Community
Participate
Working Groups
... which is not a problem, but looks ugly in a test coverage tool output. You can never get 100% coverage in a file that contains such an ITD :) A way of fixing it would be making these methods static, and calling them from a <clinit> block - unnoticable performance penalty, and only slight code size increase. I suspect that similar problem exists with other kinds of ITDs.
After more thoguths, and reading helpful comments from Alexandre and Andy in Bug Bug 103471, I decided that the best approach will be introducing a bit of knowledge of AspectJ into the tool I am working on. - small impact on the tool (basicly, if a method has "org.eclipse.aspectj.AjSynthetic" attribute, don't instrument it) - no change in AJ required - claiming AspectJ compatiblitiy will be good marketing for Coverlipse ;) resolving as INVALID - not an AJ bug.