有向量和成员的类有内存问题吗？

Question

我刚刚开始使用C++，所以如果这是一个愚蠢的问题，很抱歉。我有一个Braid类，它的成员是向量。我还没有写赋值运算符。当我对类型为Braid的对象进行大量赋值时，我遇到了内存问题：

 0  0xb7daff89 in _int_malloc () from /lib/libc.so.6
#1  0xb7db2583 in malloc () from /lib/libc.so.6
#2  0xb7f8ac59 in operator new(unsigned int) () from /usr/lib/libstdc++.so.6
#3  0x0804d05e in __gnu_cxx::new_allocator<int>::allocate (this=0xbf800204, __n=1)
    at /usr/lib/gcc/i686-pc-linux-gnu/4.4.3/../../../../include/c++/4.4.3/ext/new_allocator.h:89
#4  0x0804cb0e in std::_Vector_base<int, std::allocator<int> >::_M_allocate (this=0xbf800204, __n=1)
    at /usr/lib/gcc/i686-pc-linux-gnu/4.4.3/../../../../include/c++/4.4.3/bits/stl_vector.h:140
#5  0x0804c086 in _Vector_base (this=0xbf800204, __n=1, __a=...)
    at /usr/lib/gcc/i686-pc-linux-gnu/4.4.3/../../../../include/c++/4.4.3/bits/stl_vector.h:113
#6  0x0804b4b7 in vector (this=0xbf800204, __x=...)
    at /usr/lib/gcc/i686-pc-linux-gnu/4.4.3/../../../../include/c++/4.4.3/bits/stl_vector.h:242
#7  0x0804b234 in Braid (this=0xbf800204) at braid.h:13
#8  0x080495ed in Braid::cycleBraid (this=0xbf8001b4) at braid.cpp:191
#9  0x080497c6 in Braid::score (this=0xbf800298, b=...) at braid.cpp:251
#10 0x08049c46 in Braid::evaluateMove (this=0xbf800468, move=1, pos=0, depth=2, b=...)

我怀疑这些内存问题是因为向量正在调整大小。我想知道的是，当Braid类型的成员展开时，它的对象是否会自动展开？我正在写的代码真的很长，所以我会发布导致问题的部分。以下是代码的相关部分：

class Braid 
{
private :
  vector<int> braid; //Stores the braid.  
  int strands;
  vector < vector<bool> > history; 
  vector < vector<bool> > CM;    
public :
  Braid () : strands(0) {}
  Braid operator * (Braid);
  Braid* inputBraid(int,vector<int>);
  int printBraid();
  int printBraid(vector<vector<int>::iterator>);
  vector<int>::size_type size() const;
               .....
               .....
}

以下是导致该问题的函数:-

int Braid::evaluateMove(int move,int pos,int depth,Braid b)
{
 int netscore = 0;
 Braid curr(*this);
 curr = curr.move(move,pos);
 netscore += curr.score(b);

 while(depth > 1)
 {
   netscore += curr.evaluateMove(1,0,depth,b);
   netscore += curr.evaluateMove(2,0,depth,b);
   for(int i = 0; i < braid.size();++i)
   {
    netscore += curr.evaluateMove(3,i,depth,b);
    netscore += curr.evaluateMove(4,i,depth,b);
    netscore += curr.evaluateMove(5,i,depth,b);
    curr = curr.cycleBraid();
    netscore += curr.evaluateMove(6,0,depth,b);
   }
   --depth;
 }
 return netscore;
}

Answer 1

另一个问题：

while(depth > 1)
{
  netscore += curr.evaluateMove(1,0,depth,b);
  ....
  --depth;
}

当深度大于1时，导致无休止的递归

Answer 2

快速阅读：Braid curr(*this);导致复制。你不能用指针指向编织带吗？

Answer 3

我是唯一标记堆栈溢出的人吗？

我涂掉了无用的部分(出于演示的目的)

int Braid::evaluateMove(int move,int pos,int depth,Braid b)
{
  #int netscore = 0;
  #Braid curr(*this);
  #curr = curr.move(move,pos);
  #netscore += curr.score(b);

  while(depth > 1)
  {
    netscore += curr.evaluateMove(1,0,depth,b);
    #netscore += curr.evaluateMove(2,0,depth,b);
    #for(int i = 0; i < braid.size();++i)
    #{
    #  netscore += curr.evaluateMove(3,i,depth,b);
    #  netscore += curr.evaluateMove(4,i,depth,b);
    #  netscore += curr.evaluateMove(5,i,depth,b);
    #  curr = curr.cycleBraid();
    #  netscore += curr.evaluateMove(6,0,depth,b);
    #}
    --depth;
  }
  return netscore;
}

现在如果depth优于1..。oups

假设我这样做：

• Braid b; b.evaluateMove(1,0,2,b);
• it调用curr.evaluateMove(1,0,2,b);
• which调用curr.evaluateMove(1,0,2,b);
• which ...

系统可能会耗尽内存。

注意:为什么evaluateMove既要复制this，又要Braid (参数b__)的副本？如果我是你，我会检查我的score方法。