Hi,

I met a very strange problem with hibernate. I have a parent entity -- Employee which has two child Entities -- Comment and ManagerEmployee. Employee has one-to-many mappings to both.

When I use get(Employee.class, id), hibernate generates a huge select statement with left out join to tables of Comment and ManagerEmployee, and I got duplicated results in employee.getComments().size()

When I use find("from Employee where id = ?", id), then hibernate generates separate select statements for each Employee, Comment and ManagerEmployee tables. and there is no duplicated results in employee.getComments().size()

Did anyone meet the same problem before, can I still trust the get(class, id) method? How can get & find have different strategies when dealing with semantically same queries?

Thanks a lot for any clarify on this!

I don't know what is causing the duplicates, but the difference between session.get() and using a HQL query is that session.get() uses the fetching strategy that you have specified in your mapping files/annotations (eg. fetch="join"), but this is ignored when you use HQL. The reason that you are seeing additional select's in the second case is probably that your collections are non-lazy.

Thanks nordborg.

You solved my problem!

Just think it's better to let new Hibernate users know, depending on the mapping, Hibernate will generates different SQL queries which might in turn produce different results. get(class, id) is different from find("from class where id = ?", id).