I'd like to translate the following SQL statement into a linq query:
select COUNT(*), itemid, globalid, title, preview, previewimage, previewimage_alt, link
from (
select distinct Id, itemid, globalid, title, preview, previewimage, previewimage_alt,
(select top 1 link from LikeCounter where GlobalId=x.GlobalId) as link
from [LikeCounter] x
where PortalId=1 and LanguageId=1
) as t
GROUP BY itemid, globalid, title, preview, previewimage, previewimage_alt, link
ORDER BY COUNT(*) desc
The query is over a view that holds records of objects being "liked". Since the objects can be published in multiple places, and the view was setup to allow for filtering for a certain place, it requires a distinct before grouping the records to find out the view count (that's the reason for the additional query for the "link" column).
Is a nested SELECT statement possible in one linq statement?
The inner query is no problem:
(from x in LikeCounter
where x.PortalId==1 && x.LanguageId==1
select new {x.Id, x.ItemId, x.GlobalId, x.LanguageId, x.Title, x.Preview, x.PreviewImage_alt,
Morelink=(from y in LikeCounter
where y.GlobalId==x.GlobalId
select y.Morelink).FirstOrDefault()
}).Distinct()
But is there a way to extend this with the grouping of the distinct records, that results in just one query to the database ?
Thanks in advance for any input...
Nina
Edit:
the following query almost returns what I want -- but produces multiple queries to the SQL server:
(from y in
((from x in LikeCounter
where x.PortalId==1 && x.LanguageId==1
select new {x.Id, x.ItemId, x.GlobalId, x.LanguageId, x.Title, x.Preview, x.PreviewImage_alt,
Link=(from y in Xparo_LikeCounter
where y.GlobalId==x.GlobalId
select y.Link).FirstOrDefault()
}).Distinct())
group y by y.GlobalId into grp
select new {Data=grp, Count= grp.Count()}).OrderByDescending (x => x.Count)