Some Eclipse Foundation services are deprecated, or will be soon. Please ensure you've read this important communication.

Bug 366415

Summary: [StateMachineDiagram] Elements from redefined state machines not visualizable in redefining state machines
Product: [Modeling] Papyrus Reporter: Marc-Florian Wendland <marc-florian.wendland>
Component: CoreAssignee: Ansgar Radermacher <ansgar.radermacher>
Status: RESOLVED FIXED QA Contact:
Severity: major    
Priority: P3 CC: ansgar.radermacher
Version: 0.8.1   
Target Milestone: ---   
Hardware: All   
OS: All   
Whiteboard:

Description Marc-Florian Wendland CLA 2011-12-12 09:45:56 EST
Redefinition of behavioral descriptions allows a very efficient way of modelling by reusing already existing information, which are selectively refined in specific behavioral descriptions.

Selectively commonly means not all elements are redefined, some are reused from the already existing super class of the state machine. Currently, it is not possible to visualize already existing elements of a redefined (hence specialized) state machine in the redefining (hence specializing) state machine. Being able to visualize those elements is very important for the efficient usage of redefintion conepts. 

Currently the user has to explicitily recreate all the elements of the redefined state machine, which may cause name clashes and inconsistencies or even just time-consuming, error prone and redundant work.

In short: Element in specialized and redefined state machine must be visualizable in specializing and redefining state machines.
Comment 1 Ansgar Radermacher CLA 2013-02-08 11:00:57 EST
I'm currently testing a modification that will allow to drag&drop
elements from the extended region into a diagram, i.e. state, pseudo-states and transitions. In addition, connection points defined in the re-defined state can be dragged into the machine as well. Does this resolve to your bug report?
Comment 2 Marc-Florian Wendland CLA 2013-02-11 03:58:02 EST
Ansgar, that is great news. Seems as if it would resolve my bug report; actually, every information that can be visualized in a super state machine and region should be visualizable in the specialized state machine and extending region, even if one wants to redefine aspects or not.

Redefinition means changing something, while it should also be possible to just extend, e.g., a region, to the already existing regions, whithout modifying the semantics of the super state machine region.

And, finally, the states from a super state machine should be as completely visualized as in the super state machine, i.e., it should also show the inner structure if the user wants it to.

But, again, I guess your modification covers pretty well what I had in mind.

Will this feature be part of the Kepler release? Would be just great.

Marc-Florianb
Comment 3 Ansgar Radermacher CLA 2013-02-11 16:40:33 EST
I checked-in the modifications: revision 10100 (0.9.X) and revision 10111 (0.10.X) of file CustomStateMachineDiagramDragDropEditPolicy.java.
Thus, it would be part of both Juno SR2 and Kepler release.

Marc-Florian, can you please test.

Ansgar
Comment 4 Marc-Florian Wendland CLA 2013-02-21 06:21:08 EST
Ansgar,

tested with Kepler M5 and Papyrus nightly build (19th February, 2013). Works partially as expected.

I was able to drop Pseudo-states and states from an extended region into the extending region. 

What did not work was dropping transitions and final states into the region.
Comment 5 Ansgar Radermacher CLA 2013-04-03 07:40:29 EDT
Should be fixed in revision 10770 for the trunk and 10771 for the 0.9.X branch. Nightly builds should contain the patch tomorrow.

The new fix adds support for dropping the FinalState from an extended region. Dropping a transition from an extended region should have worked with the previous patch, but it required that source and target were already present in the diagram. Now it should also work, if this is not the case.
Comment 6 Sébastien Gérard CLA 2013-04-18 16:21:21 EDT
can we closed s fixed the task? if yes, please Ansgar, close it.
thanks.
Comment 7 Ansgar Radermacher CLA 2013-04-22 17:27:33 EDT
Bug has been fixed as stated in comment 5