[Webkit-unassigned] [Bug 49999] better V8 bindings for HTML5 ImageData

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=49999





--- Comment #2 from Vitaly Repeshko <vitalyr at chromium.org>  2010-11-24 09:00:29 PST ---
(In reply to comment #1)
> Created an attachment (id=74714)
 --> (https://bugs.webkit.org/attachment.cgi?id=74714&action=review) [details]
> proposed patch

The key part here is avoiding hitting a C++ callback when reading the "data" property on a JS wrapper of ImageData object. Setting a pure JS property on the wrapper is a good idea. Yet modifying the way the value of the property is computed is not necessary and I think can be only harmful. Instead of creating a new object template and instantiating it, we should reuse the CanvasPixelArray wraper, i.e. do something like this:
if (!wrapper.IsEmpty()) {
    v8::Handle<v8::Value> pixelArray = toV8(impl->data());
    if (!pixelArray.IsEmpty())
        wrapper->Set(v8::String::NewSymbol("data"), pixelArray, v8::ReadOnly);
}

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