6843578: Re-implement IBM doublebyte charsets
6639450: IBM949C encoder modifies state of IBM949 encoder
6569191: Cp943 io converter returns U+0000 and U+FFFD for unconvertable character
6577466: Character encoder IBM970 throws a BufferOverflowException
5065777: CharsetEncoder canEncode() methods often incorrectly return false
Summary: Re-write 11 IBM doublebyte charsets. Thanks Ulf.Zibis for the codereview!
Reviewed-by: martin