回顾C++(一)-STL中的clear

通常我们可能会写下这样的代码

std::vector<int> v1;
for (int i = 0; i < 100; ++i)
{
	v1.push_back(i);
}

cout<<"v1 before clear"<<endl;
cout<< "v1.size() = "<<v1.size()<<endl;
cout<< "v1.capacity() = "<<v1.capacity()<<endl;

v1.clear();
cout<<"v1 after clear"<<endl;
cout<< "v1.size() = "<<v1.size()<<endl;
cout<< "v1.capacity() = "<<v1.capacity()<<endl;

std::vector<int> v2;
for (int i = 0; i < 100; ++i)
{
	v2.push_back(i);
}

运行结果

v1 before clear--------------
v1.size() = 100
v1.capacity() = 128
v1 after clear--------------
v1.size() = 0
v1.capacity() = 128

clear并没有释放掉内存，而仅仅是将size置为0，如果需要立即释放掉内存，可以用一个空的容器和其交换，修改后的代码如下：

cout<<"v2 before swap"<<endl;
cout<< "v2.size() = "<<v2.size()<<endl;
cout<< "v2.capacity() = "<<v2.capacity()<<endl;

vector<int>().swap(v2);
cout<<"v2 after swap--------------"<<endl;
cout<< "v2.size() = "<<v2.size()<<endl;
cout<< "v2.capacity() = "<<v2.capacity()<<endl;

运行结果

v2 before swap--------------
v2.size() = 100
v2.capacity() = 128
v2 after swap--------------
v2.size() = 0
v2.capacity() = 0

c++11中增加了新的方法shrink_to_fit，可以释放掉多余的内存

std::vector<int> v3;
for (int i = 0; i < 100; ++i)
{
	v3.push_back(i);
}

cout<<"v3 before shrink_to_fit--------------"<<endl;
cout<< "v3.size() = "<<v3.size()<<endl;
cout<< "v3.capacity() = "<<v3.capacity()<<endl;

v3.shrink_to_fit();
cout<<"v3 after shrink_to_fit--------------"<<endl;
cout<< "v3.size() = "<<v3.size()<<endl;
cout<< "v3.capacity() = "<<v3.capacity()<<endl;

运行结果

v3 before shrink_to_fit--------------
v3.size() = 100
v3.capacity() = 128
v3 after shrink_to_fit--------------
v3.size() = 100
v3.capacity() = 100

Question?
不理解这个方法的作用，设置capacity的本意是预留内存，下次插入元素时不用涉及耗时的内存分配，这样下次插入元素肯定会触发重新分配内存。