本文共 3467 字,大约阅读时间需要 11 分钟。
队列是一种特殊的线性表,仅能在线性表的两端进行操作。
template < typename T >class Queue{public:virtual void enqueue(const T& e) = 0;virtual void dequeue() = 0;virtual T front() const = 0;virtual void clear() = 0;virtual int length() const = 0;};
顺序队列的实现:
template < typename T, int N >class StaticQueue : public Queue{protected:T m_space[N];int m_front;int m_rear;int m_length;public:StaticQueue()void enqueue(const T& e)void dequeue()T front() constvoid clear()int length() constint capacity() const};
注意事项:
StaticQueue实现要点:(循环计数法) 提高队列操作的效率(本质上时循环队列)关键操作:队满:(m_length == N) && (m_front == m_rear)
StaticQueue最终实现:template < typename T, int N >class StaticQueue : public Queue{protected:T m_space[N];int m_front;int m_rear;int m_length;public:StaticQueue() //O(1){ m_front = 0; m_rear = 0; m_length = 0;}void enqueue(const T& e) //O(1){ if(m_length < N) { m_space[m_rear] = e; m_rear = (m_rear + 1) % N; m_length++; } else { THROW_EXCEPTION(InvalidOperationException, "no space in current staticqueue..."); }}void dequeue() //O(1){ if(m_length > 0) { m_front = (m_front + 1) % N; m_length--; } else { THROW_EXCEPTION(InvalidOperationException, "no element in current staticqueue..."); }}T front() const //O(1){ if(m_length > 0) { return m_space[m_front]; } else { THROW_EXCEPTION(InvalidOperationException, "no element in current staticqueue..."); }}void clear() //O(1){ m_front = 0; m_rear = 0; m_length = 0;}int length() const //O(1){ return m_length;}int capacity() const //O(1){ return N;}bool is_empty() //O(1){ return (m_length == 0) && (m_front == m_rear);}bool is_full() //O(1){ return (m_length == N) && (m_front == m_rear);}};
顺序队列的缺陷:当数据为类类型时,StaticQueue的对象在创建时,会多次调用元素类型的构造函数,影响效率。所以我们采用链式存储结构来实现队列。
templateclass LinkQueue : public Queue {protected:LinkList m_list;public:LinkQueue(){}void enqueue(const T& e) //O(n)void dequeue() //O(1)T front() const //O(1)void clear() //O(n)int length() const //O(1)};
LinkQueue最终实现:
templateclass LinkQueue : public Queue {protected:LinkList m_list;public:LinkQueue(){}void enqueue(const T& e) //O(n){ m_list.insert(e);}void dequeue() //O(1){ if(m_list.length() > 0) { m_list.remove(0); } else { THROW_EXCEPTION(InvalidOperationException, "no elemet in current LinkQueue..."); }}T front() const //O(1){ if(m_list.length() > 0) { return m_list.get(0); } else { THROW_EXCEPTION(InvalidOperationException, "no elemet in current LinkQueue..."); }}void clear() //O(n){ while (m_list.length() > 0) { m_list.remove(0); }}int length() const //O(1){ return m_list.length();}};
LinkQueue中,入队操作由于只能操作队列的尾部(链表的最后位置),要进行遍历操作,所以时间复杂度为O(n),可以使用双向循环链表代替单链表来解决这个问题。
转载于:https://blog.51cto.com/11134889/2131758