Best practice to check if DataRow contains a certain column

Question

At the moment, when I iterate over the DataRow instances, I do this.

foreach(DataRow row in table)
  return yield new Thingy { Name = row["hazaa"] };

Sooner of later (i.e. sooner), I'll get the table to be missing the column donkey and the poo will hit the fan. After some extensive googling (about 30 seconds) I discovered the following protection syntax.

foreach(DataRow row in table)
  if(row.Table.Columns.Contains("donkey"))
    return yield new Thingy { Name = row["hazaa"] };
  else
    return null;

Now - is this the simplest syntax?! Really? I was expecting a method that gets me the field if it exists or null otherwise. Or at least a Contains method directly on the row.

Am I missing something? I'll be mapping in many fields that way so the code will look dreadfully unreadable...

Answer 1

You can create an extension method to make it cleaner:

static class DataRowExtensions
{
    public static object GetValue(this DataRow row, string column)
    {
        return row.Table.Columns.Contains(column) ? row[column] : null;
    }
}

Now call it like below:

foreach(DataRow row in table)
    return yield new Thingy { Name = row.GetValue("hazaa") };

Answer 2

As your DataTable table always has the same columns ( they won`t change for any row ) you only need to check for the columnname once.

if (table.Columns.Contains("donkey"))
{
    foreach ...
}

Answer 3

I really liked the approach taken by @Varun K. So, having that as a departing point I just wanted to put my two cents, in case it helps someone else. I simply improved it making it generic instead of just using object as a return type.

static class Extensions
{
  public static T Get<T>(this DataRow self, string column)
  {
    return self.Table.Columns.Contains(column)
      ? (T)self[column]
      : default(T);
    }
  }
}

Answer 4

To build on the answer by Varun K, use a generic type parameter:

public static T GetValue<T>(this DataRow row, string column)
{
    if (!row.Table.Columns.Contains(column))
        return default(T);

    object value = row[ColumnName];
    if (value == DBNull.Value)
        return default(T);
    return (T)value;
}

Answer 5

foreach (DataColumn item in row.Table.Columns)
{
    switch (item.ColumnName)
    {
        case "ID":
            {
                p.ID = Convert.ToInt32(row[item.ColumnName].ToString());
            }
            break;
        case "firstName":
            {
                p.firstName = row[item.ColumnName].ToString();
            }
            break;
        case "lastName":
            {
                p.lastName = row[item.ColumnName].ToString();
            }
            break;

        default:
            break;
    };
}

Answer 6

Sometimes a column name might exist, but a row does not contain the data for that column; for example, after filling DataTable using ReadXML.

A simple, fast and secure solution would be to use type checking:

if(row["columnname"].GetType() != typeof(System.DBNull)){
    //DataRow contains "columname"
}else{
    //a safe scope to set default cell data
}