Name: ooR10006 Date: 05/03/2001


Since jdk1.4.0beta-b59 Character.isIdentifierIgnorable(0xFFFF) returns true
instead of false.
The spec for isIdentifierIgnorable(char) says:
"The following Unicode characters are ignorable in a Java identifier
or a Unicode identifier:
         ISO control characters that are not whitespace
              '\u0000' through '\u0008'
              '\u000E' through '\u001B'
              '\u007F' through '\u009F'
         all characters that have the FORMAT general category value "

According to this, 0xFFFF should not be considered as ignorable character.

Character.isJavaIdentifierPart(char), Character.isUnicodeIdentifierPart(char)
also return true for 0xFFFF instead of false.

The following test shows this:

public class Test {
    public static void main(String[] args){
        System.out.println(Character.isIdentifierIgnorable((char) 0xFFFF));
        System.out.println(Character.isJavaIdentifierPart((char) 0xFFFF));
        System.out.println(Character.isUnicodeIdentifierPart((char) 0xFFFF));
    }
}

% jdk1.4.0beta-b58/solsparc/bin/java Test
false
false
false
% jdk1.4.0beta-b59/solsparc/bin/java Test
true
true
true
%

Due to this Merlin JCK tests fail:
api/java_lang/Character/index.html#attributesFullRange[Character2067]
api/java_lang/Character/index.html#attributesFullRange[Character2069]
api/java_lang/Character/index.html#attributesFullRange[Character2076]

jdk1.3.0fcsC passes the tests

======================================================================