Files
spark/src/DataStreamSuite/DataStream/Primitives/Distinct/DistinctRowEvaluator.cs
T

30 lines
1.2 KiB
C#

namespace BCG
{
public class DistinctRowEvaluator : GroupByEvaluator
{
private int myIdGenerator = 0;
public DistinctRowEvaluator (string[] columnNames) : base (null, columnNames) {}
public DistinctRowEvaluator(string colMapExpression, string[] columnNames)
: base (colMapExpression, null, columnNames) {}
public override IBlueRow? evaluate(IDataPacket packet) {
myOutputTable ??= new GroupByOutputTable(packet.Schema);
string hashString = GroupByStream.getRepString (packet, myGroupByCols);
if (packet.newRows.Count == 0)
//if there are no rows for this group we remove the group from the output table
myOutputTable.removeRow (hashString);
else if (myOutputTable.getRow (hashString) == null)
//a row is added for the first time (or first time after the removal of the group)
myOutputTable.addRow (
BlueDataUtility.getNewReplacedRow (
BlueDataUtility.extractRow (packet.newRows, 0),
myIdGenerator++, "id"),
hashString);
return myProject (myOutputTable.getRow (hashString));
}
}
}