6831794: charset EUC_TW is 12.6% of the total size of charsets.jar

6229811: Several codepoints in EUC_TW failed in roundtrip conversion
Summary: Re-write EUC_TW charset to address the size and roundtrip issue.
Reviewed-by: alanb

jdk/make/java/nio/Makefile

@@ -827,8 +827,7 @@ CHARSETMAPPING_JARFILE = $(BUILDTOOLJARDIR)/charsetmapping.jar
 
 $(FILES_gensbcs_out): $(GENCSSRC)/SingleByte-X.java $(GENCSSRC)/sbcs
 	@$(prep-target)
-	$(BOOT_JAVA_CMD) -cp $(CHARSETMAPPING_JARFILE) build.tools.charsetmapping.GenerateSBCS \
-		$(GENCSSRC) $(SCS_GEN) sbcs
+	$(BOOT_JAVA_CMD) -jar $(CHARSETMAPPING_JARFILE) $(GENCSSRC) $(SCS_GEN) sbcs
 
 # 
 # Generated file system implementation classes (Unix only)

jdk/make/sun/Makefile

@@ -58,11 +58,12 @@ ifdef OPENJDK
 else
   RENDER_SUBDIR = dcpr
 endif
+# nio need to be compiled before awt to have all charsets ready
 SUBDIRS = jar security javazic misc net audio $(RENDER_SUBDIR) image \
-	  awt splashscreen $(XAWT_SUBDIR) \
+	  nio awt splashscreen $(XAWT_SUBDIR) \
           $(HEADLESS_SUBDIR) $(DGA_SUBDIR) \
 	  font jpeg cmm applet rmi beans $(JDBC_SUBDIR) \
-	  jawt text nio launcher management $(ORG_SUBDIR) \
+	  jawt text launcher management $(ORG_SUBDIR) \
           native2ascii serialver tools jconsole tracing
 
 all build clean clobber::

jdk/make/sun/nio/FILES_java.gmk

@@ -304,7 +304,7 @@ FILES_src = \
 	sun/io/CharToByteMacUkraine.java \
 	sun/io/CharToByteTIS620.java
 
-FILES_gen_extsbcs = \
+FILES_gen_extcs = \
 	sun/nio/cs/ext/IBM037.java \
 	sun/nio/cs/ext/IBM1006.java \
 	sun/nio/cs/ext/IBM1025.java \
@@ -374,6 +374,8 @@ FILES_gen_extsbcs = \
 	sun/nio/cs/ext/MacThai.java \
 	sun/nio/cs/ext/MacTurkish.java \
 	sun/nio/cs/ext/MacUkraine.java \
-	sun/nio/cs/ext/TIS_620.java
+	sun/nio/cs/ext/TIS_620.java \
+        sun/nio/cs/ext/EUC_TWMapping.java
+
+FILES_java = $(FILES_src) $(FILES_gen_extcs)
 
-FILES_java = $(FILES_src) $(FILES_gen_extsbcs)
\ No newline at end of file

jdk/make/sun/nio/Makefile

@@ -61,14 +61,14 @@ endif # PLATFORM
 CHARSETS_JAR = $(LIBDIR)/charsets.jar
 
 # extsbcs
-FILES_genout_extsbcs = $(FILES_gen_extsbcs:%.java=$(GENSRCDIR)/%.java)
+FILES_genout_extcs = $(FILES_gen_extcs:%.java=$(GENSRCDIR)/%.java)
 
 #
 # Rules
 #
 include $(BUILDDIR)/common/Classes.gmk
 
-build: $(FILES_genout_extsbcs) $(CHARSETS_JAR)
+build: $(FILES_genout_extcs) $(CHARSETS_JAR)
 
 #
 # Extra rules to build character converters.
@@ -77,6 +77,7 @@ SERVICE_DESCRIPTION = java.nio.charset.spi.CharsetProvider
 SERVICE_DESCRIPTION_PATH = META-INF/services/$(SERVICE_DESCRIPTION)
 
 GENCSDATASRC = $(BUILDDIR)/tools/CharsetMapping
+GENCSSRCDIR = $(BUILDDIR)/tools/src/build/tools/charsetmapping
 GENCSEXT = $(GENSRCDIR)/sun/nio/cs/ext
 
 FILES_MAP = $(GENCSDATASRC)/sjis0213.map
@@ -86,16 +87,16 @@ CHARSETMAPPING_JARFILE = $(BUILDTOOLJARDIR)/charsetmapping.jar
 $(FILES_DAT): $(FILES_MAP)
 	@$(prep-target)
 	$(BOOT_JAVA_CMD) -jar $(CHARSETMAPPING_JARFILE) \
-		$(FILES_MAP) $(FILES_DAT)
+		$(FILES_MAP) $(FILES_DAT) sjis0213
 
 
-$(FILES_genout_extsbcs): $(GENCSDATASRC)/SingleByte-X.java $(GENCSDATASRC)/extsbcs
+$(FILES_genout_extcs): $(GENCSDATASRC)/SingleByte-X.java $(GENCSDATASRC)/extsbcs
 	@$(prep-target)
 	$(RM) -r $(GENCSEXT)
 	$(MKDIR) -p $(GENCSEXT)
-	$(BOOT_JAVA_CMD) -cp $(CHARSETMAPPING_JARFILE) build.tools.charsetmapping.GenerateSBCS \
-		$(GENCSDATASRC) $(GENCSEXT) extsbcs
-
+	$(BOOT_JAVA_CMD) -jar $(CHARSETMAPPING_JARFILE)	$(GENCSDATASRC) $(GENCSEXT) extsbcs
+	$(BOOT_JAVA_CMD) -jar $(CHARSETMAPPING_JARFILE) $(GENCSDATASRC) $(GENCSEXT) euctw \
+		$(GENCSSRCDIR)/GenerateEUC_TW.java
 
 $(CLASSDESTDIR)/$(SERVICE_DESCRIPTION_PATH): \
   $(SHARE_SRC)/classes/sun/nio/cs/ext/$(SERVICE_DESCRIPTION_PATH)

jdk/make/tools/CharsetMapping/Makefile

@@ -34,7 +34,7 @@ PROGRAM = charsetmapping
 include $(BUILDDIR)/common/Defs.gmk
 
 BUILDTOOL_SOURCE_ROOT = $(BUILDDIR)/tools/src
-BUILDTOOL_MAIN        = $(PKGDIR)/GenerateMapping.java
+BUILDTOOL_MAIN        = $(PKGDIR)/Main.java
 
 #
 # Build tool jar rules.

jdk/make/tools/src/build/tools/charsetmapping/CharsetMapping.java

@@ -36,7 +36,7 @@ import java.util.*;
 
 public class CharsetMapping {
     public final static char UNMAPPABLE_DECODING = '\uFFFD';
-    public final static int  UNMAPPABLE_ENCODING = -1;
+    public final static int  UNMAPPABLE_ENCODING = 0xFFFD;
 
     public static class Entry {
         public int bs;   //byte sequence reps

jdk/make/tools/src/build/tools/charsetmapping/GenerateMapping.java

@@ -27,15 +27,11 @@ package build.tools.charsetmapping;
 
 import java.io.*;
 import java.util.regex.*;
-import build.tools.charsetmapping.GenerateSBCS;
 import static build.tools.charsetmapping.CharsetMapping.*;
 
 public class GenerateMapping {
-    public static void main(String argv[]) throws IOException {
-        if (argv.length < 2) {
-            System.out.println("Usage: java GenerateMapping fMap fDat");
-            System.exit(1);
-        }
+
+    public static void genMapping(String argv[]) throws IOException {
         genDataJIS0213(new FileInputStream(argv[0]),
                        new FileOutputStream(argv[1]));
     }

jdk/make/tools/src/build/tools/charsetmapping/GenerateSBCS.java

@@ -34,11 +34,8 @@ import java.nio.charset.*;
 import static build.tools.charsetmapping.CharsetMapping.*;
 
 public class GenerateSBCS {
-    public static void main(String args[]) throws Exception {
-        if (args.length < 3) {
-            System.err.println("Usage: java GenSBCS srcDir dstDir config");
-            System.exit(1);
-        }
+
+    public static void genSBCS(String args[]) throws Exception {
 
         Scanner s = new Scanner(new File(args[0], args[2]));
         while (s.hasNextLine()) {

jdk/src/share/classes/sun/io/ByteToCharEUC_TW.java

@@ -39,30 +39,11 @@ public class ByteToCharEUC_TW extends ByteToCharConverter
     private final byte G4 = 4;
     private final byte MSB = (byte) 0x80;
     private final byte SS2 = (byte) 0x8E;
-    private final byte P2 = (byte) 0xA2;
-    private final byte P3 = (byte) 0xA3;
-
-    protected final char REPLACE_CHAR = '\uFFFD';
 
     private byte firstByte = 0, state = G0;
-    public static String unicodeCNS2, unicodeCNS3;
-    private static String unicodeCNS4, unicodeCNS5, unicodeCNS6;
-    private static String unicodeCNS7, unicodeCNS15;
-
     private int cnsPlane = 0;
-    private final static EUC_TW nioCoder = new EUC_TW();
 
-    public static String unicodeCNS1 = nioCoder.getUnicodeCNS1();
-
-        static String[] cnsChars = {
-            unicodeCNS2 = nioCoder.getUnicodeCNS2(),
-            unicodeCNS3 = nioCoder.getUnicodeCNS3(),
-            unicodeCNS4 = nioCoder.getUnicodeCNS4(),
-            unicodeCNS5 = nioCoder.getUnicodeCNS5(),
-            unicodeCNS6 = nioCoder.getUnicodeCNS6(),
-            unicodeCNS7 = nioCoder.getUnicodeCNS7(),
-            unicodeCNS15 = nioCoder.getUnicodeCNS15()
-            };
+    private EUC_TW.Decoder dec = (EUC_TW.Decoder)(new EUC_TW().newDecoder());
 
     public ByteToCharEUC_TW() {
     }
@@ -81,6 +62,7 @@ public class ByteToCharEUC_TW extends ByteToCharConverter
     }
 
     public void reset() {
+        dec.reset();
         state = G0;
         firstByte = 0;
         byteOff = charOff = 0;
@@ -95,7 +77,7 @@ public class ByteToCharEUC_TW extends ByteToCharConverter
                ConversionBufferFullException
     {
         int inputSize = 0;
-        char outputChar = (char) 0;
+        char[] c1 = new char[1];
 
         byteOff = inOff;
         charOff = outOff;
@@ -104,11 +86,12 @@ public class ByteToCharEUC_TW extends ByteToCharConverter
         while (byteOff < inEnd) {
             if (charOff >= outEnd)
                 throw new ConversionBufferFullException();
-
+            char[] outputChar = null;
             switch (state) {
             case G0:
                 if ( (input[byteOff] & MSB) == 0) {     // ASCII
-                    outputChar = (char) input[byteOff];
+                    outputChar = c1;
+                    outputChar[0] = (char) input[byteOff];
                 } else if (input[byteOff] == SS2) {     // Codeset 2
                     state = G2;
                 } else {                                // Codeset 1
@@ -119,9 +102,10 @@ public class ByteToCharEUC_TW extends ByteToCharConverter
             case G1:
                 inputSize = 2;
                 if ( (input[byteOff] & MSB) != 0) {     // 2nd byte
-                    cnsPlane = 1;
-                    outputChar = convToUnicode(firstByte,
-                                        input[byteOff], unicodeCNS1);
+                    cnsPlane = 0;
+                    outputChar = dec.toUnicode(firstByte & 0xff,
+                                               input[byteOff] & 0xff,
+                                               cnsPlane);
                 } else {                                // Error
                     badInputLength = 1;
                     throw new MalformedInputException();
@@ -154,9 +138,9 @@ public class ByteToCharEUC_TW extends ByteToCharConverter
                 break;
             case G4:
                 if ( (input[byteOff] & MSB) != 0) {     // 2nd byte
-                        outputChar = convToUnicode(firstByte,
-                                                   input[byteOff],
-                                                   cnsChars[cnsPlane - 2]);
+                    outputChar = dec.toUnicode(firstByte & 0xff,
+                                               input[byteOff] & 0xff,
+                                               cnsPlane - 1);
                 } else {                                // Error
                     badInputLength = 3;
                     throw new MalformedInputException();
@@ -166,21 +150,19 @@ public class ByteToCharEUC_TW extends ByteToCharConverter
                 break;
             }
             byteOff++;
-
-            if (outputChar != (char) 0) {
-                if (outputChar == REPLACE_CHAR) {
-                    if (subMode)                // substitution enabled
-                        outputChar = subChars[0];
-                    else {
+            if (state == G0) {
+                if (outputChar == null) {
+                    if (subMode) {               // substitution enabled
+                        outputChar = c1;
+                        outputChar[0] = subChars[0];
+                    } else {
                         badInputLength = inputSize;
                         throw new UnknownCharacterException();
                     }
                 }
-                output[charOff++] = outputChar;
-                outputChar = 0;
+                output[charOff++] = outputChar[0];
             }
         }
-
         return charOff - outOff;
     }
 
@@ -191,25 +173,4 @@ public class ByteToCharEUC_TW extends ByteToCharConverter
     public String getCharacterEncoding() {
         return "EUC_TW";
     }
-
-    protected char convToUnicode(byte byte1, byte byte2, String table)
-    {
-        int index;
-
-        if ((byte1 & 0xff) < 0xa1 || (byte2 & 0xff) < 0xa1 ||
-            (byte1 & 0xff) > 0xfe || (byte2 & 0xff) > 0xfe)
-            return REPLACE_CHAR;
-        index = (((byte1 & 0xff) - 0xa1) * 94)  + (byte2 & 0xff) - 0xa1;
-        if (index < 0 || index >= table.length())
-            return REPLACE_CHAR;
-
-        // Planes 3 and above containing zero value lead byte
-        // to accommodate surrogates for mappings which decode to a surrogate
-        // pair
-
-        if (this.cnsPlane >= 3)
-           index = (index * 2) + 1;
-
-        return table.charAt(index);
-    }
 }

jdk/src/share/classes/sun/io/CharToByteEUC_TW.java

@@ -33,19 +33,7 @@ import sun.nio.cs.ext.EUC_TW;
 
 public class CharToByteEUC_TW extends CharToByteConverter
 {
-    private final byte MSB = (byte)0x80;
-    private final byte SS2 = (byte) 0x8E;
-    private final byte P2 = (byte) 0xA2;
-    private final byte P3 = (byte) 0xA3;
-
-    private final static EUC_TW nioCoder = new EUC_TW();
-
-    private static String uniTab1 = nioCoder.getUniTab1();
-    private static String uniTab2 = nioCoder.getUniTab2();
-    private static String uniTab3 = nioCoder.getUniTab3();
-    private static String cnsTab1 = nioCoder.getCNSTab1();
-    private static String cnsTab2 = nioCoder.getCNSTab2();
-    private static String cnsTab3 = nioCoder.getCNSTab3();
+    private final EUC_TW.Encoder enc = (EUC_TW.Encoder)(new EUC_TW().newEncoder());
 
     public int flush(byte[] output, int outStart, int outEnd)
         throws MalformedInputException
@@ -59,10 +47,7 @@ public class CharToByteEUC_TW extends CharToByteConverter
     }
 
     public boolean canConvert(char ch){
-        if (((0xFF00 & ch) != 0) && (getNative(ch) != -1)){
-            return true;
-        }
-        return false;
+        return enc.canEncode(ch);
     }
 
     /**
@@ -74,9 +59,8 @@ public class CharToByteEUC_TW extends CharToByteConverter
                ConversionBufferFullException
     {
         int outputSize;
-        byte [] tmpbuf = new byte[4];
+        byte [] tmpbuf = new byte[4];;
         byte [] outputByte;
-
         byteOff = outOff;
 
         //Fixed 4122961 by bringing the charOff++ out to this
@@ -88,7 +72,7 @@ public class CharToByteEUC_TW extends CharToByteConverter
                 outputSize = 1;
                 outputByte[0] = (byte)(input[charOff] & 0x7f);
             } else {
-                outputSize = unicodeToEUC(input[charOff], outputByte);
+                outputSize = enc.toEUC(input[charOff], outputByte);
             }
 
             if (outputSize == -1) {
@@ -112,7 +96,6 @@ public class CharToByteEUC_TW extends CharToByteConverter
 
     }
 
-
     /**
      * returns the maximum number of bytes needed to convert a char
      */
@@ -120,111 +103,10 @@ public class CharToByteEUC_TW extends CharToByteConverter
         return 4;
     }
 
-
     /**
      * Return the character set ID
      */
     public String getCharacterEncoding() {
         return "EUC_TW";
     }
-
-
-    protected int getNative(char unicode) {
-        int  i,
-             cns;       // 2 chars in CNS table make 1 CNS code
-
-        if (unicode < UniTab2[0]) {
-            if ((i = searchTab(unicode, UniTab1)) == -1)
-                return -1;
-            cns = (CNSTab1[2*i] << 16) + CNSTab1[2*i+1];
-            return cns;
-        } else  if (unicode < UniTab3[0]) {
-            if ((i = searchTab(unicode, UniTab2)) == -1)
-                return -1;
-            cns = (CNSTab2[2*i] << 16) + CNSTab2[2*i+1];
-            return cns;
-        } else {
-            if ((i = searchTab(unicode, UniTab3)) == -1)
-                return -1;
-            cns = (CNSTab3[2*i] << 16) + CNSTab3[2*i+1];
-            return cns;
-        }
-    }
-
-    protected int searchTab(char code, char [] table) {
-        int     i = 0, l, h;
-
-        for (l = 0, h = table.length - 1; l < h; ) {
-                if (table[l] == code) {
-                        i = l;
-                        break;
-                }
-                if (table[h] == code) {
-                        i = h;
-                        break;
-                }
-                i = (l + h) / 2;
-                if (table[i] == code)
-                        break;
-                if (table[i] < code)
-                        l = i + 1;
-                else    h = i - 1;
-        }
-        if (code == table[i]) {
-            return i;
-        } else {
-            return -1;
-        }
-    }
-
-
-    private int unicodeToEUC(char unicode, byte ebyte[]) {
-        int cns = getNative(unicode);
-
-        if ((cns >> 16) == 0x01) {      // Plane 1
-            ebyte[0] = (byte) (((cns & 0xff00) >> 8) | MSB);
-            ebyte[1] = (byte) ((cns & 0xff) | MSB);
-            return 2;
-        }
-
-        byte cnsPlane = (byte)(cns >> 16);
-        if (cnsPlane >= (byte)0x02) {   // Plane 2
-            ebyte[0] = SS2;
-            ebyte[1] = (byte) (cnsPlane | (byte)0xA0);
-            ebyte[2] = (byte) (((cns & 0xff00) >> 8) | MSB);
-            ebyte[3] = (byte) ((cns & 0xff) | MSB);
-            return 4;
-        }
-
-        return -1;
-    }
-
-    protected int unicodeToEUC(char unicode) {
-        if (unicode <= 0x7F) { // ASCII falls into EUC_TW CS0
-            return unicode;
-        }
-
-        int cns = getNative(unicode);
-        int plane = cns >> 16;
-        int euc = (cns & 0x0000FFFF) | 0x00008080;
-
-        if (plane == 1) {
-            return euc;
-        } else if (plane == 2) {
-            return ((SS2 << 24) & 0xFF000000) | ((P2 << 16) & 0x00FF0000) |
-                euc;
-        } else if (plane == 3) {
-            return ((SS2 << 24) & 0xFF000000) | ((P3 << 16) & 0x00FF0000) |
-                euc;
-        }
-
-        return -1;
-    }
-
-    private char [] UniTab1 = uniTab1.toCharArray();
-    private char [] UniTab2 = uniTab2.toCharArray();
-    private char [] UniTab3 = uniTab3.toCharArray();
-    private char [] CNSTab1 = cnsTab1.toCharArray();
-    private char [] CNSTab2 = cnsTab2.toCharArray();
-    private char [] CNSTab3 = cnsTab3.toCharArray();
 }

jdk/src/share/classes/sun/nio/cs/ext/ISO2022.java

@@ -389,8 +389,8 @@ abstract class ISO2022
     protected static class Encoder extends CharsetEncoder {
         private final Surrogate.Parser sgp = new Surrogate.Parser();
         private final byte SS2 = (byte)0x8e;
-        private final byte P2 = (byte)0xA2;
-        private final byte P3 = (byte)0xA3;
+        private final byte PLANE2 = (byte)0xA2;
+        private final byte PLANE3 = (byte)0xA3;
         private final byte MSB = (byte)0x80;
 
         protected final byte maximumDesignatorLength = 4;
@@ -460,32 +460,32 @@ abstract class ISO2022
                 ebyte[index++] = (byte)(convByte[0] & 0x7f);
                 ebyte[index++] = (byte)(convByte[1] & 0x7f);
             } else {
-                if((convByte[0] == SS2) && (convByte[1] == P2)) {
-                    if (!SS2DesDefined) {
-                        newSS2DesDefined = true;
-                        ebyte[0] = ISO_ESC;
-                        tmpByte = SS2Desig.getBytes();
-                        System.arraycopy(tmpByte, 0, ebyte, 1, tmpByte.length);
-                        index = tmpByte.length+1;
-                    }
-                    ebyte[index++] = ISO_ESC;
-                    ebyte[index++] = ISO_SS2_7;
-                    ebyte[index++] = (byte)(convByte[2] & 0x7f);
-                    ebyte[index++] = (byte)(convByte[3] & 0x7f);
-                }
-                if((convByte[0] == SS2)&&(convByte[1] == 0xA3))
-                {
-                    if(!SS3DesDefined){
-                        newSS3DesDefined = true;
-                        ebyte[0] = ISO_ESC;
-                        tmpByte = SS3Desig.getBytes();
-                        System.arraycopy(tmpByte, 0, ebyte, 1, tmpByte.length);
-                        index = tmpByte.length+1;
+                if(convByte[0] == SS2) {
+                    if (convByte[1] == PLANE2) {
+                        if (!SS2DesDefined) {
+                            newSS2DesDefined = true;
+                            ebyte[0] = ISO_ESC;
+                            tmpByte = SS2Desig.getBytes();
+                            System.arraycopy(tmpByte, 0, ebyte, 1, tmpByte.length);
+                            index = tmpByte.length+1;
+                        }
+                        ebyte[index++] = ISO_ESC;
+                        ebyte[index++] = ISO_SS2_7;
+                        ebyte[index++] = (byte)(convByte[2] & 0x7f);
+                        ebyte[index++] = (byte)(convByte[3] & 0x7f);
+                    } else if (convByte[1] == PLANE3) {
+                        if(!SS3DesDefined){
+                            newSS3DesDefined = true;
+                            ebyte[0] = ISO_ESC;
+                            tmpByte = SS3Desig.getBytes();
+                            System.arraycopy(tmpByte, 0, ebyte, 1, tmpByte.length);
+                            index = tmpByte.length+1;
+                        }
+                        ebyte[index++] = ISO_ESC;
+                        ebyte[index++] = ISO_SS3_7;
+                        ebyte[index++] = (byte)(convByte[2] & 0x7f);
+                        ebyte[index++] = (byte)(convByte[3] & 0x7f);
                     }
-                    ebyte[index++] = ISO_ESC;
-                    ebyte[index++] = ISO_SS3_7;
-                    ebyte[index++] = (byte)(convByte[2] & 0x7f);
-                    ebyte[index++] = (byte)(convByte[3] & 0x7f);
                 }
             }
             return index;

jdk/src/share/classes/sun/nio/cs/ext/ISO2022_CN.java

@@ -105,17 +105,19 @@ public class ISO2022_CN
         private char cnsDecode(byte byte1, byte byte2, byte SS) {
             byte1 |= MSB;
             byte2 |= MSB;
-            if (SS == ISO_SS2_7) {
-                return cnsDecoder.convToUnicode(byte1, byte2,
-                                                cnsDecoder.unicodeCNS2);
-
-            } else { //SS == ISO_SS3_7
-                char[] outSurr = cnsDecoder.convToSurrogate(byte1, byte2,
-                                                            cnsDecoder.unicodeCNS3);
-                if (outSurr == null || outSurr[0] != '\u0000')
-                    return REPLACE_CHAR;
-                return outSurr[1];
-            }
+            int p = 0;
+            if (SS == ISO_SS2_7)
+                p = 1;    //plane 2, index -- 1
+            else if (SS == ISO_SS3_7)
+                p = 2;    //plane 3, index -- 2
+            else
+                return REPLACE_CHAR;  //never happen.
+            char[] ret = cnsDecoder.toUnicode(byte1 & 0xff,
+                                              byte2 & 0xff,
+                                              p);
+            if (ret == null || ret.length == 2)
+                return REPLACE_CHAR;
+            return ret[0];
         }
 
         private char SODecode(byte byte1, byte byte2, byte SOD) {
@@ -125,9 +127,12 @@ public class ISO2022_CN
                 return gb2312Decoder.decodeDouble(byte1 & 0xff,
                                                   byte2 & 0xff);
             } else {    // SOD == SODesigCNS
-                return cnsDecoder.convToUnicode(byte1,
-                                                byte2,
-                                                cnsDecoder.unicodeCNS1);
+                char[] ret = cnsDecoder.toUnicode(byte1 & 0xff,
+                                                  byte2 & 0xff,
+                                                  0);
+                if (ret == null)
+                    return REPLACE_CHAR;
+                return ret[0];
             }
         }
 

jdk/src/solaris/classes/sun/awt/motif/X11CNS11643.java

 /*
- * Copyright 2001-2005 Sun Microsystems, Inc.  All Rights Reserved.
+ * Copyright 2001-2008 Sun Microsystems, Inc.  All Rights Reserved.
  * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
  *
  * This code is free software; you can redistribute it and/or modify it
@@ -66,16 +66,19 @@ public abstract class X11CNS11643 extends Charset {
             super(cs);
             this.plane = plane;
         }
+
+        private byte[] bb = new byte[4];
         public boolean canEncode(char c) {
             if (c <= 0x7F) {
                 return false;
             }
-            int p = getNative(c) >> 16;
-            if (p == 1 && plane == 0 ||
-                p == 2 && plane == 2 ||
-                p == 3 && plane == 3)
-                return true;
-            return false;
+            int nb = toEUC(c, bb);
+            if (nb == -1)
+                return false;
+            int p = 0;
+            if (nb == 4)
+                p = (bb[1] & 0xff) - 0xa0;
+            return (p == plane);
         }
 
         public boolean isLegalReplacement(byte[] repl) {
@@ -93,19 +96,26 @@ public abstract class X11CNS11643 extends Charset {
             try {
                 while (sp < sl) {
                     char c = sa[sp];
-                    if (c >= '\uFFFE' || c <= '\u007f')
-                        return CoderResult.unmappableForLength(1);
-                    int cns = getNative(c);
-                    int p = cns >> 16;
-                    if (p == 1 && plane == 0 ||
-                        p == 2 && plane == 2 ||
-                        p == 3 && plane == 3) {
-                        if (dl - dp < 2)
-                            return CoderResult.OVERFLOW;
-                        da[dp++] = (byte) ((cns  >> 8) & 0x7f);
-                        da[dp++] = (byte) (cns & 0x7f);
-                        sp++;
-                        continue;
+                    if ( c > '\u007f'&& c < '\uFFFE') {
+                        int nb = toEUC(c, bb);
+                        if (nb != -1) {
+                            int p = 0;
+                            if (nb == 4)
+                                p = (bb[1] & 0xff) - 0xa0;
+                            if (p == plane) {
+                                if (dl - dp < 2)
+                                    return CoderResult.OVERFLOW;
+                                if (nb == 2) {
+                                    da[dp++] = (byte)(bb[0] & 0x7f);
+                                    da[dp++] = (byte)(bb[1] & 0x7f);
+                                } else {
+                                    da[dp++] = (byte)(bb[2] & 0x7f);
+                                    da[dp++] = (byte)(bb[3] & 0x7f);
+                                }
+                                sp++;
+                                continue;
+                            }
+                        }
                     }
                     return CoderResult.unmappableForLength(1);
                 }
@@ -118,23 +128,17 @@ public abstract class X11CNS11643 extends Charset {
     }
 
     private class Decoder extends EUC_TW.Decoder {
+        int plane;
         private String table;
         protected Decoder(Charset cs, int plane) {
             super(cs);
-            switch (plane) {
-            case 0:
-                table = unicodeCNS1;
-                break;
-            case 2:
-                table = unicodeCNS2;
-                break;
-            case 3:
-                table = unicodeCNS3;
-                break;
-            default:
+            if (plane == 0)
+                this.plane = plane;
+            else if (plane == 2 || plane == 3)
+                this.plane = plane - 1;
+            else
                 throw new IllegalArgumentException
                     ("Only planes 1, 2, and 3 supported");
-            }
         }
 
         //we only work on array backed buffer.
@@ -142,33 +146,26 @@ public abstract class X11CNS11643 extends Charset {
             byte[] sa = src.array();
             int sp = src.arrayOffset() + src.position();
             int sl = src.arrayOffset() + src.limit();
-            assert (sp <= sl);
-            sp = (sp <= sl ? sp : sl);
 
             char[] da = dst.array();
             int dp = dst.arrayOffset() + dst.position();
             int dl = dst.arrayOffset() + dst.limit();
-            assert (dp <= dl);
-            dp = (dp <= dl ? dp : dl);
 
             try {
                 while (sp < sl) {
                     if ( sl - sp < 2) {
                         return CoderResult.UNDERFLOW;
                     }
-                    byte b1 = sa[sp];
-                    byte b2 = sa[sp + 1];
-                    char c = convToUnicode((byte)(b1 | 0x80),
-                                           (byte)(b2 | 0x80),
-                                           table);
-                    if (c == replacement().charAt(0)
-                        //to keep the compatibility with b2cX11CNS11643
-                        /*|| c == '\u0000'*/) {
+                    int b1 = (sa[sp] & 0xff) | 0x80;
+                    int b2 = (sa[sp + 1] & 0xff) | 0x80;
+                    char[] cc = toUnicode(b1, b2, plane);
+                    // plane3 has non-bmp characters(added), x11cnsp3
+                    // however does not support them
+                    if (cc == null || cc.length == 2)
                         return CoderResult.unmappableForLength(2);
-                    }
                     if (dl - dp < 1)
                         return CoderResult.OVERFLOW;
-                    da[dp++] = c;
+                    da[dp++] = cc[0];
                     sp +=2;
                 }
                 return CoderResult.UNDERFLOW;

jdk/test/sun/nio/cs/TestISO2022CNDecoder.java

@@ -22,7 +22,7 @@
  */
 
 /* @test
-   @bug 4779029 4924625 6392664
+   @bug 4779029 4924625 6392664 6730652
    @summary Test decoding of various permutations of valid ISO-2022-CN byte sequences
  */
 
@@ -387,6 +387,12 @@ public class TestISO2022CNDecoder
     };
     private static  CoderResult test15_result = CoderResult.unmappableForLength(4);
 
+    private static boolean encodeTest6730652 () throws Exception {
+        //sample p3 codepoints
+        String strCNSP3 = "\u4e28\u4e36\u4e3f\u4e85\u4e05\u4e04\u5369\u53b6\u4e2a\u4e87\u4e49\u51e2\u56b8\u56b9\u56c4\u8053\u92b0";
+        return strCNSP3.equals(new String(strCNSP3.getBytes("x-ISO-2022-CN-CNS"), "x-ISO-2022-CN-CNS"));
+    }
+
     /**
      * Main program to test ISO2022CN conformance
      *
@@ -430,6 +436,9 @@ public class TestISO2022CNDecoder
         pass &= decodeTest(test13_bytes, test13_chars, "escapes13");
         pass &= decodeResultTest(test14_bytes, test14_result, "escapes14");
         pass &= decodeResultTest(test15_bytes, test15_result, "escapes15");
+
+        pass &= encodeTest6730652 ();
+
         // PASS/FAIL status is what the whole thing is about.
         //
         if (! pass) {