Skip to main content


Eclipse Community Forums
Forum Search:

Search      Help    Register    Login    Home
Home » Modeling » UML2 » Missing property in UML::Class::Kernel::PackageableElement
Missing property in UML::Class::Kernel::PackageableElement [message #477891] Mon, 27 October 2008 09:32 Go to next message
Eclipse UserFriend
Originally posted by: florianwendland.freenet.de

Hi all,

i've found out, that the relationship between Package and PackageableElement
isn't realized with respect to the UML Superstructure (or may be it is
wanted).
In the spec, there is a containent relationship between Package and
PackageableElement with both ends being navigable. See:

Package:
+ packagedElement : PackageableElement [0..*]

PackageableElement : NamedElement
+ owningPackage : Package [0..1]

The first accessor was generated in the code, but the second one in class
PackageableElementImpl wasn't. Instead of, there is getOwner() : Element
method, which returns of course the owningPackage, but it is not very
precisly defined.

Is this a bug or a feature :)

Marc-Florian
Re: Missing property in UML::Class::Kernel::PackageableElement [message #477907 is a reply to message #477891] Wed, 29 October 2008 19:06 Go to previous messageGo to next message
james bruck is currently offline james bruckFriend
Messages: 1724
Registered: July 2009
Senior Member
Hi Marc-Florian,

I think you are referring to figure 7.14 in the spec. Other than that one
figure, I can't see owning package in the description of PackageableElement
or mentioned anywhere else in the spec. This may be some oversight. I'll
dig into this a bit more and raise an issue with the OMG.

Thanks,
-James.

"Marc-Florian Wendland" <florianwendland@freenet.de> wrote in message
news:ge41nt$ifi$1@build.eclipse.org...
> Hi all,
>
> i've found out, that the relationship between Package and
> PackageableElement isn't realized with respect to the UML Superstructure
> (or may be it is wanted).
> In the spec, there is a containent relationship between Package and
> PackageableElement with both ends being navigable. See:
>
> Package:
> + packagedElement : PackageableElement [0..*]
>
> PackageableElement : NamedElement
> + owningPackage : Package [0..1]
>
> The first accessor was generated in the code, but the second one in class
> PackageableElementImpl wasn't. Instead of, there is getOwner() : Element
> method, which returns of course the owningPackage, but it is not very
> precisly defined.
>
> Is this a bug or a feature :)
>
> Marc-Florian
>
Re: Missing property in UML::Class::Kernel::PackageableElement [message #477913 is a reply to message #477891] Thu, 30 October 2008 18:52 Go to previous messageGo to next message
Kenn Hussey is currently offline Kenn HusseyFriend
Messages: 1620
Registered: July 2009
Senior Member
Marc-Florian,

Which version of the specification are you referencing, and which figure?
Figure 7.14 (also mentioned by James) shows the
PackageableElement::owningPackage property as being non-navigable; when this
is mapped to Java, the property is lost (as are all non-navigable
properties)...

Kenn

"Marc-Florian Wendland" <florianwendland@freenet.de> wrote in message
news:ge41nt$ifi$1@build.eclipse.org...
> Hi all,
>
> i've found out, that the relationship between Package and
> PackageableElement isn't realized with respect to the UML Superstructure
> (or may be it is wanted).
> In the spec, there is a containent relationship between Package and
> PackageableElement with both ends being navigable. See:
>
> Package:
> + packagedElement : PackageableElement [0..*]
>
> PackageableElement : NamedElement
> + owningPackage : Package [0..1]
>
> The first accessor was generated in the code, but the second one in class
> PackageableElementImpl wasn't. Instead of, there is getOwner() : Element
> method, which returns of course the owningPackage, but it is not very
> precisly defined.
>
> Is this a bug or a feature :)
>
> Marc-Florian
>
Re: Missing property in UML::Class::Kernel::PackageableElement [message #477919 is a reply to message #477913] Mon, 03 November 2008 09:16 Go to previous messageGo to next message
Eclipse UserFriend
Originally posted by: florianwendland.freenet.de

Kenn and James,

indeed, i'm referencing the figure 7.14 in the uml 2.1 spec. But why is the
owningPackage-end non-navigable? See the association between Package and
Type. There are no differences on the ends at Package-side, so how do you
see, that it is non-navigable? Or depends this on the arrow at the
packagedElement-end? Does that mean, that you can navigate from Package only
to PackageableElement but not vice versa? Because this arrow is the only
difference, i can recognize...

Surely, if it is non-navigable, than it won't be translated into code.

Thanks a lot
Marc-Florian


"Kenn Hussey" <Kenn.Hussey@embarcadero.com> schrieb im Newsbeitrag
news:gecvps$968$1@build.eclipse.org...
> Marc-Florian,
>
> Which version of the specification are you referencing, and which figure?
> Figure 7.14 (also mentioned by James) shows the
> PackageableElement::owningPackage property as being non-navigable; when
> this is mapped to Java, the property is lost (as are all non-navigable
> properties)...
>
> Kenn
>
> "Marc-Florian Wendland" <florianwendland@freenet.de> wrote in message
> news:ge41nt$ifi$1@build.eclipse.org...
>> Hi all,
>>
>> i've found out, that the relationship between Package and
>> PackageableElement isn't realized with respect to the UML Superstructure
>> (or may be it is wanted).
>> In the spec, there is a containent relationship between Package and
>> PackageableElement with both ends being navigable. See:
>>
>> Package:
>> + packagedElement : PackageableElement [0..*]
>>
>> PackageableElement : NamedElement
>> + owningPackage : Package [0..1]
>>
>> The first accessor was generated in the code, but the second one in class
>> PackageableElementImpl wasn't. Instead of, there is getOwner() : Element
>> method, which returns of course the owningPackage, but it is not very
>> precisly defined.
>>
>> Is this a bug or a feature :)
>>
>> Marc-Florian
>>
>
>
Re: Missing property in UML::Class::Kernel::PackageableElement [message #477920 is a reply to message #477919] Mon, 03 November 2008 14:08 Go to previous messageGo to next message
Eclipse UserFriend
Originally posted by: cdamus.zeligsoft.com

Hi, Marc-Florian,

If it helps, OCL expressions are able to navigate the non-navigable
association ends, and even unnamed ends where they are not ambiguous
(though there are few). The MDT OCL implementation uses the UML
CacheAdapter's inverse-reference tracking to implement this campability.

Cheers,

Christian

Marc-Florian Wendland wrote:
> Kenn and James,
>
> indeed, i'm referencing the figure 7.14 in the uml 2.1 spec. But why is the
> owningPackage-end non-navigable? See the association between Package and
> Type. There are no differences on the ends at Package-side, so how do you
> see, that it is non-navigable? Or depends this on the arrow at the
> packagedElement-end? Does that mean, that you can navigate from Package only
> to PackageableElement but not vice versa? Because this arrow is the only
> difference, i can recognize...
>
> Surely, if it is non-navigable, than it won't be translated into code.
>
> Thanks a lot
> Marc-Florian
>
>
> "Kenn Hussey" <Kenn.Hussey@embarcadero.com> schrieb im Newsbeitrag
> news:gecvps$968$1@build.eclipse.org...
>> Marc-Florian,
>>
>> Which version of the specification are you referencing, and which figure?
>> Figure 7.14 (also mentioned by James) shows the
>> PackageableElement::owningPackage property as being non-navigable; when
>> this is mapped to Java, the property is lost (as are all non-navigable
>> properties)...
>>
>> Kenn
>>
>> "Marc-Florian Wendland" <florianwendland@freenet.de> wrote in message
>> news:ge41nt$ifi$1@build.eclipse.org...
>>> Hi all,
>>>
>>> i've found out, that the relationship between Package and
>>> PackageableElement isn't realized with respect to the UML Superstructure
>>> (or may be it is wanted).
>>> In the spec, there is a containent relationship between Package and
>>> PackageableElement with both ends being navigable. See:
>>>
>>> Package:
>>> + packagedElement : PackageableElement [0..*]
>>>
>>> PackageableElement : NamedElement
>>> + owningPackage : Package [0..1]
>>>
>>> The first accessor was generated in the code, but the second one in class
>>> PackageableElementImpl wasn't. Instead of, there is getOwner() : Element
>>> method, which returns of course the owningPackage, but it is not very
>>> precisly defined.
>>>
>>> Is this a bug or a feature :)
>>>
>>> Marc-Florian
>>>
>>
>
>
Re: Missing property in UML::Class::Kernel::PackageableElement [message #477924 is a reply to message #477919] Mon, 03 November 2008 17:57 Go to previous message
Kenn Hussey is currently offline Kenn HusseyFriend
Messages: 1620
Registered: July 2009
Senior Member
Marc-Florian,

Yes, it's the absence of the arrowhead that indicates that the property is
non-navigable.

Kenn

"Marc-Florian Wendland" <florianwendland@freenet.de> wrote in message
news:gemfct$h38$1@build.eclipse.org...
> Kenn and James,
>
> indeed, i'm referencing the figure 7.14 in the uml 2.1 spec. But why is
> the owningPackage-end non-navigable? See the association between Package
> and Type. There are no differences on the ends at Package-side, so how do
> you see, that it is non-navigable? Or depends this on the arrow at the
> packagedElement-end? Does that mean, that you can navigate from Package
> only to PackageableElement but not vice versa? Because this arrow is the
> only difference, i can recognize...
>
> Surely, if it is non-navigable, than it won't be translated into code.
>
> Thanks a lot
> Marc-Florian
>
>
> "Kenn Hussey" <Kenn.Hussey@embarcadero.com> schrieb im Newsbeitrag
> news:gecvps$968$1@build.eclipse.org...
>> Marc-Florian,
>>
>> Which version of the specification are you referencing, and which figure?
>> Figure 7.14 (also mentioned by James) shows the
>> PackageableElement::owningPackage property as being non-navigable; when
>> this is mapped to Java, the property is lost (as are all non-navigable
>> properties)...
>>
>> Kenn
>>
>> "Marc-Florian Wendland" <florianwendland@freenet.de> wrote in message
>> news:ge41nt$ifi$1@build.eclipse.org...
>>> Hi all,
>>>
>>> i've found out, that the relationship between Package and
>>> PackageableElement isn't realized with respect to the UML Superstructure
>>> (or may be it is wanted).
>>> In the spec, there is a containent relationship between Package and
>>> PackageableElement with both ends being navigable. See:
>>>
>>> Package:
>>> + packagedElement : PackageableElement [0..*]
>>>
>>> PackageableElement : NamedElement
>>> + owningPackage : Package [0..1]
>>>
>>> The first accessor was generated in the code, but the second one in
>>> class PackageableElementImpl wasn't. Instead of, there is getOwner() :
>>> Element method, which returns of course the owningPackage, but it is not
>>> very precisly defined.
>>>
>>> Is this a bug or a feature :)
>>>
>>> Marc-Florian
>>>
>>
>>
>
>
Re: Missing property in UML::Class::Kernel::PackageableElement [message #627116 is a reply to message #477891] Wed, 29 October 2008 19:06 Go to previous message
james bruck is currently offline james bruckFriend
Messages: 1724
Registered: July 2009
Senior Member
Hi Marc-Florian,

I think you are referring to figure 7.14 in the spec. Other than that one
figure, I can't see owning package in the description of PackageableElement
or mentioned anywhere else in the spec. This may be some oversight. I'll
dig into this a bit more and raise an issue with the OMG.

Thanks,
-James.

"Marc-Florian Wendland" <florianwendland@freenet.de> wrote in message
news:ge41nt$ifi$1@build.eclipse.org...
> Hi all,
>
> i've found out, that the relationship between Package and
> PackageableElement isn't realized with respect to the UML Superstructure
> (or may be it is wanted).
> In the spec, there is a containent relationship between Package and
> PackageableElement with both ends being navigable. See:
>
> Package:
> + packagedElement : PackageableElement [0..*]
>
> PackageableElement : NamedElement
> + owningPackage : Package [0..1]
>
> The first accessor was generated in the code, but the second one in class
> PackageableElementImpl wasn't. Instead of, there is getOwner() : Element
> method, which returns of course the owningPackage, but it is not very
> precisly defined.
>
> Is this a bug or a feature :)
>
> Marc-Florian
>
Re: Missing property in UML::Class::Kernel::PackageableElement [message #627122 is a reply to message #477891] Thu, 30 October 2008 18:52 Go to previous message
Kenn Hussey is currently offline Kenn HusseyFriend
Messages: 1620
Registered: July 2009
Senior Member
Marc-Florian,

Which version of the specification are you referencing, and which figure?
Figure 7.14 (also mentioned by James) shows the
PackageableElement::owningPackage property as being non-navigable; when this
is mapped to Java, the property is lost (as are all non-navigable
properties)...

Kenn

"Marc-Florian Wendland" <florianwendland@freenet.de> wrote in message
news:ge41nt$ifi$1@build.eclipse.org...
> Hi all,
>
> i've found out, that the relationship between Package and
> PackageableElement isn't realized with respect to the UML Superstructure
> (or may be it is wanted).
> In the spec, there is a containent relationship between Package and
> PackageableElement with both ends being navigable. See:
>
> Package:
> + packagedElement : PackageableElement [0..*]
>
> PackageableElement : NamedElement
> + owningPackage : Package [0..1]
>
> The first accessor was generated in the code, but the second one in class
> PackageableElementImpl wasn't. Instead of, there is getOwner() : Element
> method, which returns of course the owningPackage, but it is not very
> precisly defined.
>
> Is this a bug or a feature :)
>
> Marc-Florian
>
Re: Missing property in UML::Class::Kernel::PackageableElement [message #627128 is a reply to message #477913] Mon, 03 November 2008 09:16 Go to previous message
Eclipse UserFriend
Originally posted by: florianwendland.freenet.de

Kenn and James,

indeed, i'm referencing the figure 7.14 in the uml 2.1 spec. But why is the
owningPackage-end non-navigable? See the association between Package and
Type. There are no differences on the ends at Package-side, so how do you
see, that it is non-navigable? Or depends this on the arrow at the
packagedElement-end? Does that mean, that you can navigate from Package only
to PackageableElement but not vice versa? Because this arrow is the only
difference, i can recognize...

Surely, if it is non-navigable, than it won't be translated into code.

Thanks a lot
Marc-Florian


"Kenn Hussey" <Kenn.Hussey@embarcadero.com> schrieb im Newsbeitrag
news:gecvps$968$1@build.eclipse.org...
> Marc-Florian,
>
> Which version of the specification are you referencing, and which figure?
> Figure 7.14 (also mentioned by James) shows the
> PackageableElement::owningPackage property as being non-navigable; when
> this is mapped to Java, the property is lost (as are all non-navigable
> properties)...
>
> Kenn
>
> "Marc-Florian Wendland" <florianwendland@freenet.de> wrote in message
> news:ge41nt$ifi$1@build.eclipse.org...
>> Hi all,
>>
>> i've found out, that the relationship between Package and
>> PackageableElement isn't realized with respect to the UML Superstructure
>> (or may be it is wanted).
>> In the spec, there is a containent relationship between Package and
>> PackageableElement with both ends being navigable. See:
>>
>> Package:
>> + packagedElement : PackageableElement [0..*]
>>
>> PackageableElement : NamedElement
>> + owningPackage : Package [0..1]
>>
>> The first accessor was generated in the code, but the second one in class
>> PackageableElementImpl wasn't. Instead of, there is getOwner() : Element
>> method, which returns of course the owningPackage, but it is not very
>> precisly defined.
>>
>> Is this a bug or a feature :)
>>
>> Marc-Florian
>>
>
>
Re: Missing property in UML::Class::Kernel::PackageableElement [message #627129 is a reply to message #477919] Mon, 03 November 2008 14:08 Go to previous message
Eclipse UserFriend
Originally posted by: cdamus.zeligsoft.com

Hi, Marc-Florian,

If it helps, OCL expressions are able to navigate the non-navigable
association ends, and even unnamed ends where they are not ambiguous
(though there are few). The MDT OCL implementation uses the UML
CacheAdapter's inverse-reference tracking to implement this campability.

Cheers,

Christian

Marc-Florian Wendland wrote:
> Kenn and James,
>
> indeed, i'm referencing the figure 7.14 in the uml 2.1 spec. But why is the
> owningPackage-end non-navigable? See the association between Package and
> Type. There are no differences on the ends at Package-side, so how do you
> see, that it is non-navigable? Or depends this on the arrow at the
> packagedElement-end? Does that mean, that you can navigate from Package only
> to PackageableElement but not vice versa? Because this arrow is the only
> difference, i can recognize...
>
> Surely, if it is non-navigable, than it won't be translated into code.
>
> Thanks a lot
> Marc-Florian
>
>
> "Kenn Hussey" <Kenn.Hussey@embarcadero.com> schrieb im Newsbeitrag
> news:gecvps$968$1@build.eclipse.org...
>> Marc-Florian,
>>
>> Which version of the specification are you referencing, and which figure?
>> Figure 7.14 (also mentioned by James) shows the
>> PackageableElement::owningPackage property as being non-navigable; when
>> this is mapped to Java, the property is lost (as are all non-navigable
>> properties)...
>>
>> Kenn
>>
>> "Marc-Florian Wendland" <florianwendland@freenet.de> wrote in message
>> news:ge41nt$ifi$1@build.eclipse.org...
>>> Hi all,
>>>
>>> i've found out, that the relationship between Package and
>>> PackageableElement isn't realized with respect to the UML Superstructure
>>> (or may be it is wanted).
>>> In the spec, there is a containent relationship between Package and
>>> PackageableElement with both ends being navigable. See:
>>>
>>> Package:
>>> + packagedElement : PackageableElement [0..*]
>>>
>>> PackageableElement : NamedElement
>>> + owningPackage : Package [0..1]
>>>
>>> The first accessor was generated in the code, but the second one in class
>>> PackageableElementImpl wasn't. Instead of, there is getOwner() : Element
>>> method, which returns of course the owningPackage, but it is not very
>>> precisly defined.
>>>
>>> Is this a bug or a feature :)
>>>
>>> Marc-Florian
>>>
>>
>
>
Re: Missing property in UML::Class::Kernel::PackageableElement [message #627133 is a reply to message #477919] Mon, 03 November 2008 17:57 Go to previous message
Kenn Hussey is currently offline Kenn HusseyFriend
Messages: 1620
Registered: July 2009
Senior Member
Marc-Florian,

Yes, it's the absence of the arrowhead that indicates that the property is
non-navigable.

Kenn

"Marc-Florian Wendland" <florianwendland@freenet.de> wrote in message
news:gemfct$h38$1@build.eclipse.org...
> Kenn and James,
>
> indeed, i'm referencing the figure 7.14 in the uml 2.1 spec. But why is
> the owningPackage-end non-navigable? See the association between Package
> and Type. There are no differences on the ends at Package-side, so how do
> you see, that it is non-navigable? Or depends this on the arrow at the
> packagedElement-end? Does that mean, that you can navigate from Package
> only to PackageableElement but not vice versa? Because this arrow is the
> only difference, i can recognize...
>
> Surely, if it is non-navigable, than it won't be translated into code.
>
> Thanks a lot
> Marc-Florian
>
>
> "Kenn Hussey" <Kenn.Hussey@embarcadero.com> schrieb im Newsbeitrag
> news:gecvps$968$1@build.eclipse.org...
>> Marc-Florian,
>>
>> Which version of the specification are you referencing, and which figure?
>> Figure 7.14 (also mentioned by James) shows the
>> PackageableElement::owningPackage property as being non-navigable; when
>> this is mapped to Java, the property is lost (as are all non-navigable
>> properties)...
>>
>> Kenn
>>
>> "Marc-Florian Wendland" <florianwendland@freenet.de> wrote in message
>> news:ge41nt$ifi$1@build.eclipse.org...
>>> Hi all,
>>>
>>> i've found out, that the relationship between Package and
>>> PackageableElement isn't realized with respect to the UML Superstructure
>>> (or may be it is wanted).
>>> In the spec, there is a containent relationship between Package and
>>> PackageableElement with both ends being navigable. See:
>>>
>>> Package:
>>> + packagedElement : PackageableElement [0..*]
>>>
>>> PackageableElement : NamedElement
>>> + owningPackage : Package [0..1]
>>>
>>> The first accessor was generated in the code, but the second one in
>>> class PackageableElementImpl wasn't. Instead of, there is getOwner() :
>>> Element method, which returns of course the owningPackage, but it is not
>>> very precisly defined.
>>>
>>> Is this a bug or a feature :)
>>>
>>> Marc-Florian
>>>
>>
>>
>
>
Previous Topic:Constraint Attachments
Next Topic:Importing Superstructure.uml to Rational Software Architect
Goto Forum:
  


Current Time: Thu Apr 02 17:12:27 GMT 2020

Powered by FUDForum. Page generated in 0.02536 seconds
.:: Contact :: Home ::.

Powered by: FUDforum 3.0.2.
Copyright ©2001-2010 FUDforum Bulletin Board Software

Back to the top