Re: t0028-working-tree-encoding.sh failing on musl based systems (Alpine Linux)

From: "brian m. carlson" <sandals@crustytoothpaste.net>
To: Kevin Daudt <me@ikke.info>
Cc: Rich Felker <dalias@libc.org>, Kevin Daudt <git@lists.ikke.info>,
	git@vger.kernel.org, larsxschneider@gmail.com
Subject: Re: t0028-working-tree-encoding.sh failing on musl based systems (Alpine Linux)
Date: Fri, 8 Feb 2019 13:51:38 +0000	[thread overview]
Message-ID: <20190208135137.GE11927@genre.crustytoothpaste.net> (raw)
In-Reply-To: <20190208115511.GA30779@alpha>

[-- Attachment #1: Type: text/plain, Size: 2085 bytes --]

On Fri, Feb 08, 2019 at 12:55:11PM +0100, Kevin Daudt wrote:
> On Fri, Feb 08, 2019 at 11:45:02AM +0000, brian m. carlson wrote:
> > On Fri, Feb 08, 2019 at 01:04:03AM -0500, Rich Felker wrote:
> > [..]
> > > In any case, this test seems mainly relevant to Windows users wanting
> > > to store source files in UTF-16LE with BOM. This doesn't really make
> > > sense to do on a Linux/musl system, so I'm not sure any action is
> > > needed here from either side.
> > 
> > I do know that some people use CIFS or the like to share repositories
> > between Unix and Windows. However, I agree that such people aren't
> > likely to use UTF-16 on Unix systems. The working tree encoding
> > functionality also supports other encodings which musl may or may not
> > support.
> > 
> > If you and your users are comfortable with the fact that the test (and
> > the corresponding functionality) won't work as expected with UTF-16,
> > then I agree that no action is needed.
> 
> So would you suggest that we just skip this test on Alpine Linux?

That's not exactly what I said. If Alpine Linux users are never going to
use this functionality and don't care that it's broken, then that's a
fine solution.

As originally mentioned, musl could change its libiconv to write a BOM,
which would make it compatible with other known iconv implementations.

There's also the possibility of defining NO_ICONV. That basically means
that your system won't support encodings, and then this test shouldn't
matter.

Finally, you could try applying a patch to the test to make it write the
BOM for UTF-16 since your iconv doesn't. I expect that the test will
fail again later on once you've done that, though.

I don't have musl so I can't test a patch, but theoretically, you could
use a Makefile variable and lazy test prerequisite to control the
behavior of the code and test, and we could apply a patch. I'll try to
work something up later today which might work and which you could test.
-- 
brian m. carlson: Houston, Texas, US
OpenPGP: https://keybase.io/bk2204

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 868 bytes --]