我有HashSet，每个存储T，我编写了一个测试应用程序来比较我能想到的不同关系算法，但是我对我得到的结果并不满意。
是否存在更有效的方法来实现对象关系，那么我正在测试一次？
using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using System.Threading.Tasks;
using System.Diagnostics;
namespace LINQTests
{
class Program
{
static void Main(string[] args)
{
HashSet<User> UserTable = new HashSet<User>();
HashSet<UserProperty> UserPropertyTable = new HashSet<UserProperty>();
#region Lets create some dummy data.
Console.WriteLine("Please wait while creating dummy data, this can take a while...");
Console.WriteLine("");
int rows = 1000000;
for(int x = 0; x < rows; x++)
{
Random rnd = new Random();
// Add new user.
User user = new User(string.Format("Joachim-{0}", x));
if(!UserTable.Add(user))
{
// throw new Exception.
}
else
{
UserProperty age = new UserProperty(user, "Age", rnd.Next(25, 30).ToString());
if(!UserPropertyTable.Add(age))
{
// throw new Exception.
}
UserProperty sex = new UserProperty(user, "Sex", "Male");
if (!UserPropertyTable.Add(sex))
{
// throw new Exception.
}
UserProperty location = new UserProperty(user, "Location", "Norway");
if (!UserPropertyTable.Add(location))
{
// throw new Exception.
}
}
}
#endregion
#region Lets do some query tests.
IEnumerable<User> Users;
Stopwatch stopwatch = new Stopwatch();
int matches = 0;
// Lets find all users who are of age 29.
Console.WriteLine("Finding all users who are of age 29");
Console.WriteLine("");
Console.WriteLine("---------------------------------------------------");
Console.WriteLine("{0,-20} | {1,6} | {2,9}", "Search Strategy", "Found", "Time");
Console.WriteLine("---------------------------------------------------");
// Join test.
stopwatch.Start();
Users = (from user in UserTable
join property in UserPropertyTable on user.Id equals property.UserId
where property.Key == "Age" && property.Value == "29"
select user);
matches = Users.Count();
stopwatch.Stop();
Console.WriteLine("{0,-20} | {1,6} | {2,6} ms.", "Joining Tables", matches, stopwatch.ElapsedMilliseconds);
// Dictionary test.
stopwatch.Restart();
var DictionarySearch = (from t in UserPropertyTable where t.Key == "Age" && t.Value == "29" select t).ToDictionary(x => x.UserId);
Users = (from t in UserTable where DictionarySearch.ContainsKey(t.Id) select t);
matches = Users.Count();
stopwatch.Stop();
Console.WriteLine("{0,-20} | {1,6} | {2,6} ms.", "Dictionary Contain", matches, stopwatch.ElapsedMilliseconds);
// HashSet test.
stopwatch.Restart();
var HashsetSearch = new HashSet<Guid>(from t in UserPropertyTable where t.Key == "Age" && t.Value == "29" select t.UserId);
Users = (from t in UserTable where HashsetSearch.Contains(t.Id) select t);
matches = Users.Count();
stopwatch.Stop();
Console.WriteLine("{0,-20} | {1,6} | {2,6} ms.", "HashSet Contain", matches, stopwatch.ElapsedMilliseconds);
// Following takes so long that we wont run them!
//// Array test.
//stopwatch.Restart();
//var ArrayMatch = (from t in UserPropertyTable where t.Key == "Age" && t.Value == "29" select t.UserId).ToArray();
//Users = (from t in UserTable where ArrayMatch.Contains(t.Id) select t);
//matches = Users.Count();
//stopwatch.Stop();
//Console.WriteLine("{0,-20} | {1,6} | {2,6} ms.", "Array Contain", matches, stopwatch.ElapsedMilliseconds);
//// List test.
//stopwatch.Restart();
//var ListMatch = (from t in UserPropertyTable where t.Key == "Age" && t.Value == "29" select t.UserId).ToList();
//Users = (from t in UserTable where ListMatch.Contains(t.Id) select t);
//matches = Users.Count();
//stopwatch.Stop();
//Console.WriteLine("{0,-20} | {1,6} | {2,6} ms.", "List Contain", matches, stopwatch.ElapsedMilliseconds);
Console.WriteLine("---------------------------------------------------");
#endregion
Console.WriteLine("");
Console.WriteLine("Hit return to exit...");
Console.Read();
}
}
public class User
{
public User(string UserName)
{
this.Id = Guid.NewGuid();
this.UserName = UserName;
}
public Guid Id { get; set; }
public string UserName { get; set; }
public override bool Equals(object obj)
{
User other = obj as User;
if (other == null)
return false;
return this.Id == other.Id;
}
public override int GetHashCode()
{
return Id.GetHashCode();
}
}
public class UserProperty
{
public UserProperty(User user, string key, string value)
{
this.Id = Guid.NewGuid();
this.UserId = user.Id;
this.Key = key;
this.Value = value;
}
public Guid Id { get; private set; }
public Guid UserId {get; private set;}
public string Key { get; set; }
public string Value { get; set; }
public override bool Equals(object obj)
{
UserProperty other = obj as UserProperty;
if (other == null)
return false;
return this.UserId == other.UserId && this.Key == other.Key;
}
public override int GetHashCode()
{
return string.Format("{0}-{1}", this.UserId, this.Key).GetHashCode();
}
}
}
答案 0 :(得分：2)
您可以执行以下操作：
UserId.GHC() * 37 ^ Name.GHC()应该这样做。property.Key == "Age" && property.Value == "29"谓词？是否需要为每次执行运行？您可以尝试使用整数值而不是字符串来优化它。也许您可以预先计算UserPropertyTable的索引，以便您可以在固定时间内获得匹配的元素。根据您的实施程度，这应该为您提供1-10倍的加速。如果你需要比这更快，我需要先问一些问题。通常，您可以找到仅适用于特定情况的专业技巧。
答案 1 :(得分：1)
这将使linq / join与其他方法相当：
var properties = UserPropertyTable
.Where(p=>p.Key == "Age" && p.Value == "29")
.ToArray();
Users = (from user in UserTable
join property in properties
on user.Id equals property.UserId
select user);
这是我能达到的最快（~2X）：
var filteredUserIds = new HashSet<Guid>(
UserPropertyTable
.Where(p=>p.Key == "Age" && p.Value == "29")
.Select(p=>p.UserId));
Users = (from user in UserTable
where filteredUserIds.Contains(user.Id)
select user);
输出
--------------------------------------------------- Search Strategy | Found | Time --------------------------------------------------- My method | 210366 | 157 ms. Dictionary Contain | 210366 | 325 ms. HashSet Contain | 210366 | 325 ms. ---------------------------------------------------