Re: git-bugreport-2021-01-06-1209.txt (git can't deal with special characters)

From: "Torsten Bögershausen" <tboegi@web.de>
To: Daniel Troger <random_n0body@icloud.com>
Cc: git@vger.kernel.org
Subject: Re: git-bugreport-2021-01-06-1209.txt (git can't deal with special characters)
Date: Wed, 6 Jan 2021 15:21:37 +0100	[thread overview]
Message-ID: <20210106142137.gxd2a4vnrsjj4jq5@tb-raspi4> (raw)
In-Reply-To: <A102844A-9501-4A86-854D-E3B387D378AA@icloud.com>

On Wed, Jan 06, 2021 at 12:35:26PM +0100, Daniel Troger wrote:
> Thank you for filling out a Git bug report!
> Please answer the following questions to help us understand your issue.
>
> What did you do before the bug happened? (Steps to reproduce your issue)
> I ran `git restore -p .`
>
> What did you expect to happen? (Expected behavior)
> git restore to open in interactive mode and letting me select data to discard
>
> What happened instead? (Actual behavior)
> I got an error message:
> ```
> me@iMac:[redacted]/paulbrunngård-springyard/src$ git restore -p .
> BUG: pathspec.c:495: error initializing pathspec_item
> Cannot close git diff-index --cached --numstat --summary HEAD -- :(,prefix:44)[redacted]/paulbrunngård-springyard/src/ () at /usr/local/Cellar/git/2.29.2/libexec/git-core/git-add--interactive line 183.
> ```
>
> What's different between what you expected and what actually happened?
> The main problem is that the command didn't do anything. It basically printed an error message instead of changing a file in the way I wanted.
>
> Anything else you want to add:
> Yeah. So pretty obviously the problem is the "å" in the filename. The interesting thing is that the folder with the "å" in the filename seems to exist twice, with differently encoded "å"s. But their content IS AUTOMATICALLY PERFECTLY SYNCED. And the only tool which recognizes them as two different folders is git. ls in the terminal shows them as one, finder shows them as one, even "glob" in php shows them as one.
> This is what git seems to see them at (taken from git status):
>
> `modified:   "paulbrunnga\314\212rd-springyard/`
> And further down:
> ```
> Untracked files:
>   (use "git add <file>..." to include in what will be committed)
> 	"paulbrunng\303\245rd-springyard/"
> ```
> Here is a zip containing the folder (emptied): https://arxius.io/f/7ec7e1c7
> I tried creating a new git repo, extracting the zip in it, going into the folder and running `git add empty` and could reproduce that git sees it as two different folders.
>
>
> Please review the rest of the bug report below.
> You can delete any lines you don't wish to share.
>
>
> [System Info]
> git version:
> git version 2.29.2
> cpu: x86_64
> no commit associated with this build
> sizeof-long: 8
> sizeof-size_t: 8
> shell-path: /bin/sh
> uname: Darwin 17.7.0 Darwin Kernel Version 17.7.0: Sun Jun  2 20:31:42 PDT 2019; root:xnu-4570.71.46~1/RELEASE_X86_64 x86_64
> compiler info: clang: 10.0.0 (clang-1000.11.45.5)
> libc info: no libc information available
> $SHELL (typically, interactive shell): /bin/bash
>
>
> [Enabled Hooks]
>

This uses the decomposed form of "å":
> `modified:   "paulbrunnga\314\212rd-springyard/`
...
And here is the precomposed form og the "å"
> Untracked files:
>   (use "git add <file>..." to include in what will be committed)
> 	"paulbrunng\303\245rd-springyard/"
> ```

Both are lowercase, right ?

What does
git config -l | grep unicode
tell you ?

I have
core.precomposeunicode=true

and that what we need on a Mac (to handle precomposed and decomposed better)