从实体映射到 EF Core 中的 DTO 后，Distinct 如何在 IQueryable 上工作？

Question

I have a query similar to this that I am running against our database:

var result = await _context.MyEntities
    .Select(x => new SubEntityDto { Id = x.SubEntity.Id })
    .Distinct()
    .ToListAsync();

The entity SubEntity can be the same entity set on multiple MyEntity .

Since I want a list of SubEntityDto with no duplicates, I run .Distinct() on the resulting IQueryable<SubEntityDto> . This is the SubEntityDto class:

public class SubEntityDto
{
    public int Id { get; set; }
}

And this works, but I don't know how it manages to make the list distinct when working with the DTOs. Doesn't .Distinct() use the .Equals method in this scenario? And doesn't .Equals default to reference equality, which checks whether two instances are the same instance?

If I load the list from the database, and then do Distinct() , it doesn't work anymore, like this:

var subEntities = await _context.MyEntities
    .Select(x => new SubEntityDto { Id = x.SubEntity.Id })
    .ToListAsync();

var distinctSubEntities = subEntities.Distinct().ToList(); // Not distinct.

I'm thinking that it somehow manages to do this when creating the SQL for the SQL query, but can anyone tell me what is happening? I'm puzzled about the fact that it seems to keep track of the entites after they're mapped to DTOs.

Answer 1

As mentioned in the comments to your question, when you create your query it gets translated into SQL query by your provider. So in the first case it is Queryable.Distinct() , however, in the second case you have already run your query against your database by calling ToListAsync() and result of it already translated to DTO's in memory (objects). Therefore Enumerable.Distinct() is run.

You can also use your debugger to see the created query

You can see the created query by your provider by putting a breakpoint and then investigating the contents of the variable DebugView→Query→Text Visualizer will show you the result.