> > You can easily > > > compare class/method signatures, but what about actual > code? How do > you > > > check if the body of the generated DO.equals() has the > code you're > > > expecting?! > > I think this is beyond reasonable scope of automated regression > testing. > > I'd say, write a unit test for the generated class. > > Yup, agree. In fact a good way of testing the generated code > is to actually call the code.
So. To begin with something simple. I will check in a junit part so that everybody can add unit tests when they see something wrong in a generation of samples. I don't mean calling a jboss/... that run the bean but call the CMP/Session/... Class directly. We can even use simple Reflection at the beginning to verify an interface have a correct definition, throws the correct exceptions, etc.... I leave XML to Dmitri. The things that breaks often is inheritance in Class and deployment descriptors. Unit testing Xdoclet classes themselves would be good but later. > > Ara. Vincent "Petit � petit l'oiseau fait son nid" _______________________________________________ Xdoclet-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/xdoclet-devel
