matiasvara / bcoolexamples Goto Github PK

When a new operator is defined, the proposed DSE are not filter by the interface. This become a problem if there is DSE with the same name. In such case, only one is proposed.

The order of the parameters in the CCSL relation is wrong. For instance, in the following Figure, the order of the parameters in the Causality relation are inverted

For now, the order of the parameters should be manually corrected. We work for fixing the bug.

It should be possible to match elements by sing an external method in java. This would make possible to match events by relying on a dictionary.

In some specific case, the transformation fails to get the NSUri of the imported interfaces, and then, the generated qvto is wrong or directly empty. This problem depends on how the ECL specification imports the metamodel (.ecore). We give a partial list of tips to avoid this problem:

• The URI in the ECL must be the first line, with no commented lines before.
• The URI must be imported as "platform:/plugin...'. Using 'resource' may get an exception.

This list is partial, so that don't hesitate to other solutions.

The following problem was observed if we have regenerated the qvto and we tray to regenerate the CCSL:

"Java method invocation error 'public java.util.Set fr.inria.aoste.timesquare.ecl.qvto.helper.EclQVToHelper.getLibObjects(java.lang.String)'"

In the second eclipse workspace, we get the following error:

This exception is shown in the first eclipse workspace:

A temporal solution is to manually edit the qvto and replace the lines where TFSMMoC.lib is imported by replacing the "resource" to "plugin". Then, the line becomes:

'platform:/plugin/org.gemoc.sample.tfsm.moc.lib/ccsl/TFSMMoC.ccslLib'

I wanted to run the examples mentioned on the GMCO webpage but I seem to be missing some bundles/dependencies.

Steps to reproduce:

1. Download the newest version of GMCO studio --> V3.3.0 - 2021/07/28
2. Clone this repository.
3. File < Import < Existing Projects into Workspace to load the cloned samples.

Here is a screenshot of some errors:

Can you give me a hint on how to resolve these errors? I would really like to try the BCOooLExamples myself.

Thanks in advance!

Sometime the specification of a coordination rule can become complex. In such cases, I want to reuse a coordination rule to avoid rewritten same code. First, coordination rules must have a name, second, it must be possible to define that two coordination rules are the same.

Matias

The qvto generation fails when a matching correspondence includes an expression like: "s.name = "something"". In this case, there is a bug in the translation from a GExpression to an String. The bug is not fixed yet.

In some cases, in the generated qvto, the metamodels are not correctly imported. This occurs when the .ecl imports an .ecore. The transfo should check if the import is a .ecore and then return the corresponding NSURI.