Locale::Maketext::Lexicon - Use other catalog formats in Maketext

This document describes version 0.53 of Locale::Maketext::Lexicon, released December 5, 2005.

As part of a localization class, automatically glob for available lexicons:

    package Hello::I18N;
    use base 'Locale::Maketext';
    use Locale::Maketext::Lexicon {
        '*' => [Gettext => '/usr/local/share/locale/*/LC_MESSAGES/hello.mo'],
        _decode => 1,   # decode lexicon entries into utf8-strings
    };

Explicitly specify languages, during compile- or run-time:

    package Hello::I18N;
    use base 'Locale::Maketext';
    use Locale::Maketext::Lexicon {
        de => [Gettext => 'hello_de.po'],
        fr => [
            Gettext => 'hello_fr.po',
            Gettext => 'local/hello/fr.po',
        ],
    };
    # ... incrementally add new lexicons
    Locale::Maketext::Lexicon->import({
        de => [Gettext => 'local/hello/de.po'],
    })

Alternatively, as part of a localization subclass:

    package Hello::I18N::de;
    use base 'Hello::I18N';
    use Locale::Maketext::Lexicon (Gettext => \*DATA);
    __DATA__
    # Some sample data
    msgid ""
    msgstr ""
    "Project-Id-Version: Hello 1.3.22.1\n"
    "MIME-Version: 1.0\n"
    "Content-Type: text/plain; charset=iso8859-1\n"
    "Content-Transfer-Encoding: 8bit\n"

    #: Hello.pm:10
    msgid "Hello, World!"
    msgstr "Hallo, Welt!"

    #: Hello.pm:11
    msgid "You have %quant(%1,piece) of mail."
    msgstr "Sie haben %quant(%1,Poststueck,Poststuecken)."

This module provides lexicon-handling modules to read from other localization formats, such as Gettext, Msgcat, and so on.

If you are unfamiliar with the concept of lexicon modules, please consult Locale::Maketext and <http://www.autrijus.org/webl10n/> first.

A command-line utility xgettext.pl is also installed with this module, for extracting translatable strings from source files. The CWimport() function accepts two forms of arguments:

(format => source ... )

This form takes any number of argument pairs (usually one); source may be a file name, a filehandle, or an array reference. For each such pair, it pass the contents specified by the second argument to Locale::Maketext::Lexicon::BIformat->parse as a plain list, and export its return value as the CW%Lexicon hash in the calling package. In the case that there are multiple such pairs, the lexicon defined by latter ones overrides earlier ones.

{ language => [ format, source ... ] ... }

This form accepts a hash reference. It will export a CW%Lexicon into the subclasses specified by each language, using the process described above. It is designed to alleviate the need to set up a separate subclass for each localized language, and just use the catalog files. This module will convert the language arguments into lowercase, and replace all CW- with CW_, so CWzh_TW and CWzh-tw will both map to the CWzh_tw subclass. If language begins with CW_, it is taken as an option that controls how lexicons are parsed. See Options for a list of available options. The CW* is a special language; it must be used in conjunction with a filename that also contains CW*; all matched files with a valid language code in the place of CW* will be automatically prepared as a lexicon subclass. If there is multiple CW* in the filename, the last one is used as the language name.

If set to a true value, source entries will be converted into utf8-strings (available in Perl 5.6.1 or later). This feature needs the Encode or Encode::compat module. Currently, only the CWGettext backend supports this option. This option only has effect when CW_decode is set to true. It specifies an encoding to store lexicon entries, instead of utf8-strings. If CW_encoding is set to CWlocale, the encoding from the current locale setting is used.

If you wish to override how sources specified in different data types are handled, please use a subclass that overrides CWlexicon_get_CITYPECW. XXX: not documented well enough yet. Patches welcome.

When you attempt to localize an entry missing in the lexicon, Maketext will throw an exception by default. To inhibit this behaviour, override the CW_AUTO key in your language subclasses, for example:

    $Hello::I18N::en::Lexicon{_AUTO} = 1; # autocreate missing keys

Thanks to Jesse Vincent for suggesting this module to be written. Thanks also to Sean M. Burke for coming up with Locale::Maketext in the first place, and encouraging me to experiment with alternative Lexicon syntaxes. Thanks also to Yi Ma Mao for providing the MO file parsing subroutine, as well as inspiring me to implement file globbing and transcoding support. See the AUTHORS file in the distribution for a list of people who have sent helpful patches, ideas or comments.

xgettext.pl for extracting translatable strings from common template systems and perl source files. Locale::Maketext, Locale::Maketext::Lexicon::Auto, Locale::Maketext::Lexicon::Gettext, Locale::Maketext::Lexicon::Msgcat, Locale::Maketext::Lexicon::Tie

Autrijus Tang <autrijus@autrijus.org>

Copyright 2002, 2003, 2004, 2005 by Autrijus Tang <autrijus@autrijus.org>. This program is free software; you can redistribute it and/or modify it under the same terms as Perl itself. See <http://www.perl.com/perl/misc/Artistic.html>