-
Bug
-
Resolution: Fixed
-
P3
-
1.4.1
-
None
-
mantis
-
generic
-
generic
-
Verified
Name: nl37777 Date: 06/03/2002
Some of the character encodings that are newly supported
through the nio APIs in Hopper have names that are inconsistent with the
Charset specification. The specification says: "If a charset has more
than one [IANA] registry name then its canonical name must be the
MIME-preferred name and the other names in the registry must be valid
aliases. If a supported charset is not listed in the IANA registry then
its canonical name must begin with one of the strings 'X-' or 'x-'."
The following charsets are not listed in the IANA registry: EUC-CN,
euc-jp-linux, EUC-TW, ISCII91, ISO-2022-CN-CNS, ISO-2022-CN-GB, JIS0208,
Johab, windows-949, windows-950.
The following charsets are not listed, but there are entries in the IANA
registry that may possibly be correct equivalents: JIS0201 (JIS_X0201?),
JIS0212 (JIS_X0212-1990?).
windows-936 is listed in the IANA registry as an alias of GBK, while the
J2RE treats GBK and windows-936 as separate encodings, since the Windows
code page 936 has diverged from the PRC national standard GBK. According
to the Charset specification, windows-936 must be an alias for GBK, and
the Windows version therefore can't use that name.
======================================================================