Generators[python]

Generators.md

Generators are implemented with syntax very similar to functions, but instead of returning values, a yield statement is executed to indicate each element in the series

Consider the goal of computing all factors of a positive integer. A traditional function might return a list contating all factors, implemeted as follows:

def factors(n):            # traditional function that computes factors
  results = []             # store factors in a new list
  for k in range(1, n+1):
    if n % k == 0:         # divides evenly, thus k is a factor
      results.append(k)    # add k to the list of factors
  return results           # return the entire list

In contrast, an implementation of a generator for computing those factors could be implemented as follows

def factors(n):             # generator that computes factors
  for k in range(1, n+1): 
    if n % k == 0:          # divides evenly, thus k is a factor
      yield k               # yield this factor as next result

when the flow of control naturally reaches the end of the procedure or a zero-argument return statement, a StopIteration exception is automatically raised.

An iterator can relly on multiple yield statements in different constructs, with the generated series determined by the natural flow of control. For example, a more efficient version of the above generator:

def factors(n):                        # generator that computes factors
 k = 1
 while k * k < n:                      # while k < sqrt(n)
   if n % k == 0:
     yield k
     yield n // k
   k += 1
 if k * k == n:                        # special case if n is a perfect square
   yield k