Guys,
In order to provide support for object / entity level support of hidden()
and disabled(), I have had to make changes to the metamodel and
related.
Specifically, adding:
Consent isVisible(AuthenticationSession session);
and
Consent isUsable(AuthenticationSession session);
to
oai.core.metamodel.adapter.ObjectAdapter.
Of course, this had some knock-on effects on derived classes (most
specifically, PojoAdapter and the metamodel wrapper.
The metamodel wrapper also had to be modified to actually check for
the new behaviour.
Anyway - looking at how the HTML Viewer *still* needs to be changed
(the whole concept of making an entire entity read-only is new), I'm left
wondering if Dan's suggested kludge might be the better approach!
The details:
I have provided support for checking if an object is read-only/visible.
Dan's suggestion was to implement this by manually forcing each
field/action to be read-only/visible.
My changes:
An object can specify if it is hidden or read-only with a single pair of
methods.
Advantage:
Using my additions, viewers can check an object first, before checking
individual fields/properties.
Disadvantage:
Viewers have to be modified(?).
Using Dan's suggestion (of forcing the behaviour onto each field):
Make each field/property check to the see if owning entity is read-
only/visible.
Advantage:
Should work with existing viewers out of the box.
Disadvantage:
Kludge. Seemed inefficient to me. The same "validate" method will be
called multiple times - 1 for each method/property in the entity.
Opinions, please?
