Andrew,

I'd probably do it using a stored procedure that either takes a few parameters (if the number of PKs is going to be very small and limited) or a single string parameter (that would consist of a comma delimited list of PKs). I guess you could also pass in the WHERE clause like this:

DataTable.Select["pk IN (@pk1, @pk2, @pk3, @pk4)"];

HTH,
Chad

>Hi,
>
>How can I get a subset of rows from a DataTable when all I have is a collection of primary keys? The rows that I want from the DataTable have no common data/keys so I unfortunately can't just use DataTable.Select("someColumn = @someValue"). From an SQL perspective the question be : I have a fistfull of primary key values and I want to retrieve the rows that have those primary key values)
>
>The two ways that I have thought of and partly rejected are :
>
>1. use DataTable.Select(string filter) and pass in an OR filter string that is built up of the primary keys that I want rows for. For example :
>
>DataTable.Select["pk = @pk1 OR pk = @pk2 OR pk = @pk3 OR ...").
>
>This seems like a brute force solution that is open to all kinds of problems and I would like to avoid it if I could.
>
>2. Iterate through the collection of PKs and build up a collection of the rows that I want. For example :
>
>RowCollection rowsThatIWant = new RowCollection();
>foreach(object PK in PKCollection)
>{
> DataRow foundRow = DataTable.Rows.Find(PK);
> rowsThatIWant.Add(foundRow);
>}
>
>I am afraid that this solution could take a long time if the collection of PKs is big (execution time is an important feature for our application)
>
>Any better solutions would be greatly appreciated,
>
>Andrew