[great-users] Transformation idioms

[Tivadar Szemethy]

Sandeep, Joe

> Joe,
>
> This did come up in prior discussions wrt GREAT. The idea there was to 
> create templatized patterns that could be instantiated and reused in 
> different transformation contexts. I am not sure to what extent it has 
> been implemented - perhaps one of the core developers could comment -
>
(I'm not a GR developer, I'm just chiming in if I might)

This can _almost_ be done even with the current GReAT, I tried it once.
(and I'm not sure if what's  preventing you from doing it is actually a 
bug or a feature).

The idea is that you define your generic patterns on MgaObjects of a 
very simplistic metamodel, and then use the
LibraryUpdater to customize the patterns for your metamodel (i.e. 
redirect MgaObj references). This part  works.

What does not work though is that you cannot create generic connections, 
i.e. the pattern matcher does not recognize
"Any" connections associated with MgaObjects. I can think of this as a 
bug, since I'm not sure what ConnectorAny was
intended for (if not for this - I know that connection inheritance is a 
shady business in GReAT anyway, so I'm not sure).


Tivadar


>  
>
> Thanks,
>
> Sandeep
>
>  
>
> --
>
> Sandeep Neema
>
> Research Scientist,
>
> Institute of Software Integrated Systems, Vanderbilt University
>
> Phone: 615-343-9996
>
> Email: sandeep.k.neema at vanderbilt.edu 
> <mailto:sandeep.k.neema at vanderbilt.edu>
>
> ------------------------------------------------------------------------
>
> *From:* great-users-bounces at list.isis.vanderbilt.edu 
> [mailto:great-users-bounces at list.isis.vanderbilt.edu] *On Behalf Of 
> *Joe Porter
> *Sent:* Tuesday, May 09, 2006 9:20 AM
> *To:* great-users
> *Subject:* [great-users] Transformation idioms
>
>  
>
> I have a question:
>
> Looking at the approach and capabilities of the GReAT toolchain, it 
> seems like you would want to have some common algorithmic idioms.  I'm 
> thinking of things like finding disjoint subgraphs or spanning trees.  
> One of the examples also mentioned topological sort. I'm stretching a 
> little here since I haven't used the tool, but this is a journey of 
> understanding.   Is there a way to do that sort of thing generically 
> for different metamodels, or are things like that even useful in 
> practice?  Or rather, have I missed something regarding the approach?
>
> Hopefully all of you that use GReAT on a regular basis can steer me in 
> the right direction.
>
> Thanks,
> -Joe Porter
> Southwest Research Institute
>
>------------------------------------------------------------------------
>
>_______________________________________________
>great-users mailing list
>great-users at list.isis.vanderbilt.edu
>http://list.isis.vanderbilt.edu/mailman/listinfo/great-users
>  
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://list.isis.vanderbilt.edu/pipermail/great-users/attachments/20060509/4286a727/attachment.htm