Rails - Jul 2005 - 10 Easy Steps To Make Your Rails Apps Multilingual (I18N)

Some time ago I wrote the manual about using Gettext to translate your 
Rails app. It is still available under 
http://manuals.rubyonrails.com/read/book/16 and holds some good 
introductions and broader explanations about the subject.

Recently I wanted to have my own weblog. I toyed around with a couple of 
PHP solutions, but I hated the prospect of having to maintain yet 
another pile of badly written PHP scripts. I already have enough of 
those :). So I turned to Typo[1]. But there was a problem. Typo doesn''t
support *any* kind of translation as of 2005-07-31. So I talked to 
Tobias Luettke who is the maintainer of this wonderful blogging software 
(and all around nice guy) and asked if he was interested in getting a 
patch to gettextify typo. He said he would rather not integrate with 
Ruby-GetText[2] because it has binary dependencies with Iconv (which is 
optional) and Gettext. So I looked into integrating Ri18n[3] which is 
still very young but also excellent. The author has really thought more 
about nicer integration into rails apps which is a declared goal.

So, instead of just making a patch for Typo I ended up writing a little 
guide on how to apply this strategy to every rails app. Now there is no 
excuse anymore for not making your rails apps translatable. Shame on you 
if you don''t ;). I know that a lot of people would like to see some 
support for translation (and broader localization) in Rails itself. So I 
thought in the meantime I would just post this to the Mailinglist and 
ask for feedback. If there are no problems I will also post it on the 
Wiki or the manuals section.

As always before you start make sure that you use UTF-8 everywhere (also 
in the Database, MySQL 4.1.x supports this).

Here we go.

10 Easy Steps To Make Your Rails Apps Multilingual (I18N)

1) Read the USAGE of ri18n on 
http://ri18n.berlios.de/rdoc/files/USAGE.html. Repeat two times. Keep 
referring to it if you have questions along the way.

2) Check out the fresh svn trunk into the vendor dir:

   cd $RAILS_ROOT/vendor
   svn checkout svn://svn.berlios.de/ri18n/trunk ri18n

3) Now add this path to the ADDITIONAL_LOAD_PATHS array in 
$RAILS_ROOT/config/environment.rb so that it reads:

   ADDITIONAL_LOAD_PATHS.concat %w(
     # ... other entries
     vendor
     vendor/ri18n/lib
     # ... other entries
   ).map { |dir| "#{RAILS_ROOT}/#{dir}" }.select { |dir| 
File.directory?(dir) }

4) Now require necessary ri18n library files in 
$RAILS_ROOT/config/environment.rb:

   # Load included libraries.
   # ... other libraries
   require ''gettext''
   require ''i18nservice''

5) Create a locale dir in the base dir of your application:

   mkdir $RAILS_ROOT/locale

6) Ri18n has to be configured. At the end of your 
$RAILS_ROOT/config/environment.rb file add:

   $KCODE = ''u'' # lets go unicode, you will have less trouble
...
   require ''jcode'' # and use the right string methods for
utf-8
   # The directory where translation files (PO files) will go
   I18nService.instance.po_dir = "#{RAILS_ROOT}/locale"
   # let''s use German everywhere
   I18nService.instance.lang = ''de''

Notice that the last line just sets your language to German on a global 
basis. If you want to decide which language to use in a more dynamic 
manner go and take a look at the old manual: 
http://manuals.rubyonrails.com/read/book/16. It should be easy to adapt 
the strategy I present there to your own needs.

7) You will want to have a way to harvest your files. Ri18n has a 
built-in harvester that is controlled by a rake task. So go ahead and 
add a rake task to the end of $RAILS_ROOT/Rakefile:

   task :gettext do
     $:.unshift ''vendor/ri18n/lib''
     require ''gettext''
     require ''i18nservice''
     I18nService.instance.po_dir = File.dirname(__FILE__) +
''/locale''
     desc ''Run the ri18n harvester on this application''
     Rake::GettextTask.new do |t|
       # add all desired langues here
       t.new_langs = [''de'', ''fr'']
       # add all desired paths here
       t.source_files =
[''{lib,app,components}/**/*.r{b,html,xml}'']
       t.verbose = true
     end
   end

8) Now you can call `rake gettext` in your $RAILS_ROOT and the harvester 
will collect all the strings in your files which looks something like this:

   dw@colinux: /home/dw/rails/typo:
   $ rake gettext
   (in /home/dw/rails/typo)
   I18n messages in app/views/articles/index.rhtml:
    * Good Morning
    * Hello
    * Good Bye

9) Well, now, you will actually have to go through your source code and 
translate it. If you forget how remember step 1. Every time you add or 
change a translation, just repeat step 8.

10) Most important: Have fun and be proud that from now on you can say 
that, yes, your apps will also (at least potentially) be understood by 
the majority of the internet population[4].


[1] http://typo.leetsoft.com/trac/
[2] http://ponx.s5.xrea.com/hiki/ruby-gettext.html
[3] http://ri18n.berlios.de/
[4] Which was about 64,2 % of the internet population in 2004
http://global-reach.biz/globstats/index.php3

Now, please tell me if you find this information useful and more 
important if you find anything to improve on.

Sascha Ebach

This is awesome. Thank you for blazing a trail.

Sascha Ebach wrote:
> So, instead of just making a patch for Typo I ended up writing a little 
> guide on how to apply this strategy to every rails app. Now there is no 
> excuse anymore for not making your rails apps translatable. Shame on you 
> if you don''t ;). I know that a lot of people would like to see
some
> support for translation (and broader localization) in Rails itself. So I 
> thought in the meantime I would just post this to the Mailinglist and 
> ask for feedback. If there are no problems I will also post it on the 
> Wiki or the manuals section.
>

Thanks Sascha, this email has just been starred.  I''m currently
working on a app that we wanted to make multi-lingual.  After i get
some basics, currently working on getting file upload integrated, I''ll
take a closer look at this.

thanks again,
andy

-- 
Andrew Stone

On Jul 30, 2005, at 5:53 PM, Sascha Ebach wrote:
> 9) Well, now, you will actually have to go through your source code  
> and translate it. If you forget how remember step 1. Every time you  
> add or change a translation, just repeat step 8.
I completely understand the importance of i18n, and support the  
effort. But how do most people on a budget translate their apps, and  
maintain the translations? It''s not like we can all afford to keep a  
polyglot on staff... and if you happen to find a willing volunteer,  
who''s to say that person will be available and willing to help later,  
when you need to change or add a translation?

Anyway, just curious how these issues are handled by those who deal  
with i18n, esp. in an open source environment.

- Jamis

Fantastic Sascha,

Lets get this integrated in typo proper then..

On 7/30/05, Sascha Ebach
<se-eFwX6J65rk9VioaHkBSlcw02NpfuEekPhC4ANOJQIlc@public.gmane.org> wrote:
> Recently I wanted to have my own weblog. I toyed around with a couple of
> PHP solutions, but I hated the prospect of having to maintain yet
> another pile of badly written PHP scripts. I already have enough of
> those :). So I turned to Typo[1]. But there was a problem. Typo
doesn''t
> support *any* kind of translation as of 2005-07-31. So I talked to
> Tobias Luettke who is the maintainer of this wonderful blogging software
> (and all around nice guy) and asked if he was interested in getting a
> patch to gettextify typo. He said he would rather not integrate with
> Ruby-GetText[2] because it has binary dependencies with Iconv (which is
> optional) and Gettext. So I looked into integrating Ri18n[3] which is
> still very young but also excellent. The author has really thought more
> about nicer integration into rails apps which is a declared goal.
> 
> So, instead of just making a patch for Typo I ended up writing a little
> guide on how to apply this strategy to every rails app. Now there is no
> excuse anymore for not making your rails apps translatable. Shame on you
> if you don''t ;). I know that a lot of people would like to see
some
> support for translation (and broader localization) in Rails itself. So I
> thought in the meantime I would just post this to the Mailinglist and
> ask for feedback. If there are no problems I will also post it on the
> Wiki or the manuals section.
-- 
Tobi
http://www.snowdevil.ca - Snowboards that don''t suck
http://typo.leetsoft.com - Open source weblog engine
http://blog.leetsoft.com - Technical weblog

Any idea how well ri18n handles charset conversions? The doc says it''s
a bit flakey, have you tried it out? I have an app that doesn''t need
translation, but I do need arbitrary charset conversion and I''d rather
avoid the dependency on Iconv as well if I could.

Cheers,
Colin

On 31/07/05, Tobias Luetke
<tobias.luetke-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
wrote:
> Fantastic Sascha,
> 
> Lets get this integrated in typo proper then..
> 
> On 7/30/05, Sascha Ebach
<se-eFwX6J65rk9VioaHkBSlcw02NpfuEekPhC4ANOJQIlc@public.gmane.org> wrote:
> 
> > Recently I wanted to have my own weblog. I toyed around with a couple
of
> > PHP solutions, but I hated the prospect of having to maintain yet
> > another pile of badly written PHP scripts. I already have enough of
> > those :). So I turned to Typo[1]. But there was a problem. Typo
doesn''t
> > support *any* kind of translation as of 2005-07-31. So I talked to
> > Tobias Luettke who is the maintainer of this wonderful blogging
software
> > (and all around nice guy) and asked if he was interested in getting a
> > patch to gettextify typo. He said he would rather not integrate with
> > Ruby-GetText[2] because it has binary dependencies with Iconv (which
is
> > optional) and Gettext. So I looked into integrating Ri18n[3] which is
> > still very young but also excellent. The author has really thought
more
> > about nicer integration into rails apps which is a declared goal.
> >
> > So, instead of just making a patch for Typo I ended up writing a
little
> > guide on how to apply this strategy to every rails app. Now there is
no
> > excuse anymore for not making your rails apps translatable. Shame on
you
> > if you don''t ;). I know that a lot of people would like to
see some
> > support for translation (and broader localization) in Rails itself. So
I
> > thought in the meantime I would just post this to the Mailinglist and
> > ask for feedback. If there are no problems I will also post it on the
> > Wiki or the manuals section.
> 
> --
> Tobi
> http://www.snowdevil.ca - Snowboards that don''t suck
> http://typo.leetsoft.com - Open source weblog engine
> http://blog.leetsoft.com - Technical weblog
> _______________________________________________
> Rails mailing list
> Rails-1W37MKcQCpIf0INCOvqR/iCwEArCW2h5@public.gmane.org
> http://lists.rubyonrails.org/mailman/listinfo/rails
>

Jamis Buck wrote:
> I completely understand the importance of i18n, and support the  effort. 
> But how do most people on a budget translate their apps, and  maintain 
> the translations? It''s not like we can all afford to keep a 
polyglot on
> staff... and if you happen to find a willing volunteer,  who''s to
say
> that person will be available and willing to help later,  when you need 
> to change or add a translation?
> 
> Anyway, just curious how these issues are handled by those who deal  
> with i18n, esp. in an open source environment.
When you say "on a budget" and "open source environment" you
probably
don''t mean the same kind of project. The translations of any OS project
will only be as good as its contributors. What is more important for me, 
personally, is that I have a way to be able to translate a web app at 
all. Clearly the maintainer(s) of the project will have to have thought 
about that upfront. That is often the not so easy part. For example if 
you have to account for your language going from right-to-left (arabic) 
or from top-to-bottom (chinese).

In the case of Typo, which I currently need for a German weblog I plan 
to start in 2 or 3 weeks, I don''t even have the chance of doing my 
translations. Unless I fork Typo which is clearly not what I want to do. 
I think this goes for a lot of other packages, too. Take Drupal[1] for 
example. The German translation usually lacks behind months until 
someone with a need to do it comes along and just translates it. And if 
you plan to use it commercially than you might not even be satisified 
with the translation quality of a package. Good to know that you can do 
it yourself or hire someone to do it for you ... My experience has been 
that the translations are second class. But if a package puts you into 
the position to change that yourself, I don''t find that so bad.

Everything would be so much easier if we only had one language, 
currency, time zone, date format, char set, ... It is just a fact of 
(developer) life that we don''t. So it would be awesome if this would be
something that would be planned for upfront by more (OS) developers. 
This is a particularly weak point of Rails (and more so Ruby). It adds 
more complexity to software development, yes, but IMO this is just part 
of the job description of a good developer. It gets better. Slowly, but 
surely.

Sascha

[1] http://www.drupal.org

Colin Fleming wrote:
> Any idea how well ri18n handles charset conversions? The doc says
it''s
> a bit flakey, have you tried it out? I have an app that doesn''t
need
> translation, but I do need arbitrary charset conversion and I''d
rather
> avoid the dependency on Iconv as well if I could.
Hm, you are going to have a hard time with charset conversion without 
Iconv. Charset conversion is really hard. And Iconv is a package which 
is easily available on many platforms. There is no easier conversion 
than Iconv. Please correct me if I am wrong.

If you want to do conversions with r18n you also have to have Iconv.

Sascha

Hi Sascha,

Great guide and I think this is just what I am looking for. I have a few
clients who depend heavily on the Asian market, but are based in
Australia, so if I can get this working well enough I should be able to
have their site translated into atleast one other language soon enough.

Dylan.
> Some time ago I wrote the manual about using Gettext to translate your
> Rails app. It is still available under
> http://manuals.rubyonrails.com/read/book/16 and holds some good
> introductions and broader explanations about the subject.
>
> So, instead of just making a patch for Typo I ended up writing a
> little guide on how to apply this strategy to every rails app. Now
> there is no excuse anymore for not making your rails apps
> translatable. Shame on you if you don''t ;). I know that a lot of
> people would like to see some support for translation (and broader
> localization) in Rails itself. So I thought in the meantime I would
> just post this to the Mailinglist and ask for feedback. If there are
> no problems I will also post it on the Wiki or the manuals section.

Tobias Luetke wrote:
> Fantastic Sascha,
> 
> Lets get this integrated in typo proper then..
Great. Tried to ping you on IRC, no luck. If you don''t catch me there I
still have a couple of questions about the patch. I will just mail you 
off list.

Sascha

agreed.  Sascha''s solution is great, but "use UTF-8
everywhere"
is never "easy step" for me.  I (and many Japanese web developers)
usually use EUC-JP(in scripts) and Shift_JIS(in HTML templates)
and ISO-2022-JP (in E-mails) in one application.

In addition, there is one more problem: ActiveRecord. Error
messages of it are not supported in Japanese. So I hacked
ActiveRecord::Base::human_attribute_name as bellow

      def human_attribute_name(attribute_key_name) #:nodoc:
        table = self.human_attribute_name_table
        if table && table[attribute_key_name]
          return table[attribute_key_name]
        end
        attribute_key_name.humanize
      end

and add ''cattr_accessor :human_attribute_name_table'' . 

But I know it''s not smart solution. We need total and
consistent solution for i18n/l10n in rails.

Regards,

Masayoshi Takahashi


2005/7/31, Colin Fleming
<colin.mailinglist-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>:
> Any idea how well ri18n handles charset conversions? The doc says
it''s
> a bit flakey, have you tried it out? I have an app that doesn''t
need
> translation, but I do need arbitrary charset conversion and I''d
rather
> avoid the dependency on Iconv as well if I could.
> 
> Cheers,
> Colin
> 
> On 31/07/05, Tobias Luetke
<tobias.luetke-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> wrote:
> > Fantastic Sascha,
> >
> > Lets get this integrated in typo proper then..
> >
> > On 7/30/05, Sascha Ebach
<se-eFwX6J65rk9VioaHkBSlcw02NpfuEekPhC4ANOJQIlc@public.gmane.org> wrote:
> >
> > > Recently I wanted to have my own weblog. I toyed around with a
couple of
> > > PHP solutions, but I hated the prospect of having to maintain yet
> > > another pile of badly written PHP scripts. I already have enough
of
> > > those :). So I turned to Typo[1]. But there was a problem. Typo
doesn''t
> > > support *any* kind of translation as of 2005-07-31. So I talked
to
> > > Tobias Luettke who is the maintainer of this wonderful blogging
software
> > > (and all around nice guy) and asked if he was interested in
getting a
> > > patch to gettextify typo. He said he would rather not integrate
with
> > > Ruby-GetText[2] because it has binary dependencies with Iconv
(which is
> > > optional) and Gettext. So I looked into integrating Ri18n[3]
which is
> > > still very young but also excellent. The author has really
thought more
> > > about nicer integration into rails apps which is a declared goal.
> > >
> > > So, instead of just making a patch for Typo I ended up writing a
little
> > > guide on how to apply this strategy to every rails app. Now there
is no
> > > excuse anymore for not making your rails apps translatable. Shame
on you
> > > if you don''t ;). I know that a lot of people would like
to see some
> > > support for translation (and broader localization) in Rails
itself. So I
> > > thought in the meantime I would just post this to the Mailinglist
and
> > > ask for feedback. If there are no problems I will also post it on
the
> > > Wiki or the manuals section.
> >
> > --
> > Tobi
> > http://www.snowdevil.ca - Snowboards that don''t suck
> > http://typo.leetsoft.com - Open source weblog engine
> > http://blog.leetsoft.com - Technical weblog
> > _______________________________________________
> > Rails mailing list
> > Rails-1W37MKcQCpIf0INCOvqR/iCwEArCW2h5@public.gmane.org
> > http://lists.rubyonrails.org/mailman/listinfo/rails
> >
> _______________________________________________
> Rails mailing list
> Rails-1W37MKcQCpIf0INCOvqR/iCwEArCW2h5@public.gmane.org
> http://lists.rubyonrails.org/mailman/listinfo/rails
>

I''ve never been able to figure this out.. Maybe you can explain?
What do Shift-JIS, EUC-JP, and the others offer that UTF-8 doesn''t?
Surely it''s not just for text art on 2ch. =)

masayoshi takahashi wrote:
> agreed.  Sascha''s solution is great, but "use UTF-8
everywhere"
> is never "easy step" for me.  I (and many Japanese web
developers)
> usually use EUC-JP(in scripts) and Shift_JIS(in HTML templates)
> and ISO-2022-JP (in E-mails) in one application.
>

What could be really cool is a gem that interfaces with one of the
online translation sites and generates a translation of each page in
real time, depending on the language chosen. I know these automated
translations aren''t all that good (and sometimes incredibly poor) but
it''s usually enough to get the overall ''idea'' of a
page''s content.

Does anyone know of such a translation site that provides a free SOAP
API? Does Google?

Ben

Ok, I''d figured that, I was just wondering if there was a pure-ruby
solution (although Iconv does integrate well with Ruby).

Thanks,
Colin

On 31/07/05, Sascha Ebach
<se-eFwX6J65rk9VioaHkBSlcw02NpfuEekPhC4ANOJQIlc@public.gmane.org>
wrote:
> Hm, you are going to have a hard time with charset conversion without
> Iconv. Charset conversion is really hard. And Iconv is a package which
> is easily available on many platforms. There is no easier conversion
> than Iconv. Please correct me if I am wrong.
> 
> If you want to do conversions with r18n you also have to have Iconv.
> 
> Sascha

Colin Fleming wrote:
> Ok, I''d figured that, I was just wondering if there was a
pure-ruby
> solution (although Iconv does integrate well with Ruby).
You could probably do one in Ruby. But something like charset conversion 
is an ideal thing to be implemented in a compiled language (c fans 
correct me if I am wrong). Essentially lots of big lookup tables. If 
implemnted in Ruby or any other dynamic language you would have a major 
performance bottleneck. Depending on what you do with it, this could be 
a problem.

Sascha

Colin,
If you are looking for a  pure ruby solution check out the  module at
http://www.logicden.com/site/downloads

----- Original Message -----
From: "Colin Fleming"
<colin.mailinglist-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
To: rails-1W37MKcQCpIf0INCOvqR/iCwEArCW2h5@public.gmane.org
Subject: Re: [Rails] 10 Easy Steps To Make Your Rails Apps Multilingual (I18N)
Date: Sun, 31 Jul 2005 19:37:23 +0200
> 
> Ok, I''d figured that, I was just wondering if there was a
pure-ruby
> solution (although Iconv does integrate well with Ruby).
> 
> Thanks,
> Colin
> 
> On 31/07/05, Sascha Ebach
<se-eFwX6J65rk9VioaHkBSlcw02NpfuEekPhC4ANOJQIlc@public.gmane.org> wrote:
> > Hm, you are going to have a hard time with charset conversion without
> > Iconv. Charset conversion is really hard. And Iconv is a package which
> > is easily available on many platforms. There is no easier conversion
> > than Iconv. Please correct me if I am wrong.
> >
> > If you want to do conversions with r18n you also have to have Iconv.
> >
> > Sascha
> _______________________________________________
> Rails mailing list
> Rails-1W37MKcQCpIf0INCOvqR/iCwEArCW2h5@public.gmane.org
> http://lists.rubyonrails.org/mailman/listinfo/rails


Antony Joseph
http://www.logicden.com
https://workeffort.dev.java.net
http://www.detroitjug.org


-- 
_______________________________________________
NEW! Lycos Dating Search. The only place to search multiple dating sites at
once.
http://datingsearch.lycos.com

+100

On Jul 30, 2005, at 8:53 PM, Sascha Ebach wrote:
> Some time ago I wrote the manual about using Gettext to translate your 
> Rails app. It is still available under 
> http://manuals.rubyonrails.com/read/book/16 and holds some good 
> introductions and broader explanations about the subject.
>
> Recently I wanted to have my own weblog. I toyed around with a couple 
> of PHP solutions, but I hated the prospect of having to maintain yet 
> another pile of badly written PHP scripts. I already have enough of 
> those :). So I turned to Typo[1]. But there was a problem. Typo 
> doesn''t support *any* kind of translation as of 2005-07-31. So I 
> talked to Tobias Luettke who is the maintainer of this wonderful 
> blogging software (and all around nice guy) and asked if he was 
> interested in getting a patch to gettextify typo. He said he would 
> rather not integrate with Ruby-GetText[2] because it has binary 
> dependencies with Iconv (which is optional) and Gettext. So I looked 
> into integrating Ri18n[3] which is still very young but also 
> excellent. The author has really thought more about nicer integration 
> into rails apps which is a declared goal.
>
> So, instead of just making a patch for Typo I ended up writing a 
> little guide on how to apply this strategy to every rails app. Now 
> there is no excuse anymore for not making your rails apps 
> translatable. Shame on you if you don''t ;). I know that a lot of 
> people would like to see some support for translation (and broader 
> localization) in Rails itself. So I thought in the meantime I would 
> just post this to the Mailinglist and ask for feedback. If there are 
> no problems I will also post it on the Wiki or the manuals section.
>
> As always before you start make sure that you use UTF-8 everywhere 
> (also in the Database, MySQL 4.1.x supports this).
>
> Here we go.
>
> 10 Easy Steps To Make Your Rails Apps Multilingual (I18N)
>
> 1) Read the USAGE of ri18n on 
> http://ri18n.berlios.de/rdoc/files/USAGE.html. Repeat two times. Keep 
> referring to it if you have questions along the way.
>
> 2) Check out the fresh svn trunk into the vendor dir:
>
>   cd $RAILS_ROOT/vendor
>   svn checkout svn://svn.berlios.de/ri18n/trunk ri18n
>
> 3) Now add this path to the ADDITIONAL_LOAD_PATHS array in 
> $RAILS_ROOT/config/environment.rb so that it reads:
>
>   ADDITIONAL_LOAD_PATHS.concat %w(
>     # ... other entries
>     vendor
>     vendor/ri18n/lib
>     # ... other entries
>   ).map { |dir| "#{RAILS_ROOT}/#{dir}" }.select { |dir| 
> File.directory?(dir) }
>
> 4) Now require necessary ri18n library files in 
> $RAILS_ROOT/config/environment.rb:
>
>   # Load included libraries.
>   # ... other libraries
>   require ''gettext''
>   require ''i18nservice''
>
> 5) Create a locale dir in the base dir of your application:
>
>   mkdir $RAILS_ROOT/locale
>
> 6) Ri18n has to be configured. At the end of your 
> $RAILS_ROOT/config/environment.rb file add:
>
>   $KCODE = ''u'' # lets go unicode, you will have less
trouble ...
>   require ''jcode'' # and use the right string methods for
utf-8
>   # The directory where translation files (PO files) will go
>   I18nService.instance.po_dir = "#{RAILS_ROOT}/locale"
>   # let''s use German everywhere
>   I18nService.instance.lang = ''de''
>
> Notice that the last line just sets your language to German on a 
> global basis. If you want to decide which language to use in a more 
> dynamic manner go and take a look at the old manual: 
> http://manuals.rubyonrails.com/read/book/16. It should be easy to 
> adapt the strategy I present there to your own needs.
>
> 7) You will want to have a way to harvest your files. Ri18n has a 
> built-in harvester that is controlled by a rake task. So go ahead and 
> add a rake task to the end of $RAILS_ROOT/Rakefile:
>
>   task :gettext do
>     $:.unshift ''vendor/ri18n/lib''
>     require ''gettext''
>     require ''i18nservice''
>     I18nService.instance.po_dir = File.dirname(__FILE__) +
''/locale''
>     desc ''Run the ri18n harvester on this application''
>     Rake::GettextTask.new do |t|
>       # add all desired langues here
>       t.new_langs = [''de'', ''fr'']
>       # add all desired paths here
>       t.source_files =
[''{lib,app,components}/**/*.r{b,html,xml}'']
>       t.verbose = true
>     end
>   end
>
> 8) Now you can call `rake gettext` in your $RAILS_ROOT and the 
> harvester will collect all the strings in your files which looks 
> something like this:
>
>   dw@colinux: /home/dw/rails/typo:
>   $ rake gettext
>   (in /home/dw/rails/typo)
>   I18n messages in app/views/articles/index.rhtml:
>    * Good Morning
>    * Hello
>    * Good Bye
>
> 9) Well, now, you will actually have to go through your source code 
> and translate it. If you forget how remember step 1. Every time you 
> add or change a translation, just repeat step 8.
>
> 10) Most important: Have fun and be proud that from now on you can say 
> that, yes, your apps will also (at least potentially) be understood by 
> the majority of the internet population[4].
>
>
> [1] http://typo.leetsoft.com/trac/
> [2] http://ponx.s5.xrea.com/hiki/ruby-gettext.html
> [3] http://ri18n.berlios.de/
> [4] Which was about 64,2 % of the internet population in 2004
> http://global-reach.biz/globstats/index.php3
>
> Now, please tell me if you find this information useful and more 
> important if you find anything to improve on.
>
> Sascha Ebach
> _______________________________________________
> Rails mailing list
> Rails-1W37MKcQCpIf0INCOvqR/iCwEArCW2h5@public.gmane.org
> http://lists.rubyonrails.org/mailman/listinfo/rails
>
>
___________________
Ben Jackson
Diretor de Desenvolvimento

+55 (21) 9997-0593
ben-p14LI7ZcAE/pVLaUnt/cCQC/G2K4zDHf@public.gmane.org
http://www.incomumdesign.com

I know this is a bad practice, but I have tons of ''legacy'' PHP
stuff that
uses iso-8859-1 encoding. How can I use this module effectively with this 
encoding for now? Specifying the content type of the .po file doesn''t
seem
to have any effect (The harvester aborts when it finds a bad character).

Joshua Sierles


_______________________________________________
Rails mailing list
Rails-1W37MKcQCpIf0INCOvqR/iCwEArCW2h5@public.gmane.org
http://lists.rubyonrails.org/mailman/listinfo/rails

Joshua Sierles wrote:
> I know this is a bad practice, but I have tons of
''legacy'' PHP stuff
> that uses iso-8859-1 encoding. How can I use this module effectively 
> with this encoding for now? Specifying the content type of the .po file 
> doesn''t seem to have any effect (The harvester aborts when it
finds a
> bad character).
Your best bet is to convert the files with iconv on the command line.

(something like that, untested)
$iconv -f "iso-8859-1" -t "UTF-8" legacy.php >
new.utf8.php

As always, make backup before you start.

Sascha

Thanks Sascha, for posting this. At first I was skeptical, but your  
steps were, in fact, easy.

One thing I''ve noticed, however. When I add more text to the app, run  
"rake gettext" and edit my PO file, I''ve got to restart
WEBrick, or I
just see the msgid. IOW, if I add <%= _(''greeting'') %>
to an rhtml
file and msgid="greeting"...msgstr="hello" in en.po, I see
greeting
until I restart WEBrick and then I see hello.

Is it really necessary to restart WEBrick every time I add strings?

Thanks again for posting this.


david

Le 05-07-30 à 17:53, Sascha Ebach a écrit :
> Some time ago I wrote the manual about using Gettext to translate  
> your Rails app. It is still available under http:// 
> manuals.rubyonrails.com/read/book/16 and holds some good  
> introductions and broader explanations about the subject.
>
> Recently I wanted to have my own weblog. I toyed around with a  
> couple of PHP solutions, but I hated the prospect of having to  
> maintain yet another pile of badly written PHP scripts. I already  
> have enough of those :). So I turned to Typo[1]. But there was a  
> problem. Typo doesn''t support *any* kind of translation as of  
> 2005-07-31. So I talked to Tobias Luettke who is the maintainer of  
> this wonderful blogging software (and all around nice guy) and  
> asked if he was interested in getting a patch to gettextify typo.  
> He said he would rather not integrate with Ruby-GetText[2] because  
> it has binary dependencies with Iconv (which is optional) and  
> Gettext. So I looked into integrating Ri18n[3] which is still very  
> young but also excellent. The author has really thought more about  
> nicer integration into rails apps which is a declared goal.
>
> So, instead of just making a patch for Typo I ended up writing a  
> little guide on how to apply this strategy to every rails app. Now  
> there is no excuse anymore for not making your rails apps  
> translatable. Shame on you if you don''t ;). I know that a lot of  
> people would like to see some support for translation (and broader  
> localization) in Rails itself. So I thought in the meantime I would  
> just post this to the Mailinglist and ask for feedback. If there  
> are no problems I will also post it on the Wiki or the manuals  
> section.
>
> As always before you start make sure that you use UTF-8 everywhere  
> (also in the Database, MySQL 4.1.x supports this).
>
> Here we go.
>
> 10 Easy Steps To Make Your Rails Apps Multilingual (I18N)
>
> 1) Read the USAGE of ri18n on http://ri18n.berlios.de/rdoc/files/ 
> USAGE.html. Repeat two times. Keep referring to it if you have  
> questions along the way.
>
> 2) Check out the fresh svn trunk into the vendor dir:
>
>   cd $RAILS_ROOT/vendor
>   svn checkout svn://svn.berlios.de/ri18n/trunk ri18n
>
> 3) Now add this path to the ADDITIONAL_LOAD_PATHS array in  
> $RAILS_ROOT/config/environment.rb so that it reads:
>
>   ADDITIONAL_LOAD_PATHS.concat %w(
>     # ... other entries
>     vendor
>     vendor/ri18n/lib
>     # ... other entries
>   ).map { |dir| "#{RAILS_ROOT}/#{dir}" }.select { |dir|  
> File.directory?(dir) }
>
> 4) Now require necessary ri18n library files in $RAILS_ROOT/config/ 
> environment.rb:
>
>   # Load included libraries.
>   # ... other libraries
>   require ''gettext''
>   require ''i18nservice''
>
> 5) Create a locale dir in the base dir of your application:
>
>   mkdir $RAILS_ROOT/locale
>
> 6) Ri18n has to be configured. At the end of your $RAILS_ROOT/ 
> config/environment.rb file add:
>
>   $KCODE = ''u'' # lets go unicode, you will have less
trouble ...
>   require ''jcode'' # and use the right string methods for
utf-8
>   # The directory where translation files (PO files) will go
>   I18nService.instance.po_dir = "#{RAILS_ROOT}/locale"
>   # let''s use German everywhere
>   I18nService.instance.lang = ''de''
>
> Notice that the last line just sets your language to German on a  
> global basis. If you want to decide which language to use in a more  
> dynamic manner go and take a look at the old manual: http:// 
> manuals.rubyonrails.com/read/book/16. It should be easy to adapt  
> the strategy I present there to your own needs.
>
> 7) You will want to have a way to harvest your files. Ri18n has a  
> built-in harvester that is controlled by a rake task. So go ahead  
> and add a rake task to the end of $RAILS_ROOT/Rakefile:
>
>   task :gettext do
>     $:.unshift ''vendor/ri18n/lib''
>     require ''gettext''
>     require ''i18nservice''
>     I18nService.instance.po_dir = File.dirname(__FILE__) +
''/locale''
>     desc ''Run the ri18n harvester on this application''
>     Rake::GettextTask.new do |t|
>       # add all desired langues here
>       t.new_langs = [''de'', ''fr'']
>       # add all desired paths here
>       t.source_files =
[''{lib,app,components}/**/*.r{b,html,xml}'']
>       t.verbose = true
>     end
>   end
>
> 8) Now you can call `rake gettext` in your $RAILS_ROOT and the  
> harvester will collect all the strings in your files which looks  
> something like this:
>
>   dw@colinux: /home/dw/rails/typo:
>   $ rake gettext
>   (in /home/dw/rails/typo)
>   I18n messages in app/views/articles/index.rhtml:
>    * Good Morning
>    * Hello
>    * Good Bye
>
> 9) Well, now, you will actually have to go through your source code  
> and translate it. If you forget how remember step 1. Every time you  
> add or change a translation, just repeat step 8.
>
> 10) Most important: Have fun and be proud that from now on you can  
> say that, yes, your apps will also (at least potentially) be  
> understood by the majority of the internet population[4].
>
>
> [1] http://typo.leetsoft.com/trac/
> [2] http://ponx.s5.xrea.com/hiki/ruby-gettext.html
> [3] http://ri18n.berlios.de/
> [4] Which was about 64,2 % of the internet population in 2004
> http://global-reach.biz/globstats/index.php3
>
> Now, please tell me if you find this information useful and more  
> important if you find anything to improve on.
>
> Sascha Ebach
> _______________________________________________
> Rails mailing list
> Rails-1W37MKcQCpIf0INCOvqR/iCwEArCW2h5@public.gmane.org
> http://lists.rubyonrails.org/mailman/listinfo/rails
>

David Geary wrote:
> Thanks Sascha, for posting this. At first I was skeptical, but your  
> steps were, in fact, easy.
Yeah, thanks, there are a couple of other things that really don''t work
right. I will contact the author and keep you posted.
> 
> One thing I''ve noticed, however. When I add more text to the app,
run
> "rake gettext" and edit my PO file, I''ve got to restart
WEBrick, or I
> just see the msgid. IOW, if I add <%= _(''greeting'')
%> to an rhtml  file
> and msgid="greeting"...msgstr="hello" in en.po, I see
greeting  until I
> restart WEBrick and then I see hello.
> 
> Is it really necessary to restart WEBrick every time I add strings?
When you are in development mode it should work. I''ll put it on my list
and check for it. But if I remeber correctly it worked on Debian / lighttpd.

Sascha

2005/7/31, Wilson
<defiler-ifvz4xmYPRU@public.gmane.org>:
> I''ve never been able to figure this out.. Maybe you can explain?
> What do Shift-JIS, EUC-JP, and the others offer that UTF-8
doesn''t?
They offer legacy and stable environments.
We must use Shift JIS as HTML for mobile agent,
and should use ISO-2022-JP as e-mail. If we use UTF-8,
we should use encoding converter, but converting japanese
encoding is troublesome(ex. windows GAIJI is used in
HTML which has charset=Shift_JIS).

In addition, We use many tools that don''t support UTF-8. It''s
not
impossible to change all tools, but...

P.S.
A few days ago, I met some Japanese railers, and I asked
about this problem. They advised me to use UTF-8. uummmmm...
> Surely it''s not just for text art on 2ch. =)
(^_^;;


Regards,

Masayoshi Takahashi
> masayoshi takahashi wrote:
> > agreed.  Sascha''s solution is great, but "use UTF-8
everywhere"
> > is never "easy step" for me.  I (and many Japanese web
developers)
> > usually use EUC-JP(in scripts) and Shift_JIS(in HTML templates)
> > and ISO-2022-JP (in E-mails) in one application.
> >
> _______________________________________________
> Rails mailing list
> Rails-1W37MKcQCpIf0INCOvqR/iCwEArCW2h5@public.gmane.org
> http://lists.rubyonrails.org/mailman/listinfo/rails
>

Actually I meant the po files themselves. The rake gettext command
fails because there are accented characters in my .po files.

Joshua

On 8/2/05, Sascha Ebach
<se-eFwX6J65rk9VioaHkBSlcw02NpfuEekPhC4ANOJQIlc@public.gmane.org>
wrote:
> 
> Joshua Sierles wrote:
> > I know this is a bad practice, but I have tons of
''legacy'' PHP stuff
> > that uses iso-8859-1 encoding. How can I use this module effectively
> > with this encoding for now? Specifying the content type of the .po
file
> > doesn''t seem to have any effect (The harvester aborts when it
finds a
> > bad character).
> 
> Your best bet is to convert the files with iconv on the command line.
> 
> (something like that, untested)
> $iconv -f "iso-8859-1" -t "UTF-8" legacy.php >
new.utf8.php
> 
> As always, make backup before you start.
> 
> Sascha
> _______________________________________________
> Rails mailing list
> Rails-1W37MKcQCpIf0INCOvqR/iCwEArCW2h5@public.gmane.org
> http://lists.rubyonrails.org/mailman/listinfo/rails
>

_______________________________________________
Rails mailing list
Rails-1W37MKcQCpIf0INCOvqR/iCwEArCW2h5@public.gmane.org
http://lists.rubyonrails.org/mailman/listinfo/rails

Joshua Sierles wrote:
> Actually I meant the po files themselves. The rake gettext command
> fails because there are accented characters in my .po files.
Haven''t tested that, yet, but the harvester fails on a couple of other 
things, too. For example:

_(''"test"'')
becomes
msgid ""test""
which should be
msgid "\"test\""


I will contact the author about this today.

Sascha

Hi,

Thanks for this good i18N guide.

I wonder if it would be possible to use some caching system together
with the before_filter, to avoid translation on the fly for each
request. Maybe it''s not worth?
-- 
Jean-Christophe Michel

Jean-Christophe Michel wrote:
> Hi,
> 
> Thanks for this good i18N guide.
> 
> I wonder if it would be possible to use some caching system together
> with the before_filter, to avoid translation on the fly for each
> request. Maybe it''s not worth?
I wonder what you mean by that. Can you explain a little more? What 
exactly should be cached?

Sascha

Sascha Ebach wrote:
>> I wonder if it would be possible to use some caching system together
>> with the before_filter, to avoid translation on the fly for each
>> request. Maybe it''s not worth?
> 
> I wonder what you mean by that. Can you explain a little more? What
> exactly should be cached?
Maybe I''m wrong. I thought string translation through _() calls was
made
at runtime, while the static strings translated with gettext don''t
depend on the request. So my reaction was: to optimize this, have a
cache after the translation. Ruby files with strings translated could be
cached as long as no modification is made to the source or to the
translation files.

-- 
Jean-Christophe Michel

Jean-Christophe Michel wrote:
> Maybe I''m wrong. I thought string translation through _() calls
was made
> at runtime, while the static strings translated with gettext don''t
> depend on the request. So my reaction was: to optimize this, have a
> cache after the translation. Ruby files with strings translated could be
> cached as long as no modification is made to the source or to the
> translation files.
Ok, now I understand. Yes, I guess a simple marshal would work for that. 
I suggest you make a feature request. At the moment the online way I 
know to do that is to mail the author Denis Mertz. You can find the 
email on the bottom of http://ri18n.berlios.de/.

Sascha

Hello Sascha,

great work with that I18N-Guide.

I tried this out yesterday and nearly everything works fine for me.

I just have one problem:

The command "rake gettext" seems to generate the po-files with
ascii-encoding not utf-8.

At least if I try to open these po-files with poEdit I get the error message
"cannot convert from encoding ascii".

Furthermore if use german characters in the like ''ß'' or
''ü'' in the de.po-file I get  an error message when the rails
environment is loaded.

"./script/../config/../vendor/gems/ri18n-0.0.3/lib/ri18n/po.rb:22:in
`iconv'':Iconv::IllegalSequence: "ß %i einen Wurm"..."

What is wrong? Help is as always very appreciated!

Govinda

This is my configuration (I did everything according to 10 easy steps manual):

environment.rb
----------------------
require ''gettext''
require ''i18nservice''

$KCODE = ''UTF8''     # lets go unicode, you will have less
trouble ...
require ''jcode''     # and use the right string methods for
utf-8

# The directory where translation files (PO files) will go
I18nService.instance.po_dir = "#{RAILS_ROOT}/locale"

# let''s use German everywhere
I18nService.instance.lang = ''de''

rakefile
---------------------------
task :gettext do
  $:.unshift ''vendor/gems/ri18n-0.0.3/lib''
  require ''gettext''
  require ''i18nservice''
  I18nService.instance.po_dir = File.dirname(__FILE__) +
''/locale''
  desc ''Run the ri18n harvester on this application''
  Rake::GettextTask.new do |t|
    # add all desired langues here
    t.new_langs = [''de'', ''en'']
    # add all desired paths here
    t.source_files =
[''{lib,app,components}/**/*.r{b,html,xml}'']
    t.verbose = true
  end
end

Testclass "Duck" to do initial I18N tests
-----------------------------------------
class Duck

def talk
 puts _(''I talk like a duck'')
 puts _(''Quack Quack !!'')
end

def walk
 puts _(''I walk like a duck'')
end

def eat_worms
 0.upto(4) do |i|
  puts n_(''I ate %i worm'', ''I ate %i worms'',
i)
 end
end

end

de.po
------------------------------------------------
# SOME DESCRIPTIVE TITLE.
# Copyright (C) YEAR Free Software Foundation, Inc.
# This file is distributed under the same license as the PACKAGE package.
# FIRST AUTHOR <EMAIL@ADDRESS>, YEAR.
#
#, fuzzy
msgid ""
msgstr ""
"Project-Id-Version: PACKAGE VERSION\n"
"PO-Revision-Date: YEAR-MO-DA HO:MI+ZONE\n"
"POT-Creation-Date: 2002-12-10 22:11+0100\n"
"Last-Translator: FULL NAME <EMAIL@ADDRESS>\n"
"Language-Team: LANGUAGE
<LL-2l5V8CiXCHs@public.gmane.org>\n"
"MIME-Version: 1.0\n"
"Content-Type: text/plain; charset=ASCII\n"
"Content-Transfer-Encoding: 8bit\n"
"Plural-Forms: nplurals=2; plural=n == 0;\n"

msgid "I ate %i worm"
msgid_plural "I ate %i worms"
msgstr[0] "Ich aß %i einen Wurm"
msgstr[1] "Ich aß %i Würmer"

msgid "I talk like a duck"
msgstr "Ich spreche wie eine Ente"

msgid "I walk like a duck"
msgstr "Ich lauf wie eine Ente"

msgid "Quack Quack !!"
msgstr "Quack Quack (Deutsch)"

en.po
-----------------------------------------------------------------
# SOME DESCRIPTIVE TITLE.
# Copyright (C) YEAR Free Software Foundation, Inc.
# This file is distributed under the same license as the PACKAGE package.
# FIRST AUTHOR <EMAIL@ADDRESS>, YEAR.
#
#, fuzzy
msgid ""
msgstr ""
"Project-Id-Version: PACKAGE VERSION\n"
"PO-Revision-Date: YEAR-MO-DA HO:MI+ZONE\n"
"POT-Creation-Date: 2002-12-10 22:11+0100\n"
"Last-Translator: FULL NAME <EMAIL@ADDRESS>\n"
"Language-Team: LANGUAGE
<LL-2l5V8CiXCHs@public.gmane.org>\n"
"MIME-Version: 1.0\n"
"Content-Type: text/plain; charset=ASCII\n"
"Content-Transfer-Encoding: 8bit\n"
"Plural-Forms: nplurals=2; plural=n == 0;\n"

msgid "I ate %i worm"
msgid_plural "I ate %i worms"
msgstr[0] "I ate %i worm"
msgstr[1] "I ate %i worms"

msgid "I talk like a duck"
msgstr "I talk like a duck"

msgid "I walk like a duck"
msgstr "I walk like a duck"

msgid "Quack Quack !!"
msgstr "Quack Quack (Englisch)"

application.rb
------------------------------------------
# start: utf8-support
# Content-Type HTTP header so that Apache and other web servers don’t serve up
some incorrect Content-Type

before_filter :set_default_content_type

  def set_content_type(content_type)
    @response.headers["Content-Type"] = content_type
  end

  def set_default_content_type
    set_content_type("text/html; charset=utf-8")
  end

# end: utf8-support

Mit WEB.DE FreePhone mit hoechster Qualitaet ab 0 Ct./Min.   
weltweit telefonieren! http://freephone.web.de/?mc=021201  

_______________________________________________
Rails mailing list
Rails-1W37MKcQCpIf0INCOvqR/iCwEArCW2h5@public.gmane.org
http://lists.rubyonrails.org/mailman/listinfo/rails

Govinda Pfister wrote:
> Hello Sascha,
> 
> great work with that I18N-Guide.
> 
> I tried this out yesterday and nearly everything works fine for me.
> 
> I just have one problem:
> 
> The command "rake gettext" seems to generate the po-files with 
> ascii-encoding not utf-8.
> 
> At least if I try to open these po-files with poEdit I get the error 
> message "cannot convert from encoding ascii".
> 
> Furthermore if use german characters in the like ''ß'' or
''ü'' in the
> de.po-file I get  an error message when the rails environment is loaded.
> 
> "./script/../config/../vendor/gems/ri18n-0.0.3/lib/ri18n/po.rb:22:in 
> `iconv'':Iconv::IllegalSequence: "ß %i einen
Wurm"..."
> 
> What is wrong? Help is as always very appreciated!
Yes, that was my mistake. I forgot to specify the $KCODE in the rake 
task. ri18n decides which charset to use by looking at it. So this is 
how it is supposed to look:

task :gettext do
   $:.unshift ''vendor/ri18n/lib''
   # don''t forget the $KCODE
   $KCODE = ''u''
   require ''gettext''
   require ''i18nservice''
   I18nService.instance.po_dir = File.dirname(__FILE__) +
''/locale''
   desc ''Run the ri18n harvester on this application''
   Rake::GettextTask.new do |t|
     # add all desired langues here
     t.new_langs = [''de'', ''fr'']
     # add all desired paths here
     t.source_files =
[''{lib,app,components}/**/*.r{b,html,xml}'']
     t.verbose = true
   end
end

Sascha Ebach

Hello,

now I do not get the error message any more, but an irb session (the rails
console) shows that something is still not working properly:

irb session
---------

>> ente = Duck.new
=> #
>> ente.eat_worms
Ich aß 0  Würmer
Ich aß 1 einen Wurm
Ich aß 2  Würmer
Ich aß 3  Würmer
Ich aß 4  Würmer
=> 0
>> 

The translation should be "Ich aß 2 Würmer".

This is would is in the de.po file, too. Opening this file with Scite shows the
translation with the right german characters. The UTF-8 format of the po-file is
recognized by Scite, too.

What could be wrong?

Govinda

rails-1W37MKcQCpIf0INCOvqR/iCwEArCW2h5@public.gmane.org schrieb am 10.08.05
12:58:56:

Govinda Pfister wrote:
> Hello Sascha,
> 
> great work with that I18N-Guide.
> 
> I tried this out yesterday and nearly everything works fine for me.
> 
> I just have one problem:
> 
> The command "rake gettext" seems to generate the po-files
with
> ascii-encoding not utf-8.
> 
> At least if I try to open these po-files with poEdit I get the error 
> message "cannot convert from encoding ascii".
> 
> Furthermore if use german characters in the like ''ß''
or ''ü'' in the
> de.po-file I get an error message when the rails environment is loaded.
> 
>
"./script/../config/../vendor/gems/ri18n-0.0.3/lib/ri18n/po.rb:22:in
> `iconv'':Iconv::IllegalSequence: "ß %i einen
Wurm"..."
> 
> What is wrong? Help is as always very appreciated!

Yes, that was my mistake. I forgot to specify the $KCODE in the rake 
task. ri18n decides which charset to use by looking at it. So this is 
how it is supposed to look:

task :gettext do
$:.unshift ''vendor/ri18n/lib''
# don''t forget the $KCODE
$KCODE = ''u''
require ''gettext''
require ''i18nservice''
I18nService.instance.po_dir = File.dirname(__FILE__) +
''/locale''
desc ''Run the ri18n harvester on this application''
Rake::GettextTask.new do |t|
# add all desired langues here
t.new_langs = [''de'', ''fr'']
# add all desired paths here
t.source_files = [''{lib,app,components}/**/*.r{b,html,xml}'']
t.verbose = true
end
end

Sascha Ebach
_______________________________________________
Rails mailing list
Rails-1W37MKcQCpIf0INCOvqR/iCwEArCW2h5@public.gmane.org
http://lists.rubyonrails.org/mailman/listinfo/rails

Verschicken Sie romantische, coole und witzige Bilder per SMS!   
Jetzt bei WEB.DE FreeMail: http://f.web.de/?mc=021193  

_______________________________________________
Rails mailing list
Rails-1W37MKcQCpIf0INCOvqR/iCwEArCW2h5@public.gmane.org
http://lists.rubyonrails.org/mailman/listinfo/rails

On Wednesday 10 August 2005 14:28, Govinda Pfister
wrote:
> now I do not get the error message any more, but an irb session (the
> rails console) shows that something is still not working properly:
> irb session
> Ich aß 2  Würmer
> The translation should be "Ich aß 2 Würmer".
Is your console set to display UTF-8 correctly?

Michael

-- 
Michael Schuerig               Most people would rather die than think.
mailto:michael-q5aiKMLteq4b1SvskN2V4Q@public.gmane.org                          
In fact, they do.
http://www.schuerig.de/michael/                      --Bertrand Russell

Govinda Pfister wrote:
> Hello,
> 
> now I do not get the error message any more, but an irb session (the 
> rails console) shows that something is still not working properly:
> 
> irb session
> ---------
> 
>  >> ente = Duck.new
> => #<Duck:0x83e79a8>
>  >> ente.eat_worms
> Ich aß 0  Würmer
> Ich aß 1 einen Wurm
> Ich aß 2  Würmer
> Ich aß 3  Würmer
> Ich aß 4  Würmer
> => 0
>  >>
> 
> The translation should be "Ich aß 2 Würmer".
> 
> This is would is in the de.po file, too. Opening this file with Scite 
> shows the translation with the right german characters. The UTF-8 format 
> of the po-file is recognized by Scite, too.
> 
> What could be wrong?
Nothing is wrong. Your console doesn''t support unicode output. So 
internally everything is correct, but the display on the console
doesn''t
know what to do with those characters.

Sascha