class doublelinked():
    def __init__(self, *args):
        self.__head = head = []
        self.__head += [None, head, head]
        self.__map = {}
        for i in args:
            self.append(i)

    # 插入到表尾
    def append(self, data):
        head = self.__head
        last = head[1]
        node = [data, last, head]
        head[1] = last[2] = self.__map[data] = node
        return node

    def prepend(self, data):
        head = self.__head
        last = head[2]
        node = [data, head, last]
        head[2] = last[1] = self.__map[data] = node
        return node

    def insertafter(self, node, data):
        next = node[2]
        n = [data, node, next]
        node[2] = next[1] = self.__map[data] = n
        return n

    def remove(self, node):
        data, prev, next = node
        prev[2] = next
        next[1] = prev
        node[1] = node[2] = node
        del self.__map[data]

    def removebydata(self, data):
        self.remove(self.__map[data])
    
    def __iter__(self):
        head = self.__head
        curr = head[2]
        while curr is not head:
            yield curr
            curr = curr[2]

    def __getitem__(self, data):
        return self.__map[data]

    def __reversed__(self):
        head = self.__head
        curr = head[1]
        while curr is not head:
            yield curr
            curr = curr[1]

    def __str__(self):
        return ", ".join([str(i[0]) for i in self])


dl = doublelinked('a','b')
dl.append('c')
dl.append('d')
dl.append('e')
dl.append('f')
print(dl)
print ("------------------")
dl.removebydata('b')
print(dl)

         采用一个固定有三个元素的list对象作为表结点机构，list对象的第一个元素是value，第二个元素是指向前一个表结点的引用（指针），第三个元素是指向后一个表结点的引用（指针），链表的头结点不存放数据，它作为一个哨兵，作为判断链表是否为空、链表遍历结束的标志。