拓扑排序不符合预期

这是我在Python中的图形实现 . 这是一个有向图 .

class DiGraph:
    def __init__(self):
        self.all_vertices = []
        self.vertex_map = {}
        self.size = 0

    def add(self, a, b):

        if a in self.vertex_map:
            av = self.vertex_map[a]
            if b in self.vertex_map:
                bv = self.vertex_map[b]
                av.add(bv)
            else:
                bv = Vertex(b)
                av.add(bv)
                self.vertex_map[b] = bv
                self.all_vertices.append(bv)
                self.size += 1
        else:
            av = Vertex(a)
            self.size += 1
            if b in self.vertex_map:
                bv = self.vertex_map[b]
                av.add(bv)
            else:
                bv = Vertex(b)
                av.add(bv)
                self.vertex_map[b] = bv
                self.all_vertices.append(bv)
                self.size += 1
            self.vertex_map[a] = av
            self.all_vertices.append(av)

    def __sizeof__(self):
        return self.size

    def print(self):
        for v in self.all_vertices:
            print(v.data, end='->')
            for n in v.neighbors:
                print(n.data, end=', ')
            print()


class Vertex:
    def __init__(self, data):
        self.data = data
        self.neighbors = []
        self.connections = 0

    def add(self, item):
        self.neighbors.append(item)
        self.connections += 1

这是拓扑排序的代码

def top_sort(g):
    stack = []
    visited = set()

    for v in g.all_vertices:
        if v not in visited:
            top_sort_util(v, visited, stack)

    for ele in stack:
        print(ele, end=' ')
    print()


def top_sort_util(v, visited, stack):
    visited.add(v)
    for n in v.neighbors:
        if n in visited:
            continue
        top_sort_util(n, visited, stack)
    stack.append(n)

这是来电图 .

def main():
    graph = DiGraph()
    graph.add(1, 2)
    graph.add(1, 3)
    graph.add(3, 4)
    graph.add(3, 5)
    graph.add(2, 6)
    graph.add(2, 7)
    graph.add(2, 8)

    top_sort(graph)


if __name__ == '__main__':
    main()

这是我得到的错误消息，

stack.append(n)
UnboundLocalError: local variable 'n' referenced before assignment

在调试代码时，我可以看到在行stack.append（n）上没有任何内容被附加，虽然递归折叠，但调用堆栈不会进入遍历top_sort_util内的邻居的下一次迭代 . 似乎无法理解这里的逻辑错误 . 任何帮助赞赏 .