How can I choose one record over the other?

Question

I have a list that has some duplicates.

Row#    Lineid  ItemDescItemId  RoadTax VehicleId   Amount
1   122317  None    -1  26.63   -78603  300
2   122317  None    -2  17.75   -78603  200
3   122317  None    -1  22.19   -78602  250
4   122317  Deli    -2  17.75   -78603  200

In this case, Row 2 is a duplicate of Row 4, since the LineId, RoadTax, Amount and VehicleId match. However, I want to keep the line with an item description and eliminate line # 2. So my output list looks like this:

Row#    Lineid  ItemDesc ItemId RoadTax VehicleId   Amount
1   122317  None    -1  26.63   -78603  300
3   122317  None    -1  22.19   -78602  250
4   122317  Deli    -2  17.75   -78603  200

I wrote a IEqualityComparer class based on an example on MSDN. The class looks like this:

  public class RoadTaxComparer : IEqualityComparer<RoadTaxDto>
        {
            // Items are equal if ItemId / VehicleId / RoadTax are equal.
            public bool Equals(RoadTaxDto x, RoadTaxDto y)
            {

                //Check whether the compared objects reference the same data. 
                if (Object.ReferenceEquals(x, y)) return true;

                //Check whether any of the compared objects is null. 
                if (Object.ReferenceEquals(x, null) || Object.ReferenceEquals(y, null))
                    return false;

                //Check whether the products' properties are equal. 
                return x.VehicleId == y.VehicleId && x.ItemId == y.ItemId && x.RoadTax == y.RoadTax && x.Amount == y.Amount;
            }

            // If Equals() returns true for a pair of objects  
            // then GetHashCode() must return the same value for these objects. 

            public int GetHashCode(RoadTaxDto roadTaxDto)
            {
                //Check whether the object is null 
                if (Object.ReferenceEquals(roadTaxDto, null)) return 0;

                //Get hash code for the VehicleId. 
                int hashVehicleId = roadTaxDto.VehicleId.GetHashCode();

                //Get hash code for the ItemId field. 
                int hashCodeItemId = roadTaxDto.ItemId.GetHashCode();

                //Calculate the hash code for the QuoteTaxDto. 
                return hashVehicleId ^ hashCodeItemId;
            }

        }

The RoadTaxDto structure looks like this:

class RoadTaxDto
{
public int LineId {get;set}
public string ItemDesc {get;set;}
public int VehicleId {get;set;}
public decimal RoadTax {get;set;}
public int VehicleId {get;set;}
public decimal Amount {get;set;}
}

I use the following command to eliminate duplicates.

List<RoadTaxDto> mergedList = RoadTaxes.Union(RoadTaxes, new RoadTaxComparer()).ToList();

When I run a comparer on it, I am not guaranteed that row 2 is eliminated. So how can I ensure that if a record has a duplicate, the record that says "None" will always get eliminated from the list.

Mark Toman · Answer 1 · 2013-08-02T17:44:00.110

1

I would move GetHashCode() to the RoadTaxDto and then do this:

foreach (var g in list.GroupBy(i => i.GetHashCode()))
    list2.Add(
        g.FirstOrDefault(i => i.ItemDesc != "None") ?? 
        g.First());

edited Aug 02 '13 at 17:44

answered Aug 02 '13 at 17:11

Mark Toman

3,090
2
17
18

This is an interesting approach. I will try it out and let you know. – abhi Aug 02 '13 at 18:17

Icarus · Accepted Answer · 2013-08-02T17:20:08.193

A pure "SQLish" approach wouldn't work for you?

Something like this:

var list = new [] { 
    new RoadTaxDto {LineId=122317,ItemDesc="None", ItemId=-1,RoadTax =26.63M  , VehicleId=-78603  ,Amount=300},
    new RoadTaxDto {LineId=122317,ItemDesc="None", ItemId=-2,RoadTax =17.75M  , VehicleId=-78603  ,Amount=200},
    new RoadTaxDto {LineId=122317,ItemDesc="None", ItemId=-1,RoadTax =22.19M  , VehicleId=-78602  ,Amount=250},
    new RoadTaxDto {LineId=122317,ItemDesc="Deli", ItemId=-2,RoadTax =17.75M , VehicleId=-78603  ,Amount=200}
};

var query = (from c in list join x in list 
    on new { c.LineId, c.ItemId , c.VehicleId, c.Amount,c.RoadTax} 
equals new {x.LineId, x.ItemId, x.VehicleId,x.Amount,x.RoadTax}
select new RoadTaxDto {
   LineId = c.LineId,
   ItemDesc = x.ItemDesc!="None"? x.ItemDesc:c.ItemDesc,
   VehicleId=c.VehicleId,
   Amount=c.Amount,
   RoadTax=c.RoadTax,
   ItemId=c.ItemId
}
).GroupBy(x => new { x.LineId, x.RoadTax, x.Amount, x.VehicleId} )
 .Select(grp => grp.Last());

Prints:

LineId  ItemDesc    VehicleId   ItemId  RoadTax Amount
122317  None        -78603      -1      26.63   300
122317  Deli        -78603      -2      17.75   200
122317  None        -78602      -1      22.19   250

With this approach, I would have to get rid of the RoadTaxComparer, but this works. Thank You! — abhi, Aug 02 '13 at 18:14

How can I choose one record over the other?

2 Answers2