我猜我做了一些非常愚蠢的事情，但enqueue中的字符串“elem”不会分配给字符串数组“heap”。这是否与通过引用传递的const有关？
#include "pqueue-heap.h"
using namespace std;
HeapPQueue::HeapPQueue() {
heap = new string[1000];
}
HeapPQueue::~HeapPQueue() {
delete[] heap;
}
string HeapPQueue::peek() const {
// placeholder so method compiles..
// replace with your own implementation
return "";
}
string HeapPQueue::extractMin() {
// placeholder so method compiles..
// replace with your own implementation
return peek();
}
void HeapPQueue::enqueue(const string& elem) {
logSize++;
heap[logSize] = elem;
bubbleUp(logSize);
}
void HeapPQueue::bubbleUp(int i) {
if (i / 2 == 0) return;
string insert = heap[i];
string node = heap[i / 2];
if (insert >= node) return;
heap[i] = node;
heap[i / 2] = insert;
bubbleUp(i / 2);
}
HeapPQueue *HeapPQueue::merge(HeapPQueue *one, HeapPQueue *two) {
// placeholder so method compiles..
// replace with your own implementation
return new HeapPQueue();
}
logSize受到保护，此处在pqueue头文件中提及;
/**
* File: pqueue.h
* --------------
* Defines the interface that all PQueues must implement. VectorPQueue,
* HeapPQueue, and BinomialHeapPQueue all subclass PQueue so that they
* all agree on the same interface.
*/
#ifndef _pqueue_
#define _pqueue_
#include <string>
/**
* Defines the PQueue container type for strings. The priority queue
* holds all enqueued strings, but rather than making them accessible
* in a FIFO manner, extractMin produces the alphabetically smaller strings
* before the alphabetically larger ones, regardless of insertion order.
*
* The generalization of this would be a templated priority queue, useful
* in a large number of algorithmic settings (optimization problems, simulations,
* shortest path problems, network routing, etc.)
*/
class PQueue {
public:
/**
* Defines a small set of constants that can be used to identify
* the particular implementation of interest. They're most vital to
* the construction of a PQueue using the factory method, as with
*
* PQueue *pq = PQueue::createPQueue(PQueue::BinomialHeap);
*
* Enumerated constants are programmatically easier to manage than
* type names, and it's easy to internally dispatch on enumerated
* type constants than it is to on type names.
*/
enum PQueueType {
UnsortedVector, LinkedList, Heap, BinomialHeap
};
/**
* Convenience function that gives us a string name for the
* PQueue represented by type.
*/
static std::string typeToName(PQueueType type);
public:
/**
* Manages the initialization of the material managed at the PQueue,
* base class, level. We assume that all subclasses, regardless of
* implementation, track their logical size using the logSize
* parameter held at the PQueue class level. By doing so, we can
* rely on a single implementation of the isEmpty() and size() methods
* that never have to be overridden. Construction at this level is
* so obvious that we just inline the implementation.
*/
PQueue() { logSize = 0; }
/**
* Disposes of any external resources held at the PQueue level.
* Nothing, save for an internal int, is managed at the PQueue level,
* so the destructor is inlined here.
*/
virtual ~PQueue() {}
static PQueue *createPQueue(PQueueType type);
static PQueue *merge(PQueue *one, PQueue *two);
/**
* Returns the number of elements inside the PQueue. This
* method should not be overridden, and subclasses properly
* maintain the value of logSize so that the implementation
* provided here is always correct.
*/
int size() const { return logSize; }
/**
* Returns true if and only if the PQueue is empty. Self-explanatory,
* save for the fact that isEmpty should not be overridden.
*/
bool isEmpty() const { return size() == 0; }
/**
* Inserts the provided string into the priority queue as the
* implementation sees fit. The virtual keyword, paired with the
* = 0, is C++ notation stating that no sensible implementation
* exists at this level, and that concrete subclasses should
* provide one.
*/
virtual void enqueue(const std::string& elem) = 0;
/**
* Identifies, removes, and returns the lowest-priority element currently
* in the priority queue. The behavior is undefined if called against
* an empty queue. The virtual keyword, paired with the
* = 0, is C++ notation stating that no sensible implementation
* exists at this level, and that concrete subclasses should
* provide one.
*/
virtual std::string extractMin() = 0;
/**
* Returns a copy of the lowest-priority item currently within
* the queue. The virtual keyword, paired with the
* = 0, is C++ notation stating that no sensible implementation
* exists at this level, and that concrete subclasses should
* provide one.
*/
virtual std::string peek() const = 0;
protected:
/**
* Maintains a copy of the size of the priority queue, regardless of the
* subclass's implementation. Care much be taken to ++ and -- this
* field (visible to all suclasses because of the protected access modifier)
* with each call to enqueue and extractMin, and that the logSize is properly
* updated with each merge.
*/
int logSize;
};
#endif
这是heappqueue的头文件;
#ifndef _binary_heap_pqueue_
#define _binary_heap_pqueue_
#include "pqueue.h"
#include <string>
class HeapPQueue : public PQueue {
public:
HeapPQueue();
~HeapPQueue();
static HeapPQueue *merge(HeapPQueue *one, HeapPQueue *two);
void enqueue(const std::string& elem);
std::string extractMin();
std::string peek() const;
private:
std::string *heap;
void bubbleUp(int logSize);
// provide data methods and helper methods to
// help realize the binary heap-backed PQueue
};
#endif
这是班级入队的方式;
template <typename Iterable>
static PQueue *buildPQueue(PQueue::PQueueType pqtype, Iterable iter, int size) {
PQueue *pq = PQueue::createPQueue(pqtype);
int count = 0;
foreach (string key in iter) {
pq->enqueue(key);
count++;
if (count == size) break;
}
cout << "+ Inserted " << pq->size() << " words." << endl;
return pq;
}
以下是类型结构的布局方式;
static const struct {
PQueue::PQueueType type;
int reasonableTestSize;
} testParameters[] = {
{ PQueue::UnsortedVector, 10000},
{ PQueue::LinkedList, 10000},
{ PQueue::Heap, INT_MAX},
{ PQueue::BinomialHeap, INT_MAX}
};
答案 0 :(得分：2)
您显示的构造函数不会初始化 logSize。这可能至少是问题的一部分。
此外，heap的大小固定，enqueue()不会检查是否有溢出。
答案 1 :(得分：0)
你没有使用堆的插槽0 。而是尝试增加logSize。 此外，在访问数组堆之前必须进行某种边界检查。
void HeapPQueue::enqueue(const string& elem) {
heap[logSize] = elem;
bubbleUp(logSize);
logSize++;
}