CRUD over aggregate root using traditional ado.net

Question

Can anyone show me simple CRUD statements for aggregate root using traditional ado.net? Thanks in advance!

Answer 1

(This is written on the assumption that a GUID or some non-database generated primary key is used) Also alot of boiler code such as connection management etc... should be moved to a base class for Repository. If Order is the aggregate root, one possibly should make OrderLineRepo private to the assembly

public class OrderRepository : Repository

{
    public void Save(Order order)
    {
        if(order.IsDirty)
        {
                    //sets up connection if required, command and sql
            ICommand command = BuildCommandForSave(order);  
            command.Execute();
            OrderLineRepository orderLineRepo = GetOrderLineRepo();
            foreach(OrderLine line in order.OrderLines)
            {
                orderLineRepo.Save(line);
            }
        }
    }
}

However I'd stress that this is really a simple naive implementation, and that I'd personally utilize an ORM like nHibernate for my persistence if doing DDD as the requirements for a good well tested persistence layer are non-trivial

Also this assumes that the IsDirty function takes children into account - we would also require a means to see if the order is new/edited, not just dirty