Based off the backtracking algorithm in 'The Algorithm Design Manual' but written in python.

backtracking-perms.py

__author__ = 'cody'


finished = False # found all solutions yet?

def backtrack(a, k):
    if is_a_solution(a, k):
        return process_solution(a)
    else:
        candidates = construct_candidates(a, k)
       # print(candidates)
        for candidate in candidates:
           # make_move(a,k,data_input)
           backtrack(a + [candidate], k)
           # unmake_move(a,k,data_input)
           if finished:
               return # terminate early


def construct_candidates(a, k):
    return [candidate for candidate in k if candidate not in a ]

def process_solution(a):
    print("permutation", a)
    
def is_a_solution(a, k):
    return len(a) == 3

def generate_permutations():
    backtrack([],['red', 'green', 'blue'])

generate_permutations()