Sky.Data.Csv

A simple and extremely fast CSV reader and writer which supports reading and writing raw CSV values and typed objects.

Usage

It's really simple to use. Following is an example:

using Sky.Data.Csv;

//reader and writer sample
using(var reader = CsvReader.Create("path-to-file")){
    using(var writer = CsvWriter.Create("path-to-file")){
        foreach(var row in reader){
            writer.WriteRow(row);
        }
    }
}

The Create static method from both CsvReader and CsvWriter can accept a second parameter specifying some options like buffer size, encoding, whether or not to overwrite the existing file when writing a CSV file and whether or not to use cache when reading a CSV file, etc.

The UseCache option in the CsvReaderSettings class is really useful when most of the rows in the CSV file are duplicate. This option helps a lot to improve the performance.

Also both CsvReaderSettings and CsvWriterSettings support a Separator option that you can specify the Cell Separator Character when reading or writing CSV files.

Support reading and writing objects

If you use the generic version of CsvReader<T> and CsvWriter<T>, the Create static method will also accept a parameter of type IDataResolver<T> which supports CsvReader and CsvWriter to read and write typed objects.

The AbstractDataResolver<T> provides a base type with some basic implementation and you can create subclasses of it.

public class Student
{
    public String Name { get; set; }
    public String Address { get; set; }

    public Int32 Height { get; set; }
    public DateTime Birthday { get; set; }
}

public class StudentResolver : AbstractDataResolver<Student>
{
    public override Student Deserialize(List<String> data)
    {
        var culture = CultureInfo.InvariantCulture;
        return new Student
        {
            Name = data[0],
            Height = Int32.Parse(data[1]),
            Birthday = DateTime.ParseExact(data[2], "yyyy-MM-dd", culture),
            Address = data[3],
        };
    }

    public override List<String> Serialize(Student data)
    {
        return new List<String>
        {
            data.Name,
            data.Height.ToString(),
            data.Birthday.ToString("yyyy-MM-dd"),
            data.Address,
        };
    }
}

static void Main(String[] args)
{
    var dataResolver = new StudentResolver();
    var csvPath = "path-to-csv-file";
    using (var reader = CsvReader<Student>.Create(csvPath, dataResolver))
    {
        foreach (var student in reader)
        {
            Console.WriteLine(student.Address);
        }
    }
}

Run tests in Linux

In earlier versions of the source codes, the test project only supports Windows OS. Now it also supports Linux.

If you want to run the test codes in Linux (tested on Ubuntu 18.04 LTS x64), firstly change the current directory to Sky.Data.Csv.Test and then run following commands in the terminal:

Sky.Data.Csv.Test$ dotnet run -c Release --project Sky.Data.Csv.Test.Core.csproj

Target

This is a simple but fast implementation of CsvReader and CsvWriter.

This CsvReader supports all four formats saved by the newest version of Excel, that's comma seperated, ms dos, macintosh and comma seperated UTF8.

By default, the CsvReaderSettings and CsvWriterSettings uses Encoding.Default as the default encoding and comma (,) as the field separator. You can also specify these settings (and some other settings, many settings are supported) by creating a new setting object and pass it to the Create static method.

License

MIT Licensed.