这是受Haskell实现使用列表理解的启发：

def permutation(list):
    if len(list) == 0:
        return [[]]
    else:
        return [[x] + ys for x in list for ys in permutation(delete(list, x))]

def delete(list, item):
    lc = list[:]
    lc.remove(item)
    return lc

首先，导入itertools：

import itertools

排列（顺序重要）：

print(list(itertools.permutations([1,2,3,4], 2)))

[(1, 2), (1, 3), (1, 4),
(2, 1), (2, 3), (2, 4),
(3, 1), (3, 2), (3, 4),
(4, 1), (4, 2), (4, 3)]

组合（顺序无关紧要）：

print(list(itertools.combinations('123', 2)))

[('1', '2'), ('1', '3'), ('2', '3')]

笛卡尔积（具有多个可迭代项）：

print(list(itertools.product([1,2,3], [4,5,6])))

[(1, 4), (1, 5), (1, 6),
(2, 4), (2, 5), (2, 6),
(3, 4), (3, 5), (3, 6)]

笛卡尔积（具有一个可迭代的和自身）：

print(list(itertools.product([1,2], repeat=3)))

[(1, 1, 1), (1, 1, 2), (1, 2, 1), (1, 2, 2),
(2, 1, 1), (2, 1, 2), (2, 2, 1), (2, 2, 2)]

人们确实可以对每个排列的第一个元素进行迭代，正如tzwen的答案。然而，这样编写此解决方案更有效：

def all_perms(elements):
    if len(elements) <= 1:
        yield elements  # Only permutation possible = no permutation
    else:
        # Iteration over the first element in the result permutation:
        for (index, first_elmt) in enumerate(elements):
            other_elmts = elements[:index]+elements[index+1:]
            for permutation in all_perms(other_elmts): 
                yield [first_elmt] + permutation

这个解决方案大约快了30%，显然是因为递归以len（元素）<=1而不是0结尾。它的内存效率也高得多，因为它使用了一个生成器函数（通过yield），就像Riccardo Reyes的解决方案一样。

注意，该算法具有n个阶乘时间复杂度，其中n是输入列表的长度

打印跑步结果：

global result
result = [] 

def permutation(li):
if li == [] or li == None:
    return

if len(li) == 1:
    result.append(li[0])
    print result
    result.pop()
    return

for i in range(0,len(li)):
    result.append(li[i])
    permutation(li[:i] + li[i+1:])
    result.pop()    

例子：

permutation([1,2,3])

输出：

[1, 2, 3]
[1, 3, 2]
[2, 1, 3]
[2, 3, 1]
[3, 1, 2]
[3, 2, 1]

这是初始排序后生成排列的渐近最优方式O（n*n！）。

有n个！最多进行一次置换，且具有下一次置换（..），以O（n）时间复杂度运行

在3个步骤中，

找到最大的j，使a[j]可以增加以最小可行量增加a[j]找到扩展新a[0..j]的字典最少方法

'''
Lexicographic permutation generation

consider example array state of [1,5,6,4,3,2] for sorted [1,2,3,4,5,6]
after 56432(treat as number) ->nothing larger than 6432(using 6,4,3,2) beginning with 5
so 6 is next larger and 2345(least using numbers other than 6)
so [1, 6,2,3,4,5]
'''
def hasNextPermutation(array, len):
    ' Base Condition '
    if(len ==1):
        return False
    '''
    Set j = last-2 and find first j such that a[j] < a[j+1]
    If no such j(j==-1) then we have visited all permutations
    after this step a[j+1]>=..>=a[len-1] and a[j]<a[j+1]

    a[j]=5 or j=1, 6>5>4>3>2
    '''
    j = len -2
    while (j >= 0 and array[j] >= array[j + 1]):
        j= j-1
    if(j==-1):
        return False
    # print(f"After step 2 for j {j}  {array}")
    '''
    decrease l (from n-1 to j) repeatedly until a[j]<a[l]
    Then swap a[j], a[l]
    a[l] is the smallest element > a[j] that can follow a[l]...a[j-1] in permutation
    before swap we have a[j+1]>=..>=a[l-1]>=a[l]>a[j]>=a[l+1]>=..>=a[len-1]
    after swap -> a[j+1]>=..>=a[l-1]>=a[j]>a[l]>=a[l+1]>=..>=a[len-1]

    a[l]=6 or l=2, j=1 just before swap [1, 5, 6, 4, 3, 2] 
    after swap [1, 6, 5, 4, 3, 2] a[l]=5, a[j]=6
    '''
    l = len -1
    while(array[j] >= array[l]):
        l = l-1
    # print(f"After step 3 for l={l}, j={j} before swap {array}")
    array[j], array[l] = array[l], array[j]
    # print(f"After step 3 for l={l} j={j} after swap {array}")
    '''
    Reverse a[j+1...len-1](both inclusive)

    after reversing [1, 6, 2, 3, 4, 5]
    '''
    array[j+1:len] = reversed(array[j+1:len])
    # print(f"After step 4 reversing {array}")
    return True

array = [1,2,4,4,5]
array.sort()
len = len(array)
count =1
print(array)
'''
The algorithm visits every permutation in lexicographic order
generating one by one
'''
while(hasNextPermutation(array, len)):
    print(array)
    count = count +1
# The number of permutations will be n! if no duplicates are present, else less than that
# [1,4,3,3,2] -> 5!/2!=60
print(f"Number of permutations: {count}")

def pzip(c, seq):
    result = []
    for item in seq:
        for i in range(len(item)+1):
            result.append(item[i:]+c+item[:i])
    return result


def perm(line):
    seq = [c for c in line]
    if len(seq) <=1 :
        return seq
    else:
        return pzip(seq[0], perm(seq[1:]))