Skip to main content

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
Re: [] OCL UML 2.5 alignment


After rebasing, I had to do some rework to make CS2AS test cases pass in my development branch. My transformations were working on ConstructorExp/Part and a couple of usages on renamed properties (Property::composite)

I can only say that I dislike that isIsXXXX. I don't think it's "only" related to different genmodel technology usage. The properties names have been changed to include that "is" prefix but... I can't add value/don't recall how to improve that so I don't really have anything else to contribute to the discussion.

[Branches pushed to the repo]

--On 18/03/2015 10:14, Horacio Hoyos Rodriguez wrote:

Epsilon doesn't have (yet) any support for static analysis so it does not
detect any changes in the names of attributes/methods in the underlying

I will take a look at your changes today.


Horacio Hoyos Rodríguez
EngD Student
University of York

-----Original Message-----
[] On Behalf Of Ed Willink
Sent: 17 March 2015 13:49
To: QVTD developers mailing list
Cc: MDT OCL mailing list
Subject: [] OCL UML 2.5 alignment


I have finally closed the gap in the unmodified UML(2.5).xmi to
Pivot.ecore to *.java transformation chain, so that we no longer rely on a
UML genmodel for modified UML 2.2 models. The resulting Pivot.ecore is
identical, or rather the control file for the QVTo transformations has
sufficient directives to make it the same.

However Ecore genmodels are less powerful/confusing than UML genmodels.
This means that isXXX attributes now have an isIsXXX() rather than
isXXX() accessor. This is a breaking change that affects tightly coupled
QVTd plugins.

Currently the master branches contain the UML 2.2-based models, and
ewillink/master contains the new UML 2.5-based models.

I need to promote them for M6, so please try to rebase on ewillink/master
and let me know of any problems. In the absence of any problems, I will
promote on Thursday morning.

The final commit on QVTd GIT ewillink/master fixes isIsXXX() in Epsilon
scripts. These changes were prompted by a JUnit test failure. Why didn't
the Epsilon editor detect these changes, and are the changes correct?


          Ed Willink

_______________________________________________ mailing list
To change your delivery options, retrieve your password, or unsubscribe
from this list, visit
_______________________________________________ mailing list
To change your delivery options, retrieve your password, or unsubscribe from this list, visit

Back to the top