[PATCH v2] mailmap: consistently normalize brian m. carlson's name

From: Jonathan Nieder <jrnieder@gmail.com>
To: "brian m. carlson" <sandals@crustytoothpaste.net>
Cc: Junio C Hamano <gitster@pobox.com>, git@vger.kernel.org
Subject: [PATCH v2] mailmap: consistently normalize brian m. carlson's name
Date: Mon, 24 Sep 2018 10:39:02 -0700	[thread overview]
Message-ID: <20180924173902.GB138072@aiede.svl.corp.google.com> (raw)
In-Reply-To: <20180917222158.GY432229@genre.crustytoothpaste.net>

v2.18.0-rc0~70^2 (mailmap: update brian m. carlson's email address,
2018-05-08) changed the mailmap to map

  sandals@crustytoothpaste.ath.cx
   -> brian m. carlson <sandals@crustytoothpaste.net>

instead of

  sandals@crustytoothpaste.net
    -> brian m. carlson <sandals@crustytoothpaste.ath.cx>

That means the mapping

  Brian M. Carlson <sandals@crustytoothpaste.ath.cx>
    -> brian m. carlson <sandals@crustytoothpaste.net>

is redundant, so we can remove it.  More importantly, it means that
the identity "Brian M. Carlson <sandals@crustytoothpaste.net>" used in
some commits is not normalized any more.  Add a mapping for it.

Noticed while updating Debian's Git packaging, which uses "git
shortlog --no-merges" to produce a list of changes in each version,
grouped by author's (normalized) name.

Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
---
Hi,

brian m. carlson wrote:

> I think this commit message makes sense.  I apparently still fail to
> understand how the .mailmap format works, so I can't tell you if the
> patch is correct.

Thanks for looking it over.  What would it take to make the patch make
sense, too? ;-)

Most mailmap entries are of the form

	Some Name <someemail@example.com>

which means "Wherever you see the email address someemail@example.com,
canonicalize the author's name to Some Name".  We can use that:

	brian m. carlson <sandals@crustytoothpaste.net>

When we see sandals@crustytoothpaste.ath.cx, we also want to
canonicalize the email address.  For that, we can do

	brian m. carlson <sandals@crustytoothpaste.net> <sandals@crustytoothpaste.ath.cx>

There's only one person who has used these email addresses, so we
don't have to do matching by name.  If we wanted to tighten the name
normalization to match by name, I think we'd do something like

	brian m. carlson <sandals@crustytoothpaste.net> Brian M. Carlson

but I can't get that to seem to have any effect when I test with the
"git check-mailmap" command --- for example, "git check-mailmap 'Dana
How <random.email@example.com>'" does not map and "git check-mailmap
'Random Name <danahow@gmail.com>'" maps to 'Dana L. How
<danahow@gmail.com>'.

The even tighter matching used in v1

	brian m. carlson <sandals@crustytoothpaste.net> Brian M. Carlson <sandals@crustytoothpaste.net>

does work, but it's unnecessary complexity.  We don't need it.

How about this?

Changes since v1:
- loosened the matching to only look at email and ignore name
- no other changes

 .mailmap | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/.mailmap b/.mailmap
index f165222a78..bef3352b0d 100644
--- a/.mailmap
+++ b/.mailmap
@@ -25,7 +25,7 @@ Ben Walton <bdwalton@gmail.com> <bwalton@artsci.utoronto.ca>
 Benoit Sigoure <tsunanet@gmail.com> <tsuna@lrde.epita.fr>
 Bernt Hansen <bernt@norang.ca> <bernt@alumni.uwaterloo.ca>
 Brandon Casey <drafnel@gmail.com> <casey@nrlssc.navy.mil>
-brian m. carlson <sandals@crustytoothpaste.net> Brian M. Carlson <sandals@crustytoothpaste.ath.cx>
+brian m. carlson <sandals@crustytoothpaste.net>
 brian m. carlson <sandals@crustytoothpaste.net> <sandals@crustytoothpaste.ath.cx>
 Bryan Larsen <bryan@larsen.st> <bryan.larsen@gmail.com>
 Bryan Larsen <bryan@larsen.st> <bryanlarsen@yahoo.com>
-- 
2.19.0.444.g18242da7ef

