Remove obsolete aliases that broke 'locale -a'
Commit Message
I'm attaching the proposed patch, as it contains a mixture of UTF-8 and Latin-1
(which is part of the bug!) and there may be issues decoding it. (Extract it as
binary and then inspect the bytes carefully using your loupe. :-)
Comments
On 05/14/2015 02:41 AM, Paul Eggert wrote:
> I'm attaching the proposed patch, as it contains a mixture of UTF-8
> and Latin-1 (which is part of the bug!) and there may be issues
> decoding it. (Extract it as binary and then inspect the bytes
> carefully using your loupe. :-)
Looks good to me, modulo messed up data in ChangeLog.
> 0001-Remove-obsolete-aliases-that-broke-locale-a.patch
>
>
> From 30f91a08a481c1f5f263429544abfeee1285904a Mon Sep 17 00:00:00 2001
> From: Paul Eggert <eggert@cs.ucla.edu>
> Date: Wed, 13 May 2015 18:33:45 -0700
> Subject: [PATCH] Remove obsolete aliases that broke 'locale -a'
> MIME-Version: 1.0
> Content-Type: text/plain; charset=UTF-8
> Content-Transfer-Encoding: 8bit
>
> [BZ #18412]
> * intl/locale.alias: Remove obsolete aliases "bokmål" and "français"
> which caused 'locale -a' to output Latin-1 data in UTF-8 locales,
> breaking some applications that use 'locale -a' output.
> ---
> ChangeLog | 7 +++++++
> intl/locale.alias | 9 +++++++--
> 2 files changed, 14 insertions(+), 2 deletions(-)
>
> diff --git a/ChangeLog b/ChangeLog
> index 6d96ce2..5b96578 100644
> --- a/ChangeLog
> +++ b/ChangeLog
> @@ -1,3 +1,10 @@
> +2015-05-13 Paul Eggert <eggert@cs.ucla.edu>
> +
> + [BZ #18412]
> + * intl/locale.alias: Remove obsolete aliases "bokmål" and "français"
This looks like your ChangeLog entry is not in UTF-8?
The ChangeLog is in UTF-8.
> + which caused 'locale -a' to output Latin-1 data in UTF-8 locales,
> + breaking some applications that use 'locale -a' output.
> +
> 2015-05-13 Roland McGrath <roland@hack.frob.com>
>
> * sysdeps/nacl/fdopendir.c: New file.
> diff --git a/intl/locale.alias b/intl/locale.alias
> index ab1cb7a..4053df6 100644
> --- a/intl/locale.alias
> +++ b/intl/locale.alias
> @@ -24,8 +24,14 @@
> # backward compatibility. Nobody should rely on the names defined here.
> # Locales should always be specified by their full name.
>
> +# Note: This file used to contain the lines:
> +# bokmål nb_NO.ISO-8859-1
> +# français fr_FR.ISO-8859-1
> +# but these have been commented out since they cause 'locale -a' to output
> +# text encoded in Latin-1, which breaks applications in UTF-8 locales. See:
> +# https://sourceware.org/bugzilla/show_bug.cgi?id=18412
Could you please convert this file to UTF-8 instead of ISO-8859?
> +
> bokmal nb_NO.ISO-8859-1
> -bokmål nb_NO.ISO-8859-1
> catalan ca_ES.ISO-8859-1
> croatian hr_HR.ISO-8859-2
> czech cs_CZ.ISO-8859-2
> @@ -36,7 +42,6 @@ dutch nl_NL.ISO-8859-1
> eesti et_EE.ISO-8859-1
> estonian et_EE.ISO-8859-1
> finnish fi_FI.ISO-8859-1
> -français fr_FR.ISO-8859-1
> french fr_FR.ISO-8859-1
> galego gl_ES.ISO-8859-1
> galician gl_ES.ISO-8859-1
> -- 2.1.0
Cheers,
Carlos.
On Thu, 14 May 2015, Carlos O'Donell wrote:
> Could you please convert this file to UTF-8 instead of ISO-8859?
Do we want to do this more generally for sources, except for tests of
other character sets? Files *.[chS] with other character sets in, apart
from such tests, appear to include: string/strverscmp.c
sysdeps/i386/fpu/e_atanh.S sysdeps/i386/fpu/e_atanhl.S
sysdeps/i386/fpu/e_log10.S sysdeps/i386/fpu/e_log10f.S
sysdeps/i386/fpu/e_log10l.S sysdeps/i386/fpu/e_log2.S
sysdeps/i386/fpu/e_log2f.S sysdeps/i386/fpu/e_log2l.S
sysdeps/i386/fpu/e_pow.S sysdeps/i386/fpu/e_powf.S
sysdeps/i386/fpu/e_powl.S sysdeps/i386/fpu/s_asinh.S
sysdeps/i386/fpu/s_asinhf.S sysdeps/i386/fpu/s_asinhl.S
sysdeps/i386/fpu/s_log1p.S sysdeps/i386/fpu/s_log1pf.S
sysdeps/i386/fpu/s_log1pl.S sysdeps/ia64/fpu/s_tanf.S
sysdeps/x86_64/fpu/e_log10l.S sysdeps/x86_64/fpu/e_log2l.S
sysdeps/x86_64/fpu/e_powl.S sysdeps/x86_64/fpu/s_log1pl.S
On 14 May 2015 16:53, Joseph Myers wrote:
> On Thu, 14 May 2015, Carlos O'Donell wrote:
> > Could you please convert this file to UTF-8 instead of ISO-8859?
>
> Do we want to do this more generally for sources, except for tests of
> other character sets?
yes, i think we should normalize everything to UTF-8 where appropriate
-mike
From 30f91a08a481c1f5f263429544abfeee1285904a Mon Sep 17 00:00:00 2001
From: Paul Eggert <eggert@cs.ucla.edu>
Date: Wed, 13 May 2015 18:33:45 -0700
Subject: [PATCH] Remove obsolete aliases that broke 'locale -a'
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
[BZ #18412]
* intl/locale.alias: Remove obsolete aliases "bokmål" and "français"
which caused 'locale -a' to output Latin-1 data in UTF-8 locales,
breaking some applications that use 'locale -a' output.
---
ChangeLog | 7 +++++++
intl/locale.alias | 9 +++++++--
2 files changed, 14 insertions(+), 2 deletions(-)
@@ -1,3 +1,10 @@
+2015-05-13 Paul Eggert <eggert@cs.ucla.edu>
+
+ [BZ #18412]
+ * intl/locale.alias: Remove obsolete aliases "bokmål" and "français"
+ which caused 'locale -a' to output Latin-1 data in UTF-8 locales,
+ breaking some applications that use 'locale -a' output.
+
2015-05-13 Roland McGrath <roland@hack.frob.com>
* sysdeps/nacl/fdopendir.c: New file.
@@ -24,8 +24,14 @@
# backward compatibility. Nobody should rely on the names defined here.
# Locales should always be specified by their full name.
+# Note: This file used to contain the lines:
+# bokmål nb_NO.ISO-8859-1
+# français fr_FR.ISO-8859-1
+# but these have been commented out since they cause 'locale -a' to output
+# text encoded in Latin-1, which breaks applications in UTF-8 locales. See:
+# https://sourceware.org/bugzilla/show_bug.cgi?id=18412
+
bokmal nb_NO.ISO-8859-1
-bokmål nb_NO.ISO-8859-1
catalan ca_ES.ISO-8859-1
croatian hr_HR.ISO-8859-2
czech cs_CZ.ISO-8859-2
@@ -36,7 +42,6 @@ dutch nl_NL.ISO-8859-1
eesti et_EE.ISO-8859-1
estonian et_EE.ISO-8859-1
finnish fi_FI.ISO-8859-1
-français fr_FR.ISO-8859-1
french fr_FR.ISO-8859-1
galego gl_ES.ISO-8859-1
galician gl_ES.ISO-8859-1
--
2.1.0