[Fetch] Always use utf-8 for decoding in text()

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

Index: third_party/WebKit/Source/core/html/parser/TextResourceDecoder.h
diff --git a/third_party/WebKit/Source/core/html/parser/TextResourceDecoder.h b/third_party/WebKit/Source/core/html/parser/TextResourceDecoder.h
index 2539cc8a68577c5e20da2b757637e3e510931e27..d736de9aaf2b9fe151dce7c7d87b3a2d9c866927 100644
--- a/third_party/WebKit/Source/core/html/parser/TextResourceDecoder.h
+++ b/third_party/WebKit/Source/core/html/parser/TextResourceDecoder.h
@@ -48,7 +48,13 @@ public:
 
     static PassOwnPtr<TextResourceDecoder> create(const String& mimeType, const WTF::TextEncoding& defaultEncoding = WTF::TextEncoding(), bool usesEncodingDetector = false)

[kouhei (in TOK), 2015/12/14 02:00:22]

Most of the create() methods use exactly same params as c-tors.
Would you change "bool usesEncodingDetector" to EncodingDecodingOption?

[hiroshige, 2015/12/14 05:44:02]

I'd like to enforce mimeType == "plain/text" && defaultEncoding ==
UTF8Enconding() if EncodingDecodingOption is AlwaysUseUTF8ForText, so
AlwaysUseUTF8ForText shouldn't specified from create().
Changing the last param to EncodingDetectionOption and placing RELEASE_ASSERT or
ASSERT(encodingDetectionOption != AlwaysUseUTF8ForText) would also work (but I
slightly prefer the current Patch Set).
How do you think?

     {
-        return adoptPtr(new TextResourceDecoder(mimeType, defaultEncoding, usesEncodingDetector));
+        return adoptPtr(new TextResourceDecoder(mimeType, defaultEncoding, usesEncodingDetector ? UseAllAutoDetection : UseContentAndBOMBasedDetection));
+    }
+    // Corresponds to utf-8 decode in Encoding spec:
+    // https://encoding.spec.whatwg.org/#utf-8-decode.
+    static PassOwnPtr<TextResourceDecoder> createAlwaysUseUTF8ForText()
+    {
+        return adoptPtr(new TextResourceDecoder("plain/text", UTF8Encoding(), AlwaysUseUTF8ForText));
     }
     ~TextResourceDecoder();
 
@@ -73,7 +79,28 @@ public:
     size_t checkForBOM(const char*, size_t);
 
 private:
-    TextResourceDecoder(const String& mimeType, const WTF::TextEncoding& defaultEncoding, bool usesEncodingDetector);
+
+    // TextResourceDecoder does three kind 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().
+    enum EncodingDetectionOption {
+        // Use 1. + 2. + 3.
+        UseAllAutoDetection,
+
+        // Use 1. + 2.
+        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);
@@ -96,7 +123,7 @@ private:
     bool m_checkedForMetaCharset;
     bool m_useLenientXMLDecoding; // Don't stop on XML decoding errors.
     bool m_sawError;
-    bool m_usesEncodingDetector;
+    EncodingDetectionOption m_encodingDetectionOption;
 
     OwnPtr<HTMLMetaCharsetParser> m_charsetParser;
 };