Using Entity Framework Core 2.1 and an SQLite database, I get different behaviour from a LINQ GroupJoin if I use it after another Join. It is not clear if this is a bug or if there is something I am overlooking.
I have created a minimal VS2017 project to reproduce this, which can be found here:
https://gitlab.com/haddoncd/EntityFrameworkGroupJoinIssue
In this example, I get one object for each row in the Blog table, each containing multiple PostTitles:
db.Blogs
.GroupJoin(
db.Posts,
s => s.BlogId,
p => p.BlogId,
(s, ps) => new
{
s.BlogId,
s.BlogTitle,
PostTitles = ps.Select(p => p.PostTitle),
}
)
.OrderBy(x => x.BlogId))
But in this example, the results are "flattened" - I get one object for each Blog-Post pair like you would from a conventional SQL outer join:
db.Subscriptions
.Join(
db.Blogs,
s => s.BlogId,
b => b.BlogId,
(s, b) => new
{
s.SubscriptionId,
b.BlogId,
b.BlogTitle,
}
)
.GroupJoin(
db.Posts,
x => x.BlogId,
p => p.BlogId,
(x, ps) => new
{
x.SubscriptionId,
x.BlogTitle,
PostTitles = ps.Select(p => p.PostTitle),
}
)
.OrderBy(x => x.SubscriptionId))