[
Date Prev][
Date Next][
Thread Prev][
Thread Next][
Date Index][
Thread Index]
[
List Home]
|
Re: [aspectj-users] No warning for class not found in parameter type
|
Hi Kevin,
So there were classes with the join points in on the classpath that needed promoting to the inpath? To put out a warning, wouldn't AspectJ have had to completely analyse everything on the classpath for potential join points to see if this kind of mistake had been made? That would be rather expensive I think. I guess that is one reason why the inpath exists in the first place, so that we don't analyse/weave more classes than necessary. I can see it was awkward to debug for you, I'm just having trouble working out what AspectJ could do better that wouldn't kill performance.
cheers,
Andy.
2009/3/11 Kevin Roll
<kroll@xxxxxxxxxxxx>
I just spent hours tracking down a problem which turned out to be a peculiarity in AspectJ. I have a pointcut
pointcut setter(PropertyChangeSupported b): execution(void set*(*)) && this(b);
which I use to pick out setters of a bean. The advice that uses this pointcut was not being applied when the parameter to the method was a non-primitive type. After much investigation I determined that the classes at issue were not specified in the inpath; adding them solved the problem. I use a tightly controlled inpath to minimize the amount of time that the AspectJ compiler spends when running. The behavior of simply ignoring those methods was very unexpected to me; it seems that some sort of warning should have been issued. Thoughts?
_______________________________________________
aspectj-users mailing list
aspectj-users@xxxxxxxxxxx
https://dev.eclipse.org/mailman/listinfo/aspectj-users
