Eclipse Community Forums
Forum Search:

Search      Help    Register    Login    Home
Home » Modeling » OCL » Re: OCL Recursive Functions
Re: OCL Recursive Functions [message #3107] Wed, 31 January 2007 10:34 Go to next message
Chris Lenz is currently offline Chris LenzFriend
Messages: 214
Registered: July 2009
Senior Member
Is there some information what is planned in the j2se 5.0 support and
also in the uml support features.

Our research team on university of innsbruck are fokusing on model
validation and we are also developing some tools, it would be intresting
to cooperate in some parts ;-).
Chris
Christian W. Damus schrieb:
> Hi, Chris,
>
> This bug is still open, but it hasn't been forgotten :-)
>
> The patch that you attached will be obsoleted by the J2SE 5.0 support and
> UML support features, which are in progress and scheduled for completion in
> this (M5) milestone. Also, to address
> https://bugs.eclipse.org/bugs/show_bug.cgi?id=154371 and to better align
> with the UML model of constraints, the means by which constraints are
> defined is being considerably reworked.
>
> If this patch is still working for you, then that's great. OCL 1.1 will
> address this problem and others.
>
> Cheers,
>
> Christian
>
>
> Chris Lenz wrote:
>
>> https://bugs.eclipse.org/bugs/show_bug.cgi?id=152018
>>
>> The Bug is still open, maybe someone could look to the patch applied.
>>
>> It works now for serveral month very good for me.
>>
>> Is it possible to undefine Operations.
>>
>> try{
>> feature = TypeUtil.defineOperation(
>> contextEnv.getContextClassifier(),
>> unquote(operCS.getSimpleNameCS().getValue()),
>> params, type);
>> expression = oclExpressionCS(defExpr.getExpressionCS(), contextEnv);
>> }catch (SemanticException e) {
>> //undefine the operation
>> throw e;
>> }
>>
>>
>
> <snip>
Re: OCL Recursive Functions [message #3167 is a reply to message #3107] Wed, 31 January 2007 15:40 Go to previous message
Eclipse UserFriend
Originally posted by: cdamus.ca.ibm.com

Hi, Chris,

I will be adding this information to the "What's New in M5" Wiki entry when
the milestone date nears, as well as in the OCL SDK Documentation, but can
easily summarize here.

The J2SE 5.0 support consists of refactoring the OCL API to make it generic,
removing the dependency on the Ecore metamodel. This new API is defined in
the org.eclipse.ocl plug-in. Now, instead of OCL assuming that there are
EPackages containing EClasses that have EOperations and
EStructuralFeatures, etc., these are represented by generic type variables.
The Ecore binding (org.eclipse.ocl.ecore) substitutes EPackage,
EClassifier, EOperation, etc. for these type variables; the UML binding
(org.eclipse.ocl.uml) substitutes Package, Classifier, Operation, etc.

This refactoring is essentially a clone of the previous Ecore-centric API
because it isn't the kind of change that can maintain backward
compatibility (not like introducing generic type parameters into an API
that was already generic in Java 1.4 terms).

The OCL Abstract Syntax model (OCL.ecore) is also regenerated using the J2SE
5.0 codegen option (actually, it's a new metamodel because it couldn't be
morphed in situ and maintain compatibility for the same reason).

The UML support just allows parsing OCL on UML models as on Ecore models,
primarily for the purpose of defining Constraints in UML models. This is
provided by the UMLEnvironment in the org.eclipse.ocl.uml plug-in.

Oh, and getting back to the original subject of this thread; those recursive
operation and attribute definitions are now supported in the latest 1.1
integration build. This works for the deprecated 1.0 API
(org.eclipse.emf.ocl) as well as the new, because they use the same
(internal) parser implementation.

Cheers,

Christian

Chris Lenz wrote:

> Is there some information what is planned in the j2se 5.0 support and
> also in the uml support features.
>
> Our research team on university of innsbruck are fokusing on model
> validation and we are also developing some tools, it would be intresting
> to cooperate in some parts ;-).
> Chris
> Christian W. Damus schrieb:
>> Hi, Chris,
>>
>> This bug is still open, but it hasn't been forgotten :-)
>>
>> The patch that you attached will be obsoleted by the J2SE 5.0 support and
>> UML support features, which are in progress and scheduled for completion
>> in
>> this (M5) milestone. Also, to address
>> https://bugs.eclipse.org/bugs/show_bug.cgi?id=154371 and to better align
>> with the UML model of constraints, the means by which constraints are
>> defined is being considerably reworked.
>>
>> If this patch is still working for you, then that's great. OCL 1.1 will
>> address this problem and others.
>>
>> Cheers,
>>
>> Christian
>>
>>
>> Chris Lenz wrote:
>>
>>> https://bugs.eclipse.org/bugs/show_bug.cgi?id=152018
>>>
>>> The Bug is still open, maybe someone could look to the patch applied.
>>>
>>> It works now for serveral month very good for me.
>>>
>>> Is it possible to undefine Operations.
>>>
>>> try{
>>> feature = TypeUtil.defineOperation(
>>> contextEnv.getContextClassifier(),
>>> unquote(operCS.getSimpleNameCS().getValue()),
>>> params, type);
>>> expression = oclExpressionCS(defExpr.getExpressionCS(), contextEnv);
>>> }catch (SemanticException e) {
>>> //undefine the operation
>>> throw e;
>>> }
>>>
>>>
>>
>> <snip>
Previous Topic:[Announce] MDT OCL 1.1.0 I200701291654 is available
Next Topic:[Announce] MDT OCL 1.1.0 I200702011837 is available
Goto Forum:
  


Current Time: Mon Dec 22 22:04:33 GMT 2014

Powered by FUDForum. Page generated in 0.01500 seconds
.:: Contact :: Home ::.

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