# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved.
# SPDX-License-Identifier: MIT-0
AWSTemplateFormatVersion: "2010-09-09"
Transform: AWS::Serverless-2016-10-31
Description: "App-s3-athena-dataprocessing"

Mappings:
  S3Locations:
    DataStore1:
      Current:    "datasources/datastore1/current/"
      Historical: "datasources/datastore1/historical/"
    DataStore2:
      Current:    "datasources/datastore2/current/"
      Historical: "datasources/datastore2/historical/"
    DataStore3:
      Current:    "datasources/datastore3/current/"
      Historical: "datasources/datastore3/historical/"


Parameters:
  ParameterServiceCode:
    Type: String
    Description: "Code representing this service"
  ParameterEnvCode:
    Type: String
    Description: "Code to represent the environment for uniqueness"


Resources:
############### Test Infrastructure ###############################################
  # Define a common IAM role to be used for all components of this app
  ApplicationRole:
    Type: AWS::IAM::Role
    Properties:
      AssumeRolePolicyDocument: 
        Version: "2012-10-17"
        Statement: 
          - 
            Effect: "Allow"
            Principal: 
              Service: 
                - "states.amazonaws.com"
                - "lambda.amazonaws.com"
                - "glue.amazonaws.com"
            Action: 
              - "sts:AssumeRole"
      Policies:
        - PolicyName: AppPolicy
          PolicyDocument:
            Version: 2012-10-17
            Statement:
              -
                Effect: Allow
                Action:
                - events:PutTargets
                - events:PutRule
                - events:DescribeRule
                - states:StartExecution
                - xray:PutTraceSegments
                - xray:PutTelemetryRecords
                - xray:GetSamplingRules
                - xray:GetSamplingTargets
                - logs:CreateLogDelivery
                - logs:GetLogDelivery
                - logs:UpdateLogDelivery
                - logs:DeleteLogDelivery
                - logs:ListLogDeliveries
                - logs:PutResourcePolicy
                - logs:DescribeResourcePolicies
                - logs:DescribeLogGroups
                - cloudwatch:PutMetricData
                Resource: '*'
              -
                Effect: Allow
                Action:
                - lambda:InvokeFunction
                Resource: '*'
              -
                Effect: Allow
                Action:
                - s3:*
                Resource: "*"
              - Effect: Allow
                Action: athena:*
                Resource: '*'
              - Effect: Allow
                Action:
                  - glue:*
                  - iam:ListRolePolicies
                  - iam:GetRole
                  - iam:GetRolePolicy
                Resource: '*'
      ManagedPolicyArns: 
        - arn:aws:iam::aws:policy/service-role/AWSLambdaBasicExecutionRole

  # An S3 bucket to store data being processed
  DataBucket:
    Type: AWS::S3::Bucket
    DeletionPolicy: Retain
    UpdateReplacePolicy: Retain
    Properties:
      BucketName : !Join ["",["serivcedata-",!Ref ParameterServiceCode, "-",!Ref ParameterEnvCode]]
      AccessControl: Private
      BucketEncryption:
        ServerSideEncryptionConfiguration:
          - ServerSideEncryptionByDefault:
              SSEAlgorithm: AES256
      VersioningConfiguration:
        Status: Enabled
  
  AthenaResultsBucket:
    Type: AWS::S3::Bucket
    DeletionPolicy: Retain
    UpdateReplacePolicy: Retain
    Properties:
      BucketName : !Join ["",["serivcedata-",!Ref ParameterServiceCode, "-",!Ref ParameterEnvCode,"-athena-results"]]
      AccessControl: Private
      BucketEncryption:
        ServerSideEncryptionConfiguration:
          - ServerSideEncryptionByDefault:
              SSEAlgorithm: AES256 
      VersioningConfiguration:
        Status: Enabled  

  LogGroupStateMachines:
    Type: AWS::Logs::LogGroup
    Properties:
      LogGroupName: !Join ["",["/aws/states/serivcedata-",!Ref ParameterServiceCode, "-",!Ref ParameterEnvCode, "-StateMachineLogs"]]
  
  StateMachineInitializeAccount:
    Type: AWS::Serverless::StateMachine
    Properties:
      DefinitionUri: statemachines/initialize_account.asl.json
      DefinitionSubstitutions:
        LambdaCreateS3Folder: !GetAtt LambdaCreateS3Folder.Arn
        ParameterFolderDataStore1Current: !FindInMap ["S3Locations","DataStore1","Current"]
        ParameterFolderDataStore1Historical: !FindInMap ["S3Locations","DataStore1","Historical"]
        ParameterFolderDataStore2Current: !FindInMap ["S3Locations","DataStore2","Current"]
        ParameterFolderDataStore2Historical: !FindInMap ["S3Locations","DataStore2","Historical"]
        ParameterFolderDataStore3Current: !FindInMap ["S3Locations","DataStore3","Current"]
        ParameterFolderDataStore3Historical:  !FindInMap ["S3Locations","DataStore3","Historical"]
        LambdaListGlueCrawlers: !GetAtt LambdaListGlueCrawlers.Arn
        GlueCrawlerPrefix: !Join ["",[!Ref ParameterServiceCode, "-",!Ref ParameterEnvCode]]
        LambdaStartGlueCrawler: !GetAtt LambdaStartGlueCrawler.Arn
      Tracing:
        Enabled: true
      Role: !GetAtt ApplicationRole.Arn
      Logging:
        Destinations:
          - CloudWatchLogsLogGroup:
              LogGroupArn: !GetAtt  LogGroupStateMachines.Arn
        IncludeExecutionData: TRUE
        Level: "ALL"
      Type: "STANDARD"
      Name: !Join ["",[!Ref ParameterServiceCode, "-",!Ref ParameterEnvCode,"-InitializeAccount"]]

  LambdaCreateS3Folder:
    Type: AWS::Serverless::Function
    Properties:
      CodeUri: functions/create_s3_folder/
      Handler: app.lambda_handler
      Runtime: python3.8
      Environment:
        Variables:
          bucketname: !Ref DataBucket
      Role: !GetAtt ApplicationRole.Arn
      Timeout: 30

  LambdaStartGlueCrawler:
    Type: AWS::Serverless::Function
    Properties:
      CodeUri: functions/start_glue_crawler/
      Handler: app.lambda_handler
      Runtime: python3.8
      Role: !GetAtt ApplicationRole.Arn
      Timeout: 30
  
  LambdaListGlueCrawlers:
    Type: AWS::Serverless::Function
    Properties:
      CodeUri: functions/list_glue_crawlers/
      Handler: app.lambda_handler
      Runtime: python3.8
      Role: !GetAtt ApplicationRole.Arn
      Timeout: 30


  GlueCrawlerDataStore1Current:
    Type: AWS::Glue::Crawler
    Properties: 
      DatabaseName: !Join ["",[!Ref ParameterServiceCode, "-",!Ref ParameterEnvCode]]
      Name: !Join ["",[!Ref ParameterServiceCode, "-",!Ref ParameterEnvCode,"-DataStore1-Current"]]
      Role: !GetAtt ApplicationRole.Arn
      Targets: 
        S3Targets:
          - Path: !Join ["",["s3://",!Ref DataBucket ,"/",!FindInMap ["S3Locations","DataStore1","Current"]]]
      TablePrefix: "datastore1_"
  GlueCrawlerDataStore1Historical:
    Type: AWS::Glue::Crawler
    Properties: 
      DatabaseName: !Join ["",[!Ref ParameterServiceCode, "-",!Ref ParameterEnvCode]]
      Name: !Join ["",[!Ref ParameterServiceCode, "-",!Ref ParameterEnvCode,"-DataStore1-Historical"]]
      Role: !GetAtt ApplicationRole.Arn
      Targets: 
        S3Targets:
          - Path: !Join ["",["s3://",!Ref DataBucket ,"/",!FindInMap ["S3Locations","DataStore1","Historical"]]]
      TablePrefix: "datastore1_"

  GlueCrawlerDataStore2Current:
    Type: AWS::Glue::Crawler
    Properties: 
      DatabaseName: !Join ["",[!Ref ParameterServiceCode, "-",!Ref ParameterEnvCode]]
      Name: !Join ["",[!Ref ParameterServiceCode, "-",!Ref ParameterEnvCode,"-DataStore2-Current"]]
      Role: !GetAtt ApplicationRole.Arn
      Targets: 
        S3Targets:
          - Path: !Join ["",["s3://",!Ref DataBucket ,"/",!FindInMap ["S3Locations","DataStore2","Current"]]]
      TablePrefix: "datastore2_"
  GlueCrawlerDataStore2Historical:
    Type: AWS::Glue::Crawler
    Properties: 
      DatabaseName: !Join ["",[!Ref ParameterServiceCode, "-",!Ref ParameterEnvCode]]
      Name: !Join ["",[!Ref ParameterServiceCode, "-",!Ref ParameterEnvCode,"-DataStore2-Historical"]]
      Role: !GetAtt ApplicationRole.Arn
      Targets: 
        S3Targets:
          - Path: !Join ["",["s3://",!Ref DataBucket ,"/",!FindInMap ["S3Locations","DataStore2","Historical"]]]
      TablePrefix: "datastore2_"

  GlueCrawlerDataStore3Current:
    Type: AWS::Glue::Crawler
    Properties: 
      DatabaseName: !Join ["",[!Ref ParameterServiceCode, "-",!Ref ParameterEnvCode]]
      Name: !Join ["",[!Ref ParameterServiceCode, "-",!Ref ParameterEnvCode,"-DataStore3-Current"]]
      Role: !GetAtt ApplicationRole.Arn
      Targets: 
        S3Targets:
          - Path: !Join ["",["s3://",!Ref DataBucket ,"/",!FindInMap ["S3Locations","DataStore3","Current"]]]
      TablePrefix: "datastore3_"
  GlueCrawlerDataStore3Historical:
    Type: AWS::Glue::Crawler
    Properties: 
      DatabaseName: !Join ["",[!Ref ParameterServiceCode, "-",!Ref ParameterEnvCode]]
      Name: !Join ["",[!Ref ParameterServiceCode, "-",!Ref ParameterEnvCode,"-DataStore3-Historical"]]
      Role: !GetAtt ApplicationRole.Arn
      Targets: 
        S3Targets:
          - Path: !Join ["",["s3://",!Ref DataBucket ,"/",!FindInMap ["S3Locations","DataStore3","Historical"]]]
      TablePrefix: "datastore3_"
   

Outputs:
  DataBucket:
    Description: "Bucket containing data"
    Value: !Ref DataBucket
  StateMachineInitializeAccount:
    Description: "StateMachine for setting up the account"
    Value: !Ref StateMachineInitializeAccount