@pobox.com>
>         Signed-off-by: Patrick Steinhardt <ps@pks.im>
>         Signed-off-by: Junio C Hamano <gitster@pobox.com>
>     $ cat >>.mailmap <<\EOF
>     Junio C Hamano <no-such-user@google.com> <gitster@pobox.com>
>     EOF
>     $ git show --pretty=fuller -s | grep '@'
>     Author:     Patrick Steinhardt <ps@pks.im>
>     Commit:     Junio C Hamano <no-such-user@google.com>
>         Signed-off-by: Patrick Steinhardt <ps@pks.im>
>         Signed-off-by: Junio C Hamano <gitster@pobox.com>

We do apply them in some places, like shortlog:

  $ cat >>.mailmap <<\EOF
  Junio C Hamano <no-such-user@google.com> <gitster@pobox.com>
  EOF
  $ git shortlog -1 -ne --group=trailer:signed-off-by 3f051fc9
  Junio C Hamano <no-such-user@google.com> (1):
        doc: patch-id: --verbatim locks in --stable
  
  Kristoffer Haugsbakk <code@khaugsbakk.name> (1):
        doc: patch-id: --verbatim locks in --stable

As you noted, we don't necessarily know that a given trailer represents an ident. But we try to parse them as one (not just for mailmap, but for stuff like "-e"), and if that doesn't work leave it as an opaque string. That all comes from 56d5dde752 (shortlog: parse trailer idents, 2020-09-27).

I _thought_ we added support for %(trailers) to handle mailmap, too, but...

> Or perhaps your "Yeah" is about "we've talkined about doing it"?

...yeah, I think this is the case. There's a proof of concept at:

  https://lore.kernel.org/git/YW8A5FznqLYs7MqH@coredump.intra.peff.net/

but nobody ever took it further (I do still run with it as part of my daily build, which probably enhanced my confusion ;) ).

-Peff