Unicode and Win32

[João Garcia]

Hi People,

two posts from yesterday made me re-think the way the patch (submitted 
for review yesterday) should develop. Many thanks to Jeff and Fernando!

http://gcc.gnu.org/ml/java/2003-05/msg00337.html
http://gcc.gnu.org/ml/java/2003-05/msg00339.html

If the Win32 wide-character (Unicode)follows the same convention of the 
Java Unicode character, it would be nonsense to use a conversion at 
all... (yes... I know I have resisted to this idea yesterday... I was 
falling assleep)

Can anyone cast more light over this subject?


Since I cannot access the registry in libjava (because of the 
configuration of the compiler), does anyone knows how can I get OS 
version information (to distinguish bettween win9x and NT/2000/XP)?
There should be an w32API function to do that...

Besides that, and quoting MS about GetACP():
"Remarks
Windows 2000/XP: The return value for the Indic languages is 0, because 
they are Unicode only."
URI:
http://msdn.microsoft.com/library/default.asp?url=/library/en-us/intl/nls_21bk.asp

I wouldn't know what an Indic language is! But I think that they are 
referring to the four code-pages that the patch does not support. But 
Japan is in the Pacific, is it not?

If GetACP() returns 0, my patch will default to CP1252, so it should be 
equivalent to the original UTF-8 conversion (everything works fine if 
the codes are bellow 128).

João