Here’s the problem :

While implementing a C# wrapper for an online API (Discogs) I’ve been faced to a dilemma : quite often the responses returned have mostly similar members and while modeling these responses to classes, some questions surfaces on which way to go would be the best.

Example :

Querying for a ‘release’ or a ‘master’ will return an object that contains an array of ‘artist’, however these ‘artists’ do not exactly have the same members.

Currently I decided to represent these ‘artists’ as a single ‘Artist’ class, against having respective ‘ReleaseArtist’ and ‘MasterArtist’ classes which soon becomes very confusing even though another problem arises : when a category (master or release) does not return these members, they will be null. Though it might sound confusing as well I find it less confusing than the former situation as I’ve tackled the problem by simply not showing null members when visualizing these objects.

Is this the right approach to follow ?

An example of these differences :

public class Artist
{
    public List<Alias> Aliases { get; set; }
    public string DataQuality { get; set; }
    public List<Image> Images { get; set; }
    public string Name { get; set; }
    public List<string> NameVariations { get; set; }
    public string Profile { get; set; }
    public string Realname { get; set; }
    public string ReleasesUrl { get; set; }
    public string ResourceUrl { get; set; }
    public string Uri { get; set; }
    public List<string> Urls { get; set; }
}

public class ReleaseArtist
{
    public string Join { get; set; }
    public string Name { get; set; }
    public string Anv { get; set; }
    public string Tracks { get; set; }
    public string Role { get; set; }
    public string ResourceUrl { get; set; }
    public int Id { get; set; }
}

Edit :

I decided to keep things as simple as they can be : separated.

There are few weird names but it’ll always be less confusing than the previous situations.