Language Localization (l18n, L10n) - Multilingual Output

[davidfilmer]

I'm developing a script (web cgi) for which I would like to offer multilanguage support. I would like some advise on how I might (better) approach this task. The script will be fairly small and usage (on a non-profit org website) will be fairly light, so efficiency concerns aren't terribly important.

What I'm presently thinking is that I won't hard-code any output in my script, but will populate (from a database or whatever) a hash which would look this:
    $lang{'yes'} = 'sí';
    $lang{'no'} = 'ningún';
    $lang{'The answer is'} = 'La respuesta es';

Then, in my script, I would output something like:
    print "$lang{'The answer is'} $lang{'yes'}.";
and it would print:
    La respuesta es sí.
assuming I had populated the hash from a Spanish database. (In the English version, the hash values are simply the same as the hash keys.)

Does this seem a good approach? It seems a bit clumsy, but a better idea doesn't come to mind (though I can think of a couple of ideas that I think are much worse). A better suggestion would be most welcome!

[adrianh]

You might want to take a look at Locale::Maketext. Very useful for this sort of thing. More useful tips can be found in Writing multilingual perl applications?

[waswas-fng]

You will run into problems where the nouns adverbsetc change location in languages. For example some languages would be "The red car". where others could be "The car red". I think it may be more worth while to use a full template system for completly changing the output text to the translated languages.

-Waswas