Say, can't you just leverage HtmlUnit or HttpUnit? They give you an
environment with Rhino + all usual web browser objects.
Attila.
On 2007.11.30., at 15:39, Luca wrote:
> Hello everyone,
> I started a small project for a "light" JavaScript parser using the
> Rhino technology. My idea is to have a fake JavaScript parser that can
> execute the code inside a page and give me the final result of the
> execution as raw text.
>
> I started by writing some Java Classes which correspond to Javascript
> common objects, like Document, Window, Location, and then hooking them
> one by one such as:
>
> ScriptableObject.putProperty(scope, "window", jsWindow);
>
> where jswindow belongs to the custom Java class JavaScriptWindow I
> wrote. This worked actually pretty well in many cases; I'm trying to
> run
> it on a very small set of 485 pages containing scripts, and they
> arrive
> to an end without exceptions (even though I can't say it's a "correct"
> execution) in 50% of the cases.
>
> In another 25% of cases the script cannot complete because some
> function
> is missing (it's probably incorporated from some external script)
> but on
> the remaining 25% is due some other objects missing.
>
> For instance, executing this line I get an error on indexOf:
> navigator.userAgent.indexOf('MSIE')>=0
> even if in my class JavaScriptNavigator there is a
> public String userAgens
> field, and the class is hooked into the scope with:
> ScriptableObject.putProperty(scope, "location", jsWindow.location);
>
> I guess my whole approach is naive and I need to change it completely,
> maybe using Java classes that implement the Scriptable interface, but
> I'd like some expert advice on it. Can anybody give me a hint?
>
> Thanks,
> Luca
_______________________________________________
dev-tech-js-engine-rhino mailing list
dev-tech-js-engine-rhino@lists.mozilla.org
https://lists.mozilla.org/listinfo/dev-tech-js-engine-rhino
