«

Hi gurus and nubys,

I was just wondering why access to attributes of an element is done
just via #attributes#[].
IMO it would be nice to have Element#[] like:

def [](k)
attributes[k]
end

and ditto for #[]=

Anyway, Sean Russel is my personal XML deity so if this is not allowed
there may be a reason: what is this reason?

gabriele renzi <> wrote in message news:<>. ..
> I was just wondering why access to attributes of an element is done
> just via #attributes#[].

Element#[] is already defined, since it is inherited from Parent. #[]
references all child nodes of an element except attributes.

> if this is not allowed there may be a reason: what is this reason?

It was, primarily, a design decision: either Element#[] references
nodal children, or Element#[] references attributes. I chose the
children.

--- SER

il 26 Jul 2004 09:06:38 -0700, (Sean Russell)
ha scritto::

>gabriele renzi <> wrote in message news:<>. ..
>> I was just wondering why access to attributes of an element is done
>> just via #attributes#[].
>
>Element#[] is already defined, since it is inherited from Parent. #[]
>references all child nodes of an element except attributes.

oops, sorry I just took a look at the quick RDoc api and did not
thought of inheritance

>> if this is not allowed there may be a reason: what is this reason?
>
>It was, primarily, a design decision: either Element#[] references
>nodal children, or Element#[] references attributes. I chose the
>children.

well, sure now this makes sense, thanks a lot

gabriele renzi <> wrote in message news:<>. ..
> >Element#[] is already defined, since it is inherited from Parent. #[]
> >references all child nodes of an element except attributes.
>
> oops, sorry I just took a look at the quick RDoc api and did not
> thought of inheritance

No problem. The fact that rdoc doesn't include inherited methods in
the class signature bites everyone, sooner or later, including me.


--- SER