[great-users] specifying this complex pattern

Tivadar Szemethy tiv at isis.vanderbilt.edu
Thu Apr 6 22:01:45 CDT 2006


In such cases I draw up a "something similar" pattern, read what the 
code generator produces for it and hand-craft a guard condition  after 
pondering over the generated code for a while.

Zero-cardinality on simple associations works well for me.
On the first glance, this pattern seems doable for me (except that I 
don't understand what you mean by "unique connections").


Tivadar


Matthew J. Emerson wrote:

> I need to be able to specify a complex pattern, and have it work 
> correctly in both the GRE and the Code Generator versions of my 
> transformation. Here's the pattern I need to bind:
>
>  
>
> One object that is the source of two or more unique connections, each 
> to different destination objects, where two or more of the destination 
> objects do not have a simple association with a flag object.
>
>  
>
> Once I have matched this pattern, I need to create a new simple 
> association between any one of the destination objects and a 
> newly-created flag object.
>
>  
>
> I think that the fact that negative bindings do not work makes this 
> impossible to do. The usual work-around of using test-cases instead of 
> negative patterns will not work here because I need flag one and only 
> one destination object after I match the pattern.
>
>  
>
> So, if anyone managed to follow that, can you offer any help?
>
>  
>
> --Matt
>
>------------------------------------------------------------------------
>
>_______________________________________________
>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/20060406/71064110/attachment.htm


More information about the great-users mailing list