Item1330: Wikiwords with Umlauts arent detected
Priority: Normal
Current State: Closed
Released In: 2.0.0
Target Release: major
Installed Version and Configuration
Version
Foswiki-1.0.3, Sat, 28 Feb 2009, build 2773, Plugin API version 2.0
Configuration
- Locale: de_CH.utf8
- Charset: utf-8
- OS: Debian Stable (Lenny)
Problem Description
Wikiwords containing Umlauts aren't detected.
Examples:
(I know, these words doesnt make sense...)
Tested in foswiki-1.0.3 and trunk.foswiki.org (see
http://trunk.foswiki.org/Sandbox/TestTopic19 )
--
FlurinRindisbacher - 19 Mar 2009
If you enable locale it may partly work but there is a problem with non A-Z that we need to address.
I did not have time to find my other open bug item but it is there and it dates back to (tm)wiki days.
--
KennethLavrsen - 19 Mar 2009
hi there, adding ÜÖÄ / üöä to
UpperNational /
LowerNational in
LocalSite.cfg makes it work.
i just found that browsing the source in lib/Foswiki.pm
what about adding some characters to $regex depending on which localisation is choosen?
(for example üöä would be sufficient for most german speaking countries)
--
FlurinRindisbacher - 21 Mar 2009
You don't need utf-8 for German, and you don't need to add some trick to make this work.
What you want is this:
$Foswiki::cfg{Site}{CharSet} = 'iso-8859-15';
$Foswiki::cfg{Site}{Locale} = 'de_DE.ISO-8859-15';
$Foswiki::cfg{UseLocale} = 1;
$Foswiki::cfg{Site}{LocaleRegexes} = 1;
$Foswiki::cfg{UpperNational} = '';
$Foswiki::cfg{LowerNational} = '';
- The first line tells into which locale the utf-8 text should be recoded when needed.
- The second line tells perl which locale to use for everything, including regexp classes such as :upper: and :lower:
- The next line tells Foswiki to use these perl locales
- The 4th line tells Foswiki to use these perl locales for regexp, otherwise, they use the default 'A-Z' + UpperNational and 'a-z' + LowerNational
With that setup the
WikiWords with umlauts are rendered correctly, and you won't have your warning logs filing up with warning messages that you shouldn't set utf-8 as a site charset.
--
OlivierRaginel - 29 Apr 2009
Thanks, this solves the problem. (or the upgrade to 1.0.5? )
(For Debian Users) On my UTF-8 system, i had to generate the "de_DE.ISO-8859-15"-locale the following way:
vim /etc/locale.gen and remove the # before de_DE.ISO-8859-15
then run the command locale-gen.
--
FlurinRindisbacher - 30 Apr 2009
Even though Flurin had his problem resolved I believe we still have a todo to as a minimum document how to setup these things.
And on an international site with more than one language it may also be better to do it differently and maybe even manually list the upper/lower case chars.
--
KennethLavrsen - 14 Sep 2009
There might be a problem with the Create New Topic page. The check on wikiword is in javascript, so it will mark a german word as non-wikiword.
--
ArthurClemens - 14 Sep 2009
Thats right, creating a New Topic with äöü etc. is not accepted as
WikiWord.
--
SaschaBuechel - 22 May 2010
Works ok in the unicode core 1.2
--
JozefMojzis - 24 Jun 2015