How would I convert only a subset of column values from a DataRow into a string array using LINQ?

Question

EDIT: I'm giving the credit to Nick as his answer pointed me in the right direction. In his answer he had set all DataColumns to be retrieved as String, but not all columns in the DataTable were strings. Also Select() was appropriate to get individual rows whereas SelectMany() joined all column values for multiple rows into a single array, or one big row if you will. Here is the solution:

using (dataTable)
using (var sw = new StreamWriter(resultPath, false))
{
    DataColumn[]
          exclCols = { dataTable.Columns["registryvalue"] }, //excluded
          inclCols = dataTable.Columns //included columns
                              .Cast<DataColumn>()
                              .Except(exclCols)
                              .ToArray();

    sw.WriteLine(String.Join( //header line
                 ",", inclCols.Select(column => column.ColumnName).ToArray()));

    foreach (var row in dataTable.AsEnumerable().Select(
            dataRow => inclCols.Select(column => dataRow.Field<Object>(column))))
    {
        sw.WriteLine(String.Join(
                     ",", row.Select(r => (r == null) ? 
                     "" : r.ToString()).ToArray()));
    }
}

I'm using foreach to loop through the rows of a DataTable. There are only certain columns' values that I want from the DataRow and I want to convert this subset of values into a string array (we'll call it myStringArray). The end result is:

StreamWriter.WriteLine(String.Join(",", myStringArray))

I can accomplish writing the line without using LINQ, but then that doesn't help me better understand LINQ. Getting the subset of column names I want into an array is easy, but I just don't know how to use that array to get only those columns' data from the DataRow.

I re-read your question after posting my answer below... Do you need just one column's data or do you need to combine strings from multiple columns? — Nick Strupat
Yeah but I think you will end up with an Array of String[]. Because you are going to get a String[] for each row in your DataTable. Maybe you should clarify this in your question. — Orkun Ozen
I'm looking to get a subset of all the objects that make up a DataRow and then convert those objects into a single string array. — user1197862

Nick Strupat Nick Strupat · Accepted Answer · 2012-02-15T23:56:47

EDIT 2:

Here's how to do it with LINQ

var columnNames = new [] { "Column1", "Column2", "Column3" };
var strings = dataTable.AsEnumerable().SelectMany(dataRow => columnNames.Select(columnName => dataRow.Field<String>(columnName)));

EDIT:

I'm not sure how to do that with LINQ. It is simple to express using two foreach loops though

String [] columnNames = new String [] { "Column1", "Column2", "Column3" };
var myStrings = new List<String>();
foreach (DataRow dataRow in dataTable.Rows)
    foreach (String columnName in columnNames)
        myStrings.Add(dataRow.Field<String>(columnName));

ORIGINAL:

You'll need DataTable.AsEnumerable() extension method to use LINQ against your DataTable

Something like this will do the job...

var myStrings = from dataRow in dataTable.AsEnumerable()
                select dataRow.Field<String>("TheColumnName");

How would I convert only a subset of column values from a DataRow into a string array using LINQ?

2 Answers