更多 LINQY-ness（子选择）

Question

I need to use LINQ to build a kind of weird query that uses a sub-query.

I'm really looking for distinct records. Normally, the SQL would look like this:

select distinct col1, col2 from foo where col3 = somevalue

However, col2 just happens to be a BLOB, so I can't use distinct . So, I think that the next best SQL looks like this:

select f1.col1, f1.col2 
from foo f1 where f1.id in 
  (select distinct f2.id from foo f2 where f2.col3 = somevalue

I'm not sure what is the best way to "phrase" that second query in LINQ. Here's what I have so far, and it works, but I'm not sure whether it's optimal:

var query = from f in foo
            where f.col3 == somevalue
            select new {id = f.id};

var result = from f in foo
             join q in query on f.id equals q.id
             select new MyType() {col1 = f.col1, col2 = f.col2};

That gives me what I want, but according to SQL Manager, the resulting query is about 8% more expensive than my hand-crafted SQL sub-query. Is there a better way to write that?

Answer 1

Can you try this?

var result = from f in foo
             where query.Contains(f.id)
             select new MyType() {col1 = f.col1, col2 = f.col2};