FULL PRODUCT VERSION :
java version "1.6.0_19"
Java(TM) SE Runtime Environment (build 1.6.0_19-b04)
Java HotSpot(TM) Client VM (build 16.2-b04, mixed mode, sharing)

ADDITIONAL OS VERSION INFORMATION :
Microsoft Windows [Version 6.0.6002]

A DESCRIPTION OF THE PROBLEM :
utf-8 decoder allows for directly encoded trail surrogates

For example: the sequence 0xed 0xba 0xab is decoded to "\uDEAB"


EXPECTED VERSUS ACTUAL BEHAVIOR :
EXPECTED -
I expect the behavior specified by CodingErrorAction (replace, ignore, report) to be triggered in this case. I do not expect a trail surrogate, it is not legal according to Unicode:

When a process interprets a code unit sequence which purports to be in a Unicode character encoding form, it shall treat ill-formed code unit sequences as an error condition and shall not interpret such sequences as characters.

Because surrogate code points are not Unicode scalar values, any UTF-8 byte sequence that would otherwise map to code points D800..DFFF is ill-formed.

ACTUAL -
the decoder instead accepts the invalid byte sequence (it does not matter what you set CodingErrorAction to), and converts it to a trail surrogate.

REPRODUCIBILITY :
This bug can be reproduced always.

---------- BEGIN SOURCE ----------
// decoding this should not yield the trail surrogate itself

public void test() throws Exception {
    byte[] invalid = new byte[] { (byte)0xed, (byte)0xba, (byte)0xab };
    assertFalse(new String(invalid, 0, invalid.length, "UTF-8").equals("\uDEAB"));
}
---------- END SOURCE ----------

CUSTOMER SUBMITTED WORKAROUND :
Write your own decoder.