1.1   Getting Started

Everywhere in your code where you want strings to be available in different languages you wrap them in the _() function. There are also a number of other translation functions which are documented in the API reference at http://pylonshq.com/docs/module-pylons.i18n.translation.html

In our example we want the string 'Hello' to appear in three different languages: English, French and Spanish. We also want to display the word 'Hello' in the default language. We'll then go on to use some pural words too.

Lets call our project translate_demo:

1

$ paster create -t pylons translate_demo

Now lets add a friendly controller that says hello:

1
2

$ cd translate_demo
$ paster controller hello

Edit controllers/hello.py to make use of the _() function everywhere where the string Hello appears:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15

import logging

from pylons.i18n import get_lang, set_lang

from translate_demo.lib.base import *

log = logging.getLogger(__name__)

class HelloController(BaseController):

    def index(self):
        response.write('Default: %s<br />' % _('Hello'))
        for lang in ['fr','en','es']:
            set_lang(lang)
            response.write("%s: %s<br />" % (get_lang(), _('Hello')))

When writing wrapping strings in the gettext functions, it is important not to piece sentences together manually; certain languages might need to invert the grammars. Don't do this:

1
2
3

# BAD!
msg = _("He told her ")
msg += _("not to go outside.")

but this is perfectly acceptable:

1
2

# GOOD
msg = _("He told her not to go outside")

The controller has now been internationalized, but it will raise a LanguageError until we have setup the alternative language catalogs.

GNU gettext use three types of files in the translation framework.

POT (Portable Object Template) files

The first step in the localization process. A program is used to search through your project's source code and pick out every string passed to one of the translation functions, such as _(). This list is put together in a specially-formatted template file that will form the basis of all translations. This is the .pot file.

PO (Portable Object) files

The second step in the localization process. Using the POT file as a template, the list of messages are translated and saved as a .po file.

MO (Machine Object) files

The final step in the localization process. The PO file is run through a program that turns it into an optimized machine-readable binary file, which is the .mo file. Compiling the translations to machine code makes the localized program much faster in retrieving the translations while it is running.

GNU gettext provides a suite of command line programs for extracting messages from source code and working with the associated gettext catalogs. The Babel project provides pure Python alternative versions of these tools. Unlike the GNU gettext tool xgettext', Babel supports extracting translatable strings from Python templating languages (currently Mako and Genshi).

1.2   Using Babel

To use Babel, you must first install it via easy_install. Run the command:

1

$ easy_install Babel

Pylons (as of 0.9.6) includes some sane defaults for Babel's distutils commands in the setup.cfg file.

It also includes an extraction method mapping in the setup.py file. It is commented out by default, to avoid distutils warning about it being an unrecognized option when Babel is not installed. These lines should be uncommented before proceeding with the rest of this walk through:

1
2
3
4

message_extractors = {'translate_demo': [
        ('**.py', 'python', None),
        ('templates/**.mako', 'mako', None),
        ('public/**', 'ignore', None)]},

We'll use Babel to extract messages to a .pot file in your project's i18n directory. First, the directory needs to be created. Don't forget to add it to your revision control system if one is in use:

1
2
3

$ cd translate_demo
$ mkdir translate_demo/i18n
$ svn add translate_demo/i18n

Next we can extract all messages from the project with the following command:

1
2
3
4
5
6
7

$ python setup.py extract_messages
running extract_messages
extracting messages from translate_demo/__init__.py
extracting messages from translate_demo/websetup.py
...
extracting messages from translate_demo/tests/functional/test_hello.py
writing PO template file to translate_demo/i18n/translate_demo.pot

This will create a .pot file in the i18n directory that looks something like this:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22

# Translations template for translate_demo.
# Copyright (C) 2007 ORGANIZATION
# This file is distributed under the same license as the translate_demo project.
# FIRST AUTHOR <EMAIL@ADDRESS>, 2007.
#
#, fuzzy
msgid ""
msgstr ""
"Project-Id-Version: translate_demo 0.0.0\n"
"Report-Msgid-Bugs-To: EMAIL@ADDRESS\n"
"POT-Creation-Date: 2007-08-02 18:01-0700\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"
"Generated-By: Babel 0.9dev-r215\n"

#: translate_demo/controllers/hello.py:10 translate_demo/controllers/hello.py:13
msgid "Hello"
msgstr ""

The .pot details that appear here can be customized via the extract_messages configuration in your project's setup.cfg (See the Babel Command-Line Interface Documentation for all configuration options).

Next, we'll initialize a catalog (.po file) for the Spanish language:

1
2
3
4

$ python setup.py init_catalog -l es
running init_catalog
creating catalog 'translate_demo/i18n/es/LC_MESSAGES/translate_demo.po' based on
    'translate_demo/i18n/translate_demo.pot'

Then we can edit the last line of the new Spanish .po file to add a translation of "Hello":

1
2

msgid "Hello"
msgstr "¡Hola!"

Finally, to utilize these translations in our application, we need to compile the .po file to a .mo file:

1
2
3
4
5

$ python setup.py compile_catalog
running compile_catalog
1 of 1 messages (100%) translated in 'translate_demo/i18n/es/LC_MESSAGES/translate_demo.po'
compiling catalog 'translate_demo/i18n/es/LC_MESSAGES/translate_demo.po' to
    'translate_demo/i18n/es/LC_MESSAGES/translate_demo.mo'

We can also use the update_catalog command to merge new messages from the .pot to the .po files. For example, if we later added the following line of code to the end of HelloController's index method:

1

response.write('Goodbye: %s' % _('Goodbye'))

We'd then need to re-extract the messages from the project, then run the update_catalog command:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11

$ python setup.py extract_messages
running extract_messages
extracting messages from translate_demo/__init__.py
extracting messages from translate_demo/websetup.py
...
extracting messages from translate_demo/tests/functional/test_hello.py
writing PO template file to translate_demo/i18n/translate_demo.pot
$ python setup.py update_catalog
running update_catalog
updating catalog 'translate_demo/i18n/es/LC_MESSAGES/translate_demo.po' based on
    'translate_demo/i18n/translate_demo.pot'

We'd then edit our catalog to add a translation for "Goodbye", and recompile the .po file as we did above.

For more information, see the Babel documentation and the GNU Gettext Manual.

1.3   Back To Work

Next we'll need to repeat the process of creating a .mo file for the en and fr locales:

1
2
3
4
5
6
7
8

$ python setup.py init_catalog -l en
running init_catalog
creating catalog 'translate_demo/i18n/en/LC_MESSAGES/translate_demo.po' based on
    'translate_demo/i18n/translate_demo.pot'
$ python setup.py init_catalog -l fr
running init_catalog
creating catalog 'translate_demo/i18n/fr/LC_MESSAGES/translate_demo.po' based on
    'translate_demo/i18n/translate_demo.pot'

Modify the last line of the fr catalog to look like this:

1
2
3

#: translate_demo/controllers/hello.py:10 translate_demo/controllers/hello.py:13
msgid "Hello"
msgstr "Bonjour"

Since our original messages are already in English, the en catalog can stay blank; gettext will fallback to the original.

Once you've edited these new .po files and compiled them to .mo files, you'll end up with an i18n directory containing:

1
2
3
4
5
6
7

i18n/translate_demo.pot
i18n/en/LC_MESSAGES/translate_demo.po
i18n/en/LC_MESSAGES/translate_demo.mo
i18n/es/LC_MESSAGES/translate_demo.po
i18n/es/LC_MESSAGES/translate_demo.mo
i18n/fr/LC_MESSAGES/translate_demo.po
i18n/fr/LC_MESSAGES/translate_demo.mo

1.4   Testing the Application

Start the server with the following command:

1

$ paster serve --reload development.ini

Test your controller by visiting http://localhost:5000/hello. You should see the following output:

1
2
3
4

Default: Hello
fr: Bonjour
en: Hello
es: ¡Hola!

You can now set the language used in a controller on the fly.

For example this could be used to allow a user to set which language they wanted your application to work in. You could save the value to the session object:

1
2

session['lang'] = 'en'
session.save()

then on each controller call the language to be used could be read from the session and set in your controller's __before__() method so that the pages remained in the same language that was previously set:

1
2
3

def __before__(self):
    if 'lang' in session:
        set_lang(session['lang'])

Pylons also supports defining the default language to be used in the configuration file. Set a lang variable to the desired default language in your development.ini file, and Pylons will automatically call set_lang with that language at the beginning of every request.

E.g. to set the default language to Spanish, you would add lang = es to your development.ini:

1
2
3

[app:main]
use = egg:translate_demo
lang = es

If you are running the server with the --reload option the server will automatically restart if you change the development.ini file. Otherwise restart the server manually and the output would this time be as follows:

1
2
3
4

Default: ¡Hola!
fr: Bonjour
en: Hello
es: ¡Hola!

1.5   Fallback Languages

If your code calls _() with a string that doesn't exist at all in your language catalog, the string passed to _() is returned instead.

Modify the last line of the hello controller to look like this:

1

response.write("%s %s, %s" % (_('Hello'), _('World'), _('Hi!')))

If you run the example again the output will be:

1
2
3
4

Default: ¡Hola!
fr: Bonjour World!
en: Hello World!
es: ¡Hola! World!

This is because we never provided a translation for the string 'World!' so the string itself is used.

Pylons also provides a mechanism for fallback languages, so that you can specify other languages to be used if the word is ommitted from the main language's catalog.

In this example we choose fr as the main language but es as a fallback:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14

import logging

from pylons.i18n import add_fallback, set_lang

from translate_demo.lib.base import *

log = logging.getLogger(__name__)

class HelloController(BaseController):

     def index(self):
         set_lang('fr')
         add_fallback('es')
         return "%s %s, %s" % (_('Hello'), _('World'), _('Hi!'))

If Hello is in the fr .mo file as Bonjour, World is only in es as Mundo and none of the catalogs contain Hi!, you'll get the multilingual message: Bonjour Mundo, Hi!. This is a combination of the French, Spanish and original (English in this case, as defined in our source code) words.

You can add as many fallback languages with the add_fallback() function as you like and they will be tested in the order you add them.

One case where using fallbacks in this way is particularly useful is when you wish to display content based on the languages requested by the browser in the HTTP_ACCEPT_LANGUAGE header. Typically the browser may submit a number of languages so it is useful to be add fallbacks in the order specified by the browser so that you always try to display words in the language of preference and search the other languages in order if a translation cannot be found. The languages defined in the HTTP_ACCEPT_LANGAUGE header are available in Pylons as request.languages and can be used like this:

1
2

for lang in request.languages:
    add_fallback(lang)

1.6   Translations Within Templates

You can also use the _() function within templates in exactly the same way you do in code. For example, in a Mako template:

1

${_('Hello')}

would produce the string 'Hello' in the language you had set.

Babel currently supports extracting gettext messages from Mako and Genshi templates. The Mako extractor also provides support for translator comments. Babel can be extended to extract messages from other sources via a custom extraction method plugin.

Pylons (as of 0.9.6) automatically configures a Babel extraction mapping for your Python source code and Mako templates. This is defined in your project's setup.py file:

1
2
3
4
5

message_extractors = {'translate_demo': [
        ('public/**', 'ignore', None),
        ('**.py', 'python', None),
        ('templates/**.mako', 'mako', None),
        ]},

For a project using Genshi instead of Mako, the Mako line might be replaced with:

1

('templates/**.html, 'genshi, None),

See Babel's documentation on Message Extraction for more information.

1.7   Lazy Translations

Occasionally you might come across a situation when you need to translate a string when it is accessed, not when the _() or other functions are called.

Consider this example:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18

import logging

from pylons.i18n import get_lang, set_lang

from translate_demo.lib.base import *

log = logging.getLogger(__name__)

text = _('Hello')

class HelloController(BaseController):

    def index(self):
        response.write('Default: %s<br />' % _('Hello'))
        for lang in ['fr','en','es']:
            set_lang(lang)
            response.write("%s: %s<br />" % (get_lang(), _('Hello')))
        response.write('Text: %s<br />' % text)

If we run this we get the following output:

1
2
3
4
5

Default: Hello
['fr']: Bonjour
['en']: Good morning
['es']: Hola
Text: Hello

This is because the function _('Hello') just after the imports is called when the default language is en so the variable text gets the value of the English translation even though when the string was used the default language was Spanish.

The rule of thumb in these situations is to try to avoid using the translation functions in situations where they are not executed on each request. For situations where this isn't possible, perhaps because you are working with legacy code or with a library which doesn't support internationalization, you need to use lazy translations.

If we modify the above example so that the import statements and assignment to text look like this:

1
2
3
4
5
6
7

from pylons.i18n import get_lang, lazy_gettext, set_lang

from helloworld.lib.base import *

log = logging.getLogger(__name__)

text = lazy_gettext('Hello')

then we get the output we expected:

1
2
3
4
5

Default: Hello
['fr']: Bonjour
['en']: Good morning
['es']: Hola
Text: Hola

There are lazy versions of all the standard Pylons translation functions.

There is one drawback to be aware of when using the lazy translation functions: they are not actually strings. This means that if our example had used the following code it would have failed with an error cannot concatenate 'str' and 'LazyString' objects:

1

response.write('Text: ' + text + '<br />')

For this reason you should only use the lazy translations where absolutely necessary and should always ensure they are converted to strings by calling str() or repr() before they are used in operations with real strings.

1.8   Producing a Python Egg

Finally you can produce an egg of your project which includes the translation files like this:

1

$ python setup.py bdist_egg

The setup.py automatically includes the .mo language catalogs your application needs so that your application can be distributed as an egg. This is done with the following line in your setup.py file:

1

package_data={'translate_demo': ['i18n/*/LC_MESSAGES/*.mo']},

1.9   Plural Forms

Pylons also provides the ungettext() function. It's designed for internationalizing plural words, and can be used as follows:

1
2

ungettext('There is %(num)d file here', 'There are %(num)d files here',
          n) % {'num': n}

Plural forms have a different type of entry in .pot/.po files, as described in The Format of PO Files in GNU Gettext's Manual:

1
2
3
4
5
6

#: translate_demo/controllers/hello.py:12
#, python-format
msgid "There is %(num)d file here"
msgid_plural "There are %(num)d files here"
msgstr[0] ""
msgstr[1] ""

One thing to keep in mind is that other languages don't have the same plural forms as English. While English only has 2 plural forms, singular and plural, Slovenian has 4! That means that you must use ugettext for proper pluralization. Specifically, the following will not work:

1
2
3
4
5

# BAD!
if n == 1:
    msg = _("There was no dog.")
else:
    msg = _("There were no dogs.")