[Webkit-unassigned] [Bug 75543] [GTK] Add webkit_web_view_run_javascript() to WebKit2 GTK+

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=75543





--- Comment #18 from Carlos Garcia Campos <cgarcia at igalia.com>  2012-03-21 01:02:37 PST ---
(In reply to comment #17)
> (From update of attachment 132244 [details])
> View in context: https://bugs.webkit.org/attachment.cgi?id=132244&action=review
> 
> > Source/WebKit2/UIProcess/API/gtk/WebKitJavascriptResult.cpp:38
> > +struct _WebKitJavascriptResult {
> > +    _WebKitJavascriptResult(WebKitWebView* view, WKSerializedScriptValueRef wkSerializedScriptValue)
> > +        : webView(view)
> > +        , referenceCount(1)
> > +        {
> > +            value = WKSerializedScriptValueDeserialize(wkSerializedScriptValue, webkit_web_view_get_javascript_global_context(view), 0);
> > +        }
> > +
> > +    GRefPtr<WebKitWebView> webView;
> > +    JSValueRef value;
> > +
> > +    int referenceCount;
> > +};
> 
> Instead of implementing your own reference counted type you could make this class extend from RefCounted<_WebKitJavaScriptResult> and use ref() and unref().

Is it worth converting that to a class? Is RefCounted better than our own implementation? it seems to use a mutex, while our implementation avoid the use of mutex by using atomic operations. I'll leave our current implementation for now, since it's pretty simple, and I'll write a patch to use RefCounted if it's really worth it and better.

> > Source/WebKit2/UIProcess/API/gtk/WebKitWebView.cpp:1504
> > +        g_set_error_literal(&error, WEBKIT_JAVASCRIPT_ERROR, WEBKIT_JAVASCRIPT_ERROR_SCRIPT_FAILED, _("An exception was raised in Javascript"));
> 
> Javascript -> JavaScript

Ok.

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