See SqlDataReader.NextResult (an SqlDataReader is returned from calling SqlCommand.ExecuteReader):
Advances the data reader to the next result [set], when reading the results of batch Transact-SQL statements.
Example:
string commandText = @"SELECT Id, ContactId
FROM dbo.Subscriptions;
SELECT Id, [Name]
FROM dbo.Contacts;";
List<Subscription> subscriptions = new List<Subscription>();
List<Contact> contacts = new List<Contact>();
using (SqlConnection dbConnection = new SqlConnection(@"Data Source=server;Database=database;Integrated Security=true;"))
{
dbConnection.Open();
using (SqlCommand dbCommand = dbConnection.CreateCommand())
{
dbCommand.CommandText = commandText;
using(SqlDataReader reader = dbCommand.ExecuteReader())
{
while(reader.Read())
{
subscriptions.Add(new Subscription()
{
Id = (int)reader["Id"],
ContactId = (int)reader["ContactId"]
});
}
// this advances to the next resultset
reader.NextResult();
while(reader.Read())
{
contacts.Add(new Contact()
{
Id = (int)reader["Id"],
Name = (string)reader["Name"]
});
}
}
}
}
Other examples:
Create a Stored Procedure
that has multiple selects, and fill the DataSet
.
using (SqlConnection conn = new SqlConnection(connection))
{
DataSet dataset = new DataSet();
SqlDataAdapter adapter = new SqlDataAdapter();
adapter.SelectCommand = new SqlCommand("MyProcedure", conn);
adapter.SelectCommand.CommandType = CommandType.StoredProcedure;
adapter.Fill(dataset);
return dataset;
}
The returned dataset will have a DataTable
in it's Tables array for each select in the stored procedure.
Tools like "dapper" make this very easy, regardless of whether you use ad-hoc text queries or stored procedures; for example:
using(var multi = conn.QueryMultiple(sql, args))
{
var customers = multi.Read<Customer>().AsList(); // first grid
var regionName = multi.ReadFirstOrDefault<string>(); // second grid
var addresses = multi.Read<Address>().AsList(); // third grid
// todo: use those things
}
Individual grids can also be read without buffering (as an open IEnumerable<T>
over the reader itself) via the optional parameters to Read[<T>]
.
I call a sproc and get multiple result sets with object, so you end up with a
List<List<Dictionary<string, object>>>
In the multiResultsSet each results set is then
List<Dictionary<string, object>>
They can be cast to their types and transformed into the models as needed.
After you set up the sproc command with everything needed, pass it to this:
private static List<List<Dictionary<string, object>>> ProcessReader(SqlCommand command)
{
var tables = new List<List<Dictionary<string, object>>>();
using (var reader = command.ExecuteReader())
{
do
{
var table = new List<Dictionary<string, object>>();
while (reader.Read())
table.Add(Read(reader));
tables.Add(table);
} while (reader.NextResult());
}
return tables;
}
and Read() is fairly straight forward.
private static Dictionary<string, object> Read(IDataRecord reader)
{
var row = new Dictionary<string, object>();
for (var i = 0; i < reader.FieldCount; i++)
{
var val = reader[i];
row[reader.GetName(i)] = val == DBNull.Value ? null : val;
}
return row;
}
I am using a dictionary approach. You can use Newton Json to convert it to json. This way you are not tied to types and IDataRecord
public List<List<Dictionary<string, object>>> ExecuteSqlReader(string cmd, int commandTimeout = 30, CommandType commandType = CommandType.Text)
{
var sqlCmd = new SqlCommand(cmd);
var allRecord = new List<List<Dictionary<string, object>>>();
using (var connection = GetSqlConnection(commandTimeout, commandType, ref sqlCmd))
{
using (var reader = sqlCmd.ExecuteReader())
{
if (reader.HasRows)
{
var result = new List<Dictionary<string, object>>();
while (reader.Read())
{
result = GetTableRowData(reader);
}
allRecord.Add(result);
}
while (reader.NextResult())
{
if (reader.HasRows)
{
var result = new List<Dictionary<string, object>>();
while (reader.Read())
{
result = GetTableRowData(reader);
}
allRecord.Add(result);
}
}
}
}
return allRecord;
}
If you love us? You can donate to us via Paypal or buy me a coffee so we can maintain and grow! Thank you!
Donate Us With