7096080: UTF8 update and new CESU-8 charset

7082884: Incorrect UTF8 conversion for sequence ED 31
7082883: Incorrect UTF8 conversion for sequence fc 80 80 8f bf bf
Summary: Updated UTF8 and added CESU-8 to following the latest Standard
Reviewed-by: alanb

make/java/nio/FILES_java.gmk

@@ -232,6 +232,7 @@ FILES_src = \
 	sun/nio/cs/UTF_16BE.java \
 	sun/nio/cs/UTF_16LE.java \
 	sun/nio/cs/UTF_8.java \
+	sun/nio/cs/CESU_8.java \
 	sun/nio/cs/Unicode.java \
 	sun/nio/cs/UnicodeDecoder.java \
 	sun/nio/cs/UnicodeEncoder.java \

src/share/classes/sun/nio/cs/UTF_8.java

@@ -72,7 +72,7 @@ class UTF_8 extends Unicode
         return new Encoder(this);
     }
 
-    static final void updatePositions(Buffer src, int sp,
+    private static final void updatePositions(Buffer src, int sp,
                                               Buffer dst, int dp) {
         src.position(sp - src.arrayOffset());
         dst.position(dp - dst.arrayOffset());
@@ -88,11 +88,6 @@ class UTF_8 extends Unicode
             return (b & 0xc0) != 0x80;
         }
 
-        //  [C2..DF] [80..BF]
-        private static boolean isMalformed2(int b1, int b2) {
-            return (b1 & 0x1e) == 0x0 || (b2 & 0xc0) != 0x80;
-        }
-
         //  [E0]     [A0..BF] [80..BF]
         //  [E1..EF] [80..BF] [80..BF]
         private static boolean isMalformed3(int b1, int b2, int b3) {
@@ -100,6 +95,12 @@ class UTF_8 extends Unicode
                    (b2 & 0xc0) != 0x80 || (b3 & 0xc0) != 0x80;
         }
 
+        // only used when there is only one byte left in src buffer
+        private static boolean isMalformed3_2(int b1, int b2) {
+            return (b1 == (byte)0xe0 && (b2 & 0xe0) == 0x80) ||
+                   (b2 & 0xc0) != 0x80;
+        }
+
         //  [F0]     [90..BF] [80..BF] [80..BF]
         //  [F1..F3] [80..BF] [80..BF] [80..BF]
         //  [F4]     [80..8F] [80..BF] [80..BF]
@@ -110,6 +111,16 @@ class UTF_8 extends Unicode
                    (b4 & 0xc0) != 0x80;
         }
 
+        // only used when there is less than 4 bytes left in src buffer
+        private static boolean isMalformed4_2(int b1, int b2) {
+            return (b1 == 0xf0 && b2 == 0x90) ||
+                   (b2 & 0xc0) != 0x80;
+        }
+
+        private static boolean isMalformed4_3(int b3) {
+            return (b3 & 0xc0) != 0x80;
+        }
+
         private static CoderResult lookupN(ByteBuffer src, int n)
         {
             for (int i = 1; i < n; i++) {
@@ -122,28 +133,14 @@ class UTF_8 extends Unicode
         private static CoderResult malformedN(ByteBuffer src, int nb) {
             switch (nb) {
             case 1:
-                int b1 = src.get();
-                if ((b1 >> 2) == -2) {
-                    // 5 bytes 111110xx 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx
-                    if (src.remaining() < 4)
-                        return CoderResult.UNDERFLOW;
-                    return lookupN(src, 5);
-                }
-                if ((b1 >> 1) == -2) {
-                    // 6 bytes 1111110x 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx
-                    if (src.remaining() < 5)
-                        return CoderResult.UNDERFLOW;
-                    return lookupN(src, 6);
-                }
-                return CoderResult.malformedForLength(1);
             case 2:                    // always 1
                 return CoderResult.malformedForLength(1);
             case 3:
-                b1 = src.get();
+                int b1 = src.get();
                 int b2 = src.get();    // no need to lookup b3
                 return CoderResult.malformedForLength(
                     ((b1 == (byte)0xe0 && (b2 & 0xe0) == 0x80) ||
-                     isNotContinuation(b2))?1:2);
+                     isNotContinuation(b2)) ? 1 : 2);
             case 4:  // we don't care the speed here
                 b1 = src.get() & 0xff;
                 b2 = src.get() & 0xff;
@@ -171,6 +168,7 @@ class UTF_8 extends Unicode
             return cr;
         }
 
+
         private static CoderResult malformed(ByteBuffer src,
                                              int mark, int nb)
         {
@@ -180,18 +178,36 @@ class UTF_8 extends Unicode
             return cr;
         }
 
+        private static CoderResult malformedForLength(ByteBuffer src,
+                                                      int sp,
+                                                      CharBuffer dst,
+                                                      int dp,
+                                                      int malformedNB)
+        {
+            updatePositions(src, sp, dst, dp);
+            return CoderResult.malformedForLength(malformedNB);
+        }
+
+        private static CoderResult malformedForLength(ByteBuffer src,
+                                                      int mark,
+                                                      int malformedNB)
+        {
+            src.position(mark);
+            return CoderResult.malformedForLength(malformedNB);
+        }
+
+
         private static CoderResult xflow(Buffer src, int sp, int sl,
                                          Buffer dst, int dp, int nb) {
             updatePositions(src, sp, dst, dp);
             return (nb == 0 || sl - sp < nb)
-                   ?CoderResult.UNDERFLOW:CoderResult.OVERFLOW;
+                   ? CoderResult.UNDERFLOW : CoderResult.OVERFLOW;
         }
 
         private static CoderResult xflow(Buffer src, int mark, int nb) {
-            CoderResult cr = (nb == 0 || src.remaining() < (nb - 1))
-                             ?CoderResult.UNDERFLOW:CoderResult.OVERFLOW;
             src.position(mark);
-            return cr;
+            return (nb == 0 || src.remaining() < nb)
+                   ? CoderResult.UNDERFLOW : CoderResult.OVERFLOW;
         }
 
         private CoderResult decodeArrayLoop(ByteBuffer src,
@@ -210,7 +226,6 @@ class UTF_8 extends Unicode
             // ASCII only loop
             while (dp < dlASCII && sa[sp] >= 0)
                 da[dp++] = (char) sa[sp++];
-
             while (sp < sl) {
                 int b1 = sa[sp];
                 if (b1 >= 0) {
@@ -219,13 +234,20 @@ class UTF_8 extends Unicode
                         return xflow(src, sp, sl, dst, dp, 1);
                     da[dp++] = (char) b1;
                     sp++;
-                } else if ((b1 >> 5) == -2) {
+                } else if ((b1 >> 5) == -2 && (b1 & 0x1e) != 0) {
                     // 2 bytes, 11 bits: 110xxxxx 10xxxxxx
+                    //                   [C2..DF] [80..BF]
                     if (sl - sp < 2 || dp >= dl)
                         return xflow(src, sp, sl, dst, dp, 2);
                     int b2 = sa[sp + 1];
-                    if (isMalformed2(b1, b2))
-                        return malformed(src, sp, dst, dp, 2);
+                    // Now we check the first byte of 2-byte sequence as
+                    //     if ((b1 >> 5) == -2 && (b1 & 0x1e) != 0)
+                    // no longer need to check b1 against c1 & c0 for
+                    // malformed as we did in previous version
+                    //   (b1 & 0x1e) == 0x0 || (b2 & 0xc0) != 0x80;
+                    // only need to check the second byte b2.
+                    if (isNotContinuation(b2))
+                        return malformedForLength(src, sp, dst, dp, 1);
                     da[dp++] = (char) (((b1 << 6) ^ b2)
                                        ^
                                        (((byte) 0xC0 << 6) ^
@@ -233,24 +255,37 @@ class UTF_8 extends Unicode
                     sp += 2;
                 } else if ((b1 >> 4) == -2) {
                     // 3 bytes, 16 bits: 1110xxxx 10xxxxxx 10xxxxxx
-                    if (sl - sp < 3 || dp >= dl)
+                    int srcRemaining = sl - sp;
+                    if (srcRemaining < 3 || dp >= dl) {
+                        if (srcRemaining > 1 && isMalformed3_2(b1, sa[sp + 1]))
+                            return malformedForLength(src, sp, dst, dp, 1);
                         return xflow(src, sp, sl, dst, dp, 3);
+                    }
                     int b2 = sa[sp + 1];
                     int b3 = sa[sp + 2];
                     if (isMalformed3(b1, b2, b3))
                         return malformed(src, sp, dst, dp, 3);
-                    da[dp++] = (char)
+                    char c = (char)
                         ((b1 << 12) ^
                          (b2 <<  6) ^
                          (b3 ^
                           (((byte) 0xE0 << 12) ^
                            ((byte) 0x80 <<  6) ^
                            ((byte) 0x80 <<  0))));
+                    if (Character.isSurrogate(c))
+                        return malformedForLength(src, sp, dst, dp, 3);
+                    da[dp++] = c;
                     sp += 3;
                 } else if ((b1 >> 3) == -2) {
                     // 4 bytes, 21 bits: 11110xxx 10xxxxxx 10xxxxxx 10xxxxxx
-                    if (sl - sp < 4 || dl - dp < 2)
+                    int srcRemaining = sl - sp;
+                    if (srcRemaining < 4 || dl - dp < 2) {
+                        if (srcRemaining > 1 && isMalformed4_2(b1, sa[sp + 1]))
+                            return malformedForLength(src, sp, dst, dp, 1);
+                        if (srcRemaining > 2 && isMalformed4_3(sa[sp + 2]))
+                            return malformedForLength(src, sp, dst, dp, 2);
                         return xflow(src, sp, sl, dst, dp, 4);
+                    }
                     int b2 = sa[sp + 1];
                     int b3 = sa[sp + 2];
                     int b4 = sa[sp + 3];
@@ -289,13 +324,13 @@ class UTF_8 extends Unicode
                         return xflow(src, mark, 1); // overflow
                     dst.put((char) b1);
                     mark++;
-                } else if ((b1 >> 5) == -2) {
+                } else if ((b1 >> 5) == -2 && (b1 & 0x1e) != 0) {
                     // 2 bytes, 11 bits: 110xxxxx 10xxxxxx
                     if (limit - mark < 2|| dst.remaining() < 1)
                         return xflow(src, mark, 2);
                     int b2 = src.get();
-                    if (isMalformed2(b1, b2))
-                        return malformed(src, mark, 2);
+                    if (isNotContinuation(b2))
+                        return malformedForLength(src, mark, 1);
                      dst.put((char) (((b1 << 6) ^ b2)
                                     ^
                                     (((byte) 0xC0 << 6) ^
@@ -303,24 +338,37 @@ class UTF_8 extends Unicode
                     mark += 2;
                 } else if ((b1 >> 4) == -2) {
                     // 3 bytes, 16 bits: 1110xxxx 10xxxxxx 10xxxxxx
-                    if (limit - mark < 3 || dst.remaining() < 1)
+                    int srcRemaining = limit - mark;
+                    if (srcRemaining < 3 || dst.remaining() < 1) {
+                        if (srcRemaining > 1 && isMalformed3_2(b1, src.get()))
+                            return malformedForLength(src, mark, 1);
                         return xflow(src, mark, 3);
+                    }
                     int b2 = src.get();
                     int b3 = src.get();
                     if (isMalformed3(b1, b2, b3))
                         return malformed(src, mark, 3);
-                    dst.put((char)
+                    char c = (char)
                         ((b1 << 12) ^
                          (b2 <<  6) ^
                          (b3 ^
                           (((byte) 0xE0 << 12) ^
                            ((byte) 0x80 <<  6) ^
-                               ((byte) 0x80 <<  0)))));
+                           ((byte) 0x80 <<  0))));
+                    if (Character.isSurrogate(c))
+                        return malformedForLength(src, mark, 3);
+                    dst.put(c);
                     mark += 3;
                 } else if ((b1 >> 3) == -2) {
                     // 4 bytes, 21 bits: 11110xxx 10xxxxxx 10xxxxxx 10xxxxxx
-                    if (limit - mark < 4 || dst.remaining() < 2)
+                    int srcRemaining = limit - mark;
+                    if (srcRemaining < 4 || dst.remaining() < 2) {
+                        if (srcRemaining > 1 && isMalformed4_2(b1, src.get()))
+                            return malformedForLength(src, mark, 1);
+                        if (srcRemaining > 2 && isMalformed4_3(src.get()))
+                            return malformedForLength(src, mark, 2);
                         return xflow(src, mark, 4);
+                    }
                     int b2 = src.get();
                     int b3 = src.get();
                     int b4 = src.get();
@@ -364,7 +412,7 @@ class UTF_8 extends Unicode
             return bb;
         }
 
-        // returns -1 if there is malformed byte(s) and the
+        // returns -1 if there is/are malformed byte(s) and the
         // "action" for malformed input is not REPLACE.
         public int decode(byte[] sa, int sp, int len, char[] da) {
             final int sl = sp + len;
@@ -381,11 +429,11 @@ class UTF_8 extends Unicode
                 if (b1 >= 0) {
                     // 1 byte, 7 bits: 0xxxxxxx
                     da[dp++] = (char) b1;
-                } else if ((b1 >> 5) == -2) {
+                } else if ((b1 >> 5) == -2 && (b1 & 0x1e) != 0) {
                     // 2 bytes, 11 bits: 110xxxxx 10xxxxxx
                     if (sp < sl) {
                         int b2 = sa[sp++];
-                        if (isMalformed2(b1, b2)) {
+                        if (isNotContinuation(b2)) {
                             if (malformedInputAction() != CodingErrorAction.REPLACE)
                                 return -1;
                             da[dp++] = replacement().charAt(0);
@@ -410,21 +458,33 @@ class UTF_8 extends Unicode
                             if (malformedInputAction() != CodingErrorAction.REPLACE)
                                 return -1;
                             da[dp++] = replacement().charAt(0);
-                            sp -=3;
+                            sp -= 3;
                             bb = getByteBuffer(bb, sa, sp);
                             sp += malformedN(bb, 3).length();
                         } else {
-                            da[dp++] = (char)((b1 << 12) ^
+                            char c = (char)((b1 << 12) ^
                                               (b2 <<  6) ^
                                               (b3 ^
                                               (((byte) 0xE0 << 12) ^
                                               ((byte) 0x80 <<  6) ^
                                               ((byte) 0x80 <<  0))));
+                            if (Character.isSurrogate(c)) {
+                                if (malformedInputAction() != CodingErrorAction.REPLACE)
+                                    return -1;
+                                da[dp++] = replacement().charAt(0);
+                            } else {
+                                da[dp++] = c;
+                            }
                         }
                         continue;
                     }
                     if (malformedInputAction() != CodingErrorAction.REPLACE)
                         return -1;
+                    if (sp  < sl && isMalformed3_2(b1, sa[sp])) {
+                        da[dp++] = replacement().charAt(0);
+                        continue;
+
+                    }
                     da[dp++] = replacement().charAt(0);
                     return dp;
                 } else if ((b1 >> 3) == -2) {
@@ -458,28 +518,29 @@ class UTF_8 extends Unicode
                     }
                     if (malformedInputAction() != CodingErrorAction.REPLACE)
                         return -1;
+
+                    if (sp  < sl && isMalformed4_2(b1, sa[sp])) {
+                        da[dp++] = replacement().charAt(0);
+                        continue;
+                    }
+                    sp++;
+                    if (sp  < sl && isMalformed4_3(sa[sp])) {
+                        da[dp++] = replacement().charAt(0);
+                        continue;
+                    }
                     da[dp++] = replacement().charAt(0);
                     return dp;
                 } else {
                     if (malformedInputAction() != CodingErrorAction.REPLACE)
                         return -1;
                     da[dp++] = replacement().charAt(0);
-                    sp--;
-                    bb = getByteBuffer(bb, sa, sp);
-                    CoderResult cr = malformedN(bb, 1);
-                    if (!cr.isError()) {
-                        // leading byte for 5 or 6-byte, but don't have enough
-                        // bytes in buffer to check. Consumed rest as malformed.
-                        return dp;
-                    }
-                    sp +=  cr.length();
                 }
             }
             return dp;
         }
     }
 
-    private static class Encoder extends CharsetEncoder
+    private static final class Encoder extends CharsetEncoder
                                  implements ArrayEncoder {
 
         private Encoder(Charset cs) {

src/share/classes/sun/nio/cs/standard-charsets

@@ -63,6 +63,10 @@ charset UTF-8 UTF_8
     alias UTF8				# JDK historical
     alias unicode-1-1-utf-8
 
+charset CESU-8 CESU_8
+    alias CESU8
+    alias csCESU-8
+
 charset UTF-16 UTF_16
     alias UTF_16			# JDK historical
     alias utf16

test/java/nio/charset/coders/Errors.java

@@ -23,7 +23,7 @@
 
 /* @test
  * @summary Check that error cases are replaced correctly in String/ISR/OSW
- * @bug 4457851
+ * @bug 4457851 7096080
  *
  * @build Errors Util
  * @run main Errors
@@ -193,11 +193,9 @@ public class Errors {
         t.test("\uFFFF", new byte[] { (byte)0xEF, (byte)0xBF, (byte)0xBF });
         t.test(new byte[] { X, (byte)0x7f, Y }, "x\u007Fy");
         t.test(new byte[] { X, (byte)0x80, Y }, "x\uFFFDy");
-        t.test(new byte[] { (byte)0xf0, (byte)0xf0 }, "\uFFFD");
     }
 
     public static void main(String[] args) throws Exception {
-
         test_US_ASCII(new TestString("US-ASCII"));
         test_US_ASCII(new TestStream("US-ASCII"));
 

test/sun/nio/cs/TestStringCoding.java

@@ -24,7 +24,7 @@
  */
 
 /* @test
-   @bug 6636323 6636319 7040220
+   @bug 6636323 6636319 7040220 7096080
    @summary Test if StringCoding and NIO result have the same de/encoding result
  * @run main/othervm/timeout=2000 TestStringCoding
  */
@@ -111,7 +111,8 @@ public class TestStringCoding {
         //encode unmappable surrogates
         if (enc instanceof sun.nio.cs.ArrayEncoder &&
             cs.contains(Charset.forName("ASCII"))) {
-            if (cs.name().equals("UTF-8"))    // utf8 handles surrogates
+            if (cs.name().equals("UTF-8") ||     // utf8 handles surrogates
+                cs.name().equals("CESU-8"))       // utf8 handles surrogates
                 return;
             enc.replaceWith(new byte[] { (byte)'A'});
             sun.nio.cs.ArrayEncoder cae = (sun.nio.cs.ArrayEncoder)enc;
@@ -136,7 +137,6 @@ public class TestStringCoding {
                                                        cs.name())))
                 throw new RuntimeException("encode3(surrogates) failed  -> "
                                            + cs.name());
-
             ba = new byte[str.length() - 1];
             n = cae.encode(str.toCharArray(), 0, str.length(), ba);
             if (n != 7 || !"abABABc".equals(new String(ba, 0, n,

test/sun/nio/cs/TestStringCodingUTF8.java

@@ -33,14 +33,16 @@ import java.nio.charset.*;
 
 public class TestStringCodingUTF8 {
     public static void main(String[] args) throws Throwable {
-        test();
+        test("UTF-8");
+        test("CESU-8");
         // security manager on
         System.setSecurityManager(new PermissiveSecurityManger());
-        test();
+        test("UTF-8");
+        test("CESU-8");
     }
 
-    static void test() throws Throwable {
-        Charset cs = Charset.forName("UTF-8");
+    static void test(String csn) throws Throwable {
+        Charset cs = Charset.forName(csn);
         char[] bmp = new char[0x10000];
         for (int i = 0; i < 0x10000; i++) {
             bmp[i] = (char)i;

test/sun/nio/cs/TestUTF8.java

@@ -23,7 +23,7 @@
 
 /*
  * @test
- * @bug 4486841 7040220
+ * @bug 4486841 7040220 7096080
  * @summary Test UTF-8 charset
  */
 
@@ -156,15 +156,22 @@ public class TestUTF8 {
         return 3;
     }
 
+    static int to4ByteUTF8(int uc, byte[] bb, int pos) {
+        bb[pos++] = (byte)(0xf0 | ((uc >> 18)));
+        bb[pos++] = (byte)(0x80 | ((uc >> 12) & 0x3f));
+        bb[pos++] = (byte)(0x80 | ((uc >>  6) & 0x3f));
+        bb[pos++] = (byte)(0x80 | (uc & 0x3f));
+        return 4;
+    }
+
     static void checkRoundtrip(String csn) throws Exception {
         System.out.printf("    Check roundtrip <%s>...", csn);
         char[] cc = getUTFChars();
         byte[] bb = encode(cc, csn, false);
         char[] ccO = decode(bb, csn, false);
 
-        if (!Arrays.equals(cc, ccO)) {
+        if (!Arrays.equals(cc, ccO))
             System.out.printf("    non-direct failed");
-        }
         bb = encode(cc, csn, true);
         ccO = decode(bb, csn, true);
         if (!Arrays.equals(cc, ccO)) {
@@ -180,19 +187,24 @@ public class TestUTF8 {
         System.out.println();
     }
 
-    static void check6ByteSurrs(String csn) throws Exception {
-        System.out.printf("    Check 6-byte Surrogates <%s>...%n", csn);
-        byte[] bb = new byte[(0x110000 - 0x10000) * 6];
+    static void check4ByteSurrs(String csn) throws Exception {
+        System.out.printf("    Check 4-byte Surrogates <%s>...%n", csn);
+        byte[] bb = new byte[(0x110000 - 0x10000) * 4];
         char[] cc = new char[(0x110000 - 0x10000) * 2];
         int bpos = 0;
         int cpos = 0;
         for (int i = 0x10000; i < 0x110000; i++) {
             Character.toChars(i, cc, cpos);
-            bpos += to3ByteUTF8(cc[cpos], bb, bpos);
-            bpos += to3ByteUTF8(cc[cpos + 1], bb, bpos);
+            bpos += to4ByteUTF8(i, bb, bpos);
             cpos += 2;
         }
+        checkSurrs(csn, bb, cc);
+    }
+
 
+    static void checkSurrs(String csn, byte[] bb, char[] cc)
+        throws Exception
+    {
         char[] ccO = decode(bb, csn, false);
         if (!Arrays.equals(cc, ccO)) {
             System.out.printf("    decoding failed%n");
@@ -201,14 +213,30 @@ public class TestUTF8 {
         if (!Arrays.equals(cc, ccO)) {
             System.out.printf("    decoding(direct) failed%n");
         }
-        // new String(bb, csn).getBytes(csn) will not return
-        // the 6 bytes surrogates as in bb, so only test
-        // toCharArray() here.
         if (!Arrays.equals(cc, new String(bb, csn).toCharArray())) {
             System.out.printf("    String.toCharArray() failed");
         }
+        if (!Arrays.equals(bb, new String(cc).getBytes(csn))) {
+            System.out.printf("    String.getBytes() failed");
+        }
+    }
+
+    static void check6ByteSurrs(String csn) throws Exception {
+        System.out.printf("    Check 6-byte Surrogates <%s>...%n", csn);
+        byte[] bb = new byte[(0x110000 - 0x10000) * 6];
+        char[] cc = new char[(0x110000 - 0x10000) * 2];
+        int bpos = 0;
+        int cpos = 0;
+        for (int i = 0x10000; i < 0x110000; i++) {
+            Character.toChars(i, cc, cpos);
+            bpos += to3ByteUTF8(cc[cpos], bb, bpos);
+            bpos += to3ByteUTF8(cc[cpos + 1], bb, bpos);
+            cpos += 2;
+        }
+        checkSurrs(csn, bb, cc);
     }
 
+
     static void compare(String csn1, String csn2) throws Exception {
         System.out.printf("    Diff <%s> <%s>...%n", csn1, csn2);
         char[] cc = getUTFChars();
@@ -266,6 +294,10 @@ public class TestUTF8 {
         {1, (byte)0xFF, (byte)0xFF, (byte)0xFF }, // all ones
         {1, (byte)0xE0, (byte)0xC0, (byte)0x80 }, // invalid second byte
         {1, (byte)0xE0, (byte)0x80, (byte)0xC0 }, // invalid first byte
+        {1, (byte)0xE0, (byte)0x41,},             // invalid second byte & 2 bytes
+        {3, (byte)0xED, (byte)0xAE, (byte)0x80 }, // 3 bytes surrogate
+        {3, (byte)0xED, (byte)0xB0, (byte)0x80 }, // 3 bytes surrogate
+
 
         // Four-byte sequences
         {1, (byte)0xF0, (byte)0x80, (byte)0x80, (byte)0x80 }, // U+0000 zero-padded
@@ -276,8 +308,13 @@ public class TestUTF8 {
         {1, (byte)0xFF, (byte)0xFF, (byte)0xFF, (byte)0xFF }, // all ones
         {1, (byte)0xF0, (byte)0x80, (byte)0x80, (byte)0x80},  // invalid second byte
         {1, (byte)0xF0, (byte)0xC0, (byte)0x80, (byte)0x80 }, // invalid second byte
+        {1, (byte)0xF0, (byte)41 },                           // invalid second byte
+                                                              // & only 2 bytes
+
         {2, (byte)0xF0, (byte)0x90, (byte)0xC0, (byte)0x80 }, // invalid third byte
-        {3, (byte)0xF0, (byte)0x90, (byte)0x80, (byte)0xC0 }, // invalid third byte
+        {3, (byte)0xF0, (byte)0x90, (byte)0x80, (byte)0xC0 }, // invalid forth byte
+        {2, (byte)0xF0, (byte)0x90, (byte)0x41 },             // invalid third byte
+                                                              // & 3 bytes input
 
         {1, (byte)0xF1, (byte)0xC0, (byte)0x80, (byte)0x80 }, // invalid second byte
         {2, (byte)0xF1, (byte)0x80, (byte)0xC0, (byte)0x80 }, // invalid third byte
@@ -287,30 +324,113 @@ public class TestUTF8 {
         {1, (byte)0xF5, (byte)0x80, (byte)0x80, (byte)0xC0 }, // out-range 4-byte
 
         // Five-byte sequences
-        {5, (byte)0xF8, (byte)0x80, (byte)0x80, (byte)0x80, (byte)0x80},  // invalid first byte
-        {5, (byte)0xF8, (byte)0x80, (byte)0x80, (byte)0x80, (byte)0x80 }, // U+0000 zero-padded
-        {5, (byte)0xF8, (byte)0x80, (byte)0x80, (byte)0x81, (byte)0xBF }, // U+007F zero-padded
-        {5, (byte)0xF8, (byte)0x80, (byte)0x80, (byte)0x9F, (byte)0xBF }, // U+07FF zero-padded
-        {5, (byte)0xF8, (byte)0x80, (byte)0x8F, (byte)0xBF, (byte)0xBF }, // U+FFFF zero-padded
+        {1, (byte)0xF8, (byte)0x80, (byte)0x80, (byte)0x80, (byte)0x80},  // invalid first byte
+        {1, (byte)0xF8, (byte)0x80, (byte)0x80, (byte)0x80, (byte)0x80 }, // U+0000 zero-padded
+        {1, (byte)0xF8, (byte)0x80, (byte)0x80, (byte)0x81, (byte)0xBF }, // U+007F zero-padded
+        {1, (byte)0xF8, (byte)0x80, (byte)0x80, (byte)0x9F, (byte)0xBF }, // U+07FF zero-padded
+        {1, (byte)0xF8, (byte)0x80, (byte)0x8F, (byte)0xBF, (byte)0xBF }, // U+FFFF zero-padded
+
+        {1, (byte)0xF8, (byte)0xC0, (byte)0x80, (byte)0x80, (byte)0x80},
+        {1, (byte)0xF8, (byte)0x80, (byte)0xC0, (byte)0x80, (byte)0x80 },
+        {1, (byte)0xF8, (byte)0x80, (byte)0x80, (byte)0xC1, (byte)0xBF },
+        {1, (byte)0xF8, (byte)0x80, (byte)0x80, (byte)0x9F, (byte)0xC0 },
+
+        // Six-byte sequences
+        {1, (byte)0xFC, (byte)0x80, (byte)0x80, (byte)0x80, (byte)0x80, (byte)0x80 }, // U+0000 zero-padded
+        {1, (byte)0xFC, (byte)0x80, (byte)0x80, (byte)0x80, (byte)0x81, (byte)0xBF }, // U+007F zero-padded
+        {1, (byte)0xFC, (byte)0x80, (byte)0x80, (byte)0x80, (byte)0x9F, (byte)0xBF }, // U+07FF zero-padded
+        {1, (byte)0xFC, (byte)0x80, (byte)0x80, (byte)0x8F, (byte)0xBF, (byte)0xBF }, // U+FFFF zero-padded
+        {1, (byte)0xF8, (byte)0xC0, (byte)0x80, (byte)0x80, (byte)0x80, (byte)0x80 },
+        {1, (byte)0xF8, (byte)0x80, (byte)0xC0, (byte)0x80, (byte)0x80, (byte)0x80 },
+        {1, (byte)0xF8, (byte)0x80, (byte)0x80, (byte)0xC1, (byte)0xBF, (byte)0x80 },
+        {1, (byte)0xF8, (byte)0x80, (byte)0x80, (byte)0x9F, (byte)0xC0, (byte)0x80 },
+        {1, (byte)0xF8, (byte)0x80, (byte)0x80, (byte)0x9F, (byte)0x80, (byte)0xC0 },
+    };
+
+   // The first byte is the length of malformed bytes
+    static byte[][] malformed_cesu8 = {
+        // One-byte sequences:
+        {1, (byte)0xFF },
+        {1, (byte)0xC0 },
+        {1, (byte)0x80 },
+
+        {1, (byte)0xFF, (byte)0xFF}, // all ones
+        {1, (byte)0xA0, (byte)0x80}, // 101x first byte first nibble
+
+        // Two-byte sequences:
+        {1, (byte)0xC0, (byte)0x80}, // invalid first byte
+        {1, (byte)0xC1, (byte)0xBF}, // invalid first byte
+        {1, (byte)0xC2, (byte)0x00}, // invalid second byte
+        {1, (byte)0xC2, (byte)0xC0}, // invalid second byte
+        {1, (byte)0xD0, (byte)0x00}, // invalid second byte
+        {1, (byte)0xD0, (byte)0xC0}, // invalid second byte
+        {1, (byte)0xDF, (byte)0x00}, // invalid second byte
+        {1, (byte)0xDF, (byte)0xC0}, // invalid second byte
+
+        // Three-byte sequences
+        {1, (byte)0xE0, (byte)0x80, (byte)0x80},  // 111x first byte first nibble
+        {1, (byte)0xE0, (byte)0x80, (byte)0x80 }, // U+0000 zero-padded
+        {1, (byte)0xE0, (byte)0x81, (byte)0xBF }, // U+007F zero-padded
+        {1, (byte)0xE0, (byte)0x9F, (byte)0xBF }, // U+07FF zero-padded
+
+        {1, (byte)0xE0, (byte)0xC0, (byte)0xBF }, // invalid second byte
+        {2, (byte)0xE0, (byte)0xA0, (byte)0x7F }, // invalid third byte
+        {2, (byte)0xE0, (byte)0xA0, (byte)0xC0 }, // invalid third byte
+        {1, (byte)0xFF, (byte)0xFF, (byte)0xFF }, // all ones
+        {1, (byte)0xE0, (byte)0xC0, (byte)0x80 }, // invalid second byte
+        {1, (byte)0xE0, (byte)0x80, (byte)0xC0 }, // invalid first byte
+        {1, (byte)0xE0, (byte)0x41,},             // invalid second byte & 2 bytes
+
+        // CESU-8 does not have 4, 5, 6 bytes sequenc
+        // Four-byte sequences
+        {1, (byte)0xF0, (byte)0x80, (byte)0x80, (byte)0x80 }, // U+0000 zero-padded
+        {1, (byte)0xF0, (byte)0x80, (byte)0x81, (byte)0xBF }, // U+007F zero-padded
+        {1, (byte)0xF0, (byte)0x80, (byte)0x9F, (byte)0xBF }, // U+007F zero-padded
+        {1, (byte)0xF0, (byte)0x8F, (byte)0xBF, (byte)0xBF }, // U+07FF zero-padded
+
+        {1, (byte)0xFF, (byte)0xFF, (byte)0xFF, (byte)0xFF }, // all ones
+        {1, (byte)0xF0, (byte)0x80, (byte)0x80, (byte)0x80},  // invalid second byte
+        {1, (byte)0xF0, (byte)0xC0, (byte)0x80, (byte)0x80 }, // invalid second byte
+        {1, (byte)0xF0, (byte)41 },                           // invalid second byte
+                                                              // & only 2 bytes
+        {1, (byte)0xF0, (byte)0x90, (byte)0xC0, (byte)0x80 }, // invalid third byte
+        {1, (byte)0xF0, (byte)0x90, (byte)0x80, (byte)0xC0 }, // invalid forth byte
+        {1, (byte)0xF0, (byte)0x90, (byte)0x41 },             // invalid third byte
+                                                              // & 3 bytes input
+
+        {1, (byte)0xF1, (byte)0xC0, (byte)0x80, (byte)0x80 }, // invalid second byte
+        {1, (byte)0xF1, (byte)0x80, (byte)0xC0, (byte)0x80 }, // invalid third byte
+        {1, (byte)0xF1, (byte)0x80, (byte)0x80, (byte)0xC0 }, // invalid forth byte
+        {1, (byte)0xF4, (byte)0x90, (byte)0x80, (byte)0xC0 }, // out-range 4-byte
+        {1, (byte)0xF4, (byte)0xC0, (byte)0x80, (byte)0xC0 }, // out-range 4-byte
+        {1, (byte)0xF5, (byte)0x80, (byte)0x80, (byte)0xC0 }, // out-range 4-byte
+
+        // Five-byte sequences
+        {1, (byte)0xF8, (byte)0x80, (byte)0x80, (byte)0x80, (byte)0x80},  // invalid first byte
+        {1, (byte)0xF8, (byte)0x80, (byte)0x80, (byte)0x80, (byte)0x80 }, // U+0000 zero-padded
+        {1, (byte)0xF8, (byte)0x80, (byte)0x80, (byte)0x81, (byte)0xBF }, // U+007F zero-padded
+        {1, (byte)0xF8, (byte)0x80, (byte)0x80, (byte)0x9F, (byte)0xBF }, // U+07FF zero-padded
+        {1, (byte)0xF8, (byte)0x80, (byte)0x8F, (byte)0xBF, (byte)0xBF }, // U+FFFF zero-padded
 
         {1, (byte)0xF8, (byte)0xC0, (byte)0x80, (byte)0x80, (byte)0x80},
-        {2, (byte)0xF8, (byte)0x80, (byte)0xC0, (byte)0x80, (byte)0x80 },
-        {3, (byte)0xF8, (byte)0x80, (byte)0x80, (byte)0xC1, (byte)0xBF },
-        {4, (byte)0xF8, (byte)0x80, (byte)0x80, (byte)0x9F, (byte)0xC0 },
+        {1, (byte)0xF8, (byte)0x80, (byte)0xC0, (byte)0x80, (byte)0x80 },
+        {1, (byte)0xF8, (byte)0x80, (byte)0x80, (byte)0xC1, (byte)0xBF },
+        {1, (byte)0xF8, (byte)0x80, (byte)0x80, (byte)0x9F, (byte)0xC0 },
 
         // Six-byte sequences
-        {6, (byte)0xFC, (byte)0x80, (byte)0x80, (byte)0x80, (byte)0x80, (byte)0x80 }, // U+0000 zero-padded
-        {6, (byte)0xFC, (byte)0x80, (byte)0x80, (byte)0x80, (byte)0x81, (byte)0xBF }, // U+007F zero-padded
-        {6, (byte)0xFC, (byte)0x80, (byte)0x80, (byte)0x80, (byte)0x9F, (byte)0xBF }, // U+07FF zero-padded
-        {6, (byte)0xFC, (byte)0x80, (byte)0x80, (byte)0x8F, (byte)0xBF, (byte)0xBF }, // U+FFFF zero-padded
+        {1, (byte)0xFC, (byte)0x80, (byte)0x80, (byte)0x80, (byte)0x80, (byte)0x80 }, // U+0000 zero-padded
+        {1, (byte)0xFC, (byte)0x80, (byte)0x80, (byte)0x80, (byte)0x81, (byte)0xBF }, // U+007F zero-padded
+        {1, (byte)0xFC, (byte)0x80, (byte)0x80, (byte)0x80, (byte)0x9F, (byte)0xBF }, // U+07FF zero-padded
+        {1, (byte)0xFC, (byte)0x80, (byte)0x80, (byte)0x8F, (byte)0xBF, (byte)0xBF }, // U+FFFF zero-padded
         {1, (byte)0xF8, (byte)0xC0, (byte)0x80, (byte)0x80, (byte)0x80, (byte)0x80 },
-        {2, (byte)0xF8, (byte)0x80, (byte)0xC0, (byte)0x80, (byte)0x80, (byte)0x80 },
-        {3, (byte)0xF8, (byte)0x80, (byte)0x80, (byte)0xC1, (byte)0xBF, (byte)0x80 },
-        {4, (byte)0xF8, (byte)0x80, (byte)0x80, (byte)0x9F, (byte)0xC0, (byte)0x80 },
-        {5, (byte)0xF8, (byte)0x80, (byte)0x80, (byte)0x9F, (byte)0x80, (byte)0xC0 },
+        {1, (byte)0xF8, (byte)0x80, (byte)0xC0, (byte)0x80, (byte)0x80, (byte)0x80 },
+        {1, (byte)0xF8, (byte)0x80, (byte)0x80, (byte)0xC1, (byte)0xBF, (byte)0x80 },
+        {1, (byte)0xF8, (byte)0x80, (byte)0x80, (byte)0x9F, (byte)0xC0, (byte)0x80 },
+        {1, (byte)0xF8, (byte)0x80, (byte)0x80, (byte)0x9F, (byte)0x80, (byte)0xC0 },
     };
 
-    static void checkMalformed(String csn) throws Exception {
+
+    static void checkMalformed(String csn, byte[][] malformed) throws Exception {
         boolean failed = false;
         System.out.printf("    Check malformed <%s>...%n", csn);
         Charset cs = Charset.forName(csn);
@@ -430,9 +550,12 @@ public class TestUTF8 {
 
     public static void main(String[] args) throws Exception {
         checkRoundtrip("UTF-8");
-        check6ByteSurrs("UTF-8");
-        //compare("UTF-8", "UTF-8-OLD");
-        checkMalformed("UTF-8");
+        check4ByteSurrs("UTF-8");
+        checkMalformed("UTF-8", malformed);
         checkUnderOverflow("UTF-8");
+
+        checkRoundtrip("CESU-8");
+        check6ByteSurrs("CESU-8");
+        checkMalformed("CESU-8", malformed_cesu8);
     }
 }