i18n bug regarding Arabic-Indic numbers needinfo.

[Lars Knoll]

On Tuesday 20 April 2004 00:14, Adriaan de Groot wrote:
> On Tuesday 20 April 2004 00:03, Shaheed wrote:
> > On Monday 19 April 2004 17:34, Diego Iastrubni wrote:
> > > Munzir, wrong list. I am forwarding to kde-core-devel (this was in
> > > cooker-I18N originally).
> > >
> > > > Hinid numberals are not numbers in linux!
> > > >
> > > > In kppp:
> > > > If I type the phone number in Hindi numerals e.g. ٣٦٠٦٠٦٠ . It won't
> > > > dial properly.
> >
> > Latin...
> > 0 is a number? 1
> > Bengali...
> > 0 is a number? 0
> > 1 is a number? 0
>
> This reminds me of a trick question I used to set for my students regarding
> representation of numerals; I'd ask them (on the blackboard) what ١٥ + ١٥
> was (the answer is obviously ٣٠). But before we go into this too far,
> consider: what is ٣٠ + 7? 37? ٣٧? Does QString::arg(int) produce localized
> output at all, anyway?

Yes it does if you tell it to. QString("%L1").arg(1.23) does produce localized 
output. The advantage of this scheme is that the translators can decide 
themselves if they want localized numbers or not.

See also http://doc.trolltech.com/3.3/qstring.html#arg-2

Cheers,
Lars