LINQ to DataSet case insensitive group by

Question

I have a data table and I want to perform a case insensitive group by over a column of data table (say Column1 of type string). I observed that normally LINQ to DataSet performs a case sensitive comparison. For example, if Column1 has two string values "Test" and "test", after applying group by it returns two separate rows with the values "Test" and "test", instead of one.

The query is:

var countGroupQuery = from table in dataTable.AsEnumerable()
                      group table by table.Field<string>(Column1) into groupedTable
                      select new
                      {
                          value = groupedTable.Key,
                          count = groupedTable.Count()
                      };

Is there any method to perform a case-insensitive group by so that in the above example I get only one row with one value (either "Test" or "test")? ToUpper or ToLower would actually change the values to either upper case or lower case instead of using at least one of the input values, so I don't want to use this:

group table by table.Field<string>(Column1).ToUpper() into groupedTable

Answer 1

You can't do this from a query expression, but you can do it with dot notation:

var query = dataTable.AsEnumerable()
                     .GroupBy(x => table.Field<string>(Column1),
                              StringComparer.InvariantCultureIgnoreCase)
                     .Select(groupedTable => new
                             {
                                 value = groupedTable.Key,
                                 count = groupedTable.Count()
                             });

You can even use a more complicated overload of GroupBy to do it in one call:

var query = dataTable.AsEnumerable()
                     .GroupBy(x => table.Field<string>(Column1),
                              (key, group) => { value = key, 
                                                count = group.Count() },
                              StringComparer.InvariantCultureIgnoreCase));

Obviously that's using the invariant culture - you could also use the current culture or ordinal rules.

Answer 2

This MSDN article has some information on datasets and case sensitivity..

You can control the case sensitivity of filtering, searching, and sorting by setting the dataset's CaseSensitive property.