pyinfra/config.yaml

service:
  logging_level: $LOGGING_LEVEL_ROOT|DEBUG  # Logging level for service logger
  name: $SERVICE_NAME|research  # Default service name for research service, used for prometheus metric name
  response_formatter: default  # formats analysis payloads of response messages
  upload_formatter: projecting  # formats analysis payloads of objects uploaded to storage
  # Note: This is not really the right place for this. It should be configured on a per-service basis.
  operation: $OPERATION|default
  # operation needs to be specified in deployment config for services that are called without an operation specified
  operations:
    conversion:
      input:
        multi: False
        subdir: ""
        extension: ORIGIN.pdf.gz
      output:
        subdir: "pages_as_images"
        extension: json.gz
    extraction:
      input:
        multi: False
        subdir: ""
        extension: ORIGIN.pdf.gz
      output:
        subdir: "extracted_images"
        extension: json.gz
    table_parsing:
      input:
        multi: True
        subdir: "pages_as_images"
        extension: json.gz
      output:
        subdir: "table_parses"
        extension: json.gz
    image_classification:
      input:
        multi: True
        subdir: "extracted_images"
        extension: json.gz
      output:
        subdir: ""
        extension: IMAGE_INFO.json.gz
    default:
      input:
        multi: False
        subdir: ""
        extension: in.gz
      output:
        subdir: ""
        extension: out.gz

probing_webserver:
  host: $PROBING_WEBSERVER_HOST|"0.0.0.0"  # Probe webserver address
  port: $PROBING_WEBSERVER_PORT|8080  # Probe webserver port
  mode: $PROBING_WEBSERVER_MODE|production  # webserver mode: {development, production}

rabbitmq:
  host: $RABBITMQ_HOST|localhost  # RabbitMQ host address
  port: $RABBITMQ_PORT|5672  # RabbitMQ host port
  user: $RABBITMQ_USERNAME|user  # RabbitMQ username
  password: $RABBITMQ_PASSWORD|bitnami  # RabbitMQ password
  heartbeat: $RABBITMQ_HEARTBEAT|7200  # Controls AMQP heartbeat timeout in seconds

  queues:
    input: $REQUEST_QUEUE|request_queue  # Requests to service
    output: $RESPONSE_QUEUE|response_queue  # Responses by service
    dead_letter: $DEAD_LETTER_QUEUE|dead_letter_queue  # Messages that failed to process

  callback:
    analysis_endpoint: $ANALYSIS_ENDPOINT|"http://127.0.0.1:5000"

storage:
  backend: $STORAGE_BACKEND|s3  # The type of storage to use {s3, azure}
  bucket: "STORAGE_BUCKET_NAME|STORAGE_AZURECONTAINERNAME|pyinfra-test-bucket"  # The bucket / container to pull files specified in queue requests from

  s3:
    endpoint: $STORAGE_ENDPOINT|"http://127.0.0.1:9000"
    access_key: $STORAGE_KEY|root
    secret_key: $STORAGE_SECRET|password
    region: $STORAGE_REGION|"eu-west-1"

  azure:
    connection_string: $STORAGE_AZURECONNECTIONSTRING|"DefaultEndpointsProtocol=https;AccountName=iqserdevelopment;AccountKey=4imAbV9PYXaztSOMpIyAClg88bAZCXuXMGJG0GA1eIBpdh2PlnFGoRBnKqLy2YZUSTmZ3wJfC7tzfHtuC6FEhQ==;EndpointSuffix=core.windows.net"

retry:
  tries: 3
  delay: 5
  jitter: [1, 3]