Analyzing Fuel Problem

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|

Analyzing Fuel Problem

Sean P. DeNigris
Administrator
I had an object graph that was pulling in all sorts of unrelated classes when serialized with Fuel. I was trying to find the offending object, and came up with the following:
```
(FLAnalyzer newDefault analysisFor: root) clusterization globalsBucket
```

This worked, but I was wondering if that's "the right way" to do it...

Thanks.
Cheers,
Sean
Reply | Threaded
Open this post in threaded view
|

Re: Analyzing Fuel Problem

Mariano Martinez Peck
Hi Sean,

Sorry for the delay on the answer. Yes, that's a good spot. Another thing I usually do is to put a halt in FLSerialization >> run. Just after #analysisStep for example, and then inspect the local variables. 

Also, if you want to understand WHY an object is getting in, then you can also halt in #trace and #privateTrace: from FLAnalysis. That way you monitor which objects gets pushed into the analysis stack. You can do something like

self haltIf: [ anObject whateverConditionToHaltOnDesiredObject = true ].

Finally, there are more tools for debugging this kind of issue, but I don't know their state. You can read more about them here [1].

Best regards,



On Fri, Jun 2, 2017 at 10:26 PM, Sean P. DeNigris <[hidden email]> wrote:
I had an object graph that was pulling in all sorts of unrelated classes when
serialized with Fuel. I was trying to find the offending object, and came up
with the following:
```
(FLAnalyzer newDefault analysisFor: root) clusterization globalsBucket
```

This worked, but I was wondering if that's "the right way" to do it...

Thanks.



-----
Cheers,
Sean
--
View this message in context: http://forum.world.st/Analyzing-Fuel-Problem-tp4949104.html
Sent from the Pharo Smalltalk Users mailing list archive at Nabble.com.




--