[haiku] Re: Migration from Basho to Official Localization of Haiku

  • From: Oliver Tappe <zooey@xxxxxxxxxxxxxxx>
  • To: haiku@xxxxxxxxxxxxx
  • Date: Sat, 31 Oct 2009 10:52:06 +0100

On 2009-10-30 at 23:10:28 [+0100], PulkoMandy <pulkomandy@xxxxxxxxx> wrote:
> > Things like *Emacs or irssi do have quite good methods of guessing
> > encodings.
> >
> > Maybe ICU has some, Adrien ?
> 
> Bad news : "ICU converters are not designed to perform any encoding
> autodetection." (http://userguide.icu-project.org/conversion/converters)
> At least ICU can convert :
> http://icu-project.org/apiref/icu4c/ucnv_8h.html#f4c967c5afa207d064c24e19256586b6

Hey, but that's just the converters themselves. ICU *does* cater for charset 
detection:

        http://userguide.icu-project.org/conversion/detection

I suppose this should be good enough, it just takes some thinking on which 
parts 
of that API and how exactly to wrap in liblocale.so. 
But once we have made this ICU functionality available via liblocale.so, it 
should 
be rather easy to create some generic version of sjis2utf8 with it (maybe call 
it 
'UngarbleName'?).

cheers,
        Oliver

Other related posts: