1.广度优先遍历
from collections import deque
from collections import namedtuple
def bfs(start_node, end_node, graph): # 开始节点 目标节点 图字典
node = namedtuple('node', 'name, from_node') # 使用namedtuple定义节点，用于存储前置节点
search_queue = deque() # 使用双端队列，这里当作队列使用，根据先进先出获取下一个遍历的节点
name_search = deque() # 存储队列中已有的节点名称
visited = {} # 存储已经访问过的节点
search_queue.append(node(start_node, None)) # 填入初始节点，从队列后面加入
name_search.append(start_node) # 填入初始节点名称
path = [] # 用户回溯路径
path_len = 0 # 路径长度
print('开始搜索...')
while search_queue: # 只要搜索队列中有数据就一直遍历下去
print('待遍历节点: ', name_search)
current_node = search_queue.popleft() # 从队列前边获取节点，即先进先出，这是BFS的核心
name_search.popleft() # 将名称也相应弹出
if current_node.name not in visited: # 当前节点是否被访问过
print('当前节点: ', current_node.name, end=' | ')
if current_node.name == end_node: # 退出条件，找到了目标节点，接下来执行路径回溯和长度计算
pre_node = current_node # 路径回溯的关键在于每个节点中存储的前置节点
while True: # 开启循环直到找到开始节点
if pre_node.name == start_node: # 退出条件：前置节点为开始节点
path.append(start_node) # 退出前将开始节点也加入路径，保证路径的完整性
break
else:
path.append(pre_node.name) # 不断将前置节点名称加入路径
pre_node = visited[pre_node.from_node] # 取出前置节点的前置节点，依次类推
path_len = len(path) - 1 # 获得完整路径后，长度即为节点个数-1
break
else:
visited[current_node.name] = current_node # 如果没有找到目标节点，将节点设为已访问，并将相邻节点加入搜索队列，继续找下去
for node_name in graph[current_node.name]: # 遍历相邻节点，判断相邻节点是否已经在搜索队列
if node_name not in name_search: # 如果相邻节点不在搜索队列则进行添加
search_queue.append(node(node_name, current_node.name))
name_search.append(node_name)
print('搜索完毕,最短路径为:', path[::-1], "长度为:", path_len) # 打印搜索结果
if __name__ == "__main__":
graph = dict() # 使用字典表示有向图
graph[1] = [3, 2]
graph[2] = [5]
graph[3] = [4, 7]
graph[4] = [6]
graph[5] = [6]
graph[6] = [8]
graph[7] = [8]
graph[8] = []
bfs(1, 8, graph) # 执行搜索
2.深度优先遍历
from collections import deque
from collections import namedtuple
def bfs(start_node, end_node, graph):
node = namedtuple('node', 'name, from_node')
search_stack = deque() # 这里当作栈使用
name_search = deque()
visited = {}
search_stack.append(node(start_node, None))
name_search.append(start_node)
path = []
path_len = 0
print('开始搜索...')
while search_stack:
print('待遍历节点: ', name_search)
current_node = search_stack.pop() # 使用栈模式，即后进先出，这是DFS的核心
name_search.pop()
if current_node.name not in visited:
print('当前节点: ', current_node.name, end=' | ')
if current_node.name == end_node:
pre_node = current_node
while True:
if pre_node.name == start_node:
path.append(start_node)
break
else:
path.append(pre_node.name)
pre_node = visited[pre_node.from_node]
path_len = len(path) - 1
break
else:
visited[current_node.name] = current_node
for node_name in graph[current_node.name]:
if node_name not in name_search:
search_stack.append(node(node_name, current_node.name))
name_search.append(node_name)
print('搜索完毕,路径为:', path[::-1], "长度为:", path_len) # 这里不再是最短路径，深度优先搜索无法一次给出最短路径
if __name__ == "__main__":
graph = dict()
graph[1] = [3, 2]
graph[2] = [5]
graph[3] = [4, 7]
graph[4] = [6]
graph[5] = [6]
graph[6] = [8]
graph[7] = [8]
graph[8] = []
bfs(1, 8, graph)
