| Message order in Papyrus sequence diagram xmi file [message #1701081] |
Thu, 09 July 2015 08:44  |
 Alkis Paspaliaris
Messages: 2
Registered: June 2014 |
Junior Member |
|
|
Dear all,
I would like to ask you if you could help me by explaining me how message ordering information of Papyrus sequence diagrams is represented in the corresponding xmi files.
For each lifeline, I would assume that the order of messages is shown in the xmi file in the coveredBy attribute of the respective lifeline element.
If yes, I have observed that the order there is not shown correctly. Adding a message seems to be always appended at the end of the messages list, instead of being inserted inbetween its neighbouring messages.
If no, then how can I get from the xmi file the message order for a single lifeline?
I am using just two simple lifelines exchanging messages, without any combined fragments or complex notation. My Eclipse version is Luna, and I just downloaded the Papyrus 1.1.0 version, the one that Mars uses.
Thank you a lot in advance. Any information you could provide me regarding this issue would be extremely useful.
Best regards,
Alkis
|
|
|
| Re: Message order in Papyrus sequence diagram xmi file [message #1703603 is a reply to message #1701081] |
Mon, 03 August 2015 14:34  |
Alkis Paspaliaris
Messages: 2
Registered: June 2014 |
Junior Member |
|
|
Since I haven't still got any answer, let me rephrase the question and make it more general:
Do Papyrus sequence diagrams store message order information?
As far as I have understood, they should, but I believe that it is a Papyrus bug that, when creating a new message, order of MessageOccurrenceSpecification fragments in XMI file is most frequently not respected. The same can be seen from the MessageOccurrenceSpecification fragments in the Model Explorer for the respective interaction in Papyrus IDE, order of them frequently seems random.
Another thing which I strongly believe is a bug is that, when creating a new combined fragment sometimes the lifeline information of message events in the XMI file gets lost (attribute 'coveredBy' of each lifeline, and attribute 'covered' of each MessageOccurenceSpecification fragment get blank).
In any case, I would be really grateful if someone could answer my initial question so that I know that these assumptions are correct, find a way to temporarily circumvent the problem and file a bug report about that.
Best regards,
Alkis
|
|
|
Powered by
FUDForum. Page generated in 0.02596 seconds
] 
Search
Help
Register
Login
Home
