12 Oct
2010
12 Oct
'10
8:58 a.m.
Wow, that's a hairy problem. I just read the thread a few minutes ago. I agree that
using a description does not solve the problem. But if I'm not wrong then the problem
exists for every object where
anObject copy == anObject
right? (Btw. is there a selector in Smalltalk that can test this?) On the other hand we
are talking about preventing infinite loops, right? Is there a chance that any of the
objects that meet the condition above can trigger an infinite recursion? I don't think
so. So it should be sufficient to manage the seen list only for the object that do no meet
the condition. Although the above is neither nice nor fast.
Norbert
On 12.10.2010, at 08:14, Lukas Renggli wrote:
I thought about that too, but this is not a solution:
It can be that
different objects are described with the same description instances.
Lukas
On 12 October 2010 02:46, John McKeon <p3anoman(a)gmail.com> wrote:
I've posted a mcz
to http://code.google.com/p/magritte-metamodel/issues/detail?id=7
It uses the description instead of the object in the collection of
descriptions already visited
John
On Mon, Oct 11, 2010 at 6:11 AM, Lukas Renggli <renggli(a)gmail.com> wrote:
--
Lukas Renggli
www.lukas-renggli.ch
_______________________________________________
Magritte, Pier and Related Tools ...
https://www.iam.unibe.ch/mailman/listinfo/smallwiki
That just remembers (in 'previous') and restores the currently
validated object while (in 'object') recursively walking through the
graph (which is a tree in 99% of the cases).
Lukas
On 11 October 2010 12:05, Nick Ager <nick.ager(a)gmail.com> wrote:
--
http://john-mckeon.us
_______________________________________________
Magritte, Pier and Related Tools ...
https://www.iam.unibe.ch/mailman/listinfo/smallwiki
On 11 October 2010 10:53, Lukas Renggli <renggli(a)gmail.com> wrote:
>
>> (seen includes: anObject)
>> ifTrue: [ ^ self ].
>> anObject isNil
>> ifFalse: [ seen add: anObject ].
>
> That part just tries to avoid to go into recursion if a described
> object refers to another object that descriptions refer back to the
> original object. That was requested at some point in the past, but I
> never dependent on that myself.
I understood that part but don't understand the intent of:
previous := object. object := anObject.
aBlock ensure: [ object := previous ]
_______________________________________________
Magritte, Pier and Related Tools ...
https://www.iam.unibe.ch/mailman/listinfo/smallwiki
--
Lukas Renggli
www.lukas-renggli.ch
_______________________________________________
Magritte, Pier and Related Tools ...
https://www.iam.unibe.ch/mailman/listinfo/smallwiki