Small optimization of zero hashCode

[TangBean]

Small optimization of zero hashCode: Cache if the hash has been calculated as actually being zero, enabling us to avoid recalculating this.

[patrickfav]

Hi @TangBean and thanks for the PR!

Looking at the implementation of hashCode:

        static int hashCode(byte[] byteArray, ByteOrder byteOrder) {
            int result = Arrays.hashCode(byteArray);
            result = 31 * result + (byteOrder != null ? byteOrder.hashCode() : 0);
            return result;
        }

and

Arrays.hashCode

    public static int hashCode(byte a[]) {
        if (a == null)
            return 0;

        int result = 1;
        for (byte element : a)
            result = 31 * result + element;

        return result;
    }

The internalArray can never be null, so I dont see any possible way the actual hashCode will ever be "0"?

Update: Just for the funs: ChatGPT 4 agrees with me:

[patrickfav]

Anyway, thanks for the hint about the overall feature, I found out it can be quite buggy because there are no gurantees that the inner byte array is not mutatetduring the life of an Bytes instance, so I removed the feature all toghther: #59