comprehend-semi-structured-documents-annotation-template.yml

AWSTemplateFormatVersion: '2010-09-09'
Transform: AWS::Serverless-2016-10-31

Metadata:
  AWS::ServerlessRepo::Application:
    Name: ComprehendSemiStructuredDocumentsAnnotationLambda
    Description: Deploys a Lambda which create resources to start GT SemiStructured Documents labeling job.
    Author: AWS Comprehend
    ReadmeUrl: README.md
    Labels: [serverless]
    HomePageUrl: https://aws.amazon.com/comprehend/
    SemanticVersion: 1.0.0
    SourceCodeUrl: https://github.com/aws-samples/amazon-comprehend-examples

Parameters:
  PreHumanLambdaTimeoutInSeconds:
    Type: Number
    Default: 300
    MaxValue: 900
    Description: Timeout in seconds for PreHumanLambba. Default is 300.
  ConsolidationLambdaTimeoutInSeconds:
    Type: Number
    Default: 300
    MaxValue: 900
    Description: Timeout in seconds for ConsolidationLambba. Default is 300.

Resources:
  SemiStructuredDocumentsS3Bucket:
    Type: AWS::S3::Bucket
    Properties:
      BucketName: 
        Fn::Sub: comprehend-semi-structured-documents-${AWS::Region}-${AWS::AccountId}
      VersioningConfiguration:
        Status: Enabled
      BucketEncryption: 
        ServerSideEncryptionConfiguration: 
        - ServerSideEncryptionByDefault:
            SSEAlgorithm: AES256
      CorsConfiguration:
        CorsRules:
          - AllowedHeaders: []
            AllowedMethods: [GET]
            AllowedOrigins: ['*']

  SemiStructuredDocumentsS3BucketPolicy:
    Type: AWS::S3::BucketPolicy
    Properties:
      Bucket:
        Ref: SemiStructuredDocumentsS3Bucket
      PolicyDocument:
        Version: "2012-10-17"
        Statement:
          - Sid: "S3GetObjectGT"
            Effect: "Allow"
            Principal:
              AWS:
                Fn::Sub: "arn:aws:iam::${AWS::AccountId}:root"
            Action:
              - s3:GetObject
            Resource:
              Fn::Sub: "arn:aws:s3:::${SemiStructuredDocumentsS3Bucket}/*"

  BucketAccessPolicy:
    Type: AWS::IAM::Policy
    Properties:
      Roles:
        - Ref: SageMakerExecutionRole
        - Ref: GTPreHumanTaskLambdaExecutionRole
        - Ref: GTAnnotationConsolidationLambdaExecutionRole
      PolicyName: BucketAccessPolicy
      PolicyDocument:
        Version: 2012-10-17
        Statement:
          - Effect: Allow
            Action:
              - s3:ListBucket
            Resource:
              - !Sub 'arn:aws:s3:::${SemiStructuredDocumentsS3Bucket}'
          - Effect: "Allow"
            Action:
              - s3:PutObject
              - s3:GetObject
            Resource:
              - !Sub 'arn:aws:s3:::${SemiStructuredDocumentsS3Bucket}/*'

  SageMakerExecutionRole:
    Type: AWS::IAM::Role
    Properties:
      AssumeRolePolicyDocument:
        Version: 2012-10-17
        Statement:
          - Effect: Allow
            Principal:
              Service: lambda.amazonaws.com
              AWS: !GetAtt GTAnnotationConsolidationLambdaExecutionRole.Arn
            Action: sts:AssumeRole
          - Effect: Allow
            Principal:
              Service: sagemaker.amazonaws.com
            Action: sts:AssumeRole
      ManagedPolicyArns:
        - arn:aws:iam::aws:policy/AmazonS3FullAccess
        - arn:aws:iam::aws:policy/AmazonSageMakerFullAccess
      Policies:
        - PolicyName: LambdaInvokePolicy
          PolicyDocument:
            Version: 2012-10-17
            Statement:
              - Effect: Allow
                Action:
                  - lambda:InvokeFunction
                Resource:
                  - !GetAtt GTPreHumanTaskLambdaFunction.Arn
                  - !GetAtt GTAnnotationConsolidationLambdaFunction.Arn

  GTPreHumanTaskLambdaExecutionRole:
    Type: AWS::IAM::Role
    Properties:
      AssumeRolePolicyDocument:
        Version: 2012-10-17
        Statement:
          - Effect: Allow
            Principal:
              Service: lambda.amazonaws.com
            Action: sts:AssumeRole
      ManagedPolicyArns:
        - arn:aws:iam::aws:policy/service-role/AWSLambdaBasicExecutionRole
        - arn:aws:iam::aws:policy/AmazonSageMakerFullAccess
      Policies:
        - PolicyName: TextractPolicy
          PolicyDocument:
            Version: 2012-10-17
            Statement:
              - Effect: Allow
                Action:
                  - textract:AnalyzeDocument
                  - textract:DetectDocumentText
                Resource: '*'

  GTAnnotationConsolidationLambdaExecutionRole:
    Type: AWS::IAM::Role
    Properties:
      AssumeRolePolicyDocument:
        Version: 2012-10-17
        Statement:
          - Effect: Allow
            Principal:
              Service: lambda.amazonaws.com
            Action: sts:AssumeRole
      ManagedPolicyArns:
        - arn:aws:iam::aws:policy/service-role/AWSLambdaBasicExecutionRole

  GTPreHumanTaskLambdaFunction:
    Type: AWS::Serverless::Function
    Properties:
      Handler: pre_human_task_lambda.lambda_handler
      MemorySize: 10240
      Role:
        Fn::GetAtt: [GTPreHumanTaskLambdaExecutionRole, Arn]
      CodeUri: src
      Layers:
        - !Ref PopplerBinariesLayer
      Runtime: python3.8
      Timeout: !Ref PreHumanLambdaTimeoutInSeconds

  GTAnnotationConsolidationLambdaFunction:
    Type: AWS::Serverless::Function
    Properties:
      Handler: annotation_consolidation_lambda.lambda_handler
      MemorySize: 10240
      Role: !GetAtt GTAnnotationConsolidationLambdaExecutionRole.Arn
      CodeUri: src
      Runtime: python3.8
      Timeout: !Ref ConsolidationLambdaTimeoutInSeconds

  PopplerBinariesLayer:
    Type: AWS::Serverless::LayerVersion
    Properties:
      LayerName: PopplerBinariesLayer
      Description: Poppler binaries needed to extract PDF
      ContentUri: Layers/Poppler/
      CompatibleRuntimes:
        - python3.8
      LicenseInfo: 'GPLv2'

Outputs:
  GTPreHumanTaskLambdaFunctionName:
    Description: Gt Pre Human Task Lambda Function Name.
    Value: !GetAtt GTPreHumanTaskLambdaFunction.Arn

  GTAnnotationConsolidationLambdaFunctionName:
    Description: Gt Annotation Consolidation Lambda Function Name.
    Value: !GetAtt GTAnnotationConsolidationLambdaFunction.Arn

  SemiStructuredDocumentsS3Bucket:
    Description: S3 Bucket that stores the manifests and output data for the Amazon SageMaker Ground Truth job.
    Value: !Ref SemiStructuredDocumentsS3Bucket

  SageMakerRoleARN:
    Description: IAM Role used assumed by Amazon SageMaker to invoke pre- and post-labeling functions during the Amazon SageMaker Ground Truth Job.
    Value: !GetAtt SageMakerExecutionRole.Arn

  GTPreHumanTaskLambdaRoleARN:
    Description: Role for GT Pre Human Task function lambda.
    Value: !GetAtt GTPreHumanTaskLambdaExecutionRole.Arn

  GTAnnotationConsolidationLambdaRoleARN:
    Description: Role for GT Annotation Consolidation function lambda.
    Value: !GetAtt GTAnnotationConsolidationLambdaExecutionRole.Arn