Multi mapper assumes that `splitOn` column is never 0-index in IDbDataReader

[bawkee]

Hi,

I noticed in Dapper that if the column specified in splitOn argument in its Query method is first in the SELECT statement, Dapper will throw the MultiMapException because it can't find it.

This is because the for loop in below method skips the first column due to i > 0:

From SqlMapper.cs

private static int GetNextSplit(int startIdx, string splitOn, IDataReader reader)
{
	if (splitOn == "*")
	{
		return --startIdx;
	}

	for (var i = startIdx - 1; i > 0; --i)
	{
		if (string.Equals(splitOn, reader.GetName(i), StringComparison.OrdinalIgnoreCase))
		{
			return i;
		}
	}

	throw MultiMapException(reader);
}

Now I haven't ever used dapper in any production so I may be missing something big here but I spent several hours trying to figure out any reason why would this be to no avail.

For example, consider the following tutorial I found on http://dapper-tutorial.net/:
https://dotnetfiddle.net/DPiy2b

It works because it has SELECT *, but if you change that SELECT statement to:

SELECT A.OrderID, CustomerID, EmployeeID, OrderDate, ShipperID, ProductID, Quantity

Dapper will throw the aforementioned exception.

Changing the SQL to:

SELECT A.OrderID, A.OrderID, CustomerID, EmployeeID, OrderDate, ShipperID, ProductID, Quantity

Avoids the exception, because Dapper will pick up OrderID when mapping since it starts from 1.

Again this may very well be by design, I can't tell straight away as I've only used dapper for couple of hours since 2015. If it is and I missed something obvious, I apologise for wasting everyone's time!

[MrJerB]

I think it is by design. There's no use splitting the result set at the first column is there?

[bawkee]

@MrJerB Thanks, I had a totally incorrect idea on how multimapping mechanism works and my sample DB driver (Access) didn't help at all. After some more investigation I figured it out, this SO answer helped me a lot.