Dear PerlMonks,
I'm writing application using wxWidgets on MS Windows and have problem with localization. My locale initialization code is

 undef $locale;
 $locale = Wx::Locale->new('Czech', 'cz','cz');
 $locale->AddCatalogLookupPathPrefix(filename('data/locale'));
 $locale->AddCatalog('evidence');
[download]

The file data/locale/evidence.po contain UTF-8 encoded translated texts and UTF-8 is correctly specified as charset in Content-Type:

msgid ""
msgstr ""
"Project-Id-Version: evidence VERSION\n"
"Report-Msgid-Bugs-To: \n"
"POT-Creation-Date: 2007-07-26 22:49+0200\n"
"PO-Revision-Date: YEAR-MO-DA HO:MI+ZONE\n"
"Last-Translator: FULL NAME <EMAIL@ADDRESS>\n"
"Language-Team: LANGUAGE <LL@li.org>\n"
"MIME-Version: 1.0\n"
"Content-Type: text/plain; charset=UTF-8\n"
"Content-Transfer-Encoding: 8bit\n"
[download]

but almost all accented letters are malformed. It looks like internal processing somewhere in (or behind) Wx::Locale convert all texts to Windows encoding and set the utf8 flag on...

Can somebody help me? I don't know what is going wrong. Maybe I'm using Wx::Locale in completely wrong way?

As a workaround I've wrote a wrapper routine, but it seems to me that there can be also bug in Encode module, because some chars are not converted properly (e.g. small letter c with caron - which I'm trying to fix with regexp later). Wrapper routine code is

sub _T {
 my $gettext;
 my $octets;
 my $result;

 $gettext = gettext(@_);
 $octets = Encode::encode(Wx::Locale::GetSystemEncodingName(), $gettex
+t);
 $result = Encode::decode_utf8($octets, Encode::FB_PERLQQ);
 # fix czech small letter c<
 $result =~ s/\\xC4\?/\x{10d}/g;

 return $result;
}
[download]