Add support for surrogates when serializing and deserializing for native ports

runtime/vm/unicode.h

Index: runtime/vm/unicode.h
diff --git a/runtime/vm/unicode.h b/runtime/vm/unicode.h
index 03a4b29d879898c2c83f06aac4a175d4729db209..6e6e1c038abbc1d50ce332c26d7b3340f22478a6 100644
--- a/runtime/vm/unicode.h
+++ b/runtime/vm/unicode.h
@@ -1,4 +1,4 @@
-// Copyright (c) 2011, the Dart project authors.  Please see the AUTHORS file
+// Copyright (c) 2012, the Dart project authors.  Please see the AUTHORS file
 // for details. All rights reserved. Use of this source code is governed by a
 // BSD-style license that can be found in the LICENSE file.
 
@@ -29,7 +29,8 @@ class Utf8 : AllStatic {
                                  intptr_t array_len,
                                  Type* type);
 
-  // Returns true if 'utf8_array' is a valid UTF-8 string.
+  // Returns true if 'utf8_array' is a valid UTF-8 string. UTF-8 encoded UTF-16
+  // surrogate code units are allowed.
   static bool IsValid(const uint8_t* utf8_array, intptr_t array_len);
 
   static intptr_t Length(int32_t ch);
@@ -41,7 +42,6 @@ class Utf8 : AllStatic {
   static intptr_t Decode(const uint8_t* utf8_array,
                          intptr_t array_len,
                          int32_t* ch);
-
   static bool DecodeToLatin1(const uint8_t* utf8_array,
                              intptr_t array_len,
                              uint8_t* dst,
@@ -73,6 +73,36 @@ class Utf16 : AllStatic {
 
   static const int32_t kSurrogateOffset = (0x10000 - (0xD800 << 10) - 0xDC00);
 
+  class CodePointIterator {
+   public:
+    CodePointIterator(const uint16_t* utf16_array, intptr_t array_len)
+        : utf16_array_(utf16_array),
+          array_len_(array_len),
+          index_(-1),
+          ch_(-1) {
+    }
+
+    int32_t Current() const {
+      ASSERT(index_ >= 0);
+      ASSERT(index_ < array_len_);
+      return ch_;
+    }
+
+    bool Next();
+
+    void Reset() {
+      index_ = -1;
+      ch_ = -1;
+    }
+
+   private:
+    const uint16_t* utf16_array_;
+    intptr_t array_len_;
+    intptr_t index_;
+    int32_t ch_;
+    DISALLOW_IMPLICIT_CONSTRUCTORS(CodePointIterator);
+  };
+
   // Returns the length of the code point in UTF-16 code units.
   static intptr_t Length(int32_t ch) {
     return (ch <= kMaxBmpCodepoint) ? 1 : 2;