简体繁体 English

寻找.Net排序的集合，可以访问上一个和下一个元素

[英]Looking for .Net sorted collection with access to previous and next elements

原文 2015-04-20 11:16:06 0 1 c#/ .net/ sorting/ data-structures/ sortedcollection

I'm implementing the Bentley-Ottman algorithm which requires for the sweep line (SL) a data structure with the following properties: 我正在实现Bentley-Ottman算法，该算法要求扫描线（SL）具有以下属性的数据结构：

maintain a sorted collection of T , where T is IComparable<T> , 维护T的排序集合，其中T是IComparable<T> ，
insertion of elements should be O(log count) , and should return whether the element is already inserted, 元素的插入应为O(log count) ，并应返回元素是否已插入，
deletion of elements should be O(log count) , 删除元素应为O(log count) ，
for a given element e (whether already in the collection or not) I need the previous and next element of the collection next to e in the sort order. 对于给定的元素e （无论是否已经在集合中），我需要按排序顺序在e旁边的集合的上一个和下一个元素。

SortedList<TKey, TValue> has an O(count) at insertion and deletion, since it has to move all consecutive elements in the list. SortedList<TKey, TValue>在插入和删除时具有O(count) ，因为它必须移动列表中的所有连续元素。 However, I could index it, and therefore get the previous and next elements in O(1) , once I know the index of e . 但是，我可以索引它，因此一旦知道e的索引，就可以获取O(1)的上一个和下一个元素。

SortedDictionary<TKey, TValue> and SortedSet<T> have O(log count) insertion and deletion, but I cannot find any iterator that gives me the next and previous elements. SortedDictionary<TKey, TValue>和SortedSet<T>具有O(log count)插入和删除，但是我找不到任何迭代器可以为我提供下一个和上一个元素。

Is there any implementation that gives me the full functionality? 有没有可以给我全部功能的实现方式？

And if not, what would be the fastest way to implement it? 如果不是，最快的实现方式是什么？ LinkedList<T> does not allow a binary search. LinkedList<T>不允许二进制搜索。 List<T> still has the O(count) insertion/deletion. List<T>仍然具有O(count)插入/删除。 Do I really have to implement my own balanced tree? 我真的必须实现自己的平衡树吗？

1 个解决方案

For example the TreeDictionary of the c5 collection library and nuget and github has a 例如， TreeDictionary中的C5收集库和的NuGet和github上有一个

bool TryPredecessor(K k, out KeyValuePair res) returns true if there is a precedessor of k and in that case binds the predecessor to res; 如果存在k的前置变量，则bool TryPredecessor（K k，out of KeyValuePair res）返回true，在这种情况下，将前任绑定到res； otherwise returns false and binds the default value of KeyValuePair to res. 否则返回false并将KeyValuePair的默认值绑定到res。 The predecessor of k is the entry in the sorted dictionary with the greatest key strictly less than k according to the key comparer. 根据密钥比较器，k的前身是排序字典中的最大密钥严格小于k的条目。 Throws NoSuchItemException if k does not have a predecessor entry; 如果k没有前置项，则抛出NoSuchItemException；否则，抛出NoSuchItemException。 that is, no key is less than k. 即，任何密钥都不小于k。

and a 和一个

bool TrySuccessor(K k, out KeyValuePair res) returns true if there is a successor of k and in that case binds the successor to res; 如果存在k的后继，则bool TrySuccessor（K k，out of KeyValuePair res）返回true，在这种情况下，将后继绑定到res； otherwise returns false and binds the default value of KeyValuePair to res. 否则返回false并将KeyValuePair的默认值绑定到res。 The successor of k is the entry in the sorted dictionary with the least key strictly greater than k according to the key comparer. k的后继项是排序字典中的条目，根据密钥比较器，其最小密钥严格大于k。 Throws NoSuchItemException if k does not have a successor; 如果k没有后继，则抛出NoSuchItemException；否则，抛出NoSuchItemException。 that is, no entry in the dictionary has a key that is greater than k. 也就是说，字典中的任何条目都不具有大于k的键。