Re: Re: [PEPr] Comment on Internationalisation::I18N_Gettext

On Tue, 02 Mar 2004 15:20:18 +0100, Michael Wallner wrote: >> I don't know if forcing a list of charsets in the code (files >> .inc) is a good idea, since you risk to leave out many of them. >> I'd leave it to the user. > > Well, that's in deed the problem - at all. > > The key intention for this package was providing portability to > users. The fact is that scripts that call PHPs setlocale() are not > portable yet. I see, but: a) or you list *all* the charsets/langcodes b) or you provide the users with a method to set the charsets they need by themselves. The latter option would be much more flexible, less memory-intensive and the only thing to do would be a proper doc page on how to configure it. >> Since guessHTTP() and setCharset() are already on the TODO list >> for PEAR::Translation2, what do you think about helping me with >> the gettext driver for that package? > > I already talked to Stefan N. about integration in I18N and mailed > Wolfram K. regarding this idea - I must admint I didn't have deeper > look into Translation yet because it obvoisly seemed to be DB > dependent. Translation 1.x is, Translation2 is not. There's already an *experimental* gettext driver, if you feel brave enough to extend it with a proper .mo/.po parser (and ideally a writer), then you're more than welcome! >> Your .mo/.po parsers look very nice, and could replace my very >> naive .po parser. > > A MO writer would be very interesting... :) granted. Please have a look at Translation2. There's an Admin class too, for which the db and mdb drivers are already in place, but nothing for gettext yet. Best regards, Lorenzo