Skip to main content


Eclipse Community Forums
Forum Search:

Search      Help    Register    Login    Home
Home » Modeling » EMF "Technology" (Ecore Tools, EMFatic, etc)  » Cannot create Migrator (Edapt/COPE)(emf model migration)
Cannot create Migrator (Edapt/COPE) [message #714805] Thu, 11 August 2011 14:28 Go to next message
js Missing name is currently offline js Missing nameFriend
Messages: 73
Registered: July 2009
Member
I have installed Edapt:

http://www.eclipse.org/edapt/

and downloaded the example from:

http://cope.in.tum.de/uploads/Documentation/statemachine_models.zip

I have then added a new reference to a type in the model and released the modified model change. Next I would like to create a migrator.

But when I right-click the History node and select 'Edapt' I can only select 'Check for integrity' or 'Check for breaking changes'. Where did the 'Create migrator' go as shown in the demo:

http://cope.in.tum.de/pmwiki.php?n=Demo.Main

?

Right clicking the 'Release xxx' node gives some more options. One of those is 'Create Custom Migration' but that just launches the wizard to create a custom class for specialized migration code - which should not be necessay.

Any suggestions?

Report message to a moderator

Re: Cannot create Migrator (Edapt/COPE) [message #715146 is a reply to message #714805] Fri, 12 August 2011 13:50 Go to previous messageGo to next message
Markus Herrmannsdoerfer is currently offline Markus HerrmannsdoerferFriend
Messages: 50
Registered: July 2009
Member
Hi,

Edapt is different from COPE in the sense that the migrator no longer
needs to be migrated, but the history is directly interpreted. To
register the history, you need to use the extension point
"org.eclipse.emf.edapt.migrators" from the plugin
"org.eclipse.emf.edapt.history". Then you can use the migrator from the
second Eclipse instance like before.
Sorry, but I did not yet find the time to update the documentation of
Edapt. However, I will try to improve the documentation in the next few
weeks.
Cheers,

Markus

Am 11.08.2011 16:28, schrieb js:
> I have installed Edapt:
>
> http://www.eclipse.org/edapt/
>
> and downloaded the example from:
>
> http://cope.in.tum.de/uploads/Documentation/statemachine_models.zip
>
> I have then added a new reference to a type in the model and released
> the modified model change. Next I would like to create a migrator.
>
> But when I right-click the History node and select 'Edapt' I can only
> select 'Check for integrity' or 'Check for breaking changes'. Where did
> the 'Create migrator' go as shown in the demo:
>
> http://cope.in.tum.de/pmwiki.php?n=Demo.Main
>
> ?
>
> Right clicking the 'Release xxx' node gives some more options. One of
> those is 'Create Custom Migration' but that just launches the wizard to
> create a custom class for specialized migration code - which should not
> be necessay.
>
> Any suggestions?

Report message to a moderator

Re: Cannot create Migrator (Edapt/COPE) [message #865565 is a reply to message #715146] Mon, 30 April 2012 13:11 Go to previous messageGo to next message
lucky informatik is currently offline lucky informatikFriend
Messages: 12
Registered: April 2012
Junior Member
Hi,

I have created metamodel and a model.Now I want to change metamodel and check how it works on the model.I used a reuse operation 'rename' , executed the operation and released the history. Now I was stuck at registering the history to migrate the model. I went through the Edapt tutorial.

But I was not clear with the below statements.

To register the history, you need to use the extension point
"org.eclipse.emf.edapt.migrators" from the plugin
"org.eclipse.emf.edapt.history". Then you can use the migrator from the
second Eclipse instance like before.

Could someone please explain what I have to do in the eclipse editor with the above statements or provide me with some tutorial.

Thanks,
Lucky

Report message to a moderator

Re: Cannot create Migrator (Edapt/COPE) [message #867379 is a reply to message #865565] Tue, 01 May 2012 09:00 Go to previous messageGo to next message
Markus Herrmannsdoerfer is currently offline Markus HerrmannsdoerferFriend
Messages: 50
Registered: July 2009
Member
Hi,

you have to open the plugin.xml Editor and go to the Extensions tab and
here you need to add an extension for extension point
org.eclipse.emf.edapt.migrators like shown in the attached image. If you
do not find the extension point, you have to add the plugin
org.eclipse.emf.edapt.history in the Dependencies tab.
If you have further questions, do not hesitate to ask.
Cheers,

Markus

Am 30.04.2012 15:11, schrieb lucky informatik:
> Hi,
>
> I have created metamodel and a model.Now I want to change metamodel and
> check how it works on the model.I used a reuse operation 'rename' ,
> executed the operation and released the history. Now I was stuck at
> registering the history to migrate the model. I went through the Edapt
> tutorial.
>
> But I was not clear with the below statements.
>
> To register the history, you need to use the extension point
> "org.eclipse.emf.edapt.migrators" from the plugin
> "org.eclipse.emf.edapt.history". Then you can use the migrator from the
> second Eclipse instance like before.
>
> Could someone please explain what I have to do in the eclipse editor
> with the above statements or provide me with some tutorial.
>
> Thanks,
> Lucky

Report message to a moderator

Re: Cannot create Migrator (Edapt/COPE) [message #867809 is a reply to message #867379] Tue, 01 May 2012 13:48 Go to previous messageGo to next message
lucky informatik is currently offline lucky informatikFriend
Messages: 12
Registered: April 2012
Junior Member
Thank you.Now I understood:)

Report message to a moderator

(Edapt/COPE) [message #869551 is a reply to message #867809] Thu, 03 May 2012 15:13 Go to previous messageGo to next message
lucky informatik is currently offline lucky informatikFriend
Messages: 12
Registered: April 2012
Junior Member
Hi,

I created a metamodel and a model and I changed my metamodel by adding a subclass, using Edapt operation and migrated(validated) model. My idea was to create a subclass so that I can change the superclass instance as subclass instance. For that I deleted the super class instance manually and created the same as a subclass instance.

I have heard that by using transformations(model transformation tools) the superclass instances can be transformed to subclass instances. I would like to know if there is some option available which automatically changes the superclass instances as subclass instances.

Thanks,
Lucky

[Updated on: Thu, 03 May 2012 15:25]

Report message to a moderator

Re: (Edapt/COPE) [message #869887 is a reply to message #869551] Fri, 04 May 2012 16:09 Go to previous messageGo to next message
Markus Herrmannsdoerfer is currently offline Markus HerrmannsdoerferFriend
Messages: 50
Registered: July 2009
Member
Sorry, but there is no explicit operation for that. However, you can use
MakeClassAbstract which allows you to set a subclass to which the
instances of the class that is made abstract are to be migrated. Then
you can use InlineSuperClass to inline the super class into the
subclass. I hope this fits your purpose.

Am 03.05.2012 17:13, schrieb lucky informatik:
> Hi,
>
> I created a metamodel and a model and I changed my metamodel by adding a
> subclass, using Edapt operation and migrated(validated) model. My idea
> was to create a subclass so that I can change the superclass instance as
> subclass instance. For that I deleted the super class instance and
> created the same as a subclass instance.
>
> I would like to know if there is some option available which
> automatically changes the superclass instances as subclass instances.
>
> Thanks,
> Lucky

Report message to a moderator

Re: (Edapt/COPE) [message #870497 is a reply to message #869887] Tue, 08 May 2012 12:45 Go to previous messageGo to next message
lucky informatik is currently offline lucky informatikFriend
Messages: 12
Registered: April 2012
Junior Member
Thank you very much. Please could you also clarify me whether meta-model evolution and model co-evolution be considered as one of the types of vertical inconsistency ?

Thanks,
Lucky

Report message to a moderator

Re: (Edapt/COPE) [message #870562 is a reply to message #870497] Tue, 08 May 2012 18:00 Go to previous messageGo to next message
Markus Herrmannsdoerfer is currently offline Markus HerrmannsdoerferFriend
Messages: 50
Registered: July 2009
Member
By the way: I now also created a web site which provides the
documentation for all operations:
http://www.eclipse.org/edapt/operations.php

The consistency between a model and the metamodel to which it conforms
is called vertical consistency. If a metamodel evolves, this might thus
result in a vertical inconsistency between the model and the metamodel.
To re-establish the vertical consistency, you need to co-evolve your
model. I hope this explains the terminology.


Am 08.05.2012 14:45, schrieb lucky informatik:
> Thank you very much. Please could you also clarify me whether meta-model
> evolution and model co-evolution be considered as one of the types of
> vertical inconsistency ?
>
> Thanks,
> Lucky

Report message to a moderator

Re: (Edapt/COPE) [message #870650 is a reply to message #870562] Wed, 09 May 2012 08:33 Go to previous messageGo to next message
lucky informatik is currently offline lucky informatikFriend
Messages: 12
Registered: April 2012
Junior Member
Thanks for clarifying and the website helps a lot for the beginners like me:)

Thanks,
Lucky

Report message to a moderator

Re: (Edapt/COPE) [message #882825 is a reply to message #870650] Thu, 07 June 2012 09:23 Go to previous messageGo to next message
lucky informatik is currently offline lucky informatikFriend
Messages: 12
Registered: April 2012
Junior Member
Hi,

I need some information.I have seen some classifications for metamodel changes in some literature for metamodel matching approaches like non-breaking, breakable and resolvable, breakable and non resolvable.I would like to know if these classifications are applied for operation based approaches or if there are some other classifications.Please let me know if there is some literature corresponding to these classifications.
I would also like to know if there is some theory behind in creating the operators in EDAPT.

Thanks,

Lucky

Report message to a moderator

Re: (Edapt/COPE) [message #882874 is a reply to message #882825] Thu, 07 June 2012 11:14 Go to previous messageGo to next message
Markus Herrmannsdoerfer is currently offline Markus HerrmannsdoerferFriend
Messages: 50
Registered: July 2009
Member
Hi Lucky,

have you taken a look at the publications that listed on the Edapt page:

http://www.eclipse.org/edapt/publications.php

Maybe the following publication gives you some information about
classifications of operators and provides also references to futher
literature:

An Extensive Catalog of Operators for the Coupled Evolution of
Metamodels and Models. Markus Herrmannsdoerfer, Sander Vermolen, and
Guido Wachsmuth. In Proceedings of the 3rd International Conference on
Software Language Engineering (SLE). (2010)

I hope this helps.
Cheers,

Markus


Am 07.06.2012 11:23, schrieb lucky informatik:
> Hi,
>
> I need some information.I have seen some classifications for metamodel
> changes in some literature for metamodel matching approaches like
> non-breaking, breakable and resolvable, breakable and non resolvable.I
> would like to know if these classifications are applied for operation
> based approaches or if there are some other classifications.Please let
> me know if there is some literature corresponding to these classifications.
> I would also like to know if there is some theory behind in creating the
> operators in EDAPT.
> Thanks,
>
> Lucky

Report message to a moderator

Re: (Edapt/COPE) [message #890016 is a reply to message #882874] Wed, 20 June 2012 15:11 Go to previous messageGo to next message
lucky informatik is currently offline lucky informatikFriend
Messages: 12
Registered: April 2012
Junior Member
Thanks for the information.

I have another question out of my curiosity. Suppose we make some changes to meta-model using Edapt. For example: rename or delete feature. Then we save ecore model and re-load gen-model. Then we observe some error in the source, edit or editor codes. I normally edit the java implementations to remove the errors and then migrate.
I would like to know if there is other process to automatically get rid of those erros?

Thanks,
Lucky

Report message to a moderator

Re: (Edapt/COPE) [message #891310 is a reply to message #890016] Sun, 24 June 2012 16:35 Go to previous messageGo to next message
Markus Herrmannsdoerfer is currently offline Markus HerrmannsdoerferFriend
Messages: 50
Registered: July 2009
Member
Hi Lucky,

unfortunately, at the moment, there is no such support in Edapt.
However, I have already thought about this feature and have some ideas
how this can be implemented. The idea would be that an operation also
triggers a code refactoring operation. Unfortunately, I do not have
resources to implement such a feature at the moment. If you have
resources and want to implement such a feature, we can of course
integrate it into Edapt.
Cheers,

Markus

Am 20.06.2012 17:11, schrieb lucky informatik:
> Thanks for the information.
>
> I have another question out of my curiosity. Suppose we make some
> changes to meta-model using Edapt. For example: rename or delete
> feature. Then we save ecore model and re-load gen-model. Then we observe
> some error in the source, edit or editor codes. I normally edit the java
> implementations to remove the errors and then migrate.
> I would like to know if there is other process to automatically get rid
> of those erros?
>
> Thanks,
> Lucky
>

Report message to a moderator

Re: (Edapt/COPE) [message #892712 is a reply to message #891310] Fri, 29 June 2012 10:14 Go to previous message
lucky informatik is currently offline lucky informatikFriend
Messages: 12
Registered: April 2012
Junior Member
Hi Markus,

Thanks for the reply. As you know from my questions I just started using Edapt and would like to use it for various examples. In future when I have some plans on this, I'll certainly get back to you.

Best wishes,
Lucky

Report message to a moderator

Previous Topic:my EMF model generated tests...
Next Topic:Hibernate tool for JUNO
Goto Forum:
  

[ Syndicate this forum (XML) ] [ RSS ]

Current Time: Wed Apr 24 22:31:09 GMT 2024

Powered by FUDForum. Page generated in 0.03380 seconds
.:: Contact :: Home ::.

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

Back to the top

Sign up to our Newsletter

A fresh new issue delivered monthly