CsvHelper - read in multiple columns to a single list

Question

I\'m using CSVHelper to read in lots of data

I\'m wondering if it\'s possible to read the last n columns in and transpose them to a list

Answer 1

This does the trick as a mapper.

public sealed class PersonMap : CsvClassMap<Person>
{
    private List<string> attributeColumns = 
        new List<string> { "Attribute1", "Attribute2", "Attribute3" };

    public override void CreateMap()
    {
        Map(m => m.FirstName).Name("FirstName").Index(0);
        Map(m => m.LastName).Name("LastName").Index(1);
        Map(m => m.Attributes).ConvertUsing(row =>
            attributeColumns
                .Select(column => row.GetField<string>(column))
                .Where(value => String.IsNullOrWhiteSpace(value) == false)
            );
    }
}

Then you just need something like this

using (var reader = new CsvReader(new StreamReader(filePath)))
{
    reader.Configuration.RegisterClassMap<PersonMap>();
    while (reader.Read())
    {
        var card = reader.GetRecord<Person>();
    }
}

Answer 2

I don't know this library, so following might be helpful or not.

If you already have an IEnumerable<IEnumerable<string>> which represents all records with all columns you could use this Linq query to get your List<Person> with the IList<string> Attributes:

IEnumerable<IEnumerable<string>> allRecords = .....;
IEnumerable<Person> allPersons = allRecords
.Select(rec => 
{
    var person = new Person();
    person.FirstName = rec.ElementAt(0);
    person.LastName = rec.ElementAtOrDefault(1);
    person.Attributes = rec.Skip(2).ToList();
    return person;
}).ToList();

Edit: Downloaded the library, following at least compiles, could not really test it:

IList<Person> allPersons = new List<Person>();
using (var reader = new CsvHelper.CsvReader(yourTextReader))
{
    while (reader.Read())
    {
        var person = new Person();
        person.FirstName = reader.CurrentRecord.ElementAt(0);
        person.LastName = reader.CurrentRecord.ElementAtOrDefault(1);
        person.Attributes = reader.CurrentRecord.Skip(2).ToList();
        allPersons.Add(person);
    }
}

Answer 3

You don't actually need to define the column names beforehand, you can get them from the FieldHeaders property. So in a more dynamic version of Neil's answer you could define your Mapper as such:

public sealed class PersonMap : CsvClassMap<Person> {
    public override void CreateMap() {
        Map(m => m.FirstName).Name("FirstName").Index(0);
        Map(m => m.LastName).Name("LastName").Index(1);
        Map(m => m.Attributes).ConvertUsing(row =>
            (row as CsvReader)?.FieldHeaders
                 .Where(header => header.StartsWith("Attribute"))
                 .Select(header => row.GetField<string>(header))
                 .Where(value => !string.IsNullOrWhiteSpace(value))
                 .ToList()
        );
    }
}

Answer 4

Version 3 has support for reading and writing IEnumerable properties. You can use an IList<T> property just like you have. You just need to specify the start index of the field.

Map( m => m.Attributes ).Index( 2 );