foreach字典<>值或foreach字典<> [英] foreach Dictionary<>.Values or foreach Dictionary<>

问题描述

我想了解这两种在C#中迭代Dictionary集合的样式的详细信息:

I want to know the details of this two styles of iterating Dictionary collections in C#:

Dictionary<X, Y> xydic = new Dictionary<X, Y>();

样式一:

foreach (Y y in xydic.Values) { use y }

样式二:

foreach (var it in xydic) { Y y = it.Value; use y... }

多年来我一直是C ++开发人员(现在我在C#项目中工作)，我不知道Dictionary集合如何工作，内存布局或元素如何迭代的详细信息，所以我想知道:

I've been C++ developer for years (now I'm working in a C# project) and I don't know the details of how the Dictionary collection works, the memory layout or how the elements are iterated so I'm wondering:

xydic.Values创建一个临时List<Y>吗?我没有在文档中看到有关创建临时列表的任何信息.

xydic.Values creates a temporary List<Y>? I don't see in the documentation any information about the creation of a temporary list.

如果创建了一个临时列表，这是否意味着对该集合进行了两次迭代:首先创建List<Y>，其次创建列表本身?

If a temporary list is created wouldn't this mean that the collection is iterated twice: first to create the List<Y> and second to iterate the list itself?

如果以上问题的答案是肯定的，则第二种样式应该更有效，从而使第一种样式几乎没有用，所以我认为我在某种程度上应该是错误的.

If the the answer to question above is yes, the second style should be more efficient making the first style almost useless so I think that I should be wrong in some way.

我觉得这个问题应该在某个地方回答，但我找不到答案.

I have the feeling that this question should be answered somewhere but I'm unable to locate an answer.

推荐答案

检索Dictionary<,>的.Values属性是一项O(1)操作(

Retrieving the .Values property of a Dictionary<,> is an O(1) operation (documented). The nested type Dictionary<,>.ValueCollection is a simple wrapper around the dictionary, so there is no iterating in creating it.

在调用GetEnumerator()时，您将获得嵌套的，嵌套的Dictionary<,>.ValueCollection.Enumerator结构的实例.它直接通过Dictionary<,>的private数组entries访问条目.

On calling GetEnumerator(), you get an instance of the nested, nested Dictionary<,>.ValueCollection.Enumerator struct. It acccesses the entries directly through the private array entries of the Dictionary<,>.

您可以查看源代码.

因此，您上面的样式一"是一种良好且清晰的处理方式，而没有性能开销.

So your "Style one" above is a good and clear way of doing things, with no performance overhead.

请注意，获取值的顺序是任意的.您不知道基础数组entries的组织方式，一旦Dictionary<,>具有许多插入和删除操作，然后再开始进行foreach

Note that the order in which you get the values, is arbitrary. You do not know how the underlying array entries is organized, once the Dictionary<,> has had many insertions and removals before you start foreaching it.

但是，样式一"和样式二"的顺序是相同的；两者都以相同的方式访问Dictionary<,>的私有entries数组.

However, the order you get with "Style one" and "Style two" is the same; both access the private entries array of the Dictionary<,> in the same way.

这篇关于foreach字典<>值或foreach字典<>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！