Re: [BUG] Unicode filenames handling in `git log --stat`

From: Alexander Meshcheryakov <alexander.s.m@gmail.com>
Date: Aug 9, 2022, 19:03 UTC
Message-ID: <CA+VDVVVQQ4=um_L_h=EQASPHD_oYjwZxecYewLrCAbQV_m4hwQ@mail.gmail.com>
In-Reply-To: <20220809182045.568598-1-calvinwan@google.com>

Hi Calvin,

Sure, let me demonstrate with clean git repo:

mkdir git_test; cd git_test git init touch Київ.txt Kyiv.txt Маріуполь.txt Mariupol.txt git add -A git commit -m 'foobar'

Now let's check with GNU awk `git log --stat` strings width in bytes: $ git log --stat | LC_ALL=C awk '/txt/{print length($0), $0}' 27 Kyiv.txt | 0 27 Mariupol.txt | 0 27 Київ.txt | 0 27 Маріуполь.txt | 0

And strings width in unicode characters: $ git log --stat | LC_ALL=en_US.UTF-8 awk '/txt/{print length($0), $0}' 27 Kyiv.txt | 0 27 Mariupol.txt | 0 23 Київ.txt | 0 18 Маріуполь.txt | 0

See, all lines are aligned to have length 27 bytes. But on the screen this looks distorted because length in characters differs.

On Tue, 9 Aug 2022 at 22:20, Calvin Wan <calvinwan@google.com> wrote:

Show 8 quoted lines

>
> Hi Alexander,
>
> Thank you for the report! I attempted to reproduce with the steps you
> provided, but was unable to do so. What commands would I have to run
> on a clean git repository to reproduce this?
>
> - Calvin