On Sun, Dec 6, 2009 at 6:11 PM, Adam Murdoch <[email protected]> wrote:
> > <lots of stuff that sounds interesting, but doesn't help to solve the > problem...> > > One option, for now, is to use Groovy to dynamically implement > ITestListener at runtime, based on a classpath built from the testCompile > configuration. Another is to defer the support for TestListeners with testng > until we've sorted out some of the plug-in infrastructure. I don't think we > should force a particular version of testng on the user, this feels like a > bit of a backward step. > > I must pass the name of the class to the TestNG process and let it instantiate the object. This seems to preclude using any Groovy magic to dynamically implement the listener. I could probably generate the bytecode for the class at runtime using the ASM stuff, but that's creepy (and would take me too long as I've never done that before). For now I've just commented out the TestNG support. (Sorry, Tomek.) -- John Murph Automated Logic Research Team
