Generating CIDs in Blink during MHTML serialization.

third_party/WebKit/Source/core/page/PageSerializer.h

Index: third_party/WebKit/Source/core/page/PageSerializer.h
diff --git a/third_party/WebKit/Source/core/page/PageSerializer.h b/third_party/WebKit/Source/core/page/PageSerializer.h
index b67fde2e5d9526006fe02eedaee3d1b2ce610908..89144404dd290683e956a3b4d0199d15de7e96ff 100644
--- a/third_party/WebKit/Source/core/page/PageSerializer.h
+++ b/third_party/WebKit/Source/core/page/PageSerializer.h
@@ -35,20 +35,21 @@
 #include "platform/heap/Handle.h"
 #include "platform/weborigin/KURL.h"
 #include "platform/weborigin/KURLHash.h"
-#include "wtf/HashMap.h"
 #include "wtf/ListHashSet.h"
 #include "wtf/PassOwnPtr.h"
 #include "wtf/Vector.h"
+#include "wtf/text/WTFString.h"
 
 namespace blink {
 
 class Attribute;
-class FontResource;
-class ImageResource;
 class CSSRule;
 class CSSStyleSheet;
 class CSSValue;
 class Document;
+class Element;
+class FontResource;
+class ImageResource;
 class LocalFrame;
 class Page;
 class LayoutObject;
@@ -69,9 +70,25 @@ public:
     public:
         virtual ~Delegate() { }
         virtual bool shouldIgnoreAttribute(const Attribute&) = 0;
+
+        // Method allowing the Delegate control which URLs are written into the
+        // generated html document.
+        //
+        // This method should return an html attribute value that needs to be
+        // used in place of the original link associated with the element
+        // (i.e. in place of img.src or iframe.src or object.data).
+        //
+        // This method should return a default-constructed String to indicate
+        // that no link rewriting should happen for the given element.
+        virtual String rewriteLink(const Element&) = 0;

[dcheng, 2015/12/01 04:19:31]

It might be easier to just return a bool here and return the String result in a
ref parameter.

[Łukasz Anforowicz, 2015/12/01 18:45:14]

Done.  Explicit bool does look/feel better than overloading the meaning of a
null String.

Side node: For a moment I wondered whether I should split this method into 2
(isRewritingRequired(const Element&) + String getRewrittenLink(const Element&))
but the 2 methods would have to share most of their implementation => keeping
this as a single method feels better.

     };
 
-    PageSerializer(Vector<SerializedResource>*, PassOwnPtr<Delegate>);
+    // Constructs a serializer that will write output to the given vector of
+    // SerializedResources and use the optional Delegate for controlling some
+    // serialization aspects.
+    PageSerializer(
+        Vector<SerializedResource>*,

[dcheng, 2015/12/01 04:19:31]

Nit: Vector<>&

[Łukasz Anforowicz, 2015/12/01 18:45:14]

Done.  

I guess https://www.chromium.org/blink/coding-style#TOC-Pointers-and-References
wins with https://google.github.io/styleguide/cppguide.html#Reference_Arguments
in this case :-)  Maybe one day I'll get used to the differences between
Chromium and Blink styles :->

I was wondering how deep I should go with this change.  I've decided to change
the public API (i.e. the type of the constructor parameter that you've pointed
out) but leave private implementation details intact (i.e. leaving the type of
the PageSerializer::m_resources field as a pointer).

+        Delegate*);
 
     // Initiates the serialization of the frame. All serialized content and
     // retrieved resources are added to the Vector passed to the constructor.
@@ -79,11 +96,6 @@ public:
     // Subsequent resources are images, css, etc.
     void serializeFrame(const LocalFrame&);
 
-    void registerRewriteURL(const String& from, const String& to);
-    void setRewriteURLFolder(const String&);
-
-    KURL urlForBlankFrame(const LocalFrame&);
-
     Delegate* delegate();
 
     static String markOfTheWebDeclaration(const KURL&);
@@ -108,13 +120,7 @@ private:
     Vector<SerializedResource>* m_resources;
     ListHashSet<KURL> m_resourceURLs;
 
-    using BlankFrameURLMap = WillBeHeapHashMap<RawPtrWillBeMember<const LocalFrame>, KURL>;
-    BlankFrameURLMap m_blankFrameURLs;
-    HashMap<String, String> m_rewriteURLs;
-    String m_rewriteFolder;
-    unsigned m_blankFrameCounter;
-
-    OwnPtr<Delegate> m_delegate;
+    Delegate* m_delegate;
 };
 
 } // namespace blink