Static Extension Methods Already Exist Error:

Question

I am trying to overload in my static method, I am getting an error that method already exist, how to fix this issue?

public static int GetValue(this DataTable datatable, int rowindex, string columnName)
{
return 0;
}

public static double GetValue(this DataTable datatable, int rowindex, string columnName)
{
return 0.0;
}

public static decimal GetValue(this DataTable datatable, int rowindex, string columnName)
{
return (decimal)0.0;
}

public static float GetValue(this DataTable datatable, int rowindex, string columnName)
{
return (float)0.0;
}

Vulpes · Answer

OK in that case you'll have to use: public static DataView CreateDuplicateDataView(this DataTable datatable) { try { if (datatable == null) throw new ArgumentNullException(source); var dataviewcopy = new DataView( datatable datatable.DefaultView.RowFilter datatable.DefaultView.Sort datatable.DefaultView.RowStateFilter ); dataviewcopy.AllowEdit = dataview.AllowEdit; dataviewcopy.AllowNew = dataview.AllowNew; dataviewcopy.AllowDelete = dataview.AllowDelete; return dataviewcopy; } catch (Exception ex) { throw new Exception(CreateDuplicateDataView: 
 + ex.Message); } }

Vulpes · Answer

Yes those are OK as well though (unlike the string version) they will throw an exception on a DBNull.Value. BTW if you're wondering why there doesn't seem to be a Convert.ToFloat method it's because it's actually called Convert.ToSingle given that C#'s float type is equivalent to the System.Single structure. Although in theory datatable.Rows[rowindex][columnName] should never be null (only DBNullValue) it might be better to use the Convert.ToSingle method rather than a cast because if it were null the former would return 0 rather than throw an exception.

Vulpes · Answer

Yes it's safe to do that for string values. In fact it automatically caters for DBNull.Value because when you apply ToString() to that expression you get String.Empty which is a reasonable default.

Vulpes · Answer

This version would return 0 (or some other value) if the database value was NULL but would otherwise behave as before: public static int GetIntValue(this DataTable datatable int rowindex string columnName) { try { object obj = datatable.Rows[rowindex][columnName]; if(Convert.IsDBNull(obj)) { return 0; // or perhaps some other value such as -1 } else { return Convert.ToInt32(obj); } } catch (Exception ex) { throw new Exception(GetIntValue: 
 + ex.Message); } }

Vulpes · Answer

There is one other possibility which is to use a generic method. You'd need to pass an additional parameter (of type T) so that the compiler could infer the type of T but if you made this into an 'out' parameter then you wouldn't need a return value as well: public static void GetValue (this DataTable datatable int rowindex string columnName out T value) { object obj = datatable.Rows[rowindex][columnName]; value = Convert.IsDBNull(obj) ? default(T) : (T)obj; } and call with: decimal d; dt.GetValue(rowindex columnName out d);

Vulpes · Answer

Given that the expression datatable.Rows[rowIndex][columnName] returns an object the most natural way to do this would be to have a single extension method which also returns an object and then cast that to the appropriate type: public static object GetValue(this DataTable datatable int rowindex string columnName) { return datatable.Rows[rowindex][columnName]; } If you were then getting the value of a decimal column say you'd call with: decimal d = (decimal)datatable.GetValue(rowindex columnName); or an int column with: int i = (int)datatable.GetValue(rowindex columnName); There's also the possibility that the method may return DBNull.Value if the corresponding value in the database is NULL. It would be possible to deal with that automatically by returning instead the default value of the column's type using the following revised code: public static object GetValue(this DataTable datatable int rowindex string columnName) { object obj = datatable.Rows[rowindex][columnName]; if(Convert.IsDBNull(obj)) { Type t = datatable.Columns[columnName].DataType; if( t.IsValueType) return Activator.CreateInstance(t); else return null; } return obj; } So in the first example instead of returning DBNull.Value which would cause an exception when you attempt to cast to decimal the method would return 0m which is the default value of the decimal type and there would be no exception: decimal d = (decimal)datatable.GetValue(rowindex columnName); // d == 0m

Vikram Agrawal · Answer

Hi David Well as of now whatever manish has told is absolutely right and if you want all these things then you should change your method name and here is no need to keep same name. You can name it like : GetIntValue(parameters) GetDoubleValue(parameters) GetFloatValue(parameters) GetDecimalValue(parameters) Thanks.

Manish Kumar Choudhary · Answer

Hi David Smith  Change the parameter list of one of the function so that compiler can identify the difference of those functions then your code will run fine. this DataTable datatable int rowindex string columnName this list should be different for all function.

David Smith · Answer

Ok I was just wondering. I am getting ready to post something about recordsets. Could review right fast

Vulpes · Answer

Just to avoid repeating DataView given that the type of the variable is obvious (and can be inferred by the compiler) from the constructor call on the RHS of the equals sign. If you don't like 'var' (a lot of developers don't and I have reservations about it myself) then just replace it with DataView.

David Smith · Answer

Vuples why did you use a var?

David Smith · Answer

You are passing in a dataview instead of a table the group wants to pass in a table and use the default view of the table

Vulpes · Answer

That's what I suspected :) As you've probably already gathered this isn't a straightforward matter as the DataView class doesn't have a Clone() method or similar. I think for practical purposes the following should suffice. Also you aren't limited to just copying the 'default' view as the 'this' parameter is now a DataView rather than a DataTable: public static DataView CreateDuplicateDataView(this DataView dataview) { try { if (dataview == null) throw new ArgumentNullException(source); var dataviewcopy = new DataView( dataview.Table dataView.RowFilter dataView.Sort dataView.RowStateFilter ); dataviewcopy.AllowEdit = dataview.AllowEdit; dataviewcopy.AllowNew = dataview.AllowNew; dataviewcopy.AllowDelete = dataview.AllowDelete; return dataviewcopy; } catch (Exception ex) { throw new Exception(CreateDuplicateDataView: 
 + ex.Message); } }

David Smith · Answer

I am trying to Create a duplicate dataview of the same table Vuples.

Vulpes · Answer

Well a DataView is always associated with a particular DataTable. So are you trying to create a duplicate DataView for the same table or for a clone of that table? Currently the code is doing the latter.

David Smith · Answer

Vuples is there a better way to do this below? public static DataView CreateDuplicateDataView(this DataTable datatable) { try { if (datatable == null) throw new ArgumentNullException(source); DataTable datatableCopy = datatable.Copy(); datatableCopy.DefaultView.RowFilter = datatable.DefaultView.RowFilter; return datatableCopy.DefaultView; } catch (Exception ex) { throw new Exception(CreateDuplicateDataView: 
 + ex.Message); } }

Vulpes · Answer

float is just an alias for the .NET framework type System.Single so it doesn't matter which you use. Personally I prefer to always use the C# alias for this and the other basic types though there are developers who prefer to do the opposite.

David Smith · Answer

Thanks so are saying the return type should be change from a float to a single?

David Smith · Answer

Thanks so are saying the return type should be change from a float to a single?

David Smith · Answer

Last one vuples what about the bool and float value public static bool GetBoolValue(this DataTable datatable int rowindex string columnName) { try { return Convert.ToBoolean(datatable.Rows[rowindex][columnName]); } catch (Exception ex) { throw new Exception(GetBoolValue: 
 + ex.Message); } } public static float GetFloatValue(this DataTable datatable int rowindex string columnName) { try { return (float)datatable.Rows[rowindex][columnName]; } catch (Exception ex) { throw new Exception(GetFloatValue: 
 + ex.Message); } }

David Smith · Answer

I agree it can't no robust than that. is it safe to do this below? public static string GetStringValue(this DataTable datatable int rowindex string columnName) { try { return (datatable.Rows[rowindex][columnName]).ToString(); } catch (Exception ex) { throw new Exception(GetStringValue: 
 + ex.Message); } }

Vulpes · Answer

There's no 'next version' as I don't think there's anything else which is worth catering for specifically.

David Smith · Answer

What's the next version?

David Smith · Answer

Let me see both options Default Value and Dbnull Vuples

Vulpes · Answer

Well that would catch all possible errors including conversion errors and the rowindex or columnName not existing in the DataTable so it should certainly be robust enough. The only question is whether it's over-robust in relation to DBNull values. Are you happy to thrown an exception in such cases or would you rather return a default value?

David Smith · Answer

Good point vuples. Also is there any other error handling I can add below. I want to make the function robust as possible. public static int GetIntValue(this DataTable datatable int rowindex string columnName) { try { return Convert.ToInt32(datatable.Rows[rowindex][columnName]); } catch (Exception ex) { throw new Exception(GetIntValue: 
 + ex.Message); } }

David Smith · Answer

David Vikram and Vuples thanks originally I was going to do the obj type but then I was told to do it explicitly because owner did not want to any casting. Since it's an extension class I am going just going to have both options in the class below. 1st Option: public static objectGetValue(this DataTable datatable int rowindex string columnName) { } 2nd Option: Change the name of the function public static decimal GetIntValue(this DataTable datatable int rowindex string columnName) { } public static decimal GetDoubleValue(this DataTable datatable int rowindex string columnName) { }

David Smith · Answer

I am only using rowindex is always always an integer. Get table element with given column names and row indixes. So there is not reason to convert the row index. public static int GetValue( this DataTable datatable int rowindex string columnName)

Manish Kumar Choudhary · Answer

Hi David Smith  u can just do like this(Simplest way) public static int GetValue(this DataTable datatable int rowindex string columnName) { Convert.ToInd( rowindex)//Use It as int return 0; } public static double GetValue(this DataTable datatable double rowindex string columnName) { Convert.ToInd( rowindex)//Use It as int return 0.0; } public static decimal GetValue(this DataTable datatable decimal rowindex string columnName) { Convert.ToInd( rowindex)//Use It as int return (decimal)0.0; } public static float GetValue(this DataTable datatable float rowindex string columnName) { Convert.ToInd( rowindex)//Use It as int return (float)0.0; } call the above function with just rowindex that can be converted to int easily. Visit following links for more details http://msdn.microsoft.com/en-us/library/ms229029.aspx

David Smith · Answer

I got you what is a slick way of doing below. I want to keep the same function name GetValue and just return different types public static int GetValue(this DataTable datatable int rowindex string columnName) { return 0; } public static double GetValue(this DataTable datatable int rowindex string columnName) { return 0.0; } public static decimal GetValue(this DataTable datatable int rowindex string columnName) { return (decimal)0.0; } public static float GetValue(this DataTable datatable int rowindex string columnName) { return (float)0.0; }

Manish Kumar Choudhary · Answer

Hi David Smith  Overloading means using same function name doing different things. For that Two function should have different signature otherwise compiler cant differentiate. You must have different signature.

David Smith · Answer

Is there a way to do it without changing the parameters if you notice I am returning different types