Skip to main content


Eclipse Community Forums
Forum Search:

Search      Help    Register    Login    Home
Home » Modeling » EMF » OCL to generate EMF behaviour
OCL to generate EMF behaviour [message #499427] Sun, 22 November 2009 14:14 Go to next message
Joel Rosi-Schwartz is currently offline Joel Rosi-SchwartzFriend
Messages: 624
Registered: July 2009
Location: London. England
Senior Member
Hi,

What is the status of being able to use OCL in EMF models to enable generation of behaviour in the generated code?

Many thanks,
Joel
Re: OCL to generate EMF behaviour [message #499450 is a reply to message #499427] Sun, 22 November 2009 17:11 Go to previous messageGo to next message
Ed Merks is currently offline Ed MerksFriend
Messages: 33113
Registered: July 2009
Senior Member
Joel,

https://bugs.eclipse.org/bugs/show_bug.cgi?id=255469 is complete and
waiting to be pulled into a build.

Joel Rosi-Schwartz wrote:
> Hi,
>
> What is the status of being able to use OCL in EMF models to enable
> generation of behaviour in the generated code?
>
> Many thanks,
> Joel


Ed Merks
Professional Support: https://www.macromodeling.com/
Re: OCL to generate EMF behaviour [message #499454 is a reply to message #499450] Sun, 22 November 2009 18:04 Go to previous messageGo to next message
Joel Rosi-Schwartz is currently offline Joel Rosi-SchwartzFriend
Messages: 624
Registered: July 2009
Location: London. England
Senior Member
Ed,

I have read the bug thread and I am still not clear. It appears to me that the underlying work to enable OCL based generation is ready to be put into place, but that the actual generation code is still to be done. Do I have that right?

Is the patch from Bug 255469 going to be backed into EMF 2.5 or will it only be applied to 2.6?

We are now actively working on moving ORMF forward. We plan to move the high level analysis model (which is already in EMF) to a detailed EMF model starting around Christmas time, with an aim to be generating initial code from the model early in the new year. My preference would be define as much behaviour in the model as possible from the start using OCL. How practical is it a this stage?

Thanks,
Joel
Re: OCL to generate EMF behaviour [message #499464 is a reply to message #499454] Sun, 22 November 2009 20:09 Go to previous messageGo to next message
Ed Merks is currently offline Ed MerksFriend
Messages: 33113
Registered: July 2009
Senior Member
Joel,

Comments below.

Joel Rosi-Schwartz wrote:
> Ed,
>
> I have read the bug thread and I am still not clear. It appears to me
> that the underlying work to enable OCL based generation is ready to be
> put into place, but that the actual generation code is still to be
> done. Do I have that right?
I think it will generate code that still delegates to a registered OCL
intepretted; the OCL changes to register the interpreter still need to
be committed.
>
> Is the patch from Bug 255469 going to be backed into EMF 2.5 or will
> it only be applied to 2.6?
Feature enhancement work is never done in maintenance stream.
>
> We are now actively working on moving ORMF forward. We plan to move
> the high level analysis model (which is already in EMF) to a detailed
> EMF model starting around Christmas time, with an aim to be generating
> initial code from the model early in the new year. My preference would
> be define as much behaviour in the model as possible from the start
> using OCL. How practical is it a this stage?
It sounds like a workable plan given the state of the code. We'd of
course be happy to get feedback about problems so we can fix them.
>
> Thanks,
> Joel


Ed Merks
Professional Support: https://www.macromodeling.com/
Re: OCL to generate EMF behaviour [message #499469 is a reply to message #499464] Sun, 22 November 2009 21:17 Go to previous messageGo to next message
Joel Rosi-Schwartz is currently offline Joel Rosi-SchwartzFriend
Messages: 624
Registered: July 2009
Location: London. England
Senior Member
Ed,

I will give this a try, guinea pig though I may be Wink It will be my pleasure to provide feedback and any assistance that I may be able.

Thanks,
Joel

[Updated on: Sun, 22 November 2009 21:17]

Report message to a moderator

Re: OCL to generate EMF behaviour [message #503722 is a reply to message #499427] Tue, 15 December 2009 08:50 Go to previous messageGo to next message
Joel Rosi-Schwartz is currently offline Joel Rosi-SchwartzFriend
Messages: 624
Registered: July 2009
Location: London. England
Senior Member
Hi,

I was ready to start trying this, but on looking at Bugzilla 255469 still needs to be put into an EMF build and probably more important 291365 still has to be completed for OCL. Can you please confirm that until both of these are fixed and in builds that this functionality will not be ready for testing.

While I would really like to be able to give this a go, nor do I particularly want to fight Windmills Wink


Thanks,
Joel
Re: OCL to generate EMF behaviour [message #503980 is a reply to message #503722] Wed, 16 December 2009 12:10 Go to previous messageGo to next message
Ed Merks is currently offline Ed MerksFriend
Messages: 33113
Registered: July 2009
Senior Member
Joel,

The EMF things are done, but our release notes generator hasn't been
working the last while. :-(

Best to ask about these bugzillas on the bugzillas themselves...


Joel Rosi-Schwartz wrote:
> Hi,
>
> I was ready to start trying this, but on looking at Bugzilla
> https://bugs.eclipse.org/bugs/show_bug.cgi?id=255469 still needs to be
> put into an EMF build and probably more important
> https://bugs.eclipse.org/bugs/show_bug.cgi?id=291365 still has to be
> completed for OCL. Can you please confirm that until both of these are
> fixed and in builds that this functionality will not be ready for
> testing.
>
> While I would really like to be able to give this a go, nor do I
> particularly want to fight Windmills ;)
>
>
> Thanks,
> Joel


Ed Merks
Professional Support: https://www.macromodeling.com/
Re: OCL to generate EMF behaviour [message #503985 is a reply to message #503980] Wed, 16 December 2009 17:32 Go to previous messageGo to next message
Joel Rosi-Schwartz is currently offline Joel Rosi-SchwartzFriend
Messages: 624
Registered: July 2009
Location: London. England
Senior Member
Hi Ed,


Best thing would be to rebuild the release notes generator using EMF to make it dependable Wink

So what does it mean that the "The EMF things are done"? Where can I find clues on how to add constraints to a model? I am not that fussed if I can actually generate code yet from those constraints, but it would be nice to know that I will able to in say the next three months. If this means digging into the code for hints, then can you provide a few pointers on where to start. (But I hope not, because at the moment I am busy reading CDO code;-)

Ed Merks wrote on Wed, 16 December 2009 12:10
Joel,
The EMF things are done, but our release notes generator hasn't been
working the last while. Sad



Will do.

Quote:

Best to ask about these bugzillas on the bugzillas themselves...



Thanks,
Joel
Re: OCL to generate EMF behaviour [message #504131 is a reply to message #503985] Thu, 17 December 2009 06:50 Go to previous messageGo to next message
Ed Merks is currently offline Ed MerksFriend
Messages: 33113
Registered: July 2009
Senior Member
Joel,

The bugzillas to support this were committed by Kenn and are in a
build. It's only the OCL registration that's needed. Kenn was planning
to write some helpful information about it, but like so many of us, he's
stretched for time. I think the bugzillas themselves have a bit of
information to get started...



Joel Rosi-Schwartz wrote:
> Hi Ed,
>
>
> Best thing would be to rebuild the release notes generator using EMF
> to make it dependable ;)
>
> So what does it mean that the "The EMF things are done"? Where can I
> find clues on how to add constraints to a model? I am not that fussed
> if I can actually generate code yet from those constraints, but it
> would be nice to know that I will able to in say the next three
> months. If this means digging into the code for hints, then can you
> provide a few pointers on where to start. (But I hope not, because at
> the moment I am busy reading CDO code;-)
>
> Ed Merks wrote on Wed, 16 December 2009 12:10
>> Joel,
>> The EMF things are done, but our release notes generator hasn't been
>> working the last while. :(
>
>
> Will do.
>
> Quote:
>> Best to ask about these bugzillas on the bugzillas themselves...
>
>
> Thanks,
> Joel


Ed Merks
Professional Support: https://www.macromodeling.com/
Re: OCL to generate EMF behaviour [message #504177 is a reply to message #499427] Thu, 17 December 2009 15:11 Go to previous message
Joel Rosi-Schwartz is currently offline Joel Rosi-SchwartzFriend
Messages: 624
Registered: July 2009
Location: London. England
Senior Member
I emailed Kenn Hussey about this and here is his reply

Quote:

I actually provided fairly detailed descriptions of the EMF enhancements on the New & Noteworthy page for Helios - see http://wiki.eclipse.org/EMF/New_and_Noteworthy/Helios

As it currently stands, EMF now supports the ability to "plug in" delegates for constraints/invariants, feature settings, and operation invocations. However, in order to use a particular engine (e.g. the OCL engine), delegates for that engine must be plugged in. As you've discovered, bugs are open to do this for OCL, and I attached preliminary patches at one point (which I've since updated in my development workspace) but as yet I have been unable to find the time/funding to complete the work. With any luck, that could happen in time for the next milestone...

[Updated on: Thu, 17 December 2009 10:12]

Report message to a moderator

Previous Topic:Problems with emf ecore model
Next Topic:Outputting an EObject to a string for debugging or logging
Goto Forum:
  


Current Time: Fri Mar 29 02:15:32 GMT 2024

Powered by FUDForum. Page generated in 0.04949 seconds
.:: Contact :: Home ::.

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

Back to the top