On Dec 28, 8:33 pm, "David P. Caldwell" <[EMAIL PROTECTED]> wrote:
> On Dec 27, 9:46 am, Attila Szegedi <[EMAIL PROTECTED]> wrote:
>
> > Now, Rhino will ignore getIds() if the
> > wrapped object implements java.lang.Iterable, and for..in will iterate
> > over values:
>
> > for(i in x) { java.lang.System.out.println(i); }
>
> > when "x" is a java.util.List will print the values, and will no longer
> > print indices.
>
> > Shouldn't getIds() still have priority?
>
> That would be my vote, now that you mention it. In my view, it makes
> sense to have:
>
> for each (i in x) { ... }
>
> iterate over the values, but for (i in x) should iterate over the
> indices, I'd think.
>
> > So, while we're at that, shouldn't we allow java.util.List instances
> > to be treated identically to native Java arrays, similar to
> > NativeJavaArray class? I know List is a Java 2 interface, and thus
> > didn't exist back when Rhino (and LiveConnect) were originally
> > created, but I think we should catch up with times...
>
> Not sure here. LiveConnect is quite mature -- I'd think we'd want to
> urge whoever controls the LiveConnect spec to "catch up with the
> times" and then catch up with them. But I could be persuaded.
The LiveConnect spec was created by people at Netscape. I don't know
that there's anyone who owns it in any sense now. I think Rhino is
probably the main user of JavaScript-to-Java communication.
>
> -- David P. Caldwellhttp://www.inonit.com/
_______________________________________________
dev-tech-js-engine-rhino mailing list
[email protected]
https://lists.mozilla.org/listinfo/dev-tech-js-engine-rhino
