Commit 690427c

Assume that the node to be deleted in the linked list is `d`, and the previous node of `d` is `p`, so `p.next` is `d`.

To delete `d`, just set `p.next = p.next.next`.

Because `p.next.next` is used, the loop condition should be `while (p.next != null)` instead of `while (p != null)`.

But there is no node before the `head` node, which means that the `head` node needs to be treated specially.

Is there a way to make the `head` node no longer special? In this way, there is no need to treat the `head` specially.

The way is to introduce a `dummy` node, `dummy.next = head`.

假设链表中待删除的节点是`d`，`d`的前一个节点是`p`，所以`p.next`就是`d`。 删除`d`，只需要把`p.next = p.next.next`。

因为用到了`p.next.next`，所以循环条件应为`while (p.next != null)`，而不是`while (p != null)`。

但`head`节点前面没有节点，这就意味着需要对`head`节点进行特殊处理。

是否有方法能够让`head`节点的不再特殊呢？这样就不需要特殊处理`head`了。

办法是引入`dummy`节点，`dummy.next = head`。