Sensible Extract From URL Poll Then CSV

arazzo: 1.0.1
info:
  title: Sensible Extract From URL Poll Then CSV
  summary: Asynchronously extract a document from a URL, poll until complete, then export the extraction as a CSV file.
  description: >-
    Combines the asynchronous extraction pattern with a downstream CSV export.
    The workflow submits a document URL for extraction under a chosen document
    type and config, polls the Retrieve extraction by ID endpoint until the
    status is COMPLETE, and then converts the finished extraction to a
    comma-separated values file, returning the time-limited download URL. The
    CSV export must run after the extraction completes, which the poll loop
    guarantees. Every step spells out its request inline, including the Bearer
    authorization.
  version: 1.0.0
sourceDescriptions:
- name: documentApi
  url: ../openapi/sensible-so-document-api-openapi.yml
  type: openapi
- name: getExcelFromDocumentsApi
  url: ../openapi/sensible-so-get-excel-from-documents-api-openapi.yml
  type: openapi
- name: retrieveExtractionsApi
  url: ../openapi/sensible-so-retrieve-extractions-api-openapi.yml
  type: openapi
workflows:
- workflowId: extract-from-url-poll-then-csv
  summary: Asynchronously extract a document from a URL, poll to completion, then export it as CSV.
  description: >-
    Submits a document URL for extraction, polls the extraction id until
    COMPLETE, then converts the extraction to a CSV file and returns the
    download URL.
  inputs:
    type: object
    required:
    - apiKey
    - documentType
    - configName
    - documentUrl
    properties:
      apiKey:
        type: string
        description: Sensible API key used as the Bearer token.
      documentType:
        type: string
        description: The document type to extract from.
      configName:
        type: string
        description: The config to use for extraction.
      documentUrl:
        type: string
        description: A publicly accessible or presigned URL returning the document bytes.
  steps:
  - stepId: submitExtraction
    description: >-
      Submit the document URL for asynchronous extraction under the chosen
      document type and config, and capture the returned extraction id.
    operationId: provide-a-download-url-with-config
    parameters:
    - name: Authorization
      in: header
      value: "Bearer $inputs.apiKey"
    - name: document_type
      in: path
      value: $inputs.documentType
    - name: config_name
      in: path
      value: $inputs.configName
    requestBody:
      contentType: application/json
      payload:
        document_url: $inputs.documentUrl
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      extractionId: $response.body#/id
  - stepId: pollStatus
    description: >-
      Poll the extraction by id until Sensible reports the COMPLETE status,
      retrying while the extraction is still WAITING or PROCESSING.
    operationId: retrieving-results
    parameters:
    - name: Authorization
      in: header
      value: "Bearer $inputs.apiKey"
    - name: id
      in: path
      value: $steps.submitExtraction.outputs.extractionId
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      status: $response.body#/status
    onSuccess:
    - name: extractionComplete
      type: goto
      stepId: exportCsv
      criteria:
      - context: $response.body
        condition: $.status == "COMPLETE"
        type: jsonpath
    - name: keepPolling
      type: goto
      stepId: pollStatus
      criteria:
      - context: $response.body
        condition: $.status == "WAITING" || $.status == "PROCESSING"
        type: jsonpath
  - stepId: exportCsv
    description: >-
      Convert the completed extraction to a CSV file and return the time-limited
      download URL.
    operationId: get-csv-extraction
    parameters:
    - name: Authorization
      in: header
      value: "Bearer $inputs.apiKey"
    - name: ids
      in: path
      value: $steps.submitExtraction.outputs.extractionId
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      downloadUrl: $response.body#/url
  outputs:
    extractionId: $steps.submitExtraction.outputs.extractionId
    downloadUrl: $steps.exportCsv.outputs.downloadUrl