The view is composed of three largepanels. First of all at the left of the dialog, there's a list with all the maps of the UCM. Checking a map in this list will produce a PluginBinding between the Stub and the checked map. The created PluginBinding will be shown in the tree view in the middle panel. This tree view will list all PluginBindings and all In/Out Bindings for a specific PluginBinding.
Then after creating a PluginBinding, the user will probably want to create some InBindings and OutBindings with the StartPoints and EndPoints of the plugged map. The panel at the right contains four lists. The two lists at the left lists the Ins and the Outs of the Stub. The two lists at the right list the StartPoints and the EndPoints of the plugin map. The user can then select an In with a StartPoint and then click the button between the two to create an InBinding. He can do the same thing for the OutBindings. The tree view at the middle will then get updated as well.
How everything is connected in the code?
The StubBindingDialog was first designed to be an eclipse view that would get activated after selecting a Stub. This decision was an error because we had too much things to fit in the UI at the first place to fit all that in a simple view. So the view had to be refactored into an SWT dialog.
Since the class was initially designed to be a view, it was a listener to a Stub. Each time the observed Stub changed, the view would change. The view was designed to change the currently observed Stub. When the view was refactored into a dialog, we no longer needed this mechanism because now we could observe only one Stub during all the life of the dialog.
First of all the method createDialogArea() if where all the controls are created and layed out. This dialog was done during the learning process of the developper, so it's not 100% clean... Everything works, but it's just hard to find something to modify. Breaking this out into methods for contructing each panel would have probably be much better.
Something that is essential to talk about is that each time the user perform an action with the StubBindingDialog, it is actually done directly in the model with a command (that we execute/store internally). The dialog is a listener to the model and then updates all the required controls once the model has changed. Only the required panels are updated after a change in the model. This approach violates the model view controller pattern. This should probably be changed in the future if the dialog is extended a lot to add new properties and new things to edit in a Stub.
Since everything is done using commands, this allow the user to undo and redo what he just did. The Undo/Redo buttons in the middle are for exactly that.
All the method beginning with the prefix 'handle' are events handlers for various actions the user can perform in the UI. For example handlePluginChecked will get called when a user check a Map in the map list. It then execute the right command and then the other required views will get updated accordingly.
It's important to know that if the user click the Cancel button, all the action done during the session in the dialog will get undone.