Template::Multilingual::Parser - Multilingual template parser


    use Template;
    use Template::Multilingual::Parser;
    my $parser = Template::Multilingual::Parser->new();
    my $template = Template->new(PARSER => $parser);
    $template->process('example.ttml', { language => 'en'});


This subclass of Template Toolkit's Template::Parser parses multilingual templates: templates that contain text in several languages.

      <fr>Bonjour !</fr>

Use this module directly if you have subclassed Template, otherwise you may find it easier to use Template::Multilingual.

Language codes can be any string that matches \w+, but we suggest sticking to ISO-639 which provides 2-letter codes for common languages and 3-letter codes for many others.



The new() constructor creates and returns a reference to a new parser object. A reference to a hash may be supplied as a parameter to provide configuration values.

Parser objects are typically provided as the PARSER option to the Template constructor.

Configuration values are all valid Template::Parser superclass options, and one specific to this class:


The LANGUAGE_VAR option can be used to set the name of the template variable which contains the current language. Defaults to language.

  my $parser = Template::Multilingual::Parser->new({
     LANGUAGE_VAR => 'global.language',

You will need to set this variable with the current language value at request time, usually in your Template subclass' process() method.


parse() is called by the Template Toolkit. It parses multilingual sections from the input text and translates them to Template Toolkit directives. The result is then passed to the Template::Parser superclass.


Returns a reference to an array of tokenized sections. Each section is a reference to hash with either a nolang key or a lang key.

A nolang key denotes text outside of any multilingual sections. The value is the text itself.

A lang key denotes text inside a multilingual section. The value is a reference to a hash, whose keys are language codes and values the corresponding text. For example, the following multilingual template:

  foo <t><fr>bonjour</fr><en>Hello</en></t> bar

will parse to the following sections:

  [ { nolang => 'foo ' },
    {   lang => { fr => 'bonjour', en => 'hello' } },
    { nolang => ' bar' },


This module supports language subtags to express variants, e.g. "en_US" or "en-US". Here are the rules used for language matching:

  • Exact match: the current language is found in the template

      language    template                              output
      fr          <fr>foo</fr><fr_CA>bar</fr_CA>        foo
      fr_CA       <fr>foo</fr><fr_CA>bar</fr_CA>        bar
  • Fallback to the primary language

      language    template                              output
      fr_CA       <fr>foo</fr><fr_BE>bar</fr_BE>        foo
  • Fallback to first (in alphabetical order) other variant of the primary language

      language    template                              output
      fr          <fr_FR>foo</fr_FR><fr_BE>bar</fr_BE>  bar
      fr_CA       <fr_FR>foo</fr_FR><fr_BE>bar</fr_BE>  bar


Eric Cholet, <>


Multilingual text sections cannot be used inside TT directives. The following is illegal and will trigger a TT syntax error:

    [% title = "<t><fr>Bonjour</fr><en>Hello</en></t>" %]

Use this instead:

    [% title = BLOCK %]<t><fr>Bonjour</fr><en>Hello</en></t>[% END %]

The TAG_STYLE, START_TAG and END_TAG directives are supported, but the TAGS directive is not.

Please report any bugs or feature requests to, or through the web interface at I will be notified, and then you'll automatically be notified of progress on your bug as I make changes.



ISO 639-2 Codes for the Representation of Names of Languages:


Copyright 2009 Eric Cholet, All Rights Reserved.

This program is free software; you can redistribute it and/or modify it under the same terms as Perl itself.