Make UTF-8 encoding of unpaired surrogates match Encoding standard

Source/wtf/text/TextCodecUTF8.cpp

 /*
  * Copyright (C) 2004, 2006, 2008, 2011 Apple Inc. All rights reserved.
  *
  * Redistribution and use in source and binary forms, with or without
  * modification, are permitted provided that the following conditions
  * are met:
  * 1. Redistributions of source code must retain the above copyright
  *    notice, this list of conditions and the following disclaimer.
  * 2. Redistributions in binary form must reproduce the above copyright
  *    notice, this list of conditions and the following disclaimer in the
@@ skipping 418 matching lines @@
     // Non-BMP characters take two UTF-16 code units and can take up to 4 bytes (2x).
     if (length > numeric_limits<size_t>::max() / 3)
         CRASH();
     Vector<uint8_t> bytes(length * 3);
 
     size_t i = 0;
     size_t bytesWritten = 0;
     while (i < length) {
         UChar32 character;
         U16_NEXT(characters, i, length, character);
+        // U16_NEXT will simply emit a surrogate code point if an unmatched surrogate
+        // is encountered; we must convert it to a U+FFFD (REPLACEMENT CHARACTER) here.
+        if (0xD800 <= character && character <= 0xDFFF)
+            character = replacementCharacter;
         U8_APPEND_UNSAFE(bytes.data(), bytesWritten, character);
     }
 
     return CString(reinterpret_cast<char*>(bytes.data()), bytesWritten);
 }
 
 CString TextCodecUTF8::encode(const UChar* characters, size_t length, UnencodableHandling)
 {
     return encodeCommon(characters, length);
 }
 
 CString TextCodecUTF8::encode(const LChar* characters, size_t length, UnencodableHandling)
 {
     return encodeCommon(characters, length);
 }
 
 } // namespace WTF