[Webkit-unassigned] [Bug 73215] [Qt][WK2] Split QWebPermissionRequest into QWebSecurityOrigin

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=73215





--- Comment #35 from Simon Hausmann <hausmann at webkit.org>  2012-01-23 01:15:31 PST ---
(From update of attachment 123489)
View in context: https://bugs.webkit.org/attachment.cgi?id=123489&action=review

>> Source/WebKit2/UIProcess/API/qt/qtwebsecurityorigin.cpp:43
>> +    return (scheme == QLatin1String("qrc")) ? true : (scheme == QLatin1String("file"));
> 
> That is a pretty weird way of writing that.

And shouldn't this function use the WebCore Scheme registry instead of hard-coding qrc and file?

>> Source/WebKit2/UIProcess/API/qt/qtwebsecurityorigin.cpp:52
>> +        return;
> 
> a comment would be good here, like "data: urls, inherit the locality of their owner, thus..."

Isn't there a central function in WebCore to call for that instead of hard-coding the scheme here?

> Source/WebKit2/UIProcess/API/qt/qtwebsecurityorigin.cpp:99
> +    if (containsScheme(scheme))
> +        return;
> +
> +    SchemeRegistry::registerURLSchemeAsLocal(scheme);
> +}

Why the containsScheme check? AFAICS the scheme registry uses a hashset, so the "contains" check appears redundant to me.

Why is this method exported, who is the caller?

> Source/WebKit2/UIProcess/API/qt/qtwebsecurityorigin.cpp:107
> +    if (scheme == QLatin1String("file"))
> +        return;
> +
> +    SchemeRegistry::removeURLSchemeRegisteredAsLocal(scheme);

This is again a redundant check, because the implementation in SchemeRegistry does the same checking.

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