|
|
Re: Processing the XMI of the KDM [message #910820 is a reply to message #909520] |
Mon, 10 September 2012 08:45   |
Eclipse User |
|
|
|
Hello Hugo Bruneliere thanks to reply. However, I think I didn't get it. Thus, let me trying to explain my problem to you and probably you can help much more.
My problem is the following: I have an instance of the KDM (XMI), which represents my legacy systems, my goal is remove specifics chunk of code (edges) of this XMI. These chunk of code will be every statements related to persistence (JDBC), such as ResultSet, Statement, Preparedstatements, etc. After remove these chunk of code I'll pretend to extend the KDM to deal with AspectJ, thus, afterwards these chunk of code I'll be transform to a KDM extended.
As I said in my earlier post I tried to do such goal by using XPath, reading all XMI and removing them, but this is demanding such time.
So my doubt is, how can I do that? Could you give me a simple example, because I'm new in Modisco.
Thanks guys
|
|
|
Re: Processing the XMI of the KDM [message #924368 is a reply to message #910820] |
Wed, 26 September 2012 15:46  |
Eclipse User |
|
|
|
Hi Rafael,
are you in use with any model transformation language that is able to process ecore models? E.g. ATL, QVT-O / QVT-R, ...? If this is the case, you can write your transformations and apply them to your extracted model. The transformations can be applied either with the modisco workflow or using tools from the eclipse m2m Project: http://www.eclipse.org/m2m/
An alternative approach is to load the kdm files in your java application and let your java code manipulate the in java representation of the java model.
All the best
Benjamin
|
|
|
Powered by
FUDForum. Page generated in 0.03754 seconds