Section 6 -- Execution.md

Execution

GraphQL generates a response from a request via execution.

A request for execution consists of a few pieces of information:

• The schema to use, typically solely provided by the GraphQL service.
• A Document containing GraphQL Operations and Fragments to execute.
• Optionally: The name of the Operation in the Document to execute.
• Optionally: Values for any Variables defined by the Operation.
• An initial value corresponding to the root type being executed. Conceptually, an initial value represents the "universe" of data available via a GraphQL Service. It is common for a GraphQL Service to always use the same initial value for every request.

Given this information, the result of {ExecuteRequest()} produces the response, to be formatted according to the Response section below.

Executing Requests

To execute a request, the executor must have a parsed Document (as defined in the “Query Language” part of this spec) and a selected operation name to run if the document defines multiple operations, otherwise the document is expected to only contain a single operation. The result of the request is determined by the result of executing this operation according to the "Executing Operations” section below.

ExecuteRequest(schema, document, operationName, variableValues, initialValue):

• Let {operation} be the result of {GetOperation(document, operationName)}.
• Let {coercedVariableValues} be the result of {CoerceVariableValues(schema, operation, variableValues)}.
• If {operation} is a query operation:
  • Return {ExecuteQuery(operation, schema, coercedVariableValues, initialValue)}.
• Otherwise if {operation} is a mutation operation:
  • Return {ExecuteMutation(operation, schema, coercedVariableValues, initialValue)}.
• Otherwise if {operation} is a subscription operation:
  • Return {Subscribe(operation, schema, coercedVariableValues, initialValue)}.

GetOperation(document, operationName):

• If {operationName} is {null}:
  • If {document} contains exactly one operation.
    • Return the Operation contained in the {document}.
  • Otherwise produce a query error requiring {operationName}.
• Otherwise:
  • Let {operation} be the Operation named {operationName} in {document}.
  • If {operation} was not found, produce a query error.
  • Return {operation}.

Validating Requests

As explained in the Validation section, only requests which pass all validation rules should be executed. If validation errors are known, they should be reported in the list of "errors" in the response and the request must fail without execution.

Typically validation is performed in the context of a request immediately before execution, however a GraphQL service may execute a request without immediately validating it if that exact same request is known to have been validated before. A GraphQL service should only execute requests which at some point were known to be free of any validation errors, and have since not changed.

For example: the request may be validated during development, provided it does not later change, or a service may validate a request once and memoize the result to avoid validating the same request again in the future.

Coercing Variable Values

If the operation has defined any variables, then the values for those variables need to be coerced using the input coercion rules of variable's declared type. If a query error is encountered during input coercion of variable values, then the operation fails without execution.

CoerceVariableValues(schema, operation, variableValues):

• Let {coercedValues} be an empty unordered Map.
• Let {variableDefinitions} be the variables defined by {operation}.
• For each {variableDefinition} in {variableDefinitions}:
  • Let {variableName} be the name of {variableDefinition}.
  • Let {variableType} be the expected type of {variableDefinition}.
  • Let {defaultValue} be the default value for {variableDefinition}.
  • Let {value} be the value provided in {variableValues} for the name {variableName}.
  • If {value} does not exist (was not provided in {variableValues}):
    • If {defaultValue} exists (including {null}):
      • Add an entry to {coercedValues} named {variableName} with the value {defaultValue}.
    • Otherwise if {variableType} is a Non-Nullable type, throw a query error.
    • Otherwise, continue to the next variable definition.
  • Otherwise, if {value} cannot be coerced according to the input coercion rules of {variableType}, throw a query error.
  • Let {coercedValue} be the result of coercing {value} according to the input coercion rules of {variableType}.
  • Add an entry to {coercedValues} named {variableName} with the value {coercedValue}.
• Return {coercedValues}.

Note: This algorithm is very similar to {CoerceArgumentValues()}.

Executing Operations

The type system, as described in the “Type System” section of the spec, must provide a query root object type. If mutations or subscriptions are supported, it must also provide a mutation or subscription root object type, respectively.

Query

If the operation is a query, the result of the operation is the result of executing the query’s top level selection set with the query root object type.

An initial value may be provided when executing a query.

ExecuteQuery(query, schema, variableValues, initialValue):

• Let {queryType} be the root Query type in {schema}.
• Assert: {queryType} is an Object type.
• Let {selectionSet} be the top level Selection Set in {query}.
• Let {data} be the result of running {ExecuteSelectionSet(selectionSet, queryType, initialValue, variableValues)} normally (allowing parallelization).
• Let {errors} be any field errors produced while executing the selection set.
• Return an unordered map containing {data} and {errors}.

Mutation

If the operation is a mutation, the result of the operation is the result of executing the mutation’s top level selection set on the mutation root object type. This selection set should be executed serially.

It is expected that the top level fields in a mutation operation perform side-effects on the underlying data system. Serial execution of the provided mutations ensures against race conditions during these side-effects.

ExecuteMutation(mutation, schema, variableValues, initialValue):

• Let {mutationType} be the root Mutation type in {schema}.
• Assert: {mutationType} is an Object type.
• Let {selectionSet} be the top level Selection Set in {mutation}.
• Let {data} be the result of running {ExecuteSelectionSet(selectionSet, mutationType, initialValue, variableValues)} serially.
• Let {errors} be any field errors produced while executing the selection set.
• Return an unordered map containing {data} and {errors}.

Subscription

If the operation is a subscription, the result is an event stream called the "Response Stream" where each event in the event stream is the result of executing the operation for each new event on an underlying "Source Stream".

Executing a subscription creates a persistent function on the server that maps an underlying Source Stream to a returned Response Stream.

Subscribe(subscription, schema, variableValues, initialValue):

• Let {sourceStream} be the result of running {CreateSourceEventStream(subscription, schema, variableValues, initialValue)}.
• Let {responseStream} be the result of running {MapSourceToResponseEvent(sourceStream, subscription, schema, variableValues)}
• Return {responseStream}.

Note: In large scale subscription systems, the {Subscribe} and {ExecuteSubscriptionEvent} algorithms may be run on separate services to maintain predictable scaling properties. See the section below on Supporting Subscriptions at Scale.

Event Streams

An event stream represents a sequence of discrete events over time which can be observed. As an example, a "Pub-Sub" system may produce an event stream when "subscribing to a topic", with an event occurring on that event stream for each "publish" to that topic. Event streams may produce an infinite sequence of events or may complete at any point. Event streams may complete in response to an error or simply because no more events will occur. An observer may at any point decide to stop observing an event stream by cancelling it, after which it must receive no more events from that event stream.

As an example, consider a chat application. To subscribe to new messages posted to the chat room, the client sends a request like so:

subscription NewMessages {
  newMessage(roomId: 123) {
    sender
    text
  }
}

While the client is subscribed, whenever new messages are posted to chat room with ID "123", the selection for "sender" and "text" will be evaluated and published to the client, for example:

{
  "data": {
    "newMessage": {
      "sender": "Hagrid",
      "text": "You're a wizard!"
    }
  }
}

The "new message posted to chat room" could use a "Pub-Sub" system where the chat room ID is the "topic" and each "publish" contains the sender and text.

Supporting Subscriptions at Scale

Supporting subscriptions is a significant change for any GraphQL server. Query and mutation operations are stateless, allowing scaling via cloning of GraphQL server instances. Subscriptions, by contrast, are stateful and require maintaining the GraphQL document, variables, and other context over the lifetime of the subscription.

Consider the behavior of your system when state is lost due to the failure of a single machine in a service. Durability and availability may be improved by having separate dedicated services for managing subscription state and client connectivity.

Source Stream

A Source Stream represents the sequence of events, each of which will trigger a GraphQL execution corresponding to that event. Like field value resolution, the logic to create a Source Stream is application-specific.

CreateSourceEventStream(subscription, schema, variableValues, initialValue):

• Let {subscriptionType} be the root Subscription type in {schema}.
• Assert: {subscriptionType} is an Object type.
• Let {selectionSet} be the top level Selection Set in {subscription}.
• Let {rootField} be the first top level field in {selectionSet}.
• Let {argumentValues} be the result of {CoerceArgumentValues(subscriptionType, rootField, variableValues)}.
• Let {fieldStream} be the result of running {ResolveFieldEventStream(subscriptionType, initialValue, rootField, argumentValues)}.
• Return {fieldStream}.

ResolveFieldEventStream(subscriptionType, rootValue, fieldName, argumentValues):

• Let {resolver} be the internal function provided by {subscriptionType} for determining the resolved event stream of a subscription field named {fieldName}.
• Return the result of calling {resolver}, providing {rootValue} and {argumentValues}.

Note: This {ResolveFieldEventStream} algorithm is intentionally similar to {ResolveFieldValue} to enable consistency when defining resolvers on any operation type.

Response Stream

Each event in the underlying Source Stream triggers execution of the subscription selection set using that event as a root value.

MapSourceToResponseEvent(sourceStream, subscription, schema, variableValues):

• Return a new event stream {responseStream} which yields events as follows:
• For each {event} on {sourceStream}:
  • Let {response} be the result of running {ExecuteSubscriptionEvent(subscription, schema, variableValues, event)}.
  • Yield an event containing {response}.
• When {responseStream} completes: complete this event stream.

ExecuteSubscriptionEvent(subscription, schema, variableValues, initialValue):

• Let {subscriptionType} be the root Subscription type in {schema}.
• Assert: {subscriptionType} is an Object type.
• Let {selectionSet} be the top level Selection Set in {subscription}.
• Let {data} be the result of running {ExecuteSelectionSet(selectionSet, subscriptionType, initialValue, variableValues)} normally (allowing parallelization).
• Let {errors} be any field errors produced while executing the selection set.
• Return an unordered map containing {data} and {errors}.

Note: The {ExecuteSubscriptionEvent} algorithm is intentionally similar to {ExecuteQuery} since this is how the each event result is produced.

Unsubscribe

Unsubscribe cancels the Response Stream when a client no longer wishes to receive payloads for a subscription. This may in turn also cancel the Source Stream. This is also a good opportunity to clean up any other resources used by the subscription.

Unsubscribe(responseStream)

• Cancel {responseStream}

Executing Selection Sets

To execute a selection set, the object value being evaluated and the object type need to be known, as well as whether it must be executed serially, or may be executed in parallel.

First, the selection set is turned into a grouped field set; then, each represented field in the grouped field set produces an entry into a response map.

ExecuteSelectionSet(selectionSet, objectType, objectValue, variableValues):

• Let {groupedFieldSet} be the result of {CollectFields(objectType, selectionSet, variableValues)}.
• Initialize {resultMap} to an empty ordered map.
• For each {groupedFieldSet} as {responseKey} and {fields}:
  • Let {fieldName} be the name of the first entry in {fields}. Note: This value is unaffected if an alias is used.
  • Let {fieldType} be the return type defined for the field {fieldName} of {objectType}.
  • If {fieldType} is {null}:
    • Continue to the next iteration of {groupedFieldSet}.
  • Let {responseValue} be {ExecuteField(objectType, objectValue, fields, fieldType, variableValues)}.
  • Set {responseValue} as the value for {responseKey} in {resultMap}.
• Return {resultMap}.

Note: {resultMap} is ordered by which fields appear first in the query. This is explained in greater detail in the Field Collection section below.

Normal and Serial Execution

Normally the executor can execute the entries in a grouped field set in whatever order it chooses (often in parallel). Because the resolution of fields other than top-level mutation fields must always be side effect-free and idempotent, the execution order must not affect the result, and hence the server has the freedom to execute the field entries in whatever order it deems optimal.

For example, given the following grouped field set to be executed normally:

{
  birthday {
    month
  }
  address {
    street
  }
}

A valid GraphQL executor can resolve the four fields in whatever order it chose (however of course birthday must be resolved before month, and address before street).

When executing a mutation, the selections in the top most selection set will be executed in serial order.

When executing a grouped field set serially, the executor must consider each entry from the grouped field set in the order provided in the grouped field set. It must determine the corresponding entry in the result map for each item to completion before it continues on to the next item in the grouped field set:

For example, given the following selection set to be executed serially:

{
  changeBirthday(birthday: $newBirthday) {
    month
  }
  changeAddress(address: $newAddress) {
    street
  }
}

The executor must, in serial:

• Run {ExecuteField()} for changeBirthday, which during {CompleteValue()} will execute the { month } sub-selection set normally.
• Run {ExecuteField()} for changeAddress, which during {CompleteValue()} will execute the { street } sub-selection set normally.

As an illustrative example, let's assume we have a mutation field changeTheNumber that returns an object containing one field, theNumber. If we execute the following selection set serially:

{
  first: changeTheNumber(newNumber: 1) {
    theNumber
  }
  second: changeTheNumber(newNumber: 3) {
    theNumber
  }
  third: changeTheNumber(newNumber: 2) {
    theNumber
  }
}

The executor will execute the following serially:

• Resolve the changeTheNumber(newNumber: 1) field
• Execute the { theNumber } sub-selection set of first normally
• Resolve the changeTheNumber(newNumber: 3) field
• Execute the { theNumber } sub-selection set of second normally
• Resolve the changeTheNumber(newNumber: 2) field
• Execute the { theNumber } sub-selection set of third normally

A correct executor must generate the following result for that selection set:

{
  "first": {
    "theNumber": 1
  },
  "second": {
    "theNumber": 3
  },
  "third": {
    "theNumber": 2
  }
}

Field Collection

Before execution, the selection set is converted to a grouped field set by calling {CollectFields()}. Each entry in the grouped field set is a list of fields that share a response key. This ensures all fields with the same response key (alias or field name) included via referenced fragments are executed at the same time.

As an example, collecting the fields of this selection set would collect two instances of the field a and one of field b:

{
  a {
    subfield1
  }
  ...ExampleFragment
}

fragment ExampleFragment on Query {
  a {
    subfield2
  }
  b
}

The depth-first-search order of the field groups produced by {CollectFields()} is maintained through execution, ensuring that fields appear in the executed response in a stable and predictable order.

CollectFields(objectType, selectionSet, variableValues, visitedFragments):

• If {visitedFragments} if not provided, initialize it to the empty set.
• Initialize {groupedFields} to an empty ordered map of lists.
• For each {selection} in {selectionSet}:
  • If {selection} provides the directive @skip, let {skipDirective} be that directive.
    • If {skipDirective}'s {if} argument is {true} or is a variable in {variableValues} with the value {true}, continue with the next {selection} in {selectionSet}.
  • If {selection} provides the directive @include, let {includeDirective} be that directive.
    • If {includeDirective}'s {if} argument is not {true} and is not a variable in {variableValues} with the value {true}, continue with the next {selection} in {selectionSet}.
  • If {selection} is a {Field}:
    • Let {responseKey} be the response key of {selection}.
    • Let {groupForResponseKey} be the list in {groupedFields} for {responseKey}; if no such list exists, create it as an empty list.
    • Append {selection} to the {groupForResponseKey}.
  • If {selection} is a {FragmentSpread}:
    • Let {fragmentSpreadName} be the name of {selection}.
    • If {fragmentSpreadName} is in {visitedFragments}, continue with the next {selection} in {selectionSet}.
    • Add {fragmentSpreadName} to {visitedFragments}.
    • Let {fragment} be the Fragment in the current Document whose name is {fragmentSpreadName}.
    • If no such {fragment} exists, continue with the next {selection} in {selectionSet}.
    • Let {fragmentType} be the type condition on {fragment}.
    • If {DoesFragmentTypeApply(objectType, fragmentType)} is false, continue with the next {selection} in {selectionSet}.
    • Let {fragmentSelectionSet} be the top-level selection set of {fragment}.
    • Let {fragmentGroupedFieldSet} be the result of calling {CollectFields(objectType, fragmentSelectionSet, visitedFragments)}.
    • For each {fragmentGroup} in {fragmentGroupedFieldSet}:
      • Let {responseKey} be the response key shared by all fields in {fragmentGroup}
      • Let {groupForResponseKey} be the list in {groupedFields} for {responseKey}; if no such list exists, create it as an empty list.
      • Append all items in {fragmentGroup} to {groupForResponseKey}.
  • If {selection} is an {InlineFragment}:
    • Let {fragmentType} be the type condition on {selection}.
    • If {fragmentType} is not {null} and {DoesFragmentTypeApply(objectType, fragmentType)} is false, continue with the next {selection} in {selectionSet}.
    • Let {fragmentSelectionSet} be the top-level selection set of {selection}.
    • Let {fragmentGroupedFieldSet} be the result of calling {CollectFields(objectType, fragmentSelectionSet, variableValues, visitedFragments)}.
    • For each {fragmentGroup} in {fragmentGroupedFieldSet}:
      • Let {responseKey} be the response key shared by all fields in {fragmentGroup}
      • Let {groupForResponseKey} be the list in {groupedFields} for {responseKey}; if no such list exists, create it as an empty list.
      • Append all items in {fragmentGroup} to {groupForResponseKey}.
• Return {groupedFields}.

DoesFragmentTypeApply(objectType, fragmentType):

• If {fragmentType} is an Object Type:
  • if {objectType} and {fragmentType} are the same type, return {true}, otherwise return {false}.
• If {fragmentType} is an Interface Type:
  • if {objectType} is an implementation of {fragmentType}, return {true} otherwise return {false}.
• If {fragmentType} is a Union:
  • if {objectType} is a possible type of {fragmentType}, return {true} otherwise return {false}.

Executing Fields

Each field requested in the grouped field set that is defined on the selected objectType will result in an entry in the response map. Field execution first coerces any provided argument values, then resolves a value for the field, and finally completes that value either by recursively executing another selection set or coercing a scalar value.

ExecuteField(objectType, objectValue, fieldType, fields, variableValues):

• Let {field} be the first entry in {fields}.
• Let {argumentValues} be the result of {CoerceArgumentValues(objectType, field, variableValues)}
• Let {resolvedValue} be {ResolveFieldValue(objectType, objectValue, fieldName, argumentValues)}.
• Return the result of {CompleteValue(fieldType, fields, resolvedValue, variableValues)}.

Coercing Field Arguments

Fields may include arguments which are provided to the underlying runtime in order to correctly produce a value. These arguments are defined by the field in the type system to have a specific input type: Scalars, Enum, Input Object, or List or Non-Null wrapped variations of these three.

At each argument position in a query may be a literal value or a variable to be provided at runtime.

CoerceArgumentValues(objectType, field, variableValues):

• Let {coercedValues} be an empty unordered Map.
• Let {argumentValues} be the argument values provided in {field}.
• Let {fieldName} be the name of {field}.
• Let {argumentDefinitions} be the arguments defined by {objectType} for the field named {fieldName}.
• For each {argumentDefinition} in {argumentDefinitions}:
  • Let {argumentName} be the name of {argumentDefinition}.
  • Let {argumentType} be the expected type of {argumentDefinition}.
  • Let {defaultValue} be the default value for {argumentDefinition}.
  • Let {value} be the value provided in {argumentValues} for the name {argumentName}.
  • If {value} is a Variable:
    • Let {variableName} be the name of Variable {value}.
    • Let {variableValue} be the value provided in {variableValues} for the name {variableName}.
    • If {variableValue} exists (including {null}):
      • Add an entry to {coercedValues} named {argName} with the value {variableValue}.
    • Otherwise, if {defaultValue} exists (including {null}):
      • Add an entry to {coercedValues} named {argName} with the value {defaultValue}.
    • Otherwise, if {argumentType} is a Non-Nullable type, throw a field error.
    • Otherwise, continue to the next argument definition.
  • Otherwise, if {value} does not exist (was not provided in {argumentValues}:
    • If {defaultValue} exists (including {null}):
      • Add an entry to {coercedValues} named {argName} with the value {defaultValue}.
    • Otherwise, if {argumentType} is a Non-Nullable type, throw a field error.
    • Otherwise, continue to the next argument definition.
  • Otherwise, if {value} cannot be coerced according to the input coercion rules of {argType}, throw a field error.
  • Let {coercedValue} be the result of coercing {value} according to the input coercion rules of {argType}.
  • Add an entry to {coercedValues} named {argName} with the value {coercedValue}.
• Return {coercedValues}.

Note: Variable values are not coerced because they are expected to be coerced before executing the operation in {CoerceVariableValues()}, and valid queries must only allow usage of variables of appropriate types.

Value Resolution

While nearly all of GraphQL execution can be described generically, ultimately the internal system exposing the GraphQL interface must provide values. This is exposed via {ResolveFieldValue}, which produces a value for a given field on a type for a real value.

As an example, this might accept the {objectType} Person, the {field} {"soulMate"}, and the {objectValue} representing John Lennon. It would be expected to yield the value representing Yoko Ono.

ResolveFieldValue(objectType, objectValue, fieldName, argumentValues):

• Let {resolver} be the internal function provided by {objectType} for determining the resolved value of a field named {fieldName}.
• Return the result of calling {resolver}, providing {objectValue} and {argumentValues}.

Note: It is common for {resolver} to be asynchronous due to relying on reading an underlying database or networked service to produce a value. This necessitates the rest of a GraphQL executor to handle an asynchronous execution flow.

Value Completion

After resolving the value for a field, it is completed by ensuring it adheres to the expected return type. If the return type is another Object type, then the field execution process continues recursively.

CompleteValue(fieldType, fields, result, variableValues):

• If the {fieldType} is a Non-Null type:
  • Let {innerType} be the inner type of {fieldType}.
  • Let {completedResult} be the result of calling {CompleteValue(innerType, fields, result, variableValues)}.
  • If {completedResult} is {null}, throw a field error.
  • Return {completedResult}.
• If {result} is {null} (or another internal value similar to {null} such as {undefined} or {NaN}), return {null}.
• If {fieldType} is a List type:
  • If {result} is not a collection of values, throw a field error.
  • Let {innerType} be the inner type of {fieldType}.
  • Return a list where each list item is the result of calling {CompleteValue(innerType, fields, resultItem, variableValues)}, where {resultItem} is each item in {result}.
• If {fieldType} is a Scalar or Enum type:
  • Return the result of "coercing" {result}, ensuring it is a legal value of {fieldType}, otherwise {null}.
• If {fieldType} is an Object, Interface, or Union type:
  • If {fieldType} is an Object type.
    • Let {objectType} be {fieldType}.
  • Otherwise if {fieldType} is an Interface or Union type.
    • Let {objectType} be ResolveAbstractType({fieldType}, {result}).
  • Let {subSelectionSet} be the result of calling {MergeSelectionSets(fields)}.
  • Return the result of evaluating ExecuteSelectionSet(subSelectionSet, objectType, result, variableValues) normally (allowing for parallelization).

Resolving Abstract Types

When completing a field with an abstract return type, that is an Interface or Union return type, first the abstract type must be resolved to a relevant Object type. This determination is made by the internal system using whatever means appropriate.

Note: A common method of determining the Object type for an {objectValue} in object-oriented environments, such as Java or C#, is to use the class name of the {objectValue}.

ResolveAbstractType(abstractType, objectValue):

• Return the result of calling the internal method provided by the type system for determining the Object type of {abstractType} given the value {objectValue}.

Merging Selection Sets

When more than one fields of the same name are executed in parallel, their selection sets are merged together when completing the value in order to continue execution of the sub-selection sets.

An example query illustrating parallel fields with the same name with sub-selections.

{
  me {
    firstName
  }
  me {
    lastName
  }
}

After resolving the value for me, the selection sets are merged together so firstName and lastName can be resolved for one value.

MergeSelectionSets(fields):

• Let {selectionSet} be an empty list.
• For each {field} in {fields}:
  • Let {fieldSelectionSet} be the selection set of {field}.
  • If {fieldSelectionSet} is null or empty, continue to the next field.
  • Append all selections in {fieldSelectionSet} to {selectionSet}.
• Return {selectionSet}.

Errors and Non-Nullability

If an error is thrown while resolving a field, it should be treated as though the field returned {null}, and an error must be added to the {"errors"} list in the response.

If the result of resolving a field is {null} (either because the function to resolve the field returned {null} or because an error occurred), and that field is of a Non-Null type, then a field error is thrown. The error must be added to the {"errors"} list in the response.

If the field returns {null} because of an error which has already been added to the {"errors"} list in the response, the {"errors"} list must not be further affected. That is, only one error should be added to the errors list per field.

Since Non-Null type fields cannot be {null}, field errors are propagated to be handled by the parent field. If the parent field may be {null} then it resolves to {null}, otherwise if it is a Non-Null type, the field error is further propagated to it's parent field.

If all fields from the root of the request to the source of the error return Non-Null types, then the {"data"} entry in the response should be {null}.