将std :: vector的副本附加到自身的结尾

问题描述：

我试图创建一个字符串向量的副本，并将其附加到其原始向量的结尾，即复制其内容。示例：

I am trying to make a copy of a vector of string and append it to the end of its original vector, i.e. duplicating its contents. Example:

 Input : vector<string> s = {"abc", "def"}
 Output: vector<string> s = {"abc", "def", "abc", "def"}

使用insert方法，即

I was using the insert method, i.e.

s.insert(s.end(), s.begin(), s.end());

但是，这会显示与编译器相关的结果。在，LLVM ang，它给了我预期的答案。

However, this exhibits compiler-dependent results. In, LLVM clang, it gave me the expected answer.

使用GCC它给了我

 Output: vector<string> s = {"abc", "def", "", ""}

这里是上面程序的ideone.com链接： http://ideone.com/40CH8q

Here is the ideone.com link for the program above: http://ideone.com/40CH8q

答

虽然它可以用迭代器完成，一个安全的选择是避免它们：

Although it can possibly be done with iterators, a safe alternative is to avoid them:

size_t size = v.size();  // Of course we shouldn't access .size() in the loop...
v.reserve(size * 2);     // Preallocation. Thanks @Ali for this performance hint
for (size_t i = 0; i < size; ++i)
    v.push_back(v[i]);

一般来说，使用迭代器同时修改数据结构你应该仔细阅读迭代器是无效的，当它是安全的重新使用旧的迭代器修改后。因此，使用旧方法来遍历随机访问序列有时是有意义的：使用索引变量。

In general, working with iterators while also modifying the data structure (not only its elements) is dangerous; you should read carefully when iterators are invalidated and when it's safe to reuse old iterators after a modification. Thus, it sometimes makes sense to use the "old" method to iterate through a random-access sequence: using an index variable.

将std :: vector的副本附加到自身的结尾

相关推荐