Breadth First Search in Python (宽度/广度优先搜索)

宽搜/广搜简称BFS，Dijkstra单源最短路径算法和Prim最小生成树算法均使用广搜的思想，从一点出发向外辐射直至遍历所有结点。广搜主要用于解决最短路径、层序遍历和大范围查找问题。本文主要总结在图和树（特殊的图）上的广搜。

示例1:

图的广度优先搜索（图源为AI AX AT）

代码：

import os
import sys

graph = {
	1: {2, 3},
	2: {1, 4, 5},
	3: {1, 6, 7},
	4: {2, 8},
	5: {2, 8},
	6: {3, 8, 9},
	7: {3, 9},
	8: {4, 5, 6},
	9: {6, 7}
}

def bfs(graph, s):
    queue = [s] 
    parent = {s: None}
    flag = set()
    flag.add(s)
    
    while queue:
        temp = queue.pop()
        nodes = graph[temp]    
        for vertex in nodes:
            if vertex not in flag:
                queue.append(vertex)
                flag.add(vertex)
                parent[vertex] = temp
    return parent

parent = bfs(graph, 1)
v = 8
while v:
    print(v, end=' ')
    v = parent[v]

解释：

1）此处手动输入了示例1，若实际图过大或复杂可以设置相应类

2）parent字典用于记录某结点的父结点，方便路径回溯；flag为一个集合，用于标记已经遍历过的结点（该图为无向图）；s为起始结点，v为终止结点。

示例2:

图的广度优先搜索（图源为AI AX AT） 

代码：

class TreeNode:
    def __init__(self, val):
        self.val = val
        self.left = None
        self.right = None

def levelOrder(root):
    queue = [root]
    next_queue = []
    result = []
    if root is None:
    	return []
    
    while queue:
        for root in queue:
            result.append(root.val)
            if root.left:
                next_queue.append(root.left)
            if root.right:
                next_queue.append(root.right)
        queue = next_queue
        next_queue = []
    return result

if __name__ == "__main__":
	tree = TreeNode(4)
	tree.left = TreeNode(9)
	tree.right = TreeNode(0)
	tree.left.left = TreeNode(5)
	tree.left.right = TreeNode(1)
	tree.right.left = TreeNode(3)
	tree.right.right = TreeNode(2)
	tree.left.left.left = TreeNode(10)
	tree.right.left.right = TreeNode(7)
	print(levelOrder(tree))

 解释：

1）树的广度优先搜索本质上即为树的层次遍历

树的层次遍历详见（特别注意输出方式即可）：

LeetCode in Python 102. Binary Tree Level Order Traversal (二叉树的层序遍历)-CSDN博客