Resolving element content text with character references - C++/Arabica

[Olav]

Resolving character references with C++/Arabica

I have an element that looks like this:

<PhoneNumber>&lt;NUMBER&gt;</PhoneNumber>

I would like to have the content returned as "<NUMBER>".

Not only don't I get the character references resolved, but also what
is only one Text-node child of the element-node becomes several nodes
(When I have character references)

When googeling I see posts describing the opposite problem...
Is it a question about parser-options.

On the DOM-side, this is C++ Arabica, it doesn't seem to have the
"GetData", of the Java oriented DOM, Xerces seems to have something
like GetTextValue which I don't seem to have.
....IS it just a feature missing in Arabica, or should I do it some
other way?

Thanks
Olav