Java and C# version fixed.Try it again~

In case you missed it, I approved.

I added a new test to the c# implementation. The new test is SquaresAreNotPronicTests which tests 10,000 perfect squares from (32768)^2 through (42767)^2. In retrospect, I should probably have went much larger with the squares -- I should have gone much closer to the Int64 max.

I'm thinking this new test should be implemented in java as well. I may take a swing at it -- since the Java syntax isn't far from that of c#.

I wish I had implemented this test sooner since it invalidated a couple of implementations (including one that surprised me that it invalidated.)

It looks like the initial/sample-test is just an empty template test.
Let's add a few sample tests, but otherwise it looks good.

Should the array always be sorted? If so, it should be added in the description. If not, the Java translation has a problem because my solution passed without sorting the array (I saw afterward that Hullaballoo mutate the input array to sort it. Might be a bad idea).

EDIT: looking at JS test suite, seems like this should actually be an issue!

Approved. Thanks ;-)

Approved, thanks!