yaiio / graphql-net

Convert GraphQL to IQueryable

GraphQL.Net

An implementation of GraphQL for .NET and IQueryable

docs

Description

Many of the .NET GraphQL implementations that have come out so far only seem to work in memory. For me, this isn't terribly useful since most of my data is stored in a database (and I assume that's the case for many others). This library is an implementation of the GraphQL spec that converts GraphQL queries to IQueryable. That IQueryable can then be executed using the ORM of your choice.

Here's a descriptive example, using an example from the GraphQL spec:

{
  user(id: 4) {
    id
    name
    profilePic(size: 100)
  }
}

The above GraphQL query could be translated to:

db.Users
  .Where(u => u.Id == 4)
  .Select(u => new
  {
      id = u.Id,
      name = u.Name,
      profilePic = db.ProfilePics
                     .FirstOrDefault(p => p.UserId == u.Id && p.Size == 100)
                     .Url
  })
  .FirstOrDefault();

Building a Schema

Let's assume we have an Entity Framework DbContext that looks like this:

public class TestContext : DbContext
{
    public IDbSet<User> Users { get; set; }
    public IDbSet<Account> Accounts { get; set; }
}

public class User
{
    public int Id { get; set; }
    public string Name { get; set; }

    public int AccountId { get; set; }
    public Account Account { get; set; }
}

public class Account
{
    public int Id { get; set; }
    public string Name { get; set; }
    public bool Paid { get; set; }
}

First, we create and set the default schema by providing a function that creates our context:

var schema = GraphQL<TestContext>.CreateDefaultSchema(() => new TestContext());

The default schema is required to use the helper methodGraphQL<TContext>.Execute(query), but you can execute queries against the schema without it. Next, we'll define a type in the schema and fields on that type.

var user = schema.AddType<User>();
user.AddField(u => u.Id);
user.AddField(u => u.Name);
user.AddField(u => u.Account);
user.AddField("totalUsers", (db, u) => db.Users.Count());
user.AddField("accountPaid", (db, u) => u.Account.Paid);

Fields can be defined using only a property expression, or you can specify your own fields and provide a custom resolving expression. Let's do the same for account:

schema.AddType<Account>().AddAllFields();

If we just want to expose all fields, we can use the AddAllFields helper method.

The last thing we want to do is create some queries, as fields on the schema itself. Let's add some to find users:

schema.AddListField("users", db => db.Users);
schema.AddField("user", new { id = 0 }, (db, args) => db.Users.Where(u => u.Id == args.id).FirstOrDefault());

In our first query, we want to see all users so we can just return the entire list. However, notice how in the second query we define the shape of an anonymous type new { id = 0 }. This is what is expected to be passed in from the GraphQL query. Since we've defined the shape, we can now use that in the Where clause to build our IQueryable. We use FirstOrDefault to signify that this query will return a single result.

schema.Complete();

Finally, we complete the schema when we've finished setting up. Now we're ready to execute a query.

Executing Queries

var query = @"{
user(id:1) {
    userId : id
    userName : name
    account {
        id
        paid
    }
    totalUsers
}}";

var gql = new GraphQL<TestContext>(schema);
var dict = gql.ExecuteQuery(query);
Console.WriteLine(JsonConvert.SerializeObject(dict, Formatting.Indented));

// {
//   "user": {
//     "userId": 1,
//     "userName": "Joe User",
//     "account": {
//       "id": 1,
//       "paid": true
//     },
//     "totalUsers": 2
//   }
// }

The results from executing the query are returned as a nested Dictionary<string, object> which can easily be converted to JSON and returned to the user.

Installation/NuGet

Add GraphQL.Net to your project via the Package Manager Console.

PM> Install-Package GraphQL.Net

TODO

Support directives like @skip. Support field arguments with complex types. Support enum types. Support custom primitive types (ID, datetime, etc). Introspection