Just a brief announcement: a student project here in Lugano resulted
in a parser for C# systems. It's written in C# and therefore only runs
on Windows. The nice thing is it exports to mse which can then be
loaded into Moose. Experiments with it have proven it actually works,
even for larger systems, despite the shaky nature of the whole project
(too little time). You can find the tool here:
http://www.fpcg-ks.com/ermira/PMCS.rar
Best,
Michele
Hi!
Just to share what I am currently seeing while attending a conference
on Model transformation.
People seems to enjoy the select and collect operations for sets in OCL.
For example, this is what I see on a projected slide:
context CD:Class
def allSuperclasses: self.parents->union(self.superclasses->collect(c
| c.allSuperclasses)->flatten())
Apparently, the fact that you can combine these constructs is
innovative...
Maybe in the future, we should try to target ICMT and other model-like
conference.
Cheers,
Alexandre
--
_,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:
Alexandre Bergel http://www.bergel.eu
^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
- Added new Mondrian features in FAMIXFoldergroup
- new version of FAMIXFile* classes:
- added entities in FAMIXAbstractFile: any file should knows about
the entities it defines
- added new nagivation pragmas
- convenient methods for testing: ifFolder: ifFolderIfNotFolder: ...
Cheers,
Alexandre
Dear List,
Currently, there is no connection between instances of FAMIXFile/
Folder and entities that are supposed to be defined in (e.g., a .java
files and a class, a .c file and an instance of CAModule).
I extended FAMIXAbstractFile with definedEntities to have:
FAMIXNamedEntity subclass: #FAMIXAbstractFile
instanceVariableNames: 'belongsTo entities'
Then I was unsure whether Fame should to be used. Should 'entities'
refers to an ordered collection or a FMMultivalueLink. What will be
the opposite then? sourceAnchor could be a good candidate, but I am
unsure. For now, I did without Fame.
Cheers,
Alexandre
--
_,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:
Alexandre Bergel http://www.bergel.eu
^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
Dear Moosers
As you may know, identity hashes attributed by the virtual machine
ranges up to 4096. This may be problematic when storing objects in a
dictionary. We often have to deal with high number of entities, and
using dictionary is out of question for performance reason.
One solution (inspired from Lukas) would be to use the ID that each
MooseEntity has instead of the hash.
I will try to see whether hash can be overriden in MooseEntity to
return its ID instead.
Will let you know.
Cheers,
Alexandre
--
_,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:
Alexandre Bergel http://www.bergel.eu
^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
- The MonticelloImporter now merge classes and metaclasses
- Caches have been added to OB nodes for the moose browser. Necessary
for a real use
- FAMIXFolder>>filesGroup added
Cheers,
Alexandre
--
_,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:
Alexandre Bergel http://www.bergel.eu
^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
Just to announce that external class extensions are now correctly
imported in the last version of Moose.
To explain: if you import a package A which made class extensions to
classes of a package external to the import (say Kernel), such
extensions were not taken into account.
Only class extensions made to classes of package A were taken into
account before.
Also got the count of failing tests down to 2 and of error test down
to 1.
--
Simon
hi all
I would like to know the stability of PhMondrian.
Apparently Jannik got problem and had to disable caches.
What is your experience with it.
In particular I would also like to know whether PackageBlueprint works
on
sqMoose. Would be nice to rescue this nice visualization.
In addition is distribution map (the real version) working on PhMoose
or PhMondrian?
Stef
