| resolveTemp crashes the ATL module [message #1120568] |
Sun, 29 September 2013 15:22  |
 Leonardo Montecchi
Messages: 3
Registered: September 2013 |
Junior Member |
|
|
Hello,
I'm trying to use the resolveTemp function within an ATL module. However, I'm experiencing a very strange behavior:
When calling
thisModule.resolveTemp(v, 'name');
...
If NO rule exists for 'v' creating an object named 'name', then the function is executed and returns OclUndefined, as it should
If such rule exists, then the ATL module creashes: the execution is interrupted (with not exceptions on the console) and the output file is not generated at all.
Coult you please help me, or point to some direction? I'm calling
thisModule.resolveTemp(e, 'detected').debug();
from the "do" part of a MatchedRule, and the resolveTemp is supposed to reference another MatchedRule... so it should be the right place within the module.
Thanks.
Leonardo Montecchi, PhD Student
Università degli Studi di Firenze
|
|
|
| Re: resolveTemp crashes the ATL module [message #1122366 is a reply to message #1120568] |
Tue, 01 October 2013 09:45  |
Hugo Bruneliere
Messages: 567
Registered: July 2009 |
Senior Member |
|
|
Hello,
Have you tried to call to the resolveTemp method from outside of the "do" section (i.e., in the regular matched rule assignments)?
Such "do" sections are dedicated to imperative constructs that should be use only when strictly needed.
In a great majority of cases, ATL declarative constructs (i.e., matched rules) are sufficient.
Best regards,
------------------------------------------
Hugo Bruneliere - R&D Engineer
AtlanMod research team (Inria, EMN & LINA)
Ecole des Mines de Nantes
Nantes - France
------------------------------------------
|
|
|
| Re: resolveTemp crashes the ATL module [message #1123457 is a reply to message #1122366] |
Wed, 02 October 2013 11:31 |
Leonardo Montecchi
Messages: 3
Registered: September 2013 |
Junior Member |
|
|
Hello,
thanks for your answer.
It took me a while to figure out how to do the same thing within regular matched rules, but it is possible, thanks for your suggestion.
It actually works if called from another rule, in the declarative section.
That is, instead of calling resolveTemp in rule A for finding element X of rule B, I call it in rule B for finding element Y of rule A.
However the module still crashes if I add the .debug() operation:
target <- thisModule.resolveTemp(e.refImmediateComposite(),'reset'),
--OK
target <- thisModule.resolveTemp(e.refImmediateComposite(),'reset').debug(),
--Crashes (with no visible hints)
Best regards,
Leonardo Montecchi, PhD Student
Università degli Studi di Firenze
|
|
|
| Re: resolveTemp crashes the ATL module [message #1127972 is a reply to message #1123457] |
Mon, 07 October 2013 03:36 |
Ronan B
Messages: 221
Registered: July 2009 |
Senior Member |
|
|
Hi,
In my experience using debug statements with resolveTemp can result in different behaviour compared with just calling resolveTemp. This may be related to the semantics of resolveTemp which is using the generated target model.
Regrads,
Ronan
|
|
|
Powered by
FUDForum. Page generated in 0.01514 seconds
] 
Search
Help
Register
Login
Home
