On 3 déc. 07, at 23:29, Tudor Girba wrote:
I took a look at AzadehExtraction, and in all the ExtractionRules that
are implemented you take the model as an input, but in the same time
you know in each predicate what is the exact type of entities that you
actually look for, and in the end you only return entities of the same
For example, name matching looks like:
This returns a ClassGroup.
But could this not look like this:
extract: model allClasses.
but doru the problem is that I cannot compose fully encapsulated
Of cours I can do what you are saying but first I only wanted as
input the group/model
But perhaps you have further intentions with this package and this
just the beginning. I am still looking for the use case for
heterogeneous groups :).
I do not want really hetero groups, I want polymorphic group/model
Also take a look at the ConformityStrategy hierarchy, because it is
very similar in intent with AzadehExtraction.
Azadeh can you have a look?
On Dec 3, 2007, at 12:37 PM, Stéphane Ducasse wrote:
> On 2 déc. 07, at 10:33, Tudor Girba wrote:
>> Hi Stef,
>> Could you provide just one example of such a filter?
> It is in AzadehExtraction
> From memory:
> NameMatcher new
> match: '*LAN*'
> but this could be
> NameMatcher new
> match: '*foo'
> and the implementation is
> kindSelectors := #allClasses
> kindSelectors := #allMethods
>> On Nov 30, 2007, at 2:17 PM, stephane ducasse wrote:
>>> Hi doru
>>> with azadeh we would to use group to represent the intermediate
>>> result of applying a filter
>>> on a model. Now the problem I have is that I cannot query a
>>> group as
>>> model (may be I'm wrong).
>>> Apparently a group a class does not understand allClasses.
>>> Is there a way that we can define/or get the behavior of a model
>>> on a
>>> Moose-dev mailing list
>> "We cannot reach the flow of things unless we let go."
>> Moose-dev mailing list
> Moose-dev mailing list
"In a world where everything is moving ever faster,
one might have better chances to win by moving slower."
Moose-dev mailing list