jason at archlinux.org
Thu Mar 31 13:49:12 EST 2005
> I18n isn't that hard. One of my projects is an Arabic concordancer. A
> concordancer is a simple tool that loads a large text file, say the complete
> works of Shakespeare, and then allows the user to search for a word or phrase.
> The concordance is then the target word in its context. It is used by linguists
> to understand how words behave, and especially lexographers when defining new
> words, etc.
> The problem is, the ones that exist don't cope with Arabic correctly (due to
> its right-to-left nature). I have a number of Arabic colleagues who need such a
> tool, so I've been obliging. Anyway, I can't read or speak Arabic, but creating
> such a tool, and adding i18n functionality has been really simple (in Java at
> least). Basically, all strings that would appear in your interface are stored
> in an external text file of "key = value" pairs. You have a separate file per
> language supported (these files are known as ResourceBundles and share the same
> name except they have a unique locale suffix) where the keys are the same, but
> the values are language specific. You need to load the Bundle in each class
> that needs these labels, but Java will load the appropriate language bundle
> depending on the current locale.
> 'Tis really easy. Anyway, despite this mass exodus of the forums, I hope that
> the Java experts will keep reading the Programming board, as I have some more
> Swing questions in the pipeline :P
Understood that i18n isn't that hard, but that's not what Dusty's talking
about. He's talking about voice recognition. Is it just as easy to
implement voice recognition for English as it is for Arabic? If you
implement one do you automatically get the other?
If you understand, things are just as they are. If you do not understand,
things are just as they are.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Size: 189 bytes
Desc: not available
More information about the arch-projects