Re: Ada, Gnat and Unicode

[Jano <nono@celes.unizar.es>]

Robert I. Eachus dice...
> Jano wrote:
> > Robert I. Eachus dice...
> 
> > As you may be seeing now, I want to scan a folder and transform the 
> > filenames into UTF8. That's fine for me which know that I'm getting 
> > Latin1 encoded strings from the Directory_Operations package, and any 
> > metadata entered by the user. But I was wondering what would happen to a 
> > Chinese user (not that I foresee any usage of my program in wide 
> > deployment, but when faced with the problem one *must* know ;)
> 
> Remember my advice about canonicalization.  If you get Unicode or UTF-8 
> file names from the OS, they may or may not be in a canonical form.  If 
> not, get the OS to do it for you.  And of course, this information is OS 
> specific. You won't really care what the OS's definition of canonical 
> form is, just whether the strings you are getting are in that form, and 
> if not how to call the OS to do that.

Ok, I see. In the end that's the outcome I didn't want to hear but the 
one I expected.

> Yes, it refers to source representation, but if you think about it for a 
> second, the source representation of non-Latin1 characters is an issue 
> for Character and String literals.  Otherwise the compiler doesn't care 
> what Character type you use in your program.

I was referring to that too :)

Thanks!

-- 
-------------------------
Jano
402450.at.cepsz.unizar.es
-------------------------