从DataContext.ExecuteQuery以IEnumerable检索关系表 <customobject> （）

Question

Read first: This is part of an application that does not control the database it is accessing so Entity Framework is out of the question. This is a non web based tool that is pulling existing data into memory, but the database is already made and not controlled by us.

I have a class that represents a grouping of data that I want to retrieve from multiple tables (ClassA)

public class ClassA
{
   public int ID;
   public string Detail;
   public List<string> Objects
}

The tables look something like this:

Table1

THISID(int)  | SOMETHING1(nvarchar)| SOMETHING2(nvarchar) | OTHERID(int)
1            | foo1                | bar1                 | 99
2            | foo2                | bar2                 | 98

Table 2

TABLE2ID(int) | OTHERID(int) | SOMETHING3(nvarchar)
100           | 99           | somethingA
101           | 99           | somethingB
102           | 99           | somethingC
103           | 98           | somethingA
104           | 98           | something D

I am using a datacontext like this

string connectionString = (user passed in connection string)
string query = @"SELECT t1.THISID AS ID, t1.SOMETHING1 AS Detail, t2.SOMETHING3 AS Objects
                 FROM Table1 AS t1
                 JOIN Table2 AS t2
                 ON t1.OTHERID = t2.OTHERID";
using(DataContext dc = new DataContext(connectionString))
{
   List<ClassA> aClasses = dc.ExecuteQuery<ClassA>(query);
}

This is obviously not working, but I'm not sure if there is a way to get DataContext or LINQtoSQL to treat the foreign values returned from the query like a list (or collection or enumerable). I know Entity Framework does this for you when you use it like a typical ORM, but here we are mapping a result set into an object and I'm not sure if there is a mechanism to have joined data be treated as a list in the resulting mapped object.

Any ideas on how to do this in one trip to the database? Or is it necessary to get first part from Table1 and instantiate objects, then iterate over them and fill in their Lists with an additional trip to Table2?

Answer 1

I'd be tempted to use Entity Framework. I know you specifically said you couldn't, but it sounds like your reasoning is only that you don't control the database, which just means you can't use code first. But in any event, it doesn't really matter.

You could do it in one trip to the database, but it'll (of course) require pulling back significantly more data.

In any event, you'll have to add a class.

class FlatRow
{
    public int Id { get; set; }
    public string Something1 { get; set; }
    public string Something3 { get; set; }
}

string connectionString = (user passed in connection string)
string query = @"SELECT t1.THISID AS ID, t1.SOMETHING1 AS Detail, t2.SOMETHING3 AS Objects
                 FROM Table1 AS t1
                 JOIN Table2 AS t2
                 ON t1.OTHERID = t2.OTHERID";
using(DataContext dc = new DataContext(connectionString))
{
    List<FlatRow> flat = dc.ExecuteQuery<FlatRow>(query);

    var aClasses = flat
                       .GroupBy(c => new { c.Id, c.Something1 })
                       .Select(c => new ClassA
                                    {
                                        ID = c.Key.Id,
                                        Detail = c.Key.Something1,
                                        Objects = c.Select(x => x.Something3).ToList()
                                    })
                       .ToList();
}

Basically, pull it in flat, then group it in-memory. This is what EF would do, it would just do it for you. But this isn't so bad, if it's what you need.

I'd be tempted to perf-test pulling each table in individually, though, then merging them in-memory. You'll switch to use a GroupJoin , and you'll need one spot of extra logic, but then you won't be pulling all the Something1 data redundantly for entities with multiple Something3 s.