什么使STL快速？ [英] What makes STL fast?

问题描述

如果一个实现数组类的方式是通常实现的，它的性能比它的STL等效，比如一个矢量。那么什么使STL容器/算法快速？

If one implements an array class the way it is commonly implemented, its performance is slower compared to its STL equivalent like a vector. So what makes STL containers/algorithms fast?

推荐答案

STL算法如 for_each 获取可以轻松内联的函数对象。另一方面，C使用对编译器更难以优化的函数指针。

STL algorithms like for_each take function objects that can be easily inlined. C, on the other hand, uses function pointers which are much more difficult for the compiler to optimize.

这使得一些算法，如排序，比较器函数必须被调用许多次。

This makes a big difference in some algorithms like sorting in which the comparer function must be called many times.

维基百科有一些更多信息，以备您感兴趣。

Wikipedia has some more information in case you are interested.

编辑：

至于STL的向量类，认为它必须更快，你可以找到，比如说，glibc。

As for the STL's vector class, I don't think it's necessarily faster that what you could find in, say, glibc.

这篇关于什么使STL快速？的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！