If we can't read a unicode character, write the standard "unknown" (0xFFFD) c...

base/utf_offset_string_conversions.cc

 // Copyright (c) 2009 The Chromium Authors. All rights reserved.
 // Use of this source code is governed by a BSD-style license that can be
 // found in the LICENSE file.
 
 #include "base/utf_offset_string_conversions.h"
 
 #include "base/string_piece.h"
 #include "base/utf_string_conversion_utils.h"
 
 using base::PrepareForUTF16Or32Output;
@@ skipping 18 matching lines @@
   // ICU requires 32-bit numbers.
   bool success = true;
   int32 src_len32 = static_cast<int32>(src_len);
   for (int32 i = 0; i < src_len32; i++) {
     uint32 code_point;
     size_t original_i = i;
     size_t chars_written = 0;
     if (ReadUnicodeCharacter(src, src_len32, &i, &code_point)) {
       chars_written = WriteUnicodeCharacter(code_point, output);
     } else {
-      // TODO(jungshik): consider adding 'Replacement character' (U+FFFD)
-      // in place of an invalid codepoint.
+      chars_written = WriteUnicodeCharacter(0xFFFD, output);
       success = false;
     }
     if ((output_offset != std::wstring::npos) &&
         (*offset_for_adjustment > original_i)) {
       // NOTE: ReadUnicodeCharacter() adjusts |i| to point _at_ the last
       // character read, not after it (so that incrementing it in the loop
       // increment will place it at the right location), so we need to account
       // for that in determining the amount that was read.
       if (*offset_for_adjustment <= static_cast<size_t>(i))
         output_offset = std::wstring::npos;
@@ skipping 62 matching lines @@
 
 std::wstring UTF16ToWideAndAdjustOffset(const string16& utf16,
                                         size_t* offset_for_adjustment) {
   std::wstring ret;
   UTF16ToWideAndAdjustOffset(utf16.data(), utf16.length(), &ret,
                              offset_for_adjustment);
   return ret;
 }
 
 #endif  // defined(WCHAR_T_IS_UTF32)