You don't have access to that information via fill, but it does exist.
// MissingSchemaAction.AddWithKey uses CommandBehavior.KeyInfo
SqlDataReader reader = command.ExecuteReader(CommandBehavior.KeyInfo)
DataTable table = reader.GetSchemaTable() ' this table has the information
' The "NumericPrecision" and "NumericScale" columns have that information.
Unfortunately if you get the information this way, its not obvious to then
fill the DataSet without rexecuting.
(new LoadDataReader()).Fill(dataset, reader);
internal class LoadDataReader : System.Data.Common.DbDataAdapter() {
internal int Fill(DataSet dataset, IDataReader reader) {
// the method on DbDataAdapter that takes a DataTable and a reader
is protected
// so you can only call it if your derive from DbDataAdapter and
delegate to it
base.Fill(table, "Table", reader, 0, 0);
}
}
--
This posting is provided "AS IS", with no warranties, and confers no rights.
Please do not send email directly to this alias. This alias is for newsgroup
purposes only.
"Josh" <(E-Mail Removed)> wrote in message
news:(E-Mail Removed)...
>I need to know if there is a way to determine the precision and scale of a
>column in a DataTable, when the data type is a decimal. We do not have a
>problem if there is a row that has a non-null value in it using the GetBits
>method. However, if the value is Null or even 0 I cannot get the scale.
>Let me run through it a little bit explaining along the way. We populate
>our DataSet using the following procedure:
>
>
>
> Private Function FillDataset(ByVal voCommand As SqlCommand, _
> ByVal veMissingSchemaAction As
> MissingSchemaAction) As DataSet
>
> Dim loDA As SqlDataAdapter
> Dim loDS As DataSet
>
> Try
> loDA = New SqlDataAdapter(voCommand)
> loDS = New DataSet
>
> 'MissingSchemaAction can now be set when retrieving dataset. This
> was necessary
> 'for proper sorting when going through trsADO interface.
>
> loDA.MissingSchemaAction = veMissingSchemaAction
> loDA.Fill(loDS)
>
> Return loDS
>
> Finally
> CleanUpDatabaseObjs(roDataAdapter:=loDA)
>
> End Try
>
> End Function
>
> In this case we are using MissingSchemaAction of Add. Let just say for
> this command we have 1 table, with 2 columns and 3 rows. The columns are
> defined in SQL server as a Decimal(5,3) and the values are as follows:
> Column1 Column2
> ------------------- ---------------
> Null 23.12
> 12.023 Null
> 1.555 0
>
> Column2 is not a problem for me, because I can look at the value in the
> first row and using the Decimal.GetBits method I can determine the scale
> (believe it or not the scale correctly returns 3). However, I cannot do a
> GetBits on Row 1 of Column1 because the value is Null. Furthermore I
> cannot spin through all the rows until I find a non-null/non-zero value
> because first we use some pretty big tables with many rows, performance is
> an issue and second, and probably more importantly, we cannot guarantee
> that we will always have a value, sometimes the entire column could be
> filled with Null or 0.
>
> So, is there a way to tap into the underlying table schema and determine
> how a column is defined? I am actually kind of surprised that there are
> not properties on the columns collection for precision and scale, but
> there is not, so I need another solution to figure it out.
>
> Thanks!
>
|
Similar Threads
