[Webkit-unassigned] [Bug 66877] [Chromium] Add ability to do static SerializedScriptValue deserialization

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=66877





--- Comment #3 from Greg Billock <gbillock at google.com>  2011-08-25 10:53:15 PST ---
(From update of attachment 105039)
View in context: https://bugs.webkit.org/attachment.cgi?id=105039&action=review

>> Source/WebKit/chromium/public/WebSerializedScriptValue.h:37
>> +#include "v8/include/v8.h"
> 
> this should be #include <v8.h>; however, i think you can just forward declare
> the V8 types you are using instead.  please see how this is done in other
> header files.

Thanks, Darin. I tried that, as well as #include "v8.h". Neither one compiled. Is there a gypi file I need to update or something? The example I was following was WebFrameClient.h. This file looked parallel, but obviously there's something different in the configuration, and I didn't find it. I'll just forward-declare, as it looks like that Just Works.

>> Source/WebKit/chromium/public/WebSerializedScriptValue.h:72
>> +    static v8::Handle<v8::Value> deserializeToValue(const WebString& data);
> 
> perhaps you should just have a getter on WebSerializedScriptValue that exposes
> the underlying v8 object?  this way you could also get the v8 value corresponding
> to what createInvalid() produces, or if we expose WebSerializedScriptValue through
> other interfaces, you'd be able to also get the v8 object from there.  (I see some
> other APIs that pass WebSerializedScriptValue.)
> 
> Maybe:
> 
>   WEBKIT_EXPORT v8::Handle<v8::Value> deserialize() const;

So follow WebFrame.h::frameForContext? I can definitely do that.

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.