如何使用 Python 堆实现自然语言处理算法？

from queue import PriorityQueue

# 创建一个字典，用于存储图
graph = {}
graph['pidancode.com'] = {}
graph['pidancode.com']['皮蛋编程'] = 1
graph['pidancode.com']['百度'] = 4
graph['皮蛋编程'] = {}
graph['皮蛋编程']['腾讯'] = 3
graph['皮蛋编程']['京东'] = 4
graph['百度'] = {}
graph['百度']['京东'] = 2
graph['百度']['腾讯'] = 5
graph['京东'] = {}
graph['腾讯'] = {}

# 创建一个字典，用于存储每个节点的开销
costs = {}
costs['皮蛋编程'] = 0
costs['百度'] = float('inf')
costs['腾讯'] = float('inf')
costs['京东'] = float('inf')

# 创建一个字典，用于存储每个节点的父节点
parents = {}
parents['皮蛋编程'] = None
parents['百度'] = None
parents['腾讯'] = None
parents['京东'] = None

# 创建一个优先队列，用于存储节点和它们的开销
queue = PriorityQueue()
queue.put(('皮蛋编程', 0))

while not queue.empty():
    # 获取开销最小的节点和它的开销
    node, cost = queue.get()

    # 遍历与节点相邻的所有节点
    for neighbor in graph[node].keys():
        new_cost = cost + graph[node][neighbor]

        # 如果新的开销比之前的开销小，就更新开销和父节点
        if new_cost < costs[neighbor]:
            costs[neighbor] = new_cost
            parents[neighbor] = node
            queue.put((neighbor, new_cost))

# 打印每个节点及其开销和父节点
print("节点：开销，父节点")
for node in costs.keys():
    print(node + "：" + str(costs[node]) + "，" + str(parents[node]))