Decoding the sequence
fd 93 a8 8c b9
with the charset UTF-8 with direct buffers gives

java.nio.BufferUnderflowException
	at java.nio.Buffer.nextGetIndex(Buffer.java:492)
	at java.nio.DirectByteBuffer.get(DirectByteBuffer.java:223)
	at  xxxxx .nio.cs.UTF_8$Decoder.decodeBufferLoop(UTF_8.java:339)
	at  xxxxx .nio.cs.UTF_8$Decoder.decodeLoop(UTF_8.java:378)
	at java.nio.charset.CharsetDecoder.decode(CharsetDecoder.java:562)
Posted Date : 2007-12-01 18:53:53.0

N/A

Yes.  Here's the trivial fix to UTF_8.java:

@@ -322,15 +593,15 @@
 				  ((b4 & 0x3f) << 06) |
 				  ((b5 & 0x3f) << 00));
 			    n = 5;
 			    break;
 
 			case 12: case 13:
 			    // 6 bytes, 31 bits
-			    if (src.remaining() < 4)
+			    if (src.remaining() < 5)
 				return CoderResult.UNDERFLOW;
 			    if (!isContinuation(b2 = src.get()))
 				return CoderResult.malformedForLength(1);
 			    if (!isContinuation(b3 = src.get()))
 				return CoderResult.malformedForLength(2);
 			    if (!isContinuation(b4 = src.get()))
 				return CoderResult.malformedForLength(3);
Posted Date : 2007-12-01 18:53:02.0

Here's a change to the test case FindDecoderBugs.java
that checks long UTF-8 sequences.

@@ -287,21 +287,27 @@
 		System.out.printf("Testing prefix %s%n", string(prefix));
 
 	    test(prefix);
 
 	    testExhaustively(prefix, 1);
 	    testExhaustively(prefix, 2);
 	    // Can you spare a week of CPU time?
-	    // testExhaustively(cs, tester, prefix, 3);
+	    // testExhaustively(prefix, 3);
 
 	    testRandomly(prefix, 3);
 	    testRandomly(prefix, 4);
+            if (cs == utf8) {
+                testRandomly(prefix, 5);
+                testRandomly(prefix, 6);
+            }
 	}
     }
 
+    private final static Charset utf8 = Charset.forName("UTF-8");
+
     private final static Random rnd = new Random();
     private static byte randomByte() {
 	return (byte) rnd.nextInt(0x100);
     }
     private static byte[] randomBytes(int len) {
 	byte[] a = new byte[len];
 	for (int i = 0; i < len; i++)
Posted Date : 2007-12-01 18:53:02.0