为什么这两个linq查询返回不同数量的结果？

Question

In a web application that I work with I found a slow piece of code that I wanted to speed up a bit. Original code below:

foreach (Guid g in SecondaryCustomersIds)
{
    var Customer = (from d in Db.CustomerRelationships

                    join c in Db.Customers on
                    d.PrimaryCustomerId equals c.CustomerId

                    where c.IsPrimary == true && d.SecondaryCustomerId == g
                    select c).Distinct().SingleOrDefault();
   //Add this customer to a List<>
}

I thought it might be faster to load this all into a single query, so I attempted to rewrite it as the query below:

var Customers = (from d in Db.CustomerRelationships

                 join c in Db.Customers on
                 d.PrimaryCustomerId equals c.CustomerId

                 where c.IsPrimary == true && SecondaryCustomersIds.Contains(d.SecondaryCustomerId)
                 select c).Distinct();

Which is indeed faster, but now the new query returns fewer records than the first. It seems to me that these two chunks of code are doing the same thing and should return the same number of records. Can anyone see why they would not? What am I missing here?

Answer 1

It's possible for the first query to add a null object to the list ( SingleOrDefault will return the default for the type, or null in this case, if it can't find a matching entity). Thus, for every Customer without a matching relationship, you could be adding a null object to that List<>, which would increase the count.

Answer 2

In your first scenario, does your final List<Customers> have duplicates?

You're calling Distinct , but also looping, which means you're not doing Distinct on your entire collection.

Your second example is calling Distinct on the entire collection.