Entity Framework 使用注意：Where查詢條件中用到的關聯實體不需要Include

來自博客園開發團隊開發前線最新消息：

在Entity Framework中，如果實體A關聯了實體B，你想在加載實體A的同時加載實體B。通常做法是在LINQ查詢中使用Include()。但是，如果你在查詢條件中用到了實體B，EF會自動加載實體B，這時Include不僅是多余的，而且還會增加額外的LEFT OUTER JOIN查詢，影響性能。 

請看我們在博問開發中遭遇這個問題時的一段代碼：

//For q.cnblogs.com
public class QuestionService 
{
    private IRepository<QuestionItem> _questionRepository;

    public QuestionService(IUnitOfWork unitOfWork)
        : base(unitOfWork)
    {
        _questionRepository = new Repository<QuestionItem>(unitOfWork);
    }

    public List<QuestionItem> GetUnsolvedQuestions(int pageIndex, int pageSize)
    {
        return _questionRepository.Entities
            .Include(q => q.User)
            .Where(q => q.IsActive && q.User.IsActive)
            .Skip((pageIndex - 1) * pageSize)
            .Take(pageSize)
            .ToList();
            
    }
}

public class QuestionItem
{
    public int Id { get;set; }
    public string Title { get; set; }
    public bool IsActive { get; set; }
    public int UserId { get; set; }

    public User User { get; set; }
}

public class User
{
    public int UserId { get; set; }
    public bool IsActive {get;set;}
}

在上面的代碼中，我們想在GetActiveQuestions()返回List<QuestionItem>時，QuestionItem中要包含User信息，所以在LINQ查詢使用了.Include(q => q.User)。

（特別要注意的是：這里把q.User.IsActive作為查詢條件之一）

然后我們用SQL Server Profiler發現，Entity Framework生成了如下的SQL語句：

SELECT TOP (25) 
[Filter1].[Id] AS [Id], 
[Filter1].[Title] AS [Title],
[Filter1].[UserId] AS [UserId], 
[Filter1].[UserId1] AS [UserId1], 
[Filter1].[IsActive1] AS [IsActive], 
[Filter1].[UserId2] AS [UserId1],
[Filter1].[IsActive2] AS [IsActive1]
FROM ( SELECT [Extent1].[Id] AS [Id], [Extent1].[Title] AS [Title], 
[Extent1].[UserId] AS [UserId1],[Extent1].[IsActive] AS [IsActive1],
[Extent3].[UserID] AS [UserID2], [Extent3].[IsActive] AS [IsActive2], 
row_number() OVER (ORDER BY [Extent1].[QID] DESC) AS [row_number]
    FROM   [dbo].[question_Item] AS [Extent1]
    INNER JOIN [dbo].[Users] AS [Extent2] ON [Extent1].[UserID] = [Extent2].[UserID]
    LEFT OUTER JOIN [dbo].[Users] AS [Extent3] ON [Extent1].[UserID] = [Extent3].[UserID]
    WHERE ([Extent1].[IsActive] = 1) AND ([Extent2].[IsActive] = 1) 
)  AS [Filter1]
WHERE [Filter1].[row_number] > 0
ORDER BY [Filter1].[Id] DESC

[dbo].[Users]表對應的就是User實體類，上面的SQL中[dbo].[Users]出現了兩次JOIN，由於[Users]表數據量比較大，兩次JOIN影響了執行計划，查詢耗時增加。這顯然是要避免的。

在與這個問題一陣激戰之后，我們終於找到解決方法 —— 去掉Include，就這么簡單！

從這個地方看，Entity Framework還是挺聰明的，但是由於不知道它的這個聰明之處，反而帶來了問題。

所以，代碼如人，要和她相處好，就要了解她的一切！