#41 'Scriptable' components

[Paolo Marrone]

Now it's possible to access to all the layers, from the
scripting environment, simply using jNet.getLayers(),
and to a specific layer using jNet.getLayer(layerName),
but it's very difficult to access to, for instance, a
LearningSwitch connected to a TeachingLayer, because we
should not access to that component directly using its
name, but only 'traveling' toward it, starting from the
last layer, passing throught the teacher, and finally
accessing to the wanted switch!
We could add a boolean property (named for instance
'scriptable') to all the components of the core engine,
giving it the meaning of 'visible component' from the
script engine for all the components having this
property set to true. What we must to do is simply to
'export' to the script engine all the components having
scriptable = true, using their name to identify them.
Another possibility is to avoid to add a new property,
simply exporting all the components of a neural
network, independently from the necessity to use them
into the script environment.