Problems in current implementation:

(1)Johab converters is really the clone of ms949, which is "Unified Hangul Code"
   (has full compatibility with KS C 5601-1992 and adds additional encoding
    ranges for holding additional pre-combined hanguls - the 8,822 code
    point which are needed to fully support the Johab character set)
(2)In CharacterEncoding, the ms949 is mapped to johab. Due to the (1),
   currently it causes no problem, but after we fixing (1), we also need
   to alias ms949 to it's correct converter, ms949 directly or UHC.
(3)We may need MS1361 converter which is the MS's Jahob codepage.