Hi;
I'd greatly appreciate any help/advice on how to best view MSE files
describing functional programs. I've uploaded one such file to
http://atelier.inf.unisi.ch/~baroneak/yuv.mse
Information I'm modeling includes global variable and function
definitions, and information about functions, such as which file they
are in, whether or not they are predicates (heuristically determined,
always false in the example file attached), and how many parentheses
they have (as a heuristic for function size/complexity, similar to
lines of code.)
The current mondrian script I'm using to visualize it looks like this:
------------
view := ViewRenderer new.
view interaction popupText: [:entity | 'Name: ', entity name].
view nodes: mooseModel allFiles using: RectangleShape withBorder forEach:
[:file | |fInCurFile edges | fInCurFile := mooseModel
allFunctions select:
[:curFunc | ((file name asByteString) =
(curFunc sourceAnchor copyFrom: 6 to:
curFunc sourceAnchor size))].
fInCurFile sortWith: [:f | f propertyNamed: #DIF].
view nodes: fInCurFile
using: (RectangleShape orange
width: [:entity | (entity propertyNamed: #NPAR) / 8]
height: [:entity | (entity propertyNamed: #NPAR) / 8]).
view circleLayout.
edges := Core.OrderedCollection new.
fInCurFile do: [:e | e incomingInvocations do:
[:eInv | edges add: e->(eInv invokedBy)]].
view edges: edges
from: #key
to: #value
using: (LineShape color: Color blue)].
view open.
---------
At a minimum, I'd like to be able to mouse over or click on the orange
squares representing functions and see the associated function name;
right now, the popupText only shows the filenames.
Any other advice on new visualizations, improvements to the mse
exported (I only partially understand the format), etc, would be
great.
Thanks;
Katerina Barone-Adesi
hi
when trying to access the upper and lower bounds of a TypedElement return
the element itself because there is no return sign in the methods
TypedElement>>lower, TypedElement>>upper.
Pierrick
NamedOccurrences raise an error when you want to apply it on a class.
do: DNU on sorter
in addition
Duplication view on a group of class crashes the system I get a
primitive failed on ZoomGraphicsContext.
Stef
New submission from Tudor Girba <girba(a)iam.unibe.ch>:
NamedOccurrences raise an error when you want to apply it on a class.
do: DNU on sorter (submitted by Stef)
----------
assignedto: girba
component: Mondrian
messages: 88
nosy: girba
priority: Bug
status: unread
title: NamedOccurrences raises DNU due to sort:
________________________________________________
Moose Bugs <moose-dev(a)iam.unibe.ch>
<http://macamis.unibe.ch/trackers/moose/issue66>
________________________________________________
Hello ! I had a look at the BasicCodeGenerator. I analysed the example and
then, I tried it with my own EMOF models. Of course, I got some surprising
results :)
For example, when you visit a Package, you apply the visitor ( the
BasicCodeGenerator) on its nesting package. In your example, with the EMOF
model, there is no problems because there is only one package. But in a
model with subpackages, it results in an infiny loop. So, I don't
understand why it is necessary to visit the nesting package. Any idea ?
Thank you
Sara
Due to several requests, we have extended the sub mission deadline
for FAMOOSr 2007 to May 15, 2007. Please find the revised Call for
Papers both below and attached (we apologize for multiple copies).
==========================================================
FAMOOSR 2007 - EXTENDED DEADLINE
==========================================================
Call for Contributions
Workshop on FAMIX and Moose in Reengineering
http://smallwiki.unibe.ch/moose/famoosr2007
June 25, 2007 in Zurich, Switzerland
==========================================================
Collocated with TOOLS 2007
Objects, Models, Components, Patterns
ETH Zurich, Switzerland 24-28 June 2007
Conference site: http://tools.ethz.ch
Important Dates
---------------
Submission: May 15, 2007
Notification: May 29, 2007
Camera ready: June 15, 2007
Workshop: June 25, 2007
Description
-----------
The increasing amount of data available about software systems poses
new challenges for re- and reverse engineering research, as the
proposed approaches need to scale. In this context, concerns about
meta-modeling and analysis techniques need to be augmented by
technical concerns about how to reuse and how to build upon the
efforts of previous research.
The goal of this one-day workshop is to strengthen the community of
researchers and practitioners who are working in re- and reverse
engineering, by providing a forum for building future research using
Moose and Famix as shared infrastructure. We are interested in
collecting and synchronizing the different FAMIX extensions, which
are currently in use, to facilitate better cooperation between
research groups.
We solicit experience reports (4 pages of text, IEEE format) and
position papers (2 pages, IEEE format). Experience reports will be
expected to discuss meta-modeling and software analysis using, but
not limited to, FAMIX or Moose. Position papers will be expected to
describe new directions and challenges for FAMIX or Moose. Papers may
address issues along general themes, including but not limited to:
- Analysis specific FAMIX extensions for evolution data, dynamic
traces, bugzilla entries, etc...
- Meta-modeling in reengineering tools
- Visualization techniques.
- Analysis techniques: clustering, data mining, machine learning,
pattern matching, probabilistic approaches, etc...
- Mechanisms for tool composition and rapid tool prototyping.
- Reusability of research: issues related to reusing tools written by
other researchers, mechanisms for making research results and tools
available to the community.
- Persistency and manipulation of models and meta-models.
Submissions are not limited to FAMIX and Moose or active users of
either. We welcome any idea you feel is relevant.
Please send your submissions by email to girba(a)iam.unibe.ch and
akuhn(a)iam.unibe.ch.
Organizers
----------
Tudor Girba, University of Bern, Switzerland
Adrian Kuhn, University of Bern, Switzerland
Program committee
-----------------
Juan-Carlos Cruz, UBS, Switzerland
Serge Demeyer, University of Antwerp, Belgium
Orla Greevy, University of Bern, Switzerland
Michele Lanza, University of Lugano, Switzerland
Radu Marinescu, Politehnica University of Timisoara, Romania
Kim Mens, Catholic University of Louvain, Belgium
Tom Mens, University of Mons-Hainaut, Belgium
Oscar Nierstrasz, University of Bern, Switzerland
Martin Pinzger, University of Zurich, Switzerland
Sander Tichelaar, Credit Suisse, Switzerland
Roel Wuyts, University Libre of Bruxelles, Belgium
Hello,
I 've worked on the Importer class and I 've understood how you create a
model from MSE specifications (I had a look on the method called
importMSE)
But now, I want to know how you manage to load this model in Moose. I
didn't find the class and the method that do this.
Thank you for your help,
Sara
Hi,
I had a look into the SubVersion bundle and I have some questions about it:
- I saw that there is a nice model based on hismo behind, but I was not
able to find the concept of commit in that model. In other words, I can
I know which set of files (fileVersion) belongs to the same commit-set?
- Importing the model from a remote repository takes long. Is it because
besides the log, it also downloads the actual files?
- Is it possible to create a model from a log file instead of a remote
repository?
... or maybe the best thing is that I just contact the author of
SubVersion. Does anybody know his name / email? Because I did not find
it in the bundle
Many thanks
Cheers
Marco
