So I think this is reasonable (its the pattern we followed in bzr and it worked very well there). We may well want to update list-tests to have a --subunit option too, which would suite your case better (since you want to do programmatic introspection).
So I think this is reasonable (its the pattern we followed in bzr and it worked very well there). We may well want to update list-tests to have a --subunit option too, which would suite your case better (since you want to do programmatic introspection).