Question

我有SQL 2008服务器，其中包含员工，电话和电子邮件，其中员工与电话和电子邮件处于一对多关系（因此员工可以拥有多个电话号码和电子邮件）。

现在我想用所有手机和电子邮件查询员工：

var query = (from e in db.Employee
             select new EmployeeDTO
             {
                 Id = e.Id,
                 Name = e.Name,
                 Phones = e.Phones,
                 Emails = e.Emails
             }).ToList();

但是linq将它转换为非常丑陋的sql，它使得左边的联接与电话表，但然后它在单独的查询中查询每个员工的电子邮件，如：

SELECT [t0].[EmployeeID], [t0].[Name], [t1].[Phone] (
    SELECT COUNT(*)
    FROM [dbo].[Phone] AS [t2]
    WHERE [t2].[EmployeeID] = [t0].[EmployeeID]
    ) AS [value]
FROM [dbo].[Employee] AS [t0]
LEFT OUTER JOIN [dbo].[Phone] AS [t1] ON [t1].[EmployeeID] = [t0].[EmployeeID]
ORDER BY [t0].[EmployeeID], [t1].[PhoneID]
-- Context: SqlProvider(Sql2008) Model: AttributedMetaModel Build: 3.5.30729.4926

SELECT [t0].[EmailID], [t0].[Email]
FROM [dbo].[Email] AS [t0]
WHERE [t0].[EmployeeID] = @x1
-- @x1: Input UniqueIdentifier (Size = 0; Prec = 0; Scale = 0) [1ed2e9a0-ca4b-4912-8f8e-000ed8d0b2af]
-- Context: SqlProvider(Sql2008) Model: AttributedMetaModel Build: 3.5.30729.4926

SELECT [t0].[EmailID], [t0].[Email]
FROM [dbo].[Email] AS [t0]
WHERE [t0].[EmployeeID]= @x1
-- @x1: Input UniqueIdentifier (Size = 0; Prec = 0; Scale = 0) [2203192b-05ef-4b59-b4b4-0010e26a8119]
-- Context: SqlProvider(Sql2008) Model: AttributedMetaModel Build: 3.5.30729.4926

我需要的是简单的sql：

SELECT [t0].[EmployeeID], [t0].[Name], [t1].[Phone], [t2].[Email]
FROM [dbo].[Employee] AS [t0]
LEFT OUTER JOIN [dbo].[Phone] AS [t1] ON [t1].[EmployeeID] = [t0].[EmployeeID]
LEFT OUTER JOIN [dbo].[Email] AS [t2] ON [t2].[EmployeeID] = [t0].[EmployeeID]

如何在不对结果进行分组的情况下编写此类linq查询？我知道如何进行多次左连接，但后来我需要按EmployeeID对结果进行分组。

Answer 1

你试过这样做吗？

var query = (from e in db.Employee
    join p in db.Phones on e.EmployeeID equals p.EmployeeID into tmpPhones
    from phones in tmpPhones.DefaultIfEmpty()
    join m in db.Emails on e.EmployeeID equals m.EmployeeID into tmpEmails
    from emails in tmpEmails.DefaultIfEmpty()

             select new EmployeeDTO
             {
                 Id = e.Id,
                 Name = e.Name,
                 Phones = phones,
                 Emails = emails
             }).ToList();

你需要的是LoadWith来处理多个1到多个关系，但这不起作用 -

http://social.msdn.microsoft.com/forums/en-US/linqprojectgeneral/thread/a68744f6-e1f3-4d28-82eb-87b8187ec0a8/

跳过篮球应该有效 -

 var query = (from e in db.Employee
        join p in db.Phones on e.EmployeeID equals p.EmployeeID into tmpPhones
        from phones in tmpPhones.DefaultIfEmpty()
        join m in db.Emails on e.EmployeeID equals m.EmployeeID into tmpEmails
        from emails in tmpEmails.DefaultIfEmpty()
        group e by e into eGrp

                 select new EmployeeDTO
                 {
                     Id = eGrp.Key.Id,
                     Name = eGrp.Key.Name,
                     Phones = eGrp.Select(x => x.Phones),
                     Emails = eGrp.Select(x => x.Emails)
                 }).ToList();

Linq to sql - 生成错误的sql？

1 个答案: