[inahga]

checkPrimeFactorsTooClose() doesn't check the ultimate case of factors being too close.

[aarongable]

LGTM. Wish we could do this by just changing the start point of fermat's algorithm, but I don't think there's a clean way to do that due to BigInt.Sqrt taking the floor.

[inahga]

It can be done, but it harms comprehensibility IMO

      399 -  a := new(big.Int)                                                                                                  
      400 -  a.Sqrt(n).Add(a, one)                                                                                                                                                                             
      400 +  a := new(big.Int).Sqrt(n)                                                                                          
      401 +  b2 := new(big.Int).Mul(a, a)                                                                                       
      402 +  if b2.Cmp(n) != 0 {                                                                                                
      403 +    a.Add(a, one)                                                                                                    
      404 +    b2.Mul(a, a)                                                                                                     
      405 +  }                                                                                                                  
      406  
      407    // We calculate b2 to see if it is a perfect square (i.e. b^2), and therefore
      408    // b is an integer. Specifically, b2 = a^2 - n.
      404 -  b2 := new(big.Int)                                                                                                 
      405 -  b2.Mul(a, a).Sub(b2, n)                                                                                            
      409 +  b2.Sub(b2, n)                       

[aarongable]

Yep agreed, the cleanliness is critical for code that's already this unreadable.