Added shuffle functions

[jmcwilliams403]

based on Knuth shuffle. May need to add clarification in comments for where this algorithm came from.

Seeds are given as capital Long object wrappers to allow null seeds to indicate unseeded randomness.

[tommyettinger]

Hm... The implementation of the Knuth (AKA Fisher-Yates) shuffle algorithm looks solid, but instantiating new Random objects for each shuffle creates unexpected garbage. I'm also not certain how well Random mixes its seed, since it takes a 64-bit seed but only has a period of 2 to the 48, so if setSeed() doesn't adequately mix its input, identical shuffles with different Long seeds could be produced intentionally. My preference would be to take a Random as an argument, rather than one Long, which allows for subclasses of Random to be used (my juniper library is full of those subclasses, and also already has shuffling). Using a Random as a parameter also allows libGDX's RandomXS128 to be used.

Just within digital, there are a few static methods in Hasher that can generate pseudo-random bounded ints given a long seed. Using a sequential seed within the shuffle should work with any of randomize1Bounded, randomize2Bounded, or randomize3Bounded, but it might have detectable quirks/bias if you had several arrays of the same length, like arrayA, arrayB, and arrayC, and called shuffle(arrayA, 1L) and then shuffle(arrayB, 2L), shuffle(arrayC, 3L), etc. (changing the seed only by a small amount).

So, I'm thinking I won't merge this as-is, but I can merge it and then make the changes I have in mind. Mostly, I just want to avoid creating garbage references to temporary Random objects, but there are some other things I could add too -- shuffling rectangular 2D arrays so any element can be shuffled to any position is harder than it seems, and having some support for that could be nice.