Hi!
I am analyzing Mondrian in Moose. One thing that I am bumping into, is that wrong dependencies are inferred whereas I am sure that they do not exist. This is not a new, we have to type in Smalltalk. For example:
MOEdge>>fromPositions ^ shape fromPositions
#fromPositions is implemented 3 times, by MOAbstractLayout, MOEdge, and MOLineShape. In the case of MOEdge>>fromPositions, I am sure that it does not invoke MOAbstractLayout>>fromPositions
I guess that everyone analyzing smalltalk code bumped into this very situation.
My question is how do we do in that case? One easy solution, is to annotate Mondrian with an adequate pragma:
MOEdge>>fromPositions <invoke: #fromPositions of: MOLineShape> ^ shape fromPositions
We can decline this pragma into: <doesNotInvoke: #selector of: AClass> <invoke: #fromPositions of: #(MOLineShape MOEdge)>
Simon told me Cyrille has worked on this problem. I am currently trying to use RoelTyper.
Cheers, Alexandre
Hi Alex
Did you try the different strategies provided for invocations resolution in the importer wizard? Indeed, Cyrille coded a few strategies to use RoelTyper for better method resolution (as well as inference of instance variable types). Unfortunately, there are no test for this so I don't no how reliable it is. I would love to have tests and precision/recall score for these strategies!
On 18 août 2010, at 15:21, Alexandre Bergel wrote:
Hi!
I am analyzing Mondrian in Moose. One thing that I am bumping into, is that wrong dependencies are inferred whereas I am sure that they do not exist. This is not a new, we have to type in Smalltalk. For example:
MOEdge>>fromPositions ^ shape fromPositions
#fromPositions is implemented 3 times, by MOAbstractLayout, MOEdge, and MOLineShape. In the case of MOEdge>>fromPositions, I am sure that it does not invoke MOAbstractLayout>>fromPositions
I guess that everyone analyzing smalltalk code bumped into this very situation.
My question is how do we do in that case? One easy solution, is to annotate Mondrian with an adequate pragma:
MOEdge>>fromPositions <invoke: #fromPositions of: MOLineShape> ^ shape fromPositions
We can decline this pragma into: <doesNotInvoke: #selector of: AClass> <invoke: #fromPositions of: #(MOLineShape MOEdge)>
Simon told me Cyrille has worked on this problem. I am currently trying to use RoelTyper.
Cheers, Alexandre
-- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- Simon
I will try. In the meantime, I created a new issue. http://code.google.com/p/moose-technology/issues/detail?id=435
Cheers, Alexandre
On 18 Aug 2010, at 09:42, Simon Denier wrote:
Hi Alex
Did you try the different strategies provided for invocations resolution in the importer wizard? Indeed, Cyrille coded a few strategies to use RoelTyper for better method resolution (as well as inference of instance variable types). Unfortunately, there are no test for this so I don't no how reliable it is. I would love to have tests and precision/recall score for these strategies!
On 18 août 2010, at 15:21, Alexandre Bergel wrote:
Hi!
I am analyzing Mondrian in Moose. One thing that I am bumping into, is that wrong dependencies are inferred whereas I am sure that they do not exist. This is not a new, we have to type in Smalltalk. For example:
MOEdge>>fromPositions ^ shape fromPositions
#fromPositions is implemented 3 times, by MOAbstractLayout, MOEdge, and MOLineShape. In the case of MOEdge>>fromPositions, I am sure that it does not invoke MOAbstractLayout>>fromPositions
I guess that everyone analyzing smalltalk code bumped into this very situation.
My question is how do we do in that case? One easy solution, is to annotate Mondrian with an adequate pragma:
MOEdge>>fromPositions <invoke: #fromPositions of: MOLineShape> ^ shape fromPositions
We can decline this pragma into: <doesNotInvoke: #selector of: AClass> <invoke: #fromPositions of: #(MOLineShape MOEdge)>
Simon told me Cyrille has worked on this problem. I am currently trying to use RoelTyper.
Cheers, Alexandre
-- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- Simon
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
On 18 août 2010, at 16:27, Alexandre Bergel wrote:
I will try. In the meantime, I created a new issue. http://code.google.com/p/moose-technology/issues/detail?id=435
Maybe it would be better to open an issue about testing AbstractCandidateListOperator subclasses, since they already use RoelTyper.
Cyrille, could you make a quick summary of the strategies available?
Cheers, Alexandre
On 18 Aug 2010, at 09:42, Simon Denier wrote:
Hi Alex
Did you try the different strategies provided for invocations resolution in the importer wizard? Indeed, Cyrille coded a few strategies to use RoelTyper for better method resolution (as well as inference of instance variable types). Unfortunately, there are no test for this so I don't no how reliable it is. I would love to have tests and precision/recall score for these strategies!
On 18 août 2010, at 15:21, Alexandre Bergel wrote:
Hi!
I am analyzing Mondrian in Moose. One thing that I am bumping into, is that wrong dependencies are inferred whereas I am sure that they do not exist. This is not a new, we have to type in Smalltalk. For example:
MOEdge>>fromPositions ^ shape fromPositions
#fromPositions is implemented 3 times, by MOAbstractLayout, MOEdge, and MOLineShape. In the case of MOEdge>>fromPositions, I am sure that it does not invoke MOAbstractLayout>>fromPositions
I guess that everyone analyzing smalltalk code bumped into this very situation.
My question is how do we do in that case? One easy solution, is to annotate Mondrian with an adequate pragma:
MOEdge>>fromPositions <invoke: #fromPositions of: MOLineShape> ^ shape fromPositions
We can decline this pragma into: <doesNotInvoke: #selector of: AClass> <invoke: #fromPositions of: #(MOLineShape MOEdge)>
Simon told me Cyrille has worked on this problem. I am currently trying to use RoelTyper.
Cheers, Alexandre
-- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- Simon
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- Simon
So strategies available are:
a CandidateListOperator (the class used at the origin, could maybe be removed) compute a list of possible type for a FamixInvocation's receiver when: - the receiver is a Class (so its type is then evident :)) - the receiver is 'self' - the receiver is 'super'
a CandidateListOperatorAcceptingAnyKindOfReceiver compute a list of possible type for a FamixInvocation's receiver when: - the receiver is a Class (so its type is then evident :)) - the receiver is 'self' - the receiver is 'super' - the receiver is a FAMIXAttribute should normally accept any kind of receiver, but all type are not yet supported or implemented.
a CandidateListOperatorAcceptingBasicAndVariableReceivers compute a list of possible type for a FamixInvocation's receiver when: - the receiver is a Class (so its type is then evident :)) - the receiver is 'self' - the receiver is 'super' - the receiver is a FAMIXAttribute
a CandidateListOperatorAcceptingSimpleKindsOfReceiver compute a list of possible type for a FamixInvocation's receiver when: - the receiver is a Class (so its type is then evident :)) - the receiver is 'self' - the receiver is 'super' (so does the same job than the original class)
a CandidateListOperatorNotAcceptingAnyReceiver doesn't compute any possible type.
should normally accept any kind of receiver, but all type are not yet supported or implemented. 2010/8/18 Simon Denier Simon.Denier@inria.fr
On 18 août 2010, at 16:27, Alexandre Bergel wrote:
I will try. In the meantime, I created a new issue. http://code.google.com/p/moose-technology/issues/detail?id=435
Maybe it would be better to open an issue about testing AbstractCandidateListOperator subclasses, since they already use RoelTyper.
Cyrille, could you make a quick summary of the strategies available?
Cheers, Alexandre
On 18 Aug 2010, at 09:42, Simon Denier wrote:
Hi Alex
Did you try the different strategies provided for invocations resolution
in the importer wizard? Indeed, Cyrille coded a few strategies to use RoelTyper for better method resolution (as well as inference of instance variable types). Unfortunately, there are no test for this so I don't no how reliable it is. I would love to have tests and precision/recall score for these strategies!
On 18 août 2010, at 15:21, Alexandre Bergel wrote:
Hi!
I am analyzing Mondrian in Moose. One thing that I am bumping into, is
that wrong dependencies are inferred whereas I am sure that they do not exist. This is not a new, we have to type in Smalltalk. For example:
MOEdge>>fromPositions ^ shape fromPositions
#fromPositions is implemented 3 times, by MOAbstractLayout, MOEdge, and
MOLineShape.
In the case of MOEdge>>fromPositions, I am sure that it does not invoke
MOAbstractLayout>>fromPositions
I guess that everyone analyzing smalltalk code bumped into this very
situation.
My question is how do we do in that case? One easy solution, is to annotate Mondrian with an adequate pragma:
MOEdge>>fromPositions <invoke: #fromPositions of: MOLineShape> ^ shape fromPositions
We can decline this pragma into: <doesNotInvoke: #selector of: AClass> <invoke: #fromPositions of: #(MOLineShape MOEdge)>
Simon told me Cyrille has worked on this problem. I am currently trying
to use RoelTyper.
Cheers, Alexandre
-- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- Simon
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- Simon
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
On 18 août 2010, at 17:07, Cyrille Delaunay wrote:
So strategies available are:
a CandidateListOperator (the class used at the origin, could maybe be removed) compute a list of possible type for a FamixInvocation's receiver when:
- the receiver is a Class (so its type is then evident :))
- the receiver is 'self'
- the receiver is 'super'
So now it''s the same as CandidateListOperatorAcceptingSimpleKindsOfReceiver right?
a CandidateListOperatorAcceptingAnyKindOfReceiver compute a list of possible type for a FamixInvocation's receiver when:
- the receiver is a Class (so its type is then evident :))
- the receiver is 'self'
- the receiver is 'super'
- the receiver is a FAMIXAttribute
should normally accept any kind of receiver, but all type are not yet supported or implemented.
CandidateListOperatorAcceptingBasicAndVariableReceivers CandidateListOperatorAcceptingSimpleKindsOfReceiver are nowadays the same?
Nowadays there are some tests for candidates, which imply testing CandidateListOperator. We should have independent tests for each strategy (best would be to extract the current tests for candidates, and also start testing CandidateListOperatorAcceptingBasicAndVariableReceivers which might have good precision)
a CandidateListOperatorAcceptingBasicAndVariableReceivers compute a list of possible type for a FamixInvocation's receiver when:
- the receiver is a Class (so its type is then evident :))
- the receiver is 'self'
- the receiver is 'super'
- the receiver is a FAMIXAttribute
a CandidateListOperatorAcceptingSimpleKindsOfReceiver compute a list of possible type for a FamixInvocation's receiver when:
- the receiver is a Class (so its type is then evident :))
- the receiver is 'self'
- the receiver is 'super' (so does the same job than the original class)
a CandidateListOperatorNotAcceptingAnyReceiver doesn't compute any possible type.
should normally accept any kind of receiver, but all type are not yet supported or implemented. 2010/8/18 Simon Denier Simon.Denier@inria.fr
On 18 août 2010, at 16:27, Alexandre Bergel wrote:
I will try. In the meantime, I created a new issue. http://code.google.com/p/moose-technology/issues/detail?id=435
Maybe it would be better to open an issue about testing AbstractCandidateListOperator subclasses, since they already use RoelTyper.
Cyrille, could you make a quick summary of the strategies available?
Cheers, Alexandre
On 18 Aug 2010, at 09:42, Simon Denier wrote:
Hi Alex
Did you try the different strategies provided for invocations resolution in the importer wizard? Indeed, Cyrille coded a few strategies to use RoelTyper for better method resolution (as well as inference of instance variable types). Unfortunately, there are no test for this so I don't no how reliable it is. I would love to have tests and precision/recall score for these strategies!
On 18 août 2010, at 15:21, Alexandre Bergel wrote:
Hi!
I am analyzing Mondrian in Moose. One thing that I am bumping into, is that wrong dependencies are inferred whereas I am sure that they do not exist. This is not a new, we have to type in Smalltalk. For example:
MOEdge>>fromPositions ^ shape fromPositions
#fromPositions is implemented 3 times, by MOAbstractLayout, MOEdge, and MOLineShape. In the case of MOEdge>>fromPositions, I am sure that it does not invoke MOAbstractLayout>>fromPositions
I guess that everyone analyzing smalltalk code bumped into this very situation.
My question is how do we do in that case? One easy solution, is to annotate Mondrian with an adequate pragma:
MOEdge>>fromPositions <invoke: #fromPositions of: MOLineShape> ^ shape fromPositions
We can decline this pragma into: <doesNotInvoke: #selector of: AClass> <invoke: #fromPositions of: #(MOLineShape MOEdge)>
Simon told me Cyrille has worked on this problem. I am currently trying to use RoelTyper.
Cheers, Alexandre
-- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- Simon
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- Simon
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- Simon
2010/8/18 Simon Denier Simon.Denier@inria.fr
On 18 août 2010, at 17:07, Cyrille Delaunay wrote:
So strategies available are:
a CandidateListOperator (the class used at the origin, could maybe be removed) compute a list of possible type for a FamixInvocation's receiver when:
- the receiver is a Class (so its type is then evident :))
- the receiver is 'self'
- the receiver is 'super'
So now it''s the same as CandidateListOperatorAcceptingSimpleKindsOfReceiver right?
yes
a CandidateListOperatorAcceptingAnyKindOfReceiver compute a list of possible type for a FamixInvocation's receiver when:
- the receiver is a Class (so its type is then evident :))
- the receiver is 'self'
- the receiver is 'super'
- the receiver is a FAMIXAttribute
should normally accept any kind of receiver, but all type are not yet supported or implemented.
CandidateListOperatorAcceptingBasicAndVariableReceivers CandidateListOperatorAcceptingSimpleKindsOfReceiver are nowadays the same?
yes
Nowadays there are some tests for candidates, which imply testing CandidateListOperator. We should have independent tests for each strategy (best would be to extract the current tests for candidates, and also start testing CandidateListOperatorAcceptingBasicAndVariableReceivers which might have good precision)
a CandidateListOperatorAcceptingBasicAndVariableReceivers compute a list of possible type for a FamixInvocation's receiver when:
- the receiver is a Class (so its type is then evident :))
- the receiver is 'self'
- the receiver is 'super'
- the receiver is a FAMIXAttribute
a CandidateListOperatorAcceptingSimpleKindsOfReceiver compute a list of possible type for a FamixInvocation's receiver when:
- the receiver is a Class (so its type is then evident :))
- the receiver is 'self'
- the receiver is 'super' (so does the same job than the original class)
a CandidateListOperatorNotAcceptingAnyReceiver doesn't compute any possible type.
should normally accept any kind of receiver, but all type are not yet supported or implemented. 2010/8/18 Simon Denier Simon.Denier@inria.fr
On 18 août 2010, at 16:27, Alexandre Bergel wrote:
I will try. In the meantime, I created a new issue. http://code.google.com/p/moose-technology/issues/detail?id=435
Maybe it would be better to open an issue about testing AbstractCandidateListOperator subclasses, since they already use RoelTyper.
Cyrille, could you make a quick summary of the strategies available?
Cheers, Alexandre
On 18 Aug 2010, at 09:42, Simon Denier wrote:
Hi Alex
Did you try the different strategies provided for invocations
resolution in the importer wizard? Indeed, Cyrille coded a few strategies to use RoelTyper for better method resolution (as well as inference of instance variable types). Unfortunately, there are no test for this so I don't no how reliable it is. I would love to have tests and precision/recall score for these strategies!
On 18 août 2010, at 15:21, Alexandre Bergel wrote:
Hi!
I am analyzing Mondrian in Moose. One thing that I am bumping into, is
that wrong dependencies are inferred whereas I am sure that they do not exist. This is not a new, we have to type in Smalltalk. For example:
MOEdge>>fromPositions ^ shape fromPositions
#fromPositions is implemented 3 times, by MOAbstractLayout, MOEdge,
and MOLineShape.
In the case of MOEdge>>fromPositions, I am sure that it does not
invoke MOAbstractLayout>>fromPositions
I guess that everyone analyzing smalltalk code bumped into this very
situation.
My question is how do we do in that case? One easy solution, is to annotate Mondrian with an adequate pragma:
MOEdge>>fromPositions <invoke: #fromPositions of: MOLineShape> ^ shape fromPositions
We can decline this pragma into: <doesNotInvoke: #selector of: AClass> <invoke: #fromPositions of: #(MOLineShape MOEdge)>
Simon told me Cyrille has worked on this problem. I am currently
trying to use RoelTyper.
Cheers, Alexandre
-- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- Simon
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- Simon
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- Simon
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
Yeah, this is what I thought. Maybe we could introduce CandidateListTypeInferencer that use Roel Typer?
I entered what you said on http://code.google.com/p/moose-technology/issues/detail?id=435 Maybe something can be added to the Moose Book.
Cheers, Alexandre
On 18 Aug 2010, at 11:07, Cyrille Delaunay wrote:
So strategies available are:
a CandidateListOperator (the class used at the origin, could maybe be removed) compute a list of possible type for a FamixInvocation's receiver when:
- the receiver is a Class (so its type is then evident :))
- the receiver is 'self'
- the receiver is 'super'
a CandidateListOperatorAcceptingAnyKindOfReceiver compute a list of possible type for a FamixInvocation's receiver when:
- the receiver is a Class (so its type is then evident :))
- the receiver is 'self'
- the receiver is 'super'
- the receiver is a FAMIXAttribute
should normally accept any kind of receiver, but all type are not yet supported or implemented.
a CandidateListOperatorAcceptingBasicAndVariableReceivers compute a list of possible type for a FamixInvocation's receiver when:
- the receiver is a Class (so its type is then evident :))
- the receiver is 'self'
- the receiver is 'super'
- the receiver is a FAMIXAttribute
a CandidateListOperatorAcceptingSimpleKindsOfReceiver compute a list of possible type for a FamixInvocation's receiver when:
- the receiver is a Class (so its type is then evident :))
- the receiver is 'self'
- the receiver is 'super' (so does the same job than the original class)
a CandidateListOperatorNotAcceptingAnyReceiver doesn't compute any possible type.
should normally accept any kind of receiver, but all type are not yet supported or implemented. 2010/8/18 Simon Denier Simon.Denier@inria.fr
On 18 août 2010, at 16:27, Alexandre Bergel wrote:
I will try. In the meantime, I created a new issue. http://code.google.com/p/moose-technology/issues/detail?id=435
Maybe it would be better to open an issue about testing AbstractCandidateListOperator subclasses, since they already use RoelTyper.
Cyrille, could you make a quick summary of the strategies available?
Cheers, Alexandre
On 18 Aug 2010, at 09:42, Simon Denier wrote:
Hi Alex
Did you try the different strategies provided for invocations resolution in the importer wizard? Indeed, Cyrille coded a few strategies to use RoelTyper for better method resolution (as well as inference of instance variable types). Unfortunately, there are no test for this so I don't no how reliable it is. I would love to have tests and precision/recall score for these strategies!
On 18 août 2010, at 15:21, Alexandre Bergel wrote:
Hi!
I am analyzing Mondrian in Moose. One thing that I am bumping into, is that wrong dependencies are inferred whereas I am sure that they do not exist. This is not a new, we have to type in Smalltalk. For example:
MOEdge>>fromPositions ^ shape fromPositions
#fromPositions is implemented 3 times, by MOAbstractLayout, MOEdge, and MOLineShape. In the case of MOEdge>>fromPositions, I am sure that it does not invoke MOAbstractLayout>>fromPositions
I guess that everyone analyzing smalltalk code bumped into this very situation.
My question is how do we do in that case? One easy solution, is to annotate Mondrian with an adequate pragma:
MOEdge>>fromPositions <invoke: #fromPositions of: MOLineShape> ^ shape fromPositions
We can decline this pragma into: <doesNotInvoke: #selector of: AClass> <invoke: #fromPositions of: #(MOLineShape MOEdge)>
Simon told me Cyrille has worked on this problem. I am currently trying to use RoelTyper.
Cheers, Alexandre
-- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- Simon
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- Simon
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
On 18 août 2010, at 17:31, Alexandre Bergel wrote:
Yeah, this is what I thought. Maybe we could introduce CandidateListTypeInferencer that use Roel Typer?
Once again, Roel Typer is already used in CandidateListOperatorAcceptingBasicAndVariableReceivers and CandidateListOperatorAcceptingAnyKindOfReceiver to compute possible types for attributes.
See #computeInstanceVariableCandidateListFor:
Now there is still work to do
I entered what you said on http://code.google.com/p/moose-technology/issues/detail?id=435 Maybe something can be added to the Moose Book.
Cheers, Alexandre
On 18 Aug 2010, at 11:07, Cyrille Delaunay wrote:
So strategies available are:
a CandidateListOperator (the class used at the origin, could maybe be removed) compute a list of possible type for a FamixInvocation's receiver when:
- the receiver is a Class (so its type is then evident :))
- the receiver is 'self'
- the receiver is 'super'
a CandidateListOperatorAcceptingAnyKindOfReceiver compute a list of possible type for a FamixInvocation's receiver when:
- the receiver is a Class (so its type is then evident :))
- the receiver is 'self'
- the receiver is 'super'
- the receiver is a FAMIXAttribute
should normally accept any kind of receiver, but all type are not yet supported or implemented.
a CandidateListOperatorAcceptingBasicAndVariableReceivers compute a list of possible type for a FamixInvocation's receiver when:
- the receiver is a Class (so its type is then evident :))
- the receiver is 'self'
- the receiver is 'super'
- the receiver is a FAMIXAttribute
a CandidateListOperatorAcceptingSimpleKindsOfReceiver compute a list of possible type for a FamixInvocation's receiver when:
- the receiver is a Class (so its type is then evident :))
- the receiver is 'self'
- the receiver is 'super' (so does the same job than the original class)
a CandidateListOperatorNotAcceptingAnyReceiver doesn't compute any possible type.
should normally accept any kind of receiver, but all type are not yet supported or implemented. 2010/8/18 Simon Denier Simon.Denier@inria.fr
On 18 août 2010, at 16:27, Alexandre Bergel wrote:
I will try. In the meantime, I created a new issue. http://code.google.com/p/moose-technology/issues/detail?id=435
Maybe it would be better to open an issue about testing AbstractCandidateListOperator subclasses, since they already use RoelTyper.
Cyrille, could you make a quick summary of the strategies available?
Cheers, Alexandre
On 18 Aug 2010, at 09:42, Simon Denier wrote:
Hi Alex
Did you try the different strategies provided for invocations resolution in the importer wizard? Indeed, Cyrille coded a few strategies to use RoelTyper for better method resolution (as well as inference of instance variable types). Unfortunately, there are no test for this so I don't no how reliable it is. I would love to have tests and precision/recall score for these strategies!
On 18 août 2010, at 15:21, Alexandre Bergel wrote:
Hi!
I am analyzing Mondrian in Moose. One thing that I am bumping into, is that wrong dependencies are inferred whereas I am sure that they do not exist. This is not a new, we have to type in Smalltalk. For example:
MOEdge>>fromPositions ^ shape fromPositions
#fromPositions is implemented 3 times, by MOAbstractLayout, MOEdge, and MOLineShape. In the case of MOEdge>>fromPositions, I am sure that it does not invoke MOAbstractLayout>>fromPositions
I guess that everyone analyzing smalltalk code bumped into this very situation.
My question is how do we do in that case? One easy solution, is to annotate Mondrian with an adequate pragma:
MOEdge>>fromPositions <invoke: #fromPositions of: MOLineShape> ^ shape fromPositions
We can decline this pragma into: <doesNotInvoke: #selector of: AClass> <invoke: #fromPositions of: #(MOLineShape MOEdge)>
Simon told me Cyrille has worked on this problem. I am currently trying to use RoelTyper.
Cheers, Alexandre
-- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- Simon
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- Simon
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- Simon
On Aug 18, 2010, at 3:21 PM, Alexandre Bergel wrote:
Hi!
I am analyzing Mondrian in Moose.
Excellent this is the best thing that we can do: using moose to analyse our own software. We want to do that for pharo too.
One thing that I am bumping into, is that wrong dependencies are inferred whereas I am sure that they do not exist. This is not a new, we have to type in Smalltalk. For example:
MOEdge>>fromPositions ^ shape fromPositions
#fromPositions is implemented 3 times, by MOAbstractLayout, MOEdge, and MOLineShape. In the case of MOEdge>>fromPositions, I am sure that it does not invoke MOAbstractLayout>>fromPositions
yes but how a analyser can get that?
I guess that everyone analyzing smalltalk code bumped into this very situation.
My question is how do we do in that case? One easy solution, is to annotate Mondrian with an adequate pragma:
MOEdge>>fromPositions <invoke: #fromPositions of: MOLineShape> ^ shape fromPositions
the problem with static information is that it may fast get desynchronised
We can decline this pragma into: <doesNotInvoke: #selector of: AClass> <invoke: #fromPositions of: #(MOLineShape MOEdge)>
Simon told me Cyrille has worked on this problem. I am currently trying to use RoelTyper.
Cheers, Alexandre
-- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
Hi,
On 19 Aug 2010, at 11:27, Stéphane Ducasse wrote:
On Aug 18, 2010, at 3:21 PM, Alexandre Bergel wrote:
Hi!
I am analyzing Mondrian in Moose.
Excellent this is the best thing that we can do: using moose to analyse our own software. We want to do that for pharo too.
One thing that I am bumping into, is that wrong dependencies are inferred whereas I am sure that they do not exist. This is not a new, we have to type in Smalltalk. For example:
MOEdge>>fromPositions ^ shape fromPositions
#fromPositions is implemented 3 times, by MOAbstractLayout, MOEdge, and MOLineShape. In the case of MOEdge>>fromPositions, I am sure that it does not invoke MOAbstractLayout>>fromPositions
yes but how a analyser can get that?
I guess that everyone analyzing smalltalk code bumped into this very situation.
My question is how do we do in that case? One easy solution, is to annotate Mondrian with an adequate pragma:
MOEdge>>fromPositions <invoke: #fromPositions of: MOLineShape> ^ shape fromPositions
the problem with static information is that it may fast get desynchronised
I agree. Except if you have a nice PluggableType system that can also ensure some level of checking, you do not want to have these kind of annotations in your source code.
Cheers, Doru
We can decline this pragma into: <doesNotInvoke: #selector of: AClass> <invoke: #fromPositions of: #(MOLineShape MOEdge)>
Simon told me Cyrille has worked on this problem. I am currently trying to use RoelTyper.
Cheers, Alexandre
-- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- www.tudorgirba.com
"Obvious things are difficult to teach."
-
Alexandre Bergel -
Cyrille Delaunay -
Simon Denier -
Stéphane Ducasse -
Tudor Girba