| eInvoke fails [message #991607] |
Wed, 19 December 2012 16:01  |
Eclipse User |
|
|
|
Hello,
it's me again . For my project I have to use the eInvoke-method to call an Operation for a certain EObject.
But no matter which method I want to invoke it always fails with the Exception:
java.lang.UnsupportedOperationException: eInvoke not implemented for getAllSuperClassifiers
I honestly don't have any clue what goes wrong, the internet doesn't show that much detail
for what "eInvoke not implemented for XZY" means.
Any help would be appreciated.
Greetings
Markus
|
|
|
|
|
|
|
| Re: eInvoke fails [message #991967 is a reply to message #991952] |
Thu, 20 December 2012 10:08  |
| Eclipse User |
|
|
|
Ed,
I added the method
org.eclipse.emf.codegen.ecore.genmodel.GenModel.initialize(boolean) in
2.8 and it's called by when creating new models, i.e.,
org.eclipse.emf.importer.ModelImporter.adjustGenModel(Monitor), and when
a GenModel is inferred from an Xcore model.
On 20/12/2012 6:39 AM, Ed Willink wrote:
> Hi Ed
>
> Is it easy for OCLinEcore to change this default too, or is this only
> possible once genmodel is integrated into the model model?
>
> Regards
>
> Ed Willink
>
> On 20/12/2012 14:19, Ed Merks wrote:
>> Markus,
>>
>> With Xcore-based models, it defaults to true...
>>
>> On 20/12/2012 6:05 AM, Markus Riedl wrote:
>>> Ahhh that seems to be the problem.. quite hidden in fact..
>>>
>>> Greetings
>>
>
|
|
|
| Re: eInvoke fails [message #991999 is a reply to message #991967] |
Thu, 20 December 2012 11:25 |
| Eclipse User |
|
|
|
Hi Ed
Great. So it's fixed for everything; I had a vague feeling that things
had got better here.
Regards
Ed
On 20/12/2012 15:08, Ed Merks wrote:
> Ed,
>
> I added the method
> org.eclipse.emf.codegen.ecore.genmodel.GenModel.initialize(boolean) in
> 2.8 and it's called by when creating new models, i.e.,
> org.eclipse.emf.importer.ModelImporter.adjustGenModel(Monitor), and
> when a GenModel is inferred from an Xcore model.
>
>
> On 20/12/2012 6:39 AM, Ed Willink wrote:
>> Hi Ed
>>
>> Is it easy for OCLinEcore to change this default too, or is this only
>> possible once genmodel is integrated into the model model?
>>
>> Regards
>>
>> Ed Willink
>>
>> On 20/12/2012 14:19, Ed Merks wrote:
>>> Markus,
>>>
>>> With Xcore-based models, it defaults to true...
>>>
>>> On 20/12/2012 6:05 AM, Markus Riedl wrote:
>>>> Ahhh that seems to be the problem.. quite hidden in fact..
>>>>
>>>> Greetings
>>>
>>
>
|
|
|
| Re: eInvoke fails [message #992007 is a reply to message #991999] |
Thu, 20 December 2012 11:42 |
| Eclipse User |
|
|
|
Ed,
Yes, I have to balance keeping things stable for the established user
base, while making useful improvements for the new user base. This
approach of keeping defaults such that the base is stable while setting
improved defaults for new models seemed like a good way to balance those
conflicting needs.
On 20/12/2012 8:25 AM, Ed Willink wrote:
> Hi Ed
>
> Great. So it's fixed for everything; I had a vague feeling that things
> had got better here.
>
> Regards
>
> Ed
>
>
> On 20/12/2012 15:08, Ed Merks wrote:
>> Ed,
>>
>> I added the method
>> org.eclipse.emf.codegen.ecore.genmodel.GenModel.initialize(boolean)
>> in 2.8 and it's called by when creating new models, i.e.,
>> org.eclipse.emf.importer.ModelImporter.adjustGenModel(Monitor), and
>> when a GenModel is inferred from an Xcore model.
>>
>>
>> On 20/12/2012 6:39 AM, Ed Willink wrote:
>>> Hi Ed
>>>
>>> Is it easy for OCLinEcore to change this default too, or is this
>>> only possible once genmodel is integrated into the model model?
>>>
>>> Regards
>>>
>>> Ed Willink
>>>
>>> On 20/12/2012 14:19, Ed Merks wrote:
>>>> Markus,
>>>>
>>>> With Xcore-based models, it defaults to true...
>>>>
>>>> On 20/12/2012 6:05 AM, Markus Riedl wrote:
>>>>> Ahhh that seems to be the problem.. quite hidden in fact..
>>>>>
>>>>> Greetings
>>>>
>>>
>>
>
|
|
|
|
| Re: eInvoke fails [message #992836 is a reply to message #992750] |
Sat, 22 December 2012 14:41 |
| Eclipse User |
|
|
|
Markus,
Yes, unfortunately a model needs to generate support for that and if a
model doesn't do that, your extension of it isn't going to be able to
support operation reflection either.
On 22/12/2012 5:54 AM, Markus Riedl wrote:
> It maybe default, but the genmodel which I referred to was not under
> my control (not until now), there they intentionally or not set
> Operation Reflection to false and even in IBM RSA 8.5 the default is
> false
>
>
> Greetings
|
|
|
Powered by
FUDForum. Page generated in 0.04495 seconds
Home