The test used four code pages from ftp://www.unicode.org/ .  The
page for JIS0202 is for "JIS X 0201 (1976) to Unicode 1.1".  The
code page for SHIFTJIS says that
#   This table contains the data the Unicode Consortium has on how
#       Shift-JIS (a combination of JIS 0201 and JIS 0208) maps into Unicode"
and is dated 8 March 1994.
The code page for JIS0212 is for "JIS X 0212 (1990)".
This JIS0208 test uses a code page for "JIS X 0208 (1990)".

Hex under "IN" is input to a conversion.  Hex in the "CHECK" column
is the expected output.  Hex under the "OUT" column is the actual output
using a 1.3 or 1.4 JDK.

CHECKING BYTE ARRAY TO STRING
PASS? CODE    IN    CHECK   OUT     COMMENT
FAIL  JIS0201  5C    00A5  005C  # YEN SIGN
FAIL  JIS0201  7E    203E  007E  # OVERLINE
FAIL  JIS0212  2237  007E  FF5E  # TILDE
FAIL  SJIS    5C    00A5  005C  # YEN SIGN
FAIL  SJIS    7E    203E  007E  # OVERLINE
FAIL  SJIS    815F  005C  FF3C  # REVERSE SOLIDUS
FAIL  JIS0208  2140  005C  FF3C  # REVERSE SOLIDUS

CHECKING STRING TO BYTE ARRAY
PASS? CODE    IN    CHECK   OUT     COMMENT
FAIL  JIS0212  007E  2237  7E    # TILDE
FAIL  SJIS    005C  815F  5C    # REVERSE SOLIDUS
FAIL  JIS0208  005C  2140  5C    # REVERSE SOLIDUS

The tests and code pages are attached.  These problems were also
reported in 4296969.  However, 4296969 reports a host of problems.
Not all of the "problems" reported in 4296969 are, in my
opinion, legitimate.  This bug is submitted to keep real bugs
from getting buried.

algol% /usr/j2se/bin/java -version
java version "1.3.0"
Java(TM) 2 Runtime Environment, Standard Edition (build 1.3.0-b25)
Java HotSpot(TM) Client VM (build 1.3.0-b25, mixed mode)
sqesvr% /usr/local/java/jdk1.4/solaris/bin/java -version
java version "1.4.0beta"
Java(TM) 2 Runtime Environment, Standard Edition (build 1.4.0beta-b27)
Java HotSpot(TM) Core VM (build 1.3-internal, interpreted mode)


allan.jacobs@Eng 2000-08-11