Class: GraphQL::Schema::Resolver

Inherits:
Object
  • Object
show all
Extended by:
Member::BaseDSLMethods, Member::HasArguments, Member::HasPath
Includes:
Member::GraphQLTypeNames, Member::HasPath
Defined in:
lib/graphql/schema/resolver.rb,
lib/graphql/schema/resolver/has_payload_type.rb

Overview

A class-based container for field configuration and resolution logic. It supports:

  • Arguments, via .argument(...) helper, which will be applied to the field.
  • Return type, via .type(..., null: ...), which will be applied to the field.
  • Description, via .description(...), which will be applied to the field
  • Resolution, via #resolve(**args) method, which will be called to resolve the field.
  • #object and #context accessors for use during #resolve.

Resolvers can be attached with the resolver: option in a field(...) call.

A resolver’s configuration may be overridden with other keywords in the field(...) call.

See the Resolver.field_options to see how a Resolver becomes a set of field configuration options.

See Also:

  • for a concrete subclass of `Resolver`.
  • `Resolver` is a replacement for `GraphQL::Function`

Direct Known Subclasses

Mutation, Subscription

Defined Under Namespace

Modules: HasPayloadType

Constant Summary

Constants included from Member::GraphQLTypeNames

Member::GraphQLTypeNames::Boolean, Member::GraphQLTypeNames::ID, Member::GraphQLTypeNames::Int

Instance Attribute Summary collapse

Class Method Summary collapse

Instance Method Summary collapse

Methods included from Member::HasPath

path

Methods included from Member::HasArguments

add_argument, argument, argument_class, argument_with_loads, arguments, own_arguments

Methods included from Member::BaseDSLMethods

accessible?, default_graphql_name, description, graphql_name, introspection, introspection?, mutation, name, overridden_graphql_name, to_graphql, visible?

Constructor Details

#initialize(object:, context:) ⇒ Resolver

Returns a new instance of Resolver

Parameters:



32
33
34
35
36
37
38
39
40
41
 
# File 'lib/graphql/schema/resolver.rb', line 32

def initialize(object:, context:)
  @object = object
  @context = context
  # Since this hash is constantly rebuilt, cache it for this call
  @arguments_by_keyword = {}
  self.class.arguments.each do |name, arg|
    @arguments_by_keyword[arg.keyword] = arg
  end
  @arguments_loads_as_type = self.class.arguments_loads_as_type
end

Instance Attribute Details

#context ⇒ GraphQL::Query::Context (readonly)

Returns:



47
48
49
 
# File 'lib/graphql/schema/resolver.rb', line 47

def context
  @context
end

#object ⇒ Object (readonly)

Returns The application object this field is being resolved on

Returns:

  • (Object) —

    The application object this field is being resolved on



44
45
46
 
# File 'lib/graphql/schema/resolver.rb', line 44

def object
  @object
end

Class Method Details

.argument(name, type, *rest, loads: nil, **kwargs, &block) ⇒ Object

Add an argument to this field’s signature, but also add some preparation hook methods which will be used for this argument

See Also:

  • for the signature


257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
 
# File 'lib/graphql/schema/resolver.rb', line 257

def argument(name, type, *rest, loads: nil, **kwargs, &block)
  arg_defn = super(*argument_with_loads(name, type, *rest, loads: loads, **kwargs, &block))

  own_arguments_loads_as_type[arg_defn.keyword] = loads if loads

  if loads && arg_defn.type.list?
    class_eval <<-RUBY, __FILE__, __LINE__ + 1
    def load_#{arg_defn.keyword}(values)
      argument = @arguments_by_keyword[:#{arg_defn.keyword}]
      lookup_as_type = @arguments_loads_as_type[:#{arg_defn.keyword}]
      GraphQL::Execution::Lazy.all(values.map { |value| load_application_object(argument, lookup_as_type, value) })
    end
    RUBY
  elsif loads
    class_eval <<-RUBY, __FILE__, __LINE__ + 1
    def load_#{arg_defn.keyword}(value)
      argument = @arguments_by_keyword[:#{arg_defn.keyword}]
      lookup_as_type = @arguments_loads_as_type[:#{arg_defn.keyword}]
      load_application_object(argument, lookup_as_type, value)
    end
    RUBY
  else
    class_eval <<-RUBY, __FILE__, __LINE__ + 1
    def load_#{arg_defn.keyword}(value)
      value
    end
    RUBY
  end

  arg_defn
end

.arguments_loads_as_type ⇒ Object

This method is part of a private API. You should avoid using this method if possible, as it may be removed or be changed in the future. 



290
291
292
293
 
# File 'lib/graphql/schema/resolver.rb', line 290

def arguments_loads_as_type
  inherited_lookups = superclass.respond_to?(:arguments_loads_as_type) ? superclass.arguments_loads_as_type : {}
  inherited_lookups.merge(own_arguments_loads_as_type)
end

.complexity(new_complexity = nil) ⇒ Integer, Proc

Specifies the complexity of the field. Defaults to 1

Returns:

  • (Integer, Proc) 


229
230
231
232
233
234
 
# File 'lib/graphql/schema/resolver.rb', line 229

def complexity(new_complexity = nil)
  if new_complexity
    @complexity = new_complexity
  end
  @complexity || (superclass.respond_to?(:complexity) ? superclass.complexity : 1)
end

.extras(new_extras = nil) ⇒ Object

Additional info injected into #resolve

See Also:

  • {GraphQL{GraphQL::Schema{GraphQL::Schema::Field{GraphQL::Schema::Field#extras}


183
184
185
186
187
188
189
 
# File 'lib/graphql/schema/resolver.rb', line 183

def extras(new_extras = nil)
  if new_extras
    @own_extras = new_extras
  end
  own_extras = @own_extras || []
  own_extras + (superclass.respond_to?(:extras) ? superclass.extras : [])
end

.field_options ⇒ Object 



236
237
238
239
240
241
242
243
244
245
246
247
 
# File 'lib/graphql/schema/resolver.rb', line 236

def field_options
  {
    type: type_expr,
    description: description,
    extras: extras,
    resolver_method: :resolve_with_support,
    resolver_class: self,
    arguments: arguments,
    null: null,
    complexity: complexity,
  }
end

.null(allow_null = nil) ⇒ Object

Specifies whether or not the field is nullable. Defaults to true TODO unify with #type

Parameters:

  • allow_null (Boolean) (defaults to: nil) —

    Whether or not the response can be null



194
195
196
197
198
199
200
 
# File 'lib/graphql/schema/resolver.rb', line 194

def null(allow_null = nil)
  if !allow_null.nil?
    @null = allow_null
  end

  @null.nil? ? (superclass.respond_to?(:null) ? superclass.null : true) : @null
end

.resolve_method(new_method = nil) ⇒ Symbol

Default :resolve set below.

Returns:

  • (Symbol) —

    The method to call on instances of this object to resolve the field



174
175
176
177
178
179
 
# File 'lib/graphql/schema/resolver.rb', line 174

def resolve_method(new_method = nil)
  if new_method
    @resolve_method = new_method
  end
  @resolve_method || (superclass.respond_to?(:resolve_method) ? superclass.resolve_method : :resolve)
end

.type(new_type = nil, null: nil) ⇒ Class

Call this method to get the return type of the field, or use it as a configuration method to assign a return type instead of generating one. TODO unify with #null

Parameters:

  • new_type (Class, nil) (defaults to: nil) —

    If a type definition class is provided, it will be used as the return type of the field

  • null (true, false) —

    Whether or not the field may return nil

Returns:

  • (Class) —

    The type which this field returns.



209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
 
# File 'lib/graphql/schema/resolver.rb', line 209

def type(new_type = nil, null: nil)
  if new_type
    if null.nil?
      raise ArgumentError, "required argument `null:` is missing"
    end
    @type_expr = new_type
    @null = null
  else
    if @type_expr
      GraphQL::Schema::Member::BuildType.parse_type(@type_expr, null: @null)
    elsif superclass.respond_to?(:type)
      superclass.type
    else
      nil
    end
  end
end

.type_expr ⇒ Object

A non-normalized type configuration, without null applied



250
251
252
 
# File 'lib/graphql/schema/resolver.rb', line 250

def type_expr
  @type_expr || (superclass.respond_to?(:type_expr) ? superclass.type_expr : nil)
end

Instance Method Details

#authorized?(**args) ⇒ Boolean, early_return_data

Called after arguments are loaded, but before resolving.

Override it to check everything before calling the mutation.

Parameters:

  • args (Hash) —

    The input arguments

Returns:

  • (Boolean, early_return_data) —

    If false, execution will stop (and early_return_data will be returned instead, if present.)

Raises:



130
131
132
 
# File 'lib/graphql/schema/resolver.rb', line 130

def authorized?(**args)
  true
end

#ready?(**args) ⇒ Boolean, early_return_data

Called before arguments are prepared. Implement this hook to make checks before doing any work.

If it returns a lazy object (like a promise), it will be synced by GraphQL (but the resulting value won’t be used).

Parameters:

  • args (Hash) —

    The input arguments, if there are any

Returns:

  • (Boolean, early_return_data) —

    If false, execution will stop (and early_return_data will be returned instead, if present.)

Raises:



119
120
121
 
# File 'lib/graphql/schema/resolver.rb', line 119

def ready?(**args)
  true
end

#resolve(**args) ⇒ Object

Do the work. Everything happens here.

Returns:

  • (Object) —

    An object corresponding to the return type

Raises:

  • (NotImplementedError) 


105
106
107
 
# File 'lib/graphql/schema/resolver.rb', line 105

def resolve(**args)
  raise NotImplementedError, "#{self.class.name}#resolve should execute the field's logic"
end

#resolve_with_support(**args) ⇒ Object

This method is part of a private API. You should avoid using this method if possible, as it may be removed or be changed in the future.

This method is actually called by the runtime, it does some preparation and then eventually calls the user-defined #resolve method.



53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
 
# File 'lib/graphql/schema/resolver.rb', line 53

def resolve_with_support(**args)
  # First call the ready? hook which may raise
  ready_val = if args.any?
    ready?(**args)
  else
    ready?
  end
  context.schema.after_lazy(ready_val) do |is_ready, ready_early_return|
    if ready_early_return
      if is_ready != false
        raise "Unexpected result from #ready? (expected `true`, `false` or `[false, {...}]`): [#{authorized_result.inspect}, #{ready_early_return.inspect}]"
      else
        ready_early_return
      end
    elsif is_ready
      # Then call each prepare hook, which may return a different value
      # for that argument, or may return a lazy object
      load_arguments_val = load_arguments(args)
      context.schema.after_lazy(load_arguments_val) do |loaded_args|
        # Then call `authorized?`, which may raise or may return a lazy object
        authorized_val = if loaded_args.any?
          authorized?(loaded_args)
        else
          authorized?
        end
        context.schema.after_lazy(authorized_val) do |(authorized_result, early_return)|
          # If the `authorized?` returned two values, `false, early_return`,
          # then use the early return value instead of continuing
          if early_return
            if authorized_result == false
              early_return
            else
              raise "Unexpected result from #authorized? (expected `true`, `false` or `[false, {...}]`): [#{authorized_result.inspect}, #{early_return.inspect}]"
            end
          elsif authorized_result
            # Finally, all the hooks have passed, so resolve it
            if loaded_args.any?
              public_send(self.class.resolve_method, **loaded_args)
            else
              public_send(self.class.resolve_method)
            end
          else
            nil
          end
        end
      end
    end
  end
end