> But now suppose we refactor it to look like this:
>
>    public SortedList<int> FindPrimesBelow(int max)
>    {
>      SortedList<int> candidates = InitialiseCandidates(max);
>      SortedList<int> primes = new SortedList<int>();
>
>      while (candidates.Size() > 0)
>      {
>        int prime = GetNextPrime(candidates);
>        primes.Add(prime);
>        RemovePrimeAndItsMultiples(prime, candidates, max);
>      }
>
>      return primes;
>    }
>
>    private SortedList<int> InitialiseCandidates(int max)
>    {
>      SortedList<int> candidates = new SortedList<int>();
>
>      for (int n=2; n<max; n++)
>      {
>        candidates.Add(n);
>      }
>
>      return candidates;
>    }
>
>    private int GetNextPrime(SortedList<int> candidates)
>    {
>      return candidates[0];
>    }
>
>    private int RemovePrimeAndItsMultiples(int prime, SortedList<int>
> candidates, int max)
>    {
>      for (int k=1; k*prime < max; k++)
>      {
>        candidates.Remove(k*prime);
>      }
>    }

So you're writing 3x as much code just to avoid comments? Is the above 
really that much more readable to justify the extra work writing it?

public SortedList<int> Primes(int max)
   {
     SortedList<int> candidates = new SortedList<int>();
     SortedList<int> primes = new SortedList<int>();
     for (int n=2; n<max; n++) candidates.Add(n); // initialise
     while (candidates.Size() > 0)
     {
       int p = candidates[0]; // get next prime
       primes.Add(p);
       for (int k=1; k*p < max; k++) candidates.Remove(k*p); // remove 
multiples
     }
     return primes;
   }

Post a reply to this message