Oct 16th, 2008, 04:00 AM
Step-Into problem with method interception
I use the AOP functionality (with Dynamic Proxies) of Spring. This functionality is very handy when implementing cross-cutting concern and saved me a lot of boilerplate coding (as the literature always promised me).
But when it comes to debugging AOP makes my life harder: when you step-into the intercepted method then you first meet the Spring AOP infrastructure. After finding the right positions where to step into, you get to the interceptors that are applied on the target. After you passed all interceports you finally get to the intercepted target (the place where you originally wanted to get to).
Of course you can set a breakpoint at the intercepted target and resume the execution to this breakpoint, but this isn't very intuitive when debugging. Furthermore, there is no need to debug through the AOP infrastructure or the interceptors most of the time. Application programmers normally are not interested in die AOP functionality (because it contains infrastructure cross-cutting concerns) or even don't know what AOP is. For these people debugging gets very cumbersome and they do not understand what's going on.
Summing up, I'm searching for a solution how to step over the AOP code easily. There should be a possibility how application programmers do not get in touch with AOP, while leveraging the power of AOP in the background.
Are there any thoughts or strategies on this issue?
Thanks a lot for any hints,