Re: adduser 3.112+nmu2: Please update the PO translation for the package adduser

[Sergey Alyoshin <alyoshin.s@gmail.com>]

On 11/6/10, David Prévot <david@tilapin.org> wrote:
> Le 05/11/2010 15:09, Sergey Alyoshin a écrit :
>> Данная информация корректна? [Д/н] Д
>> /* Дд for Yy (yes), not work: FAIL */
>
> Ouch, that's annoying... Could you confirm that it actually works with
> “y” and “Y” (it seems so after a few tests, but I would prefer a second
> opinion).

Y and y work just great.

>> Данная информация корректна? [Д/н] Н
>> /* Нн for Nn (no), try again: PASS */
>
> “H” is in ASCII, does it also works with “н”?

Н is not ASCII char, it is Cyrillic capital "en" (UTF-8 0xd09d).

For this test code (my second Perl code after "hello world"):

my $yesexpr = langinfo(YESEXPR());
my $noexpr = langinfo(NOEXPR());

foreach my $c ('y', 'Y', 'n', 'N', 'д', 'Д', 'н', 'Н', 'ж') {
        if ($c =~ m/$yesexpr/o) {print "$c match $yesexpr\n";}
        if ($c !~ m/$yesexpr/o) {print "$c not match $yesexpr\n";}
        if ($c =~ m/$noexpr/o) {print "$c match $noexpr\n";}
        if ($c !~ m/$noexpr/o) {print "$c not match $noexpr\n";}
}

where 'д', 'Д', 'н', 'Н', 'ж' is UTF-8 Cyrillic chars ('ж' is random
Cyrillic char).

The output is:

y match ^[ДдYy].*
y not match ^[НнNn].*
Y match ^[ДдYy].*
Y not match ^[НнNn].*
n not match ^[ДдYy].*
n match ^[НнNn].*
N not match ^[ДдYy].*
N match ^[НнNn].*
д match ^[ДдYy].*
д match ^[НнNn].*   # FAIL
Д match ^[ДдYy].*
Д match ^[НнNn].*   # FAIL
н match ^[ДдYy].*    # FAIL
н match ^[НнNn].*
Н match ^[ДдYy].*   # FAIL
Н match ^[НнNn].*
ж match ^[ДдYy].*   # FAIL, any Cyrillic char will match?
ж match ^[НнNn].*   # FAIL

>> Following bug reports seems relevant:
>> http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=442619
>
> Do you have perl-modules installed?

Yes.

>> http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=486877
>
> Do you have I18N::Langinfo installed?

I think it is installed as
/usr/lib/perl/5.10.1/I18N/Langinfo.pm

>> It seems to me like Perl regexp problem.
>
> Maybe related to UTF-8 characters. Please fill another bug for this one,
> I'm not sure to be able fix it (but since it's a localization issue, it
> is possible to fix it, if someone provide a patch).

Should I, the problem seems with Perl?