[Pharo-project] problem mixing Traits and Inheritance
Hi, Here's my problem : - I have the class NSCollectionStream which define ( for example ) #atEnd - I have the class NSSubclassOfCollectionStream ( which is a suclass of NSCollectionStream). When I use a trait in NSSubclassOfCollectionStream that required the method #atEnd, the defintion in NSCollectionStream is not used and the method #atEnd is overridden with : atEnd self explicitRequirement. I think we should take care about definitions in superclass, no?
We had a discussion about this on this mailing list some time ago. The conclusion, as far as I remember, was that yes, ideally it should detect that the method exists in a superclass and hence should not add the atEnd method to the subclass in your example. This is not that easy to implement, though (if you think through the different cases of structural changes that would need to be captured to correctly update classes using traits). The simplest solution is to not implement atEnd in the trait. Anyway, the information that this method is a requirement should be automatically provided by a tool rather than being hardcoded by the developer. Cheers, Adrian On Jul 8, 2009, at 11:08 , Cyrille Delaunay wrote:
Hi, Here's my problem : - I have the class NSCollectionStream which define ( for example ) #atEnd - I have the class NSSubclassOfCollectionStream ( which is a suclass of NSCollectionStream).
When I use a trait in NSSubclassOfCollectionStream that required the method #atEnd, the defintion in NSCollectionStream is not used and the method #atEnd is overridden with : atEnd self explicitRequirement.
I think we should take care about definitions in superclass, no? _______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
On Jul 8, 2009, at 11:31 AM, Adrian Lienhard wrote:
We had a discussion about this on this mailing list some time ago. The conclusion, as far as I remember, was that yes, ideally it should detect that the method exists in a superclass and hence should not add the atEnd method to the subclass in your example. This is not that easy to implement, though (if you think through the different cases of structural changes that would need to be captured to correctly update classes using traits). The simplest solution is to not implement atEnd in the trait. Anyway, the information that this method is a requirement should be automatically provided by a tool rather than being hardcoded by the developer.
yes after our discussion it was clear that requirement was too static for a language like smalltalk.
Cheers, Adrian
On Jul 8, 2009, at 11:08 , Cyrille Delaunay wrote:
Hi, Here's my problem : - I have the class NSCollectionStream which define ( for example ) #atEnd - I have the class NSSubclassOfCollectionStream ( which is a suclass of NSCollectionStream).
When I use a trait in NSSubclassOfCollectionStream that required the method #atEnd, the defintion in NSCollectionStream is not used and the method #atEnd is overridden with : atEnd self explicitRequirement.
I think we should take care about definitions in superclass, no? _______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
_______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
As we agree some times ago, the method
atEnd self explicitRequirement.
should not be defined. Alexandre On 8 Jul 2009, at 11:08, Cyrille Delaunay wrote:
Hi,
Here's my problem : - I have the class NSCollectionStream which define ( for example ) #atEnd - I have the class NSSubclassOfCollectionStream ( which is a suclass of NSCollectionStream).
When I use a trait in NSSubclassOfCollectionStream that required the method #atEnd, the defintion in NSCollectionStream is not used and the method #atEnd is overridden with : atEnd self explicitRequirement.
I think we should take care about definitions in superclass, no?
_______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
-- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
participants (4)
-
Adrian Lienhard -
Alexandre Bergel -
Cyrille Delaunay -
Stéphane Ducasse