Revert of UTF-8 detector for pages missing encoding info

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

 /*
     Copyright (C) 1999 Lars Knoll (knoll@mpi-hd.mpg.de)
     Copyright (C) 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2012 Apple Inc. All rights reserved.
     Copyright (C) 2005, 2006, 2007 Alexey Proskuryakov (ap@nypop.com)
 
     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 10 matching lines @@
 
 #include "core/html/parser/TextResourceDecoder.h"
 
 #include "core/HTMLNames.h"
 #include "core/dom/DOMImplementation.h"
 #include "core/html/parser/HTMLMetaCharsetParser.h"
 #include "platform/text/TextEncodingDetector.h"
 #include "wtf/StringExtras.h"
 #include "wtf/text/TextCodec.h"
 #include "wtf/text/TextEncodingRegistry.h"
-#include "wtf/text/UTF8.h"
 
 using namespace WTF;
 
 namespace blink {
 
 using namespace HTMLNames;
 
 const int minimumLengthOfXMLDeclaration = 8;
 
 static inline bool bytesEqual(const char* p, char b0, char b1, char b2, char b3, char b4)
@@ skipping 301 matching lines @@
 
     if (!m_charsetParser->checkForMetaCharset(data, length))
         return;
 
     setEncoding(m_charsetParser->encoding(), EncodingFromMetaTag);
     m_charsetParser.clear();
     m_checkedForMetaCharset = true;
     return;
 }
 
-String TextResourceDecoder::decode(const char* data, size_t len)
-{
-    size_t lengthOfBOM = 0;
-    if (!m_checkedForBOM)
-        lengthOfBOM = checkForBOM(data, len);
-
-    bool movedDataToBuffer = false;
-
-    if (m_contentType == CSSContent && !m_checkedForCSSCharset) {
-        if (!checkForCSSCharset(data, len, movedDataToBuffer))
-            return emptyString();
-    }
-
-    // We check XML declaration in HTML content only if there is enough data available
-    if (((m_contentType == HTMLContent && len >= minimumLengthOfXMLDeclaration) || m_contentType == XMLContent) && !m_checkedForXMLCharset) {
-        if (!checkForXMLCharset(data, len, movedDataToBuffer))
-            return emptyString();
-    }
-
-    const char* dataForDecode = data + lengthOfBOM;
-    size_t lengthForDecode = len - lengthOfBOM;
-
-    if (!m_buffer.isEmpty()) {
-        if (!movedDataToBuffer) {
-            size_t oldSize = m_buffer.size();
-            m_buffer.grow(oldSize + len);
-            memcpy(m_buffer.data() + oldSize, data, len);
-        }
-
-        dataForDecode = m_buffer.data() + lengthOfBOM;
-        lengthForDecode = m_buffer.size() - lengthOfBOM;
-    }
-
-    if (m_contentType == HTMLContent && !m_checkedForMetaCharset)
-        checkForMetaCharset(dataForDecode, lengthForDecode);
-
-    detectTextEncoding(data, len);
-
-    ASSERT(m_encoding.isValid());
-
-    if (!m_codec)
-        m_codec = newTextCodec(m_encoding);
-
-    String result = m_codec->decode(dataForDecode, lengthForDecode, DoNotFlush, m_contentType == XMLContent && !m_useLenientXMLDecoding, m_sawError);
-
-    m_buffer.clear();
-    return result;
-}
-
-// We use the encoding detector in following cases:
+// We use the encoding detector in two cases:
 //   1. Encoding detector is turned ON and no other encoding source is
 //      available (that is, it's DefaultEncoding).
 //   2. Encoding detector is turned ON and the encoding is set to
 //      the encoding of the parent frame, which is also auto-detected.
 //   Note that condition #2 is NOT satisfied unless parent-child frame
 //   relationship is compliant to the same-origin policy. If they're from
 //   different domains, |m_source| would not be set to EncodingFromParentFrame
 //   in the first place.
-void TextResourceDecoder::detectTextEncoding(const char* data, size_t len)
-{
-    if (!shouldDetectEncoding())
-        return;
-
-    if (WTF::Unicode::isUTF8andNotASCII(data, len)) {
-        setEncoding(UTF8Encoding(), EncodingFromContentSniffing);
-        return;
-    }
-    if (m_encodingDetectionOption == UseAllAutoDetection) {
-        WTF::TextEncoding detectedEncoding;
-        if (detectTextEncodingUniversal(data, len, m_hintEncoding, &detectedEncoding))
-            setEncoding(detectedEncoding, EncodingFromContentSniffing);
-    }
-}
-
-bool TextResourceDecoder::shouldDetectEncoding() const
+bool TextResourceDecoder::shouldAutoDetect() const
 {
     // Just checking m_hintEncoding suffices here because it's only set
     // in setHintEncoding when the source is AutoDetectedEncoding.
-    return m_source == DefaultEncoding || (m_source == EncodingFromParentFrame && m_hintEncoding);
+    return m_encodingDetectionOption == UseAllAutoDetection
+        && (m_source == DefaultEncoding || (m_source == EncodingFromParentFrame && m_hintEncoding));
+}
+
+String TextResourceDecoder::decode(const char* data, size_t len)
+{
+    size_t lengthOfBOM = 0;
+    if (!m_checkedForBOM)
+        lengthOfBOM = checkForBOM(data, len);
+
+    bool movedDataToBuffer = false;
+
+    if (m_contentType == CSSContent && !m_checkedForCSSCharset) {
+        if (!checkForCSSCharset(data, len, movedDataToBuffer))
+            return emptyString();
+    }
+
+    // We check XML declaration in HTML content only if there is enough data available
+    if (((m_contentType == HTMLContent && len >= minimumLengthOfXMLDeclaration) || m_contentType == XMLContent) && !m_checkedForXMLCharset) {
+        if (!checkForXMLCharset(data, len, movedDataToBuffer))
+            return emptyString();
+    }
+
+    const char* dataForDecode = data + lengthOfBOM;
+    size_t lengthForDecode = len - lengthOfBOM;
+
+    if (!m_buffer.isEmpty()) {
+        if (!movedDataToBuffer) {
+            size_t oldSize = m_buffer.size();
+            m_buffer.grow(oldSize + len);
+            memcpy(m_buffer.data() + oldSize, data, len);
+        }
+
+        dataForDecode = m_buffer.data() + lengthOfBOM;
+        lengthForDecode = m_buffer.size() - lengthOfBOM;
+    }
+
+    if (m_contentType == HTMLContent && !m_checkedForMetaCharset)
+        checkForMetaCharset(dataForDecode, lengthForDecode);
+
+    if (shouldAutoDetect()) {
+        WTF::TextEncoding detectedEncoding;
+        if (detectTextEncoding(data, len, m_hintEncoding, &detectedEncoding))
+            setEncoding(detectedEncoding, EncodingFromContentSniffing);
+    }
+
+    ASSERT(m_encoding.isValid());
+
+    if (!m_codec)
+        m_codec = newTextCodec(m_encoding);
+
+    String result = m_codec->decode(dataForDecode, lengthForDecode, DoNotFlush, m_contentType == XMLContent && !m_useLenientXMLDecoding, m_sawError);
+
+    m_buffer.clear();
+    return result;
 }
 
 String TextResourceDecoder::flush()
 {
     // If we can not identify the encoding even after a document is completely
     // loaded, we need to detect the encoding if other conditions for
     // autodetection is satisfied.
-    if (m_buffer.size()
+    if (m_buffer.size() && shouldAutoDetect()
         && ((!m_checkedForXMLCharset && (m_contentType == HTMLContent || m_contentType == XMLContent)) || (!m_checkedForCSSCharset && (m_contentType == CSSContent)))) {
-        detectTextEncoding(m_buffer.data(), m_buffer.size());
+        WTF::TextEncoding detectedEncoding;
+        if (detectTextEncoding(m_buffer.data(), m_buffer.size(), m_hintEncoding, &detectedEncoding))
+            setEncoding(detectedEncoding, EncodingFromContentSniffing);
     }
 
     if (!m_codec)
         m_codec = newTextCodec(m_encoding);
 
     String result = m_codec->decode(m_buffer.data(), m_buffer.size(), FetchEOF, m_contentType == XMLContent && !m_useLenientXMLDecoding, m_sawError);
     m_buffer.clear();
     m_codec.clear();
     m_checkedForBOM = false; // Skip BOM again when re-decoding.
     return result;
 }
 
 } // namespace blink