GNU bug report logs - #18258
t0251-gpt-unicode.sh fails when the C.UTF-8 locale is unavailable

Previous Next

To reply to this bug, email your comments to 18258 AT debbugs.gnu.org.

Toggle the display of automated, internal messages from the tracker.

Message #5 received at submit <at> debbugs.gnu.org (full text, mbox):

From: ludo <at> gnu.org (Ludovic Courtès)
To: bug-parted <at> gnu.org
Subject: t0251-gpt-unicode.sh fails when the C.UTF-8 locale is unavailable
Date: Wed, 13 Aug 2014 13:53:53 +0200

Hello,

t0251-gpt-unicode.sh requires the C.UTF-8 locale and fails if it isn’t
available.

What about:

  1. Exiting with 77 when the locale is unavailable?

  2. Choosing another locale, such as en_US.utf8?  As of GNU libc 2.19,
     C.UTF-8 (or any other spelling) does not exist.

Thanks,
Ludo’.

Message #8 received at 18258 <at> debbugs.gnu.org (full text, mbox):

From: ludo <at> gnu.org (Ludovic Courtès)
To: "Brian C. Lane" <bcl <at> redhat.com>
Cc: 18258 <at> debbugs.gnu.org
Subject: Re: bug#18258: t0251-gpt-unicode.sh fails when the C.UTF-8 locale is
 unavailable
Date: Wed, 13 Aug 2014 21:55:26 +0200

"Brian C. Lane" <bcl <at> redhat.com> skribis:

> See my patch here:
>
> http://lists.alioth.debian.org/pipermail/parted-devel/2014-July/004568.html

-export LC_ALL=C.UTF-8
+if [ -e "/usr/share/locale/en_US/" ]; then
+    export LC_ALL=en_US.UTF-8
+else
+    export LC_ALL=C.UTF-8
+fi

The problem is that this directory is not necessarily where locale data
resides.

One way to check whether the locale encoding is supported would be
with ‘wc -m’:

--8<---------------cut here---------------start------------->8---
$ printf 'foo\341\264\244'|LC_ALL=C wc -m
6

$ printf 'foo\341\264\244'|LC_ALL=en_US.utf8 wc -m
4
--8<---------------cut here---------------end--------------->8---

The test could look for a valid UTF-8 locale like this:

--8<---------------cut here---------------start------------->8---
found_locale=no
for locale in en_US de_DE fr_FR es_ES
do
  LC_ALL="$locale.utf8"

  # In a UTF-8 locale, the string below prints as 4 characters.
  if [ `printf 'foo\341\264\244' | wc -m` -eq 4 ]; then
    found_locale=yes
    break
  fi
done

if [ "$found_locale" != "yes" ]; then
  echo "no valid UTF-8 locale found; skipping" >&2
  exit 77
fi
--8<---------------cut here---------------end--------------->8---

Thoughts?

Thanks,
Ludo’.

Message #11 received at 18258 <at> debbugs.gnu.org (full text, mbox):

From: "Brian C. Lane" <bcl <at> redhat.com>
To: Ludovic Courtès <ludo <at> gnu.org>
Cc: 18258 <at> debbugs.gnu.org
Subject: Re: bug#18258: t0251-gpt-unicode.sh fails when the C.UTF-8 locale is
 unavailable
Date: Wed, 13 Aug 2014 13:59:44 -0700

On Wed, Aug 13, 2014 at 09:55:26PM +0200, Ludovic Courtès wrote:
> The test could look for a valid UTF-8 locale like this:
> 
> --8<---------------cut here---------------start------------->8---
> found_locale=no
> for locale in en_US de_DE fr_FR es_ES
> do
>   LC_ALL="$locale.utf8"
> 
>   # In a UTF-8 locale, the string below prints as 4 characters.
>   if [ `printf 'foo\341\264\244' | wc -m` -eq 4 ]; then
>     found_locale=yes
>     break
>   fi
> done
> 
> if [ "$found_locale" != "yes" ]; then
>   echo "no valid UTF-8 locale found; skipping" >&2
>   exit 77
> fi
> --8<---------------cut here---------------end--------------->8---
> 
> Thoughts?

Nice! I like that better.

-- 
Brian C. Lane | Anaconda Team | IRC: bcl #anaconda | Port Orchard, WA (PST8PDT)

Message #14 received at 18258 <at> debbugs.gnu.org (full text, mbox):

From: Phillip Susi <psusi <at> ubuntu.com>
To: Ludovic Courtès <ludo <at> gnu.org>, 18258 <at> debbugs.gnu.org
Subject: Re: bug#18258: t0251-gpt-unicode.sh fails when the C.UTF-8 locale
 is unavailable
Date: Tue, 19 Aug 2014 10:40:19 -0400

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 8/13/2014 7:53 AM, Ludovic Courtès wrote:
> Hello,
> 
> t0251-gpt-unicode.sh requires the C.UTF-8 locale and fails if it
> isn’t available.
> 
> What about:
> 
> 1. Exiting with 77 when the locale is unavailable?
> 
> 2. Choosing another locale, such as en_US.utf8?  As of GNU libc
> 2.19, C.UTF-8 (or any other spelling) does not exist.

How can the C locale possibly not exist on a system that is not
totally broken?  It is the default and fallback locale when you don't
have any locales installed.



-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.17 (MingW32)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/

iQEcBAEBAgAGBQJT82HTAAoJEI5FoCIzSKrwGhAIAKy0t2kj5KOWM+GO6S5G7RiK
W9nFK2kFR5iD8w31H/5a+qSDWsnFh0pbxEIy47jX/5ujfmeyCRtpI+Ox+ajnu2B/
6oqnkxq/KyOYqIRRiqx3BrF4FPIoc+QiKZpmCxwfUs8QkMzGKcawz1yz3vMk0zcW
gTi3CKo8Rtemdb/KSjaJ172BpuCzrFB0v9iQ41KXy+N1cnZbtmF9xi6otJTtwdQS
bYwMA+k6YyfA041P5k2NKOEEFOL3JVG+fbWvbfFT38q7BD8BHkcOQwpgRyp1oN+3
ijh5c/MEg3eHK63chqYa5QQmuHUDfPE/8vQ5NgkAHmNYX7dh8uZrcV6KfjBp6Ao=
=uDJU
-----END PGP SIGNATURE-----

Message #17 received at 18258 <at> debbugs.gnu.org (full text, mbox):

From: ludo <at> gnu.org (Ludovic Courtès)
To: Phillip Susi <psusi <at> ubuntu.com>
Cc: 18258 <at> debbugs.gnu.org
Subject: Re: bug#18258: t0251-gpt-unicode.sh fails when the C.UTF-8 locale is
 unavailable
Date: Tue, 19 Aug 2014 22:11:21 +0200

Phillip Susi <psusi <at> ubuntu.com> skribis:

> How can the C locale possibly not exist on a system that is not
> totally broken?  It is the default and fallback locale when you don't
> have any locales installed.

C always exists, but C.UTF-8 no.

Ludo’.

Message #20 received at 18258 <at> debbugs.gnu.org (full text, mbox):

From: mhw <at> netris.org
To: Phillip Susi <psusi <at> ubuntu.com>
Cc: Ludovic Courtès <ludo <at> gnu.org>, 18258 <at> debbugs.gnu.org
Subject: Re: bug#18258: t0251-gpt-unicode.sh fails when the C.UTF-8 locale is
 unavailable
Date: Fri, 22 Aug 2014 16:25:49 -0400

Phillip Susi <psusi <at> ubuntu.com> writes:

> On 8/13/2014 7:53 AM, Ludovic Courtès wrote:
>> Hello,
>> 
>> t0251-gpt-unicode.sh requires the C.UTF-8 locale and fails if it
>> isn’t available.
>> 
>> What about:
>> 
>> 1. Exiting with 77 when the locale is unavailable?
>> 
>> 2. Choosing another locale, such as en_US.utf8?  As of GNU libc
>> 2.19, C.UTF-8 (or any other spelling) does not exist.
>
> How can the C locale possibly not exist on a system that is not
> totally broken?  It is the default and fallback locale when you don't
> have any locales installed.

C.UTF-8 is a non-standard Debian extension.  It is not available in
upstream GNU libc, and it is absent from other popular distributions
such as Fedora and Arch.

See:

  https://bugzilla.redhat.com/show_bug.cgi?id=902094
  https://bugs.archlinux.org/task/32296
  https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=522776

      Mark

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.