#25 gettext for translations

[Appie Verschoor]

On the nucleusforum there was the sugestion to use
gettext for translations:
http://forum.nucleuscms.org/viewtopic.php?p=9363#9363
I did some quick reading, and it sounds not bad at all.
The poster; Ruben has implemented this in squirrelmail
as well. Maybe we could give this a better look and use
it to get translations. I haven't had time to dig in
deep so I have no recommandations yet.
Appie (xiffy)

[Wouter Demuynck]

Logged In: YES
user_id=315939

I did some quick reading also. Looks interesting.

If this can be used in a way that is not dependent on the
existance of the php extension, we should implement it

Pro:
- source code easier to read (since english text is there)
- language constants no longer need to be in memory (loaded
when needed)
- I assume that when a translation is not available, the
english text is used?

Con;
- ?

What happens if the original (english) text is changed after
the language files are created? is it easy to update the
lang file

[Ruben Mendes]

Logged In: YES
user_id=866238

1 - When the translation is not available the text in the
source code is used.

2 - When the original texts are changed we have to generate
a new translation file (usually with .pot extension). The
translators must then merge the changes with their current
(outdated) translation. See
http://www.squirrelmail.org/wiki/en_US/UpdatingTranslations
for instructions.

3 - Gettext is the standard way to add translation support
so there are lot's of non-english speaking people that are
familiar with it.

4 - The biggest problem I see is to convert all current
translations files to the .po getext format automatically.
It should be scriptable but it doesn't seem a trivial problem.

5 - It was not me who implemented the i18n in the core
squirrelmail code. I just help with the i18n code in some
plugins and do l10n to pt_PT.

Ruben