python non recursive solution to n queen problem

The complexity may be a little high-and I didn't pay much attention to it

I've been thinking for a long time and looking for a long time, but I don't see anything that can solve the n queen problem with python and don't call recursion because I don't quite understand recursion (especially when it comes to n layer). IQ is limited-

import copy

def check(A,x,y):
  B=[]
  flag=True
  for i in range(len(A)):
    for j in range(len(A)):
      if A[i][j]==1:
        B.append([i,j])
  for m in range(len(B)):
    p = B[m][0]
    q = B[m][1]
    if y == q or (x-p)==abs(y-q):
      flag=False
  return flag

def queen(n):
  A=[[0 for __ in range(n)] for _ in range(n)]
  answer=[]
  for _ in range(n):
    stack=[[0,_,A]]
    while stack:
      judge = 0
      obj=stack.pop(-1)
      x=obj[0]
      y=obj[1]
      array=obj[2]
      flag=check(array,x,y)
      if not flag:
        while 1:
          if check(array, x, y):
            break
          else:
            if stack:
              b=stack.pop(-1)
              x=b[0]
              y=b[1]
              array=b[2]
            else:
              judge=1
              break
      if judge==1:
        break
      array=copy.deepcopy(array)
      array[x][y]=1
      for m in range(n):
        if m!=y and m!=y-1 and m!=y+1 and x+1<n :
          stack.append([x+1,m,array])
      # print(array)
      for j in range(len(array[n-1])):
        if array[n-1][j]==1:
          answer.append(array)
  print(len(answer))
queen(8)

What is stored in answer is all the last feasible combinations
The current solution is the 8 queens problem
My idea is to use dfs to bring along each search position x, y and the chessboard to be placed, i.e. [x, y, A]
This does not cause all operations to be performed on one matrix