Creating internal transition in UML-RT? [message #1771509] |
Fri, 25 August 2017 20:48 |
reza ahmadi Messages: 47 Registered: September 2016 |
Member |
|
|
Hi,
Does it need special procedure to create an internal transition, or by simply picking "internal" as kind in UML tab (in properties window) a transition should be an internal one?
As I see unexpected behavior in a state machine with an internal transition. The transition simply does not get triggered:
Unexpected message to capsule instance prop1__TOP.trafficLight role trafficLight on port TestData protocol TestChannel signal nextTest
"trafficLight" is a capsule which its state machine has an internal transition (its kind is "internal"). That internal transition has "nextTest" as its trigger.
Thanks.
[Updated on: Fri, 25 August 2017 22:33] Report message to a moderator
|
|
|
|
|
|
|
Re: Creating internal transition in UML-RT? [message #1771713 is a reply to message #1771615] |
Tue, 29 August 2017 19:26 |
reza ahmadi Messages: 47 Registered: September 2016 |
Member |
|
|
Thanks Ernesto & Peter.
Ernesto: I did not understand "internal transitions directly owned by the state machine". I can see that by adding an internal transition to a non-composite state, it turned to a composite as a region is added to the state automatically and the internal transition is inserted inside the new region.
In any case, an internal transition on a composite state worked for me and fixed my scenario.
[Updated on: Tue, 29 August 2017 19:32] Report message to a moderator
|
|
|
Re: Creating internal transition in UML-RT? [message #1771715 is a reply to message #1771713] |
Tue, 29 August 2017 19:42 |
Ernesto Posse Messages: 438 Registered: March 2011 |
Senior Member |
|
|
In UML all elements except for the top-level model are owned by (contained in) some other element. States are owned indirectly by other States or by a State machine. Indirectly because they are owned by a "Region", but in UML-RT, as opposed to general UML, states and state machines own only one region which is why this region is hidden in the model explorer. If a state owns anything (transitions, pseudo-states, other states), then it is composite. So you can add internal transitions (or anything) to a simple state, but that will make it automatically into a composite state. A state machine also owns states, pseudo-states and transitions. Whenever you add any of these to the state machine, it is owned by the state machine. This includes internal transitions, which is just another kind of transition. My statement was that while we do support internal transitions, I think we only support internal transitions owned by a (composite) state and not by the state machine itself (more precisely, internal transitions owned by the state machine's own Region).
|
|
|
|
Powered by
FUDForum. Page generated in 0.04324 seconds