UTF-8 detector for pages missing encoding info

third_party/WebKit/Source/core/html/parser/TextResourceDecoder.h

 /*
     Copyright (C) 1999 Lars Knoll (knoll@mpi-hd.mpg.de)
     Copyright (C) 2006 Alexey Proskuryakov (ap@nypop.com)
     Copyright (C) 2006, 2008 Apple Inc. All rights reserved.
 
     This library is free software; you can redistribute it and/or
     modify it under the terms of the GNU Library General Public
     License as published by the Free Software Foundation; either
     version 2 of the License, or (at your option) any later version.
 
@@ skipping 62 matching lines @@
     {
         m_hintEncoding = encoding.name();
     }
 
     void useLenientXMLDecoding() { m_useLenientXMLDecoding = true; }
     bool sawError() const { return m_sawError; }
     size_t checkForBOM(const char*, size_t);
 
 private:
 
-    // TextResourceDecoder does three kind of encoding detection:
+    // TextResourceDecoder does four kinds of encoding detection:
     // 1. By BOM,
     // 2. By Content if |m_contentType| is not |PlainTextContext|
-    //    (e.g. <meta> tag for HTML), and
-    // 3. By detectTextEncoding().
+    //    (e.g. <meta> tag for HTML),
+    // 3. By isUTF8Encoded() to detect if the document
+    //    is of UTF-8, and
+    // 4. By detectTextEncodingUniversal().
     enum EncodingDetectionOption {
-        // Use 1. + 2. + 3.
+        // Use 1. + 2. + 4.
         UseAllAutoDetection,
 
-        // Use 1. + 2.
+        // Use 1. + 2. + 3.
         UseContentAndBOMBasedDetection,
 
         // Use None of them.
         // |m_contentType| must be |PlainTextContent| and
         // |m_encoding| must be UTF8Encoding.
         // This doesn't change encoding based on BOMs, but still processes
         // utf-8 BOMs so that utf-8 BOMs don't appear in the decoded result.
         AlwaysUseUTF8ForText
     };
 
     TextResourceDecoder(const String& mimeType, const WTF::TextEncoding& defaultEncoding, EncodingDetectionOption);
 
     enum ContentType { PlainTextContent, HTMLContent, XMLContent, CSSContent }; // PlainText only checks for BOM.
     static ContentType determineContentType(const String& mimeType);
     static const WTF::TextEncoding& defaultEncoding(ContentType, const WTF::TextEncoding& defaultEncoding);
 
     bool checkForCSSCharset(const char*, size_t, bool& movedDataToBuffer);
     bool checkForXMLCharset(const char*, size_t, bool& movedDataToBuffer);
     void checkForMetaCharset(const char*, size_t);
     bool shouldAutoDetect() const;
+    void detectTextEncoding(const char*, size_t);
 
     ContentType m_contentType;
     WTF::TextEncoding m_encoding;
     OwnPtr<TextCodec> m_codec;
     EncodingSource m_source;
     const char* m_hintEncoding;
     Vector<char> m_buffer;
     bool m_checkedForBOM;
     bool m_checkedForCSSCharset;
     bool m_checkedForXMLCharset;
     bool m_checkedForMetaCharset;
     bool m_useLenientXMLDecoding; // Don't stop on XML decoding errors.
     bool m_sawError;
     EncodingDetectionOption m_encodingDetectionOption;
 
     OwnPtr<HTMLMetaCharsetParser> m_charsetParser;
 };
 
 } // namespace blink
 
 #endif