Resolver - Amazon Serverless Application Model
SyntaxPropertiesExamples
Services or capabilities described in Amazon Web Services documentation might vary by Region. To see the differences applicable to the China Regions, see Getting Started with Amazon Web Services in China (PDF).

Resolver

Configure resolvers for the fields of your GraphQL API. Amazon Serverless Application Model (Amazon SAM) supports JavaScript pipeline resolvers.

Syntax

To declare this entity in your Amazon Serverless Application Model (Amazon SAM) template, use the following syntax.

YAML

OperationType:
  LogicalId:
    Caching: CachingConfig
    CodeUri: String
    FieldName: String
    InlineCode: String 
    MaxBatchSize: Integer
    Pipeline: List
    Runtime: Runtime
    Sync: SyncConfig

Properties

Caching

The caching configuration for the resolver that has caching activated.

Type: CachingConfig

Required: No

Amazon CloudFormation compatibility: This property is passed directly to the CachingConfig property of an AWS::AppSync::Resolver resource.

CodeUri

The resolver function code’s Amazon Simple Storage Service (Amazon S3) URI or path to a local folder.

If you specify a path to a local folder, Amazon CloudFormation requires that the file is first uploaded to Amazon S3 before deployment. You can use the Amazon SAM CLI to facilitate this process. For more information, see Using the Amazon SAM CLI to upload local files at deployment.

If neither CodeUri or InlineCode are provided, Amazon SAM will generate InlineCode that redirects the request to the first pipeline function and receives the response from the last pipeline function.

Type: String

Required: No

Amazon CloudFormation compatibility: This property is passed directly to the CodeS3Location property of an AWS::AppSync::Resolver resource.

FieldName

The name of your resolver. Specify this property to override the LogicalId value.

Type: String

Required: No

Amazon CloudFormation compatibility: This property is passed directly to the FieldName property of an AWS::AppSync::Resolver resource.

InlineCode

The resolver code that contains the request and response functions.

If neither CodeUri or InlineCode are provided, Amazon SAM will generate InlineCode that redirects the request to the first pipeline function and receives the response from the last pipeline function.

Type: String

Required: No

Amazon CloudFormation compatibility: This property is passed directly to the Code property of an AWS::AppSync::Resolver resource.

LogicalId

The unique name for your resolver. In a GraphQL schema, your resolver name should match the field name that its used for. Use that same field name for LogicalId.

Type: String

Required: Yes

Amazon CloudFormation compatibility: This property is unique to Amazon SAM and doesn’t have an Amazon CloudFormation equivalent.

MaxBatchSize

The maximum number of resolver request inputs that will be sent to a single Amazon Lambda function in a BatchInvoke operation.

Type: Integer

Required: No

Amazon CloudFormation compatibility: This property is passed directly to the MaxBatchSize property of an AWS::AppSync::Resolver resource.

OperationType

The GraphQL operation type that is associated with your resolver. For example, Query, Mutation, or Subscription. You can nest multiple resolvers by LogicalId within a single OperationType.

Type: String

Required: Yes

Amazon CloudFormation compatibility: This property is passed directly to the TypeName property of an AWS::AppSync::Resolver resource.

Pipeline

Functions linked with the pipeline resolver. Specify functions by logical ID in a list.

Type: List

Required: Yes

Amazon CloudFormation compatibility: This property is unique to Amazon SAM and doesn't have an Amazon CloudFormation equivalent. It is similar to the PipelineConfig property of an AWS::AppSync::Resolver resource.

Runtime

The runtime of your pipeline resolver or function. Specifies the name and version to use.

Type: Runtime

Required: Yes

Amazon CloudFormation compatibility: This property is unique to Amazon SAM and doesn't have an Amazon CloudFormation equivalent. It is similar to the Runtime property of an AWS::AppSync::Resolver resource.

Sync

Describes a Sync configuration for a resolver.

Specifies which Conflict Detection strategy and Resolution strategy to use when the resolver is invoked.

Type: SyncConfig

Required: No

Amazon CloudFormation compatibility: This property is passed directly to the SyncConfig property of an AWS::AppSync::Resolver resource.

Examples

Use the Amazon SAM generated resolver function code and save fields as variables

Here is the GraphQL schema for our example:

schema {
  query: Query
  mutation: Mutation
}

type Query {
  getPost(id: ID!): Post
}

type Mutation {
  addPost(author: String!, title: String!, content: String!): Post!
}

type Post {
  id: ID!
  author: String
  title: String
  content: String
}

Here is a snippet of our Amazon SAM template:

AWSTemplateFormatVersion: '2010-09-09'
Transform: AWS::Serverless-2016-10-31
...
Resources:
  MyGraphQLApi:
    Type: AWS::Serverless::GraphQLApi
    Properties:
      ...
      Functions:
        preprocessPostItem:
          ...
        createPostItem:
          ...
      Resolvers:
        Mutation:
          addPost:
            Runtime:
              Name: APPSYNC_JS
              Version: 1.0.0
            Pipeline:
            - preprocessPostItem
            - createPostItem

In our Amazon SAM template, we don’t specify CodeUri or InlineCode. At deployment, Amazon SAM automatically generates the following inline code for our resolver:

export function request(ctx) {
  return {};
}

export function response(ctx) {
  return ctx.prev.result;
}

This default resolver code redirects the request to the first pipeline function and receives the response from the last pipeline function.

In our first pipeline function, we can use the provided args field to parse the request object and create our variables. We can then use these variables within our function. Here is an example of our preprocessPostItem function:

import { util } from "@aws-appsync/utils";

export function request(ctx) {
  const author = ctx.args.author;
  const title = ctx.args.title;
  const content = ctx.args.content;
  
  // Use variables to process data
  
}

export function response(ctx) {
  return ctx.result;
}