9 Jun
2011
9 Jun
'11
10:23 a.m.
Hi,
On 9 Jun 2011, at 10:17, Simon Denier wrote:
On 8 juin 2011, at 23:12, Usman Bhatti wrote:
Nope. A Famix reference in C# would be like:
ClassA.staticMethodCall()
-> one reference to ClassA and one invocation
So a reference has some kind of semantic meaning for the program (it is "used"
during the execution, it's not just something for typechecking). If you want, ClassA
is some kind of global variable which is accessed during execution (it's like that in
Smalltalk actually).
Now the "reference" to a type in a parameter is modelled in a different manner:
you actually have a FamixParameter entity which has a #declaredType property, which points
to the referenced FamixType. Same thing with FamixAttribute. That's what I meant by
"indirect" reference (although I don't like the term): you don't reify
the dependency directly, but you can still access it through #declaredType.
Indeed, simply declaring the type of a parameter is not a reference. However, passing a
class as an argument to a method call is.
a = b; [reference to A + 2 accesses to A]
No. There is no reference to A, only an access to a and an access to b.
Ok. Now I understand. Merely, passing a type as a parameter creates a reference. It is
not important how that parameter is used in the method. One problem with this approach is that #declaredType
property is often overlooked in dependency analysis (especially since in Smalltalk models
this prop is rarely used or even set). I don't know what could be the impact of
collecting #declaredType with the other dependencies.
---> This is actually a problem in MooseChef and could be largely discussed.
That is a good point.
Cheers,
Doru
--
Simon Denier
_______________________________________________
Moose-dev mailing list
Moose-dev(a)iam.unibe.ch
https://www.iam.unibe.ch/mailman/listinfo/moose-dev
--
www.tudorgirba.com
"To lead is not to demand things, it is to make them happen."