* CRC(A000A000A000...), CRC(0B000B000B...), CRC(00C000C000C...), CRC(000D000D000D...)
     * and then to XOR them together.  The strideTable enables us to hash an int followed by 12
     * zero bytes (3 ints), while the byteTable is for advancing one byte at a time.
     * This algorithm is due to the paper "Everything we know about CRC but [are] afraid to forget"
     * by Kadatch and Jenkins, 2010.
     */

    Crc32cHasher() {
      super(16);