FAMIXMethod>>parentPackage
Locked
 
|
Hi,
I changed the behavior of FAMIXMethod>>parentPackage. In fact, when a method is in the same package than its class, its parentPackage should be nil. So, to not break the behavior, I changed the accessor like this: === parentPackage ^ parentPackage ifNil:[ self belongsTo parentPackage] === But this breaks during importation of MSE files (it breaks some tests of smalldude and eyesee). The problem is due to the class that is not already defined during the importation. So I changed the source code like that. I am not convince of the efficiency of this code. Have you another idea ? === parentPackage ^ parentPackage ifNil:[ |cl| cl := self belongsTo. cl isNil ifTrue:[nil] ifFalse:[cl parentPackage] ] === --- Jannik Laval _______________________________________________ Moose-dev mailing list [hidden email] https://www.iam.unibe.ch/mailman/listinfo/moose-dev |
|
|
Looks bad in my opinion. Why is there some dependency between MSE import and smalldude/eyesee tests? On 4 mai 2011, at 09:14, Laval Jannik wrote: > Hi, > > I changed the behavior of FAMIXMethod>>parentPackage. > > In fact, when a method is in the same package than its class, its parentPackage should be nil. > > So, to not break the behavior, I changed the accessor like this: > > === > parentPackage > ^ parentPackage ifNil:[ self belongsTo parentPackage] > === > > But this breaks during importation of MSE files (it breaks some tests of smalldude and eyesee). > The problem is due to the class that is not already defined during the importation. > So I changed the source code like that. I am not convince of the efficiency of this code. > Have you another idea ? > > === > parentPackage > > ^ parentPackage ifNil:[ > |cl| > cl := self belongsTo. > cl isNil > ifTrue:[nil] > ifFalse:[cl parentPackage] > ] > === > > --- > Jannik Laval > > > _______________________________________________ > Moose-dev mailing list > [hidden email] > https://www.iam.unibe.ch/mailman/listinfo/moose-dev -- Simon Denier _______________________________________________ Moose-dev mailing list [hidden email] https://www.iam.unibe.ch/mailman/listinfo/moose-dev |
|
|
for eyesee, I do not know.
For smallDude, a test import a model from a MSE Stream. I think that this case should be tested in MSE importer tests. Jannik On May 4, 2011, at 11:20 , Simon Denier wrote: > > Looks bad in my opinion. Why is there some dependency between MSE import and smalldude/eyesee tests? > > > On 4 mai 2011, at 09:14, Laval Jannik wrote: > >> Hi, >> >> I changed the behavior of FAMIXMethod>>parentPackage. >> >> In fact, when a method is in the same package than its class, its parentPackage should be nil. >> >> So, to not break the behavior, I changed the accessor like this: >> >> === >> parentPackage >> ^ parentPackage ifNil:[ self belongsTo parentPackage] >> === >> >> But this breaks during importation of MSE files (it breaks some tests of smalldude and eyesee). >> The problem is due to the class that is not already defined during the importation. >> So I changed the source code like that. I am not convince of the efficiency of this code. >> Have you another idea ? >> >> === >> parentPackage >> >> ^ parentPackage ifNil:[ >> |cl| >> cl := self belongsTo. >> cl isNil >> ifTrue:[nil] >> ifFalse:[cl parentPackage] >> ] >> === >> >> --- >> Jannik Laval >> >> >> _______________________________________________ >> Moose-dev mailing list >> [hidden email] >> https://www.iam.unibe.ch/mailman/listinfo/moose-dev > > -- > Simon Denier > > > > > _______________________________________________ > Moose-dev mailing list > [hidden email] > https://www.iam.unibe.ch/mailman/listinfo/moose-dev _______________________________________________ Moose-dev mailing list [hidden email] https://www.iam.unibe.ch/mailman/listinfo/moose-dev |
|
|
Hello,
There are no tests in EyeSee related to import model from MSE stream. Jannik could you send me an image that I can reproduce this error in EyeSee? (I should only run EyeSee tests in this images? regards, On Wed, May 4, 2011 at 11:56 AM, jannik.laval <[hidden email]> wrote: for eyesee, I do not know. -- Andre Hora _______________________________________________ Moose-dev mailing list [hidden email] https://www.iam.unibe.ch/mailman/listinfo/moose-dev |
|
|
I am not sure that eyesee is related.
You can take an image and see the report on hudson: On May 4, 2011, at 12:28 , Andre Hora wrote: Hello, _______________________________________________ Moose-dev mailing list [hidden email] https://www.iam.unibe.ch/mailman/listinfo/moose-dev |
|
|
Jannik, I see in Hudson that we have this error with EyeSee since #330
Apr 29, 2011 3:57:59 PM
So, I think it is not related with your change. On Wed, May 4, 2011 at 2:34 PM, jannik.laval <[hidden email]> wrote:
-- Andre Hora _______________________________________________ Moose-dev mailing list [hidden email] https://www.iam.unibe.ch/mailman/listinfo/moose-dev |
|
|
Good and bad news :)
That is a good news for my changes. The bad news is that the bug is 10 days old... Anyone know eyesee better than me to fix it ? Cheers, Jannik On May 4, 2011, at 14:53 , Andre Hora wrote: Jannik, I see in Hudson that we have this error with EyeSee since #330 Apr 29, 2011 3:57:59 PM _______________________________________________ Moose-dev mailing list [hidden email] https://www.iam.unibe.ch/mailman/listinfo/moose-dev |
|
|
It is related to the integration with Glamour. I can have a look.
ps: not 10 days, 5 days ;) On Wed, May 4, 2011 at 2:59 PM, jannik.laval <[hidden email]> wrote:
-- Andre Hora _______________________________________________ Moose-dev mailing list [hidden email] https://www.iam.unibe.ch/mailman/listinfo/moose-dev |
|
|
Fixed the bug Glamour.Tests.EyeSee.GLMEyeSeeMorphicTest.testSpawn.
For what I saw in the comments of Doru, this bug was also related to run the test in Hudson, so, could run the job in Hudson ? On Wed, May 4, 2011 at 3:08 PM, Andre Hora <[hidden email]> wrote: It is related to the integration with Glamour. I can have a look. -- Andre Hora _______________________________________________ Moose-dev mailing list [hidden email] https://www.iam.unibe.ch/mailman/listinfo/moose-dev |
|
|
In reply to this post by Andre Hora
:) iep, I would like to say 10 builds (but this is 9).
Jannik On May 4, 2011, at 15:08 , Andre Hora wrote: It is related to the integration with Glamour. I can have a look. _______________________________________________ Moose-dev mailing list [hidden email] https://www.iam.unibe.ch/mailman/listinfo/moose-dev |
|
|
In reply to this post by Laval Jannik-2
Hi Jannik,
Thanks for looking into this. There is a problem with this implementation: it breaks the contract of basic attributes in Moose. MooseNamedEntity>>parentPackage has as opposite FAMIXPackage>>childNamedEntities. This means that the contract is that if parentPackage is not nil then the entity is to be found in the childNamedEntities collection. In this case, the reason for parentPackage is that you want to use it as a unified API for querying the parent package scope. Chef already provides this API intention in the packageScope. So, we should move the computation from parentPackage in packageScope, and then have queries depend on packageScope. Cheers, Doru On 4 May 2011, at 09:14, Laval Jannik wrote: > Hi, > > I changed the behavior of FAMIXMethod>>parentPackage. > > In fact, when a method is in the same package than its class, its parentPackage should be nil. > > So, to not break the behavior, I changed the accessor like this: > > === > parentPackage > ^ parentPackage ifNil:[ self belongsTo parentPackage] > === > > But this breaks during importation of MSE files (it breaks some tests of smalldude and eyesee). > The problem is due to the class that is not already defined during the importation. > So I changed the source code like that. I am not convince of the efficiency of this code. > Have you another idea ? > > === > parentPackage > > ^ parentPackage ifNil:[ > |cl| > cl := self belongsTo. > cl isNil > ifTrue:[nil] > ifFalse:[cl parentPackage] > ] > === > > --- > Jannik Laval > > > _______________________________________________ > Moose-dev mailing list > [hidden email] > https://www.iam.unibe.ch/mailman/listinfo/moose-dev -- www.tudorgirba.com "Yesterday is a fact. Tomorrow is a possibility. Today is a challenge." _______________________________________________ Moose-dev mailing list [hidden email] https://www.iam.unibe.ch/mailman/listinfo/moose-dev |
|
|
In reply to this post by Andre Hora
Hi,
Thanks a lot, Andre. I was getting crazy with this bug because I could not reproduce it on Mac. I restarted the build and everything is green. Hurray! :) Cheers, Doru On 4 May 2011, at 15:49, Andre Hora wrote: > Fixed the bug Glamour.Tests.EyeSee.GLMEyeSeeMorphicTest.testSpawn. > For what I saw in the comments of Doru, this bug was also related to run the test in Hudson, so, could run the job in Hudson ? > > On Wed, May 4, 2011 at 3:08 PM, Andre Hora <[hidden email]> wrote: > It is related to the integration with Glamour. I can have a look. > ps: not 10 days, 5 days ;) > > > On Wed, May 4, 2011 at 2:59 PM, jannik.laval <[hidden email]> wrote: > Good and bad news :) > > That is a good news for my changes. > The bad news is that the bug is 10 days old... > Anyone know eyesee better than me to fix it ? > > Cheers, > Jannik > > On May 4, 2011, at 14:53 , Andre Hora wrote: > >> Jannik, I see in Hudson that we have this error with EyeSee since #330 Apr 29, 2011 3:57:59 PM >> So, I think it is not related with your change. >> >> On Wed, May 4, 2011 at 2:34 PM, jannik.laval <[hidden email]> wrote: >> I am not sure that eyesee is related. >> You can take an image and see the report on hudson: >> http://hudson.moosetechnology.org/job/moose-latest-dev/lastCompletedBuild/testReport/ >> >> Jannik >> >> On May 4, 2011, at 12:28 , Andre Hora wrote: >> >>> Hello, >>> >>> There are no tests in EyeSee related to import model from MSE stream. >>> Jannik could you send me an image that I can reproduce this error in EyeSee? (I should only run EyeSee tests in this images? >>> >>> regards, >>> >>> On Wed, May 4, 2011 at 11:56 AM, jannik.laval <[hidden email]> wrote: >>> for eyesee, I do not know. >>> For smallDude, a test import a model from a MSE Stream. >>> >>> I think that this case should be tested in MSE importer tests. >>> >>> Jannik >>> >>> >>> On May 4, 2011, at 11:20 , Simon Denier wrote: >>> >>> > >>> > Looks bad in my opinion. Why is there some dependency between MSE import and smalldude/eyesee tests? >>> > >>> > >>> > On 4 mai 2011, at 09:14, Laval Jannik wrote: >>> > >>> >> Hi, >>> >> >>> >> I changed the behavior of FAMIXMethod>>parentPackage. >>> >> >>> >> In fact, when a method is in the same package than its class, its parentPackage should be nil. >>> >> >>> >> So, to not break the behavior, I changed the accessor like this: >>> >> >>> >> === >>> >> parentPackage >>> >> ^ parentPackage ifNil:[ self belongsTo parentPackage] >>> >> === >>> >> >>> >> But this breaks during importation of MSE files (it breaks some tests of smalldude and eyesee). >>> >> The problem is due to the class that is not already defined during the importation. >>> >> So I changed the source code like that. I am not convince of the efficiency of this code. >>> >> Have you another idea ? >>> >> >>> >> === >>> >> parentPackage >>> >> >>> >> ^ parentPackage ifNil:[ >>> >> |cl| >>> >> cl := self belongsTo. >>> >> cl isNil >>> >> ifTrue:[nil] >>> >> ifFalse:[cl parentPackage] >>> >> ] >>> >> === >>> >> >>> >> --- >>> >> Jannik Laval >>> >> >>> >> >>> >> _______________________________________________ >>> >> Moose-dev mailing list >>> >> [hidden email] >>> >> https://www.iam.unibe.ch/mailman/listinfo/moose-dev >>> > >>> > -- >>> > Simon Denier >>> > >>> > >>> > >>> > >>> > _______________________________________________ >>> > Moose-dev mailing list >>> > [hidden email] >>> > https://www.iam.unibe.ch/mailman/listinfo/moose-dev >>> >>> >>> _______________________________________________ >>> Moose-dev mailing list >>> [hidden email] >>> https://www.iam.unibe.ch/mailman/listinfo/moose-dev >>> >>> >>> >>> -- >>> Andre Hora >>> >>> _______________________________________________ >>> Moose-dev mailing list >>> [hidden email] >>> https://www.iam.unibe.ch/mailman/listinfo/moose-dev >> >> >> _______________________________________________ >> Moose-dev mailing list >> [hidden email] >> https://www.iam.unibe.ch/mailman/listinfo/moose-dev >> >> >> >> >> -- >> Andre Hora >> >> _______________________________________________ >> Moose-dev mailing list >> [hidden email] >> https://www.iam.unibe.ch/mailman/listinfo/moose-dev > > > _______________________________________________ > Moose-dev mailing list > [hidden email] > https://www.iam.unibe.ch/mailman/listinfo/moose-dev > > > > > -- > Andre Hora > > > > > -- > Andre Hora > > _______________________________________________ > Moose-dev mailing list > [hidden email] > https://www.iam.unibe.ch/mailman/listinfo/moose-dev -- www.tudorgirba.com "Sometimes the best solution is not the best solution." _______________________________________________ Moose-dev mailing list [hidden email] https://www.iam.unibe.ch/mailman/listinfo/moose-dev |
|
|
Hello Doru,
In fact, the bug was related to size of the window that you opened the visualization. So, it was a "bit random". Maybe you have a big screen :) On Wed, May 4, 2011 at 6:25 PM, Tudor Girba <[hidden email]> wrote: Hi, -- Andre Hora _______________________________________________ Moose-dev mailing list [hidden email] https://www.iam.unibe.ch/mailman/listinfo/moose-dev |
|
|
In reply to this post by Tudor Girba
Hi Doru,
Your analysis is right. I will make changes tomorrow. Cheers, Jannik On May 4, 2011, at 18:23 , Tudor Girba wrote: > Hi Jannik, > > Thanks for looking into this. > > There is a problem with this implementation: it breaks the contract of basic attributes in Moose. MooseNamedEntity>>parentPackage has as opposite FAMIXPackage>>childNamedEntities. This means that the contract is that if parentPackage is not nil then the entity is to be found in the childNamedEntities collection. > > In this case, the reason for parentPackage is that you want to use it as a unified API for querying the parent package scope. Chef already provides this API intention in the packageScope. So, we should move the computation from parentPackage in packageScope, and then have queries depend on packageScope. > > Cheers, > Doru > > > On 4 May 2011, at 09:14, Laval Jannik wrote: > >> Hi, >> >> I changed the behavior of FAMIXMethod>>parentPackage. >> >> In fact, when a method is in the same package than its class, its parentPackage should be nil. >> >> So, to not break the behavior, I changed the accessor like this: >> >> === >> parentPackage >> ^ parentPackage ifNil:[ self belongsTo parentPackage] >> === >> >> But this breaks during importation of MSE files (it breaks some tests of smalldude and eyesee). >> The problem is due to the class that is not already defined during the importation. >> So I changed the source code like that. I am not convince of the efficiency of this code. >> Have you another idea ? >> >> === >> parentPackage >> >> ^ parentPackage ifNil:[ >> |cl| >> cl := self belongsTo. >> cl isNil >> ifTrue:[nil] >> ifFalse:[cl parentPackage] >> ] >> === >> >> --- >> Jannik Laval >> >> >> _______________________________________________ >> Moose-dev mailing list >> [hidden email] >> https://www.iam.unibe.ch/mailman/listinfo/moose-dev > > -- > www.tudorgirba.com > > "Yesterday is a fact. > Tomorrow is a possibility. > Today is a challenge." > > > > > _______________________________________________ > Moose-dev mailing list > [hidden email] > https://www.iam.unibe.ch/mailman/listinfo/moose-dev _______________________________________________ Moose-dev mailing list [hidden email] https://www.iam.unibe.ch/mailman/listinfo/moose-dev |
«
Return to Moose
|
1 view|%1 views
| Free forum by Nabble | Edit this page |