Implement "replacement" text encoding.

Source/wtf/text/TextCodecReplacement.cpp

+// Copyright 2014 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 "config.h"
+#include "wtf/text/TextCodecReplacement.h"
+
+#include "wtf/PassOwnPtr.h"
+#include "wtf/text/CString.h"
+#include "wtf/text/StringBuffer.h"
+#include "wtf/text/StringBuilder.h"
+#include "wtf/text/WTFString.h"
+#include "wtf/unicode/CharacterNames.h"
+
+namespace WTF {
+
+TextCodecReplacement::TextCodecReplacement()
+    : m_sentEOF(false)
+{
+}
+
+void TextCodecReplacement::registerEncodingNames(EncodingNameRegistrar registrar)
+{
+    // The "replacement" encoding should not be available by name directly,
+    // but is necessary during registration. It will be removed after the
+    // labels are registered.
+    registrar("replacement", "replacement");
+    registrar("csiso2022kr", "replacement");

[eseidel, 2014/04/01 23:12:16]

36 out of 100M pages surveyed had this encoding.

+    registrar("iso-2022-cn", "replacement");

[eseidel, 2014/04/01 23:12:16]

7 off 100M pages.

+    registrar("iso-2022-cn-ext", "replacement");

[eseidel, 2014/04/01 23:12:16]

No data, presumably 0 pages.

+    registrar("iso-2022-kr", "replacement");

[eseidel, 2014/04/01 23:12:16]

I suspect this is lumped in with the 36 above. :)

+}
+
+static PassOwnPtr<TextCodec> newStreamingTextDecoderReplacement(const TextEncoding&, const void*)
+{
+    return adoptPtr(new TextCodecReplacement);
+}
+
+void TextCodecReplacement::registerCodecs(TextCodecRegistrar registrar)
+{
+    registrar("replacement", newStreamingTextDecoderReplacement, 0);
+}
+
+String TextCodecReplacement::decode(const char*, size_t, FlushBehavior, bool, bool& sawError)
+{
+    sawError = true;
+    if (m_sentEOF)
+        return String();
+
+    m_sentEOF = true;
+    return String(&Unicode::replacementCharacter, 1);
+}
+
+} // namespace WTF