Index: src/java/org/apache/lucene/analysis/CharArraySet.java
===================================================================
--- src/java/org/apache/lucene/analysis/CharArraySet.java	(revision 883650)
+++ src/java/org/apache/lucene/analysis/CharArraySet.java	(working copy)
@@ -131,8 +131,11 @@
    */
   public boolean add(char[] text) {
     if (ignoreCase)
-      for(int i=0;i<text.length;i++)
-        text[i] = Character.toLowerCase(text[i]);
+      for(int i=0;i<text.length;){
+        i += Character.toChars(
+              Character.toLowerCase(
+                  Character.codePointAt(text, i)), text, i);
+      }
     int slot = getSlot(text, 0, text.length);
     if (entries[slot] != null) return false;
     entries[slot] = text;
@@ -149,9 +152,11 @@
     if (len != text2.length)
       return false;
     if (ignoreCase) {
-      for(int i=0;i<len;i++) {
-        if (Character.toLowerCase(text1[off+i]) != text2[i])
+      for(int i=0;i<len;) {
+        final int codePointAt = Character.codePointAt(text1, off+i);
+        if (Character.toLowerCase(codePointAt) != Character.codePointAt(text2, i))
           return false;
+        i += Character.charCount(codePointAt); 
       }
     } else {
       for(int i=0;i<len;i++) {
@@ -167,9 +172,11 @@
     if (len != text2.length)
       return false;
     if (ignoreCase) {
-      for(int i=0;i<len;i++) {
-        if (Character.toLowerCase(text1.charAt(i)) != text2[i])
+      for(int i=0;i<len;) {
+        final int codePointAt = Character.codePointAt(text1, i);
+        if (Character.toLowerCase(codePointAt) != Character.codePointAt(text2, i))
           return false;
+        i += Character.charCount(codePointAt);
       }
     } else {
       for(int i=0;i<len;i++) {
@@ -179,7 +186,9 @@
     }
     return true;
   }
+  
 
+
   private void rehash() {
     final int newSize = 2*entries.length;
     char[][] oldEntries = entries;
@@ -198,8 +207,10 @@
     int code = 0;
     final int stop = offset + len;
     if (ignoreCase) {
-      for (int i=offset; i<stop; i++) {
-        code = code*31 + Character.toLowerCase(text[i]);
+      for (int i=offset; i<stop;) {
+        final int codePointAt = Character.codePointAt(text, i);
+        code = code*31 + Character.toLowerCase(codePointAt);
+        i += Character.charCount(codePointAt);
       }
     } else {
       for (int i=offset; i<stop; i++) {
@@ -213,8 +224,10 @@
     int code = 0;
     int len = text.length();
     if (ignoreCase) {
-      for (int i=0; i<len; i++) {
-        code = code*31 + Character.toLowerCase(text.charAt(i));
+      for (int i=0; i<len;) {
+        int codePointAt = Character.codePointAt(text, i);
+        code = code*31 + Character.toLowerCase(codePointAt);
+        i += Character.charCount(codePointAt);
       }
     } else {
       for (int i=0; i<len; i++) {
Index: src/test/org/apache/lucene/analysis/TestCharArraySet.java
===================================================================
--- src/test/org/apache/lucene/analysis/TestCharArraySet.java	(revision 883650)
+++ src/test/org/apache/lucene/analysis/TestCharArraySet.java	(working copy)
@@ -175,4 +175,53 @@
       // expected
     }
   }
+  
+  public void testSupplementaryChars() {
+    // for reference see
+    // http://unicode.org/cldr/utility/list-unicodeset.jsp?a=[[%3ACase_Sensitive%3DTrue%3A]%26[^[\u0000-\uFFFF]]]&esc=on
+    String[] upperArr = new String[] {"Abc\ud801\udc1c",
+        "\ud801\udc1c\ud801\udc1cCDE", "A\ud801\udc1cB"};
+    String[] lowerArr = new String[] {"abc\ud801\udc44",
+        "\ud801\udc44\ud801\udc44cde", "a\ud801\udc44b"};
+    CharArraySet set = new CharArraySet(Arrays.asList(TEST_STOP_WORDS), true);
+    for (String upper : upperArr) {
+      set.add(upper);
+    }
+    for (int i = 0; i < upperArr.length; i++) {
+      assertTrue(set.contains(upperArr[i]));
+      assertTrue(set.contains(lowerArr[i]));
+    }
+    set = new CharArraySet(Arrays.asList(TEST_STOP_WORDS), false);
+    for (String upper : upperArr) {
+      set.add(upper);
+    }
+    for (int i = 0; i < upperArr.length; i++) {
+      assertTrue(set.contains(upperArr[i]));
+      assertFalse(set.contains(lowerArr[i]));
+    }
+  }
+  
+  public void testSingleHighSurrogate() {
+    String[] upperArr = new String[] {"ABC\uD800", "ABC\uD800EfG", "\uD800EfG",
+        "\uD800\ud801\udc1cB"};
+    
+    String[] lowerArr = new String[] {"abc\uD800", "abc\uD800efg", "\uD800efg",
+        "\uD800\ud801\udc44b"};
+    CharArraySet set = new CharArraySet(Arrays.asList(TEST_STOP_WORDS), true);
+    for (String upper : upperArr) {
+      set.add(upper);
+    }
+    for (int i = 0; i < upperArr.length; i++) {
+      assertTrue(set.contains(upperArr[i]));
+      assertTrue(set.contains(lowerArr[i]));
+    }
+    set = new CharArraySet(Arrays.asList(TEST_STOP_WORDS), false);
+    for (String upper : upperArr) {
+      set.add(upper);
+    }
+    for (int i = 0; i < upperArr.length; i++) {
+      assertTrue(set.contains(upperArr[i]));
+      assertFalse(set.contains(lowerArr[i]));
+    }
+  }
 }