Question

I have a first query that returns a set of entities:

var resultSet = ....query....ToList();

which would return A, B, C, D, E

The entities inside this set are organized into chains because they have a reference (prevEntityId) pointing to the same type of entity, i.e.:

A -> B -> D

C -> E

I would like to write a second query so that only A and C are now returned but I have no idea how to write it.

I would prefer your answer with linq methods (like .Where()) instead of the new linq syntax.

Thank you

Update: sorry I initially used the wrong vocabulary in my question with the term “foreign key”. Actually an entity has a direct reference to the previous entity so to elect an entity in the second query, there must be no other entity that references it. I thought it would take 2 queries but if you think it can be done in one…

Answer 1

EDIT: Okay, as the relationship is the opposite of what I expected, you’re looking for entities such that no entity has it as the previous one.

One option:

 Where(x => !fullList.Any(y => y.PrevEntityId == x.Id));

Now, if your query is actually being executed in SQL it would be worth seeing what that generates – it may well not be efficient. It pretty definitely won’t be efficient in LINQ to Objects. For LINQ to Objects, you’d be better off with:

 HashSet<string> prevIds = new HashSet<string>(fullList.Select(x => x.PrevId));
 var heads = fullList.Where(x => !prevIds.Contains(x));

If you could more information about your environment, that would help us to help you.