Question

I just need to delete repeating user info from database. My c# code is below, but I just wonder how to achieve this in SQL rather without using cursor. I think the trick starts with the fetching the first rows or the remaining rows of the whole repeating datasets separated by email.

In C#, I gather repeating emails by groups of 1000 and delete the remaining rows after skipping the first one.

List<string> top1000_emails;
do
{
  top1000_emails = sql.dbCommand.GetFirstColumn<string>(@"select top 1000 email
      from UserBase
      group by email
      having COUNT(email) > 1");

  for (int i = 0; i < top1000_emails.Count; i++)
  {
     var tmpids = sql.dbCommand.GetFirstColumn<long>("select [Id] from UserBase where email = {0}", top1000_emails[i]).Skip(1);
     sql.dbCommand.DeleteByIds<UserBase>(tmpids);
   }
} while (top1000_emails.Count > 0);

Answer 1

You can do ti simply through SQL, like this (if you have SQL Server 2005 or higher):

;WITH a AS (
    SELECT  *,
            ROW_NUMBER() OVER (PARTITION BY email ORDER BY Id) RowNum
    FROM    UserBase
)
-- deleted rows will be:
SELECT  *
--DELETE 
FROM    a
WHERE   a.RowNum <> 1