On 16/Sep/2010 14:30, Robert Muir wrote: > On Thu, Sep 16, 2010 at 9:26 AM, Tim Ellison <t.p.elli...@gmail.com> wrote: > >> Yes, I agree (though I don't know what happens when it is uppercased >> 'properly' in Turkish) >> > > in this case it would be uppercased to US-ASCİİ (note the dotted i's) > > >> Nah, the Charset name is obliged to be a strict subset of characters >> [1], so I propose we just deal with the transform directly in >> CharsetProviderImpl without asking Character. >> >> [1] >> >> http://download.oracle.com/javase/1.5.0/docs/api/java/nio/charset/Charset.html >> >> > +1, seems like now i understand why this constraint exists!
The principle works ok. I attached a patch on HARMONY-6649 to show that making a local toUpperCase() method for the charset names solves the circularity problem in bootstrapping, and does the right thing irrespective of locale. There are some compatibility issues though, since on the RI String foo="foo"; foo.toLowerCase() == foo but it doesn't if we simply use ICU's toLowerCase method :-( I also expect we need to fix String#equalsIgnoreCase() to do the right thing... Regards, Tim
