|
|
|
|
Re: Modisco parsing output [message #753522 is a reply to message #753464] |
Thu, 27 October 2011 14:58 |
Hugo Bruneliere Messages: 674 Registered: July 2009 |
Senior Member |
|
|
Hello Stephane,
XMI is the standard serialization format for models, and it is the same independently from the metamodels the models conform to (e.g.; the Java one, the UML2 one or any other metamodel).
Even if the extension name may change, there are still XMI files (I invite you to open them in a textual editor to realize this).
I think there is a confusion when you talk about generating a metamodel (i.e. a Ecore file, which is also a XMI file but with a specific extension name) from your Java code.
A metamodel describes the viewpoint from which your want to model a given system, not the view on this system (which is represented as a model that conforms to this metamodel).
So which viewpoint do you want to consider?
You can model the global architecture of the system, its detailed structure, the internal of its implementation, its provided features, its business or domain, etc.
Once you will have chosen/define this viewpoint, and so the correponding metamodel (i.e. Ecore file), you will be able to produce models (XMI files) that conforms to this metamodel and actually represent your required view of the system.
In order to do this, MoDisco provides some basic bricks that will help you to start and chain the various steps of your process.
Best regards,
Hugo
--------------------------------------------------------
Hugo Bruneliere, PhD
NaoMod team (IMT Atlantique & LS2N-CNRS)
Nantes - France
--------------------------------------------------------
|
|
|
|
Powered by
FUDForum. Page generated in 0.04096 seconds