But control flow is not enough to reason about the code one needs the full AST
control flow givres you:
- complexity
- some clone detection technique
- what else ???
nicolas
Hi doru
it was just to look at how it was done.I spent today building a first super simple ControlFlowGraph for Pharo and the idea is to have this genericfor other languages.
Stef
On 22 Feb 2018, at 10:51, Tudor Girba <tudor@tudorgirba.com> wrote:
Hi,
I am not sure I understand the use cases for why you would need FAST-pharo. I mean, what does RB not already offer? Could you detail them?
Cheers,
Doru
On Feb 22, 2018, at 10:24 AM, Nicolas Anquetil <Nicolas.Anquetil@inria.fr> wrote:
I beleive steph would like to have FAST-pharo, because FAST-core alone does not do much, it's only the infrastructure.
I am working on making FAST-pharo loadable and passing all the tests
nicolas
On 21/02/2018 10:19, Cyril Ferlicot wrote:
On Wed, Feb 21, 2018 at 8:54 AM, Stéphane Ducasse
<stephane.ducasse@inria.fr> wrote:
Hi Stef,
You may want to use the CI image
https://ci.inria.fr/moose/job/FAST-On-Moose/
I cannot execute this image with the launcher.
And I cannot execute on my mac.
Hi!
At Synectique we depend on it with this snippet:
project: 'FASTCore'
with: [ spec
className: #ConfigurationOfFASTCore;
versionString: #development;
repository: 'http://smalltalkhub.com/mc/Moose/FASTCore/main/' ].
So, I think this should work:
Metacello new
repository: 'http://smalltalkhub.com/mc/Moose/FASTCore/main/';
configuration: 'FASTCore';
version: #development;
load
For the state of FAST: Fast meta model was not valid because it had
multiple opposites for some properties and in Moose 6 or 6.1 (I don't
remember) a model validation was added because MooseQuery require a
valide meta model.
With Nicolas we worked on it to make FASTCore valid. There was at the
top of the hierarchy a method #parentNode and multiple properties
defining it as opposite, which is not valid for FAME.
What we did is that instead of pointing #parentNode for multiple
properties, we created, for each property, a new opposite property in
the right subclass of FASTEntity and we reimplemented #parentNode to
point this new property. At the end, we replaced the #parentNode at
the top of the hierarchy by a subclass responsibility. To keep it was
only for backward compatibility.
We validated the changes with the tests of one of Synectique
meta-model. Maybe the implementations for the other languages are
still not valid but I did not had the knowledge nor the time to check
it :(
--
Nicolas Anquetil
RMod team -- Inria Lille
_______________________________________________
Moose-dev mailing list
Moose-dev@list.inf.unibe.ch
https://www.list.inf.unibe.ch/listinfo/moose-dev
--
www.tudorgirba.com
www.feenk.com
"We can create beautiful models in a vacuum.
But, to get them effective we have to deal with the inconvenience of reality."
_______________________________________________
Moose-dev mailing list
Moose-dev@list.inf.unibe.ch
https://www.list.inf.unibe.ch/listinfo/moose-dev
--------------------------------------------Stéphane Ducasse03 59 35 87 52Assistant: Julie JonasFAX 03 59 57 78 50TEL 03 59 35 86 16S. Ducasse - Inria40, avenue Halley,Parc Scientifique de la Haute Borne, Bât.A, Park PlazaVilleneuve d'Ascq 59650France
_______________________________________________ Moose-dev mailing list Moose-dev@list.inf.unibe.ch https://www.list.inf.unibe.ch/listinfo/moose-dev
-- Nicolas Anquetil RMod team -- Inria Lille