afmp94/graphql-ruby

Ruby implementation of Facebook's GraphQL

graphql

A Ruby implementation of GraphQL.

• Website
• API Documentation

Installation

Install from RubyGems by adding it to your Gemfile, then bundling.

# Gemfile
gem 'graphql'

$ bundle install

Overview

Declare types & build a schema

# Declare a type...
PostType = GraphQL::ObjectType.define do
  name "Post"
  description "A blog post"

  field :id, !types.ID
  field :title, !types.String
  field :body, !types.String
  field :comments, types[!CommentType]
end

# ...and a query root
QueryType = GraphQL::ObjectType.define do
  name "Query"
  description "The query root of this schema"

  field :post do
    type PostType
    argument :id, !types.ID
    resolve -> (obj, args, ctx) { Post.find(args["id"]) }
  end
end

# Then create your schema
Schema = GraphQL::Schema.define do
  query QueryType
  max_depth 8
end

Execute queries

Execute GraphQL queries on a given schema, from a query string.

result_hash = Schema.execute(query_string)
# {
#   "data" => {
#     "post" => {
#        "id" => 1,
#        "title" => "GraphQL is nice"
#     }
#   }
# }

Use with Relay

If you're building a backend for Relay, you'll need:

• A JSON dump of the schema, which you can get by sending GraphQL::Introspection::INTROSPECTION_QUERY
• Relay-specific helpers for GraphQL, see GraphQL::Relay

Goals

• Implement the GraphQL spec & support a Relay front end
• Provide idiomatic, plain-Ruby API with similarities to reference implementation where possible
• Support Ruby on Rails and Relay

Getting Involved

• Say hi & ask questions in the #ruby channel on Slack or on Twitter!
• Report bugs by posting a description, full stack trace, and all relevant code in a GitHub issue.
• Features & patches are welcome! Consider discussing it in an issue or in the #ruby channel on Slack to make sure we're on the same page.
• Run the tests with rake test or start up guard with bundle exec guard.
• Build the site with rake site:serve, then visit localhost:4000.

Related Projects

Code

• graphql-ruby + Rails demo (src / heroku)
• graphql-batch, a batched query execution strategy
• graphql-libgraphqlparser, bindings to libgraphqlparser, a C-level parser.

Blog Posts

• Building a blog in GraphQL and Relay on Rails Introduction, Part 1, Part 2
• https://medium.com/@khor/relay-facebook-on-rails-8b4af2057152
• https://blog.jacobwgillespie.com/from-rest-to-graphql-b4e95e94c26b#.4cjtklrwt
• http://mgiroux.me/2015/getting-started-with-rails-graphql-relay/
• http://mgiroux.me/2015/uploading-files-using-relay-with-rails/

To Do

• StaticValidation improvements
  • Use catch-all type/field/argument definitions instead of terminating traversal
  • Reduce ad-hoc traversals?
  • Validators are order-dependent, is this a smell?
  • Tests for interference between validators are poor
  • Maybe this is a candidate for a rewrite?
• Add Rails-y argument validations, eg less_than: 100, max_length: 255, one_of: [...]
  • Must be customizable
• Relay:
  • GlobalNodeIdentification.to_global_id should receive the type name and object, not id. (Or, maintain the "type_name, id in, type_name, id out" pattern?)
  • Reduce duplication in ArrayConnection / RelationConnection
  • Improve API for creating edges (better RANGE_ADD support)
  • If the new edge isn't a member of the connection's objects, raise a nice error
• Missing Enum value should raise a descriptive error, not "key not found"
• args should whitelist keys -- if you request a key that isn't defined for the field, it should 💥
• Fix middleware
  • Handle out-of-bounds lookup, eg graphql-batch
  • Handle non-serial execution, eg @defer
• Support non-instance-eval .define, eg .define { |defn| ... }
• First-class promise support
  • like graphql-batch but more local
  • support promises in connection resolves
• Add immutable transformation API to AST
  • Support working with AST as data
  • Adding fields to selections (__typename can go anywhere, others are type-specific)
  • Renaming fragments from local names to unique names
• Support AST subclasses? This would be hard, I think classes are used as hash keys in many places.
• Support object deep-copy (schema, type, field, argument)? To support multiple schemas based on the same types.
• Improve the website
  • Feature the logo in the header
  • Split readme.md into index.md (a homepage with code samples) and a technical readme (how to install, link to homepage)
  • Move "Related projects" to a guide
  • Revisit guides, maybe split them into smaller, more specific pages
  • Put guide titles into the <title />