The Result<T> class is being used by methods that return a value.

var result = Result<int>.Successful(1);

var error = new Error("An error occured.");
var result = Result<int>.Failed(error);

// the error can also hold an exception and/or metadata
var error = new Error("An error occured.", new Exception());
var error = new Error("An error occured.", new Dictionary<string, object> { { "EventId", "4576" } });
var error = new Error("An error occured.", new Exception(), new Dictionary<string, object> { { "EventId", "4576" } });

An example usage of the Result<T> class.

public async Task<Result<int>> ExecuteJob()
{
    try
    {
        var jobId = ExecuteCleanupJob();

        if (jobId == 0)
        {
            // create an Error indicating the reason of failure
            var error = new Error("Cleanup job was not executed.");

            return Result<int>.Fail(error);
        }

        return Result<int>.Ok(jobId);
    }
    catch (Exception ex)
    {
        // create an Error and attach the exception
        var error = new Error("An unexpected error occured while trying execute Cleanup job.", ex);

        return Result<int>.Fail(error);
    }
}

The anatomy of a Result

A Result holds certain information about itself.

var sucessfulResult = Result.Successful();

sucessfulResult.IsSuccessful // true
sucessfulResult.IsFailed // false
sucessfulResult.Error // throws InvalidOperationException as the result is not in a failed state


var failedResult = Result.Failed(new Error("Operation failed.");

failedResult.IsSuccessful // false
failedResult.IsFailed // true
failedResult.Error // IError instance

When the Result<T> holds a return value.

var sucessfulResult = Result<int>.Successful(1);

sucessfulResult.IsSuccessful // true
sucessfulResult.IsFailed // false
sucessfulResult.Value // 1
sucessfulResult.Error // throws InvalidOperationException as the result is not in a fail state


var failedResult = Result<int>.Failed(new Error("Operation failed.");

failedResult.IsSuccessful // false
failedResult.IsFailed // true
failedResult.Value // throws InvalidOperationException as the result is not in a success state
failedResult.Error // IError instance

Extending

You can introduce your very own Error classes by extending the existing one.

The below custom NotFoundError class is being used when an application might need to return a NotFound 404 response.

public interface IHttpError : IError
{
    public HttpStatusCode HttpStatusCode { get; }
}

public sealed class NotFoundError : Error, IHttpError
{
    public HttpStatusCode HttpStatusCode => HttpStatusCode.NotFound;

    public NotFoundError(string message = "The request resource was not found.")
        : base(message)
    {
    }
}

An example of the above custom Error class.

public async Task<Result<IEnumerable<Project>>> GetProjects()
{
    try
    {
        var projects = await GetProjects();

        if (!projects.Any())
        {
            var error = new NotFoundError(); // <-- the new NotFoundError Error class

            return Result<int>.Fail(error);
        }

        return Result<int>.Ok(jobId);
    }
    catch (Exception ex)
    {
        // create an Error and attach the exception
        var error = new Error("An unexpected error occured while trying execute Cleanup job.", ex);

        return Result<int>.Fail(error);
    }
}

You can also introduce simple Error classes to better reflect your domain.

public sealed class ApplicationError : Error
{
    public ApplicationError(string eventId, string message, Exception exception, Dictionary<string, object> metadata)
        : base(message, exception, metadata)
    {
    }
}

The Error class provides 3 constructors, so you are free to use whichever suits your needs best. See definition