在 LINQ 中加入和分组

Question

这些是我的输入表：

Persons Table:
================
ID  Code
------------------------
1   Person1     # Person1: John Smith, 25, 50Kg
2   Person2     # Person2: William Brown, 30, 80Kg
3   Person3     # Person3: James Miller, 32, 73Kg

StringProperties Table:
=========================
ID  PersonID    Name        Value
----------------------------------------------
1   1           FirstName   John        # Person1: John Smith, 25, 50Kg
2   1           LastName    Smith       # Person1: John Smith, 25, 50Kg
3   2           FirstName   William     # Person2: William Brown, 30, 80Kg
4   2           LastName    Brown       # Person2: William Brown, 30, 80Kg
5   3           FirstName   James       # Person3: James Miller, 32, 73Kg
6   3           LastName    Miller      # Person3: James Miller, 32, 73Kg

NumericProperties Table:
=========================
ID  PersonID    Name    Value
-----------------------------------------
1   1           Age     25          # Person1: John Smith, 25, 50Kg
2   1           Weight  50          # Person1: John Smith, 25, 50Kg
3   2           Age     30          # Person2: William Brown, 30, 80Kg
4   2           Weight  80          # Person2: William Brown, 30, 80Kg
5   3           Age     32          # Person3: James Miller, 32, 73Kg
6   3           Weight  73          # Person3: James Miller, 32, 73Kg

我想编写生成以下结果的 LINQ 查询：

Result:
==========
Code        FirstName   LastName    Age     Weight
-----------------------------------------------------------------
Person1     John        Smith       25      50
Person2     William     Brown       30      80
Person3     James       Miller      32      73

这是我的代码，但不能正常工作：

var q = from p in db.Persons
        join s in db.StringProperties on p.ID equals s.PersonID
        join n in db.NumericProperties on p.ID equals n.PersonID
        group p by p.Code into g
        select new
        {
            g.Key,
            g
        };

Answer 1

假设我们有看起来像这样的实体：

public class Person
{
    public int ID { get; set; }
    public string Code { get; set; }
}

public class StringProperty
{
    public int ID { get; set; }
    public int PersonID { get; set; }
    public string Name { get; set; }
    public string Value { get; set; }
}

public class NumericProperty
{
    public int ID { get; set; }
    public int PersonID { get; set; }
    public string Name { get; set; }
    public int Value { get; set; }
}

数据如下：

var persons = new List<Person> 
{
    new Person { ID = 1, Code = "Person1" },
    new Person { ID = 2, Code = "Person2" },
    new Person { ID = 3, Code = "Person3" }
};

var stringProperties = new List<StringProperty>
{
    new StringProperty { ID = 1, PersonID = 1, Name = "FirstName", Value = "John" },
    new StringProperty { ID = 2, PersonID = 1, Name = "LastName", Value = "Smith" },
    new StringProperty { ID = 3, PersonID = 2, Name = "FirstName", Value = "William" },
    new StringProperty { ID = 4, PersonID = 2, Name = "LastName", Value = "Brown" },
    new StringProperty { ID = 5, PersonID = 3, Name = "FirstName", Value = "James" },
    new StringProperty { ID = 6, PersonID = 3, Name = "LastName", Value = "Miller" }
};

var numericProperties = new List<NumericProperty>
{
    new NumericProperty { ID = 1, PersonID = 1, Name = "Age", Value = 25 },
    new NumericProperty { ID = 2, PersonID = 1, Name = "Weight", Value = 50 },
    new NumericProperty { ID = 3, PersonID = 2, Name = "Age", Value = 30 },
    new NumericProperty { ID = 4, PersonID = 2, Name = "Weight", Value = 80 },
    new NumericProperty { ID = 5, PersonID = 3, Name = "Age", Value = 32 },
    new NumericProperty { ID = 6, PersonID = 3, Name = "Weight", Value = 73 }
};

我们可以像这样将人物加入字符串值 table 并同时进行旋转：

var stringValues = from p in persons
                   join s in stringProperties on p.ID equals s.PersonID
                   group s by p.Code into g 
                   select new
                   {
                       Code = g.Key,
                       FirstName = g.Where(s => s.Name == "FirstName").First().Value,
                       LastName = g.Where(s => s.Name == "LastName").First().Value,
                   };

对数值执行相同的操作：

var numericValues = from p in persons
                    join n in numericProperties on p.ID equals n.PersonID
                    group n by p.Code into g 
                    select new
                    {
                        Code = g.Key,
                        Age = g.Where(n => n.Name == "Age").First().Value,
                        Weight = g.Where(n => n.Name == "Weight").First().Value,
                    };

现在把他们一起加入：

var q = from s in stringValues
        join n in numericValues on s.Code equals n.Code
        select new 
        {
            Code = s.Code,
            FirstName = s.FirstName,
            LastName = s.LastName,
            Age = n.Age,
            Weight = n.Weight
        };

您可以在一条语句中执行此操作，但如果您将其拆分成这样会更简单。

实际上我更有可能在存储过程中执行此操作，因为它在那里更快。

Answer 2

有多种方法可以解决这个问题。这里有一对。

首先，感谢 DavidG 提供 class 和输入数据。

这是选项 1：

var query =
    from p in persons
    join s in stringProperties on p.ID equals s.PersonID into gss
    join n in numericProperties on p.ID equals n.PersonID into gns
    from fn in gss.Where(x => x.Name == "FirstName")
    from ln in gss.Where(x => x.Name == "LastName")
    from a in gns.Where(x => x.Name == "Age")
    from w in gns.Where(x => x.Name == "Weight")
    select new
    {
        p.Code,
        FirstName = fn.Value,
        LastName = ln.Value,
        Age = a.Value,
        Weight = w.Value,
    };

这是选项 2：

var query =
    from p in persons
    join s in stringProperties on p.ID equals s.PersonID into gss
    join n in numericProperties on p.ID equals n.PersonID into gns
    let sl = gss.ToLookup(x => x.Name, x => x.Value)
    let nl = gns.ToLookup(x => x.Name, x => x.Value)
    from FirstName in sl["FirstName"]
    from LastName in sl["LastName"]
    from Age in nl["Age"]
    from Weight in nl["Weight"]
    select new
    {
        p.Code,
        FirstName,
        LastName,
        Age,
        Weight,
    };

这两个都给我这个结果：

在 LINQ 中加入和分组

Join and Group in LINQ

c#

linq

join

group-by