If we are going test this, (I don't know if we already have it) would be useful to test a basic message sending. And to have�a testWellCompiledMethods (or something like that) that checks if all the methods are correctly compiled, e.g.: a method which instead of referencing the instance variable that says in it's source, references the next one. I think that something of that is actually implemented in SmallLint for VASmalltalk, but I may be wrong...

This tests could be some sort of: "if this doesn't work, it's all FUBAR" :)

+1 to Alexandre's version.

Cheers,

Mariano.

On Tue, May 11, 2010 at 5:39 PM, St�phane Ducasse <stephane.ducasse@inria.fr> wrote:

On May 11, 2010, at 9:34 PM, Alexandre Bergel wrote:

>> testMetaclassSuperclass
>> � � �"self run: #testMetaclassSuperclass"
>>
>> � � �self assert: Dictionary class superclass == Set class.
>> � � �self assert: OrderedCollection class superclass == SequenceableCollection class.
>
> What is the important is the metaclass relationship, and not really the fact that Dictionary is a subclass of Set.
> I think this is better:
> � � � self assert: Dictionary class superclass == Dictionary superclass class.
> � � � self assert: OrderedCollection class superclass == OrderedCollection superclass class.

ok
but do you think that the system would work if such test would not work?

>> I would rewrite the test as:
>
> testSuperclass
> � � � "self debug: #testSuperclass"
>
> � � � | s b |
>
> � � � s := OrderedCollection new.
> � � � b := [:cls | cls ifNotNil: [s add: cls. b value: cls superclass] ].
> � � � b value: OrderedCollection.
>
> � � � self assert: OrderedCollection allSuperclasses = s allButFirst.
> � � � self assert: OrderedCollection withAllSuperclasses = s.
>
> You even test #withAllSuperclasses in that case.

this one is more interesting

_______________________________________________
Pharo-project mailing list
Pharo-project@lists.gforge.inria.fr
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project