Name: ooR10006 Date: 12/26/2000


This bug is filed only as the reason for JCK API test exclusion. Merlin API spec is correctly
changed, so this bug may be closed.
There is a typo in jdk1.3, 1.2.2, 1.2 API specs for method Character.isIdentifierIgnorable(ch):
wrong range 0x200A through 0x200E instead of 0x202A through 0x202E is specified as the char range
for which method isIdentifierIgnorable(ch) should return true.
In fact, these should be chars in the range 0x202A through 0x202E as specified in Unicode 2.0.14
standard and implemented in jdk1.3(1.2.2, 1.2) API.

The following simple test shows this:

class test {
     public static void main(String[] args){
         for (int i = 0x200A; i <= 0x200E; ++i){
             if (Character.isIdentifierIgnorable((char)i) == false){
                 System.out.println("Error for " + Integer.toHexString(i));
             }
         }
         for (int i = 0x202A; i <= 0x202E; ++i){
             if (Character.isIdentifierIgnorable((char)i) == false){
                 System.out.println("Error for " + Integer.toHexString(i));
             }
         }
     }
}

novo81% jdk1.3/solaris/bin/javac test.java
novo81% jdk1.3/solaris/bin/java -cp . test
Error for 200a
Error for 200b
novo81%

Due to this JCK 1.3a, JCK1.3, JCK1.2.2 API test
api/java_lang/Character/index.html#attributes[Character2050],
JCK 1.2a api/java_lang/Character/index.html#attributes[Character2049]
may fail.

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