从另一个函数中释放指针
给定代码:
#include<iostream>
using namespace std;
class String
{
char *pstr;
unsigned size;
public:
String(){ pstr=0;size=0;}
String(const char *);
void show(){ cout << pstr << endl ; }
~String () { cout << "In Dtor" << endl; delete [] pstr; }
};
String::String(const char * cptr)
{
size = strlen (cptr) + 1;
cout << "String is - " << cptr << " - of size " << size - 1 << endl ;
pstr = new char [ size ] ;
for ( int i = 0 ; i < size ; i++)
pstr[ i ] = cptr [ i ];
}
int main()
{
String s("Hello World");
s.show();
s.~String();
}
输出:
String is - Hello World - of size 11
Hello World
In Dtor
----Debug Assertion Failure----
In Dtor
为什么析构函数会再次被调用?当我调用析构函数时?
什么是断言失败?
另外这个代码有效吗?
char * ptr=0;
void fun()
{
const char * p = "Hello World";
int size = strlen(p )+ 1;
cout << size << endl;
ptr = (char *)malloc(size);
for ( int i = 0 ; i < size ; i++)
ptr[ i ] = p [ i ];
cout << p << endl << ptr << endl ;
}
int main()
{
fun();
free (ptr); --> Note
}
指针可以从另一个函数中释放吗?这是我在这里试图理解的主要内容。
Given the code:
#include<iostream>
using namespace std;
class String
{
char *pstr;
unsigned size;
public:
String(){ pstr=0;size=0;}
String(const char *);
void show(){ cout << pstr << endl ; }
~String () { cout << "In Dtor" << endl; delete [] pstr; }
};
String::String(const char * cptr)
{
size = strlen (cptr) + 1;
cout << "String is - " << cptr << " - of size " << size - 1 << endl ;
pstr = new char [ size ] ;
for ( int i = 0 ; i < size ; i++)
pstr[ i ] = cptr [ i ];
}
int main()
{
String s("Hello World");
s.show();
s.~String();
}
Output:
String is - Hello World - of size 11
Hello World
In Dtor
----Debug Assertion Failure----
In Dtor
Why does the Destructor get called again? When i have invoked the destructor?
And what is an Assertion Failure?
Also is this code Valid?
char * ptr=0;
void fun()
{
const char * p = "Hello World";
int size = strlen(p )+ 1;
cout << size << endl;
ptr = (char *)malloc(size);
for ( int i = 0 ; i < size ; i++)
ptr[ i ] = p [ i ];
cout << p << endl << ptr << endl ;
}
int main()
{
fun();
free (ptr); --> Note
}
Can the pointer be freed from another function? This is the main thing i am trying to understand here.
如果你对这篇内容有疑问,欢迎到本站社区发帖提问 参与讨论,获取更多帮助,或者扫码二维码加入 Web 技术交流群。
绑定邮箱获取回复消息
由于您还没有绑定你的真实邮箱,如果其他用户或者作者回复了您的评论,将不能在第一时间通知您!
发布评论
评论(2)
您不应该手动调用析构函数 - 当
s
在最后一个 '}' 超出范围时调用它。断言失败意味着名为
assert(somecondition)
和 somecondition 的东西被调用。错误的。这是一种用于验证您的假设的技术 - 如果您的代码依赖于某些特定条件为真,并且该条件确实应该为真,除非您有错误,那么您插入一个断言。然后,您可以选择在启用断言的情况下进行编译 - 这意味着如果您的假设错误,您将收到这样的错误。对于发布版本,您经常禁用断言 - 不会为断言语句生成任何代码,并且没有额外的运行时成本。
在某些情况下,手动调用析构函数是正确的 - 在您了解并使用“placement new”之前,您将不需要它。
You should not invoke the destructor manually - it's invoked when
s
goes out of scope at the final '}'An assertion failure means that something called
assert(somecondition)
and somecondition was false. It's a technique used to validate your assumptions - if your code depends on some specific condition being true, and that condition really should be true unless you have a bug, then you insert an assert.You can then choose to compile with assertions enabled - this means that you'll get such an error if your assumption was wrong. For a release build, you often disable assertions - no code is generated for the assert statement, and there's no extra runtime cost.
There are cases when it's correct to manually invoke the destructor - You won't need this until you learn about and use "placement new".
除了埃里克已经说过的之外:
在删除手动析构函数调用后,您的代码仍然容易出现双重删除:您必须禁用复制构造函数/赋值运算符,或者正确实现它们(如果您坚持,您将需要引用计数)拥有堆分配的内存)。
In addition to what Erik has already said:
Your code would still remain prone to double deletion after you remove manual destructor call: you have to either disable copy constructor/assignment operator, or implement them correctly (you'll need reference counting if you insist on owning heap-allocated memory).