Support performance measurement for solutions

[ihac]

Three concerns:

1. compute complexity (in theory)
2. average execution time
3. average memory usage

It stills require a deeper research on which benchmark framework is best, and easy-to-use of course.

[ihac]

inline benchmarking provided by ScalaMeter (import scalameter for benchmarking) is quite convenient:

  test("compare performance by large test set") {
    val inputs = 10 to 100000 by 1000
    val raw_time = standardConfig measure {
      inputs.foreach(_.totient)
    }
    val improved_time = standardConfig measure {
      inputs.foreach(_.totient_improve)
    }

    println(s"totient: ${raw_time}, totient_improved: ${improved_time}")
    println(s"speedup: ${raw_time.value / improved_time.value}")
  }