On 2018-02-09, Jaikiran Pai wrote: > That test was introduced long back as part of [5] to test another > similar NPE. I think with the change I now have, it no longer reaches > this place in the code where the Copy task would end up getting a null > from a mapper, because various other utility classes already handle > this contract and skip that resource.
I completely agree with your fix. What this specific test failure shows, though, is that we now no longer are able to distinguish between a resource that we could copy but can't because it doesn't provide a name and the case of "all resources have been stripped out by mappers as "I don't know how to handle it". Unfortunately I don't see any way to distinguish between "all resources have been excluded by mappers" and "all resources have been up to date" and certainly not how to figure out a resource didn't have a name when we want to allow such resources if followed by a merge mapper. I'm fine with removing the test but would want us to document the change in behavior. Where a resource with a null name used to cause a build exception it will now be silently ignored by copy. Stefan --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@ant.apache.org For additional commands, e-mail: dev-h...@ant.apache.org
