Reputation: 87
I have something similar to the code below in LINQPAD using C# Statements. My goal is to get the actual SQL Insert statments not actually update the database.
I can easily delete the data after it inserts with this small sample, but I will need this for a larger push of data. I hope that I have missed something simple either in L2S or LINQPad.
Is there an easier way retrieve the SQL Insert?
var e1 = new MyEntity(){ Text = "First" };
var e2 = new MyEntity(){ Text = "Second" };
MyEntities.InsertOnSubmit(e1);
MyEntities.InsertOnSubmit(e2);
SubmitChanges();
Upvotes: 1
Views: 717
Reputation: 10418
When we did the samples for "LINQ in Action", we used the following method which gets the scheduled changes from the context:
public String GetChangeText(System.Data.Linq.DataContext context)
{
MethodInfo mi = typeof(DataContext).GetMethod("GetChangeText",
BindingFlags.NonPublic | BindingFlags.Instance);
return mi.Invoke(context, null).ToString();
}
If you want to see this in action, download the samples in LINQPad (see http://www.thinqlinq.com/Default/LINQ-In-Action-Samples-available-in-LINQPad.aspx) and check out chapter 6 example 6.29.
Upvotes: 1
Reputation: 69290
A quick-n-dirty way is to wrap everything in a transaction scope that is never commited:
using(TransactionScope ts = new TransactionScope())
{
var e1 = new MyEntity(){ Text = "First" };
var e2 = new MyEntity(){ Text = "Second" };
MyEntities.InsertOnSubmit(e1);
MyEntities.InsertOnSubmit(e2);
SubmitChanges();
// Deliberately not committing the transaction.
}
This works well for small volumes. If the data volume is large and you have full recovery model on the database the transaction log growth might become a problem.
Upvotes: 2