Top-level domain-based default encoding

third_party/WebKit/Source/core/loader/TextResourceDecoderBuilder.cpp

Index: third_party/WebKit/Source/core/loader/TextResourceDecoderBuilder.cpp
diff --git a/third_party/WebKit/Source/core/loader/TextResourceDecoderBuilder.cpp b/third_party/WebKit/Source/core/loader/TextResourceDecoderBuilder.cpp
index 077ad6d0f5013afd3cd332a3a30ec748e3eb6825..204a8ede7bba46ee6cc5b04da2d07f6a855c5bde 100644
--- a/third_party/WebKit/Source/core/loader/TextResourceDecoderBuilder.cpp
+++ b/third_party/WebKit/Source/core/loader/TextResourceDecoderBuilder.cpp
@@ -42,6 +42,71 @@ static inline bool canReferToParentFrameEncoding(const LocalFrame* frame, const
     return parentFrame && parentFrame->document()->securityOrigin()->canAccess(frame->document()->securityOrigin());
 }
 
+static const char* legacyEncodings[] = {

[esprehn, 2016/03/21 22:56:25]

I'd name this into an array of structs, so you could access the short name and
the long name separately without the + 1 below.

[Jinsuk Kim, 2016/03/22 21:51:52]

Done. Also used anonymous namespace to limit the scope of the struct defined for
this array.

+    "au", "windows-1252",
+    "az", "ISO-8859-9",
+    "bd", "windows-1252",
+    "bg", "windows-1251",
+    "br", "windows-1252",
+    "ca", "windows-1252",
+    "ch", "windows-1252",
+    "cn", "GBK",
+    "cz", "windows-1250",
+    "de", "windows-1252",
+    "dk", "windows-1252",
+    "ee", "windows-1256",
+    "eg", "windows-1257",
+    "et", "windows-1252",
+    "fi", "windows-1252",
+    "fr", "windows-1252",
+    "gb", "windows-1252",
+    "gr", "ISO-8859-7",
+    "hk", "Big5",
+    "hr", "windows-1250",
+    "hu", "ISO-8859-2",
+    "il", "windows-1255",
+    "ir", "windows-1257",
+    "is", "windows-1252",
+    "it", "windows-1252",
+    "jp", "Shift_JIS",
+    "kr", "windows-949",
+    "lt", "windows-1256",
+    "lv", "windows-1256",
+    "mk", "windows-1251",
+    "nl", "windows-1252",
+    "no", "windows-1252",
+    "pl", "ISO-8859-2",
+    "pt", "windows-1252",
+    "ro", "ISO-8859-2",
+    "rs", "windows-1251",
+    "ru", "windows-1251",
+    "se", "windows-1252",
+    "si", "ISO-8859-2",
+    "sk", "windows-1250",
+    "th", "windows-874",
+    "tr", "ISO-8859-9",
+    "tw", "Big5",
+    "tz", "windows-1252",
+    "ua", "windows-1251",
+    "us", "windows-1252",
+    "vn", "windows-1258",
+    "xa", "windows-1252",
+    "xb", "windows-1257"
+};
+
+static const WTF::TextEncoding getEncodingFromDomain(const KURL& url)
+{
+    Vector<String> tokens;
+    url.host().split(".", tokens);
+    if (!tokens.isEmpty()) {
+        auto tld = tokens.last();
+        for (size_t i = 0; i < WTF_ARRAY_LENGTH(legacyEncodings); i += 2) {
+            if (tld == legacyEncodings[i])
+                return WTF::TextEncoding(legacyEncodings[i + 1]);
+        }
+    }
+    return WTF::TextEncoding();
+}
 
 TextResourceDecoderBuilder::TextResourceDecoderBuilder(const AtomicString& mimeType, const AtomicString& encoding)
     : m_mimeType(mimeType)
@@ -56,12 +121,13 @@ TextResourceDecoderBuilder::~TextResourceDecoderBuilder()
 
 inline PassOwnPtr<TextResourceDecoder> TextResourceDecoderBuilder::createDecoderInstance(Document* document)
 {
+    const WTF::TextEncoding encodingFromDomain = getEncodingFromDomain(document->url());
     if (LocalFrame* frame = document->frame()) {
         if (Settings* settings = frame->settings())
-            return TextResourceDecoder::create(m_mimeType, settings->defaultTextEncodingName(), settings->usesEncodingDetector());
+            return TextResourceDecoder::create(m_mimeType, encodingFromDomain.isValid() ? encodingFromDomain : settings->defaultTextEncodingName(), settings->usesEncodingDetector());
     }
 
-    return TextResourceDecoder::create(m_mimeType, String());
+    return TextResourceDecoder::create(m_mimeType, encodingFromDomain);
 }
 
 inline void TextResourceDecoderBuilder::setupEncoding(TextResourceDecoder* decoder, Document* document)