"""
# @Time : 2020/10/6
# @Author : Jimou Chen
"""
class Point:
def __init__(self, x, y, step):
self.x = x
self.y = y
self.step = step
queue = []
dx = [0, 1, 0, -1]
dy = [1, 0, -1, 0]
graph = [[0 for i in range(100)] for j in range(100)]
flags = [[0 for i in range(100)] for j in range(100)]
flag = 0
if __name__ == '__main__':
m, n = map(int, input().split())
for i in range(m):
l = list(map(int, input().split()))
for j in range(n):
graph[i][j] = l[j]
start_x, start_y = map(int, input().split())
end_x, end_y = map(int, input().split())
start_point = Point(start_x, start_y, 0)
queue.append(start_point)
flags[start_x][start_y] = 1
while len(queue) != 0:
x = queue[0].x
y = queue[0].y
if x == end_x and y == end_y:
flag = 1
print('step = ', queue[0].step)
break
for k in range(0, 4):
tx = x + dx[k]
ty = y + dy[k]
if graph[tx][ty] == 1 and flags[tx][ty] == 0:
# 入队
temp = Point(tx, ty, queue[0].step + 1)
queue.append(temp)
flags[tx][ty] = 1
queue.pop(0)
if flag == 0:
print('no ans')
- 相比之下，得到的结果一定是最小值
- 不用回溯，在这里的速度比dfs快
