arazzo: 1.0.1
info:
  title: Microsoft Office Discover and Download a OneDrive File
  summary: Resolve the user's drive, browse the root, inspect an item, and download its content.
  description: >-
    The onboarding path for any OneDrive integration on Microsoft Graph. The
    workflow resolves the signed-in user's drive and its quota, lists the items
    in the drive root, inspects the target item to confirm it is a file rather
    than a folder, and then downloads the item's binary content. Every step
    spells out its request inline so the flow can be read and executed without
    opening the underlying OpenAPI description.
  version: 1.0.0
sourceDescriptions:
- name: microsoftGraphApi
  url: ../openapi/microsoft-office-openapi.yml
  type: openapi
workflows:
- workflowId: drive-file-discovery-download
  summary: Locate a file in the OneDrive root by name and download its content.
  description: >-
    Resolves the drive, enumerates root children, verifies the chosen item is a
    file, and retrieves its content stream. Graph answers the content request
    with a 302 redirect to a pre-authenticated download URL, or with the bytes
    directly, so both outcomes are treated as success.
  inputs:
    type: object
    required:
    - itemId
    properties:
      itemId:
        type: string
        description: >-
          The driveItem id to download. Resolve it from the root listing
          returned by the browseRoot step when it is not known ahead of time.
  steps:
  - stepId: resolveDrive
    description: >-
      Resolve the signed-in user's OneDrive to confirm access and capture the
      drive id, type, and remaining quota before transferring any content.
    operationId: getMyDrive
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      driveId: $response.body#/id
      driveType: $response.body#/driveType
      quota: $response.body#/quota
  - stepId: browseRoot
    description: >-
      List the items in the root of the drive so the caller can resolve a
      target file by name when only a human-readable name is known.
    operationId: listMyDriveRootChildren
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      rootItems: $response.body#/value
      firstItemId: $response.body#/value/0/id
      firstItemName: $response.body#/value/0/name
  - stepId: inspectItem
    description: >-
      Read the target driveItem's metadata to confirm it exists and carries a
      file facet, which distinguishes a downloadable file from a folder.
    operationId: getMyDriveItem
    parameters:
    - name: item_id
      in: path
      value: $inputs.itemId
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      itemId: $response.body#/id
      itemName: $response.body#/name
      itemSize: $response.body#/size
      file: $response.body#/file
      lastModifiedDateTime: $response.body#/lastModifiedDateTime
    onSuccess:
    - name: isFile
      type: goto
      stepId: downloadContent
      criteria:
      - context: $response.body
        condition: $.file
        type: jsonpath
    - name: isFolder
      type: end
      criteria:
      - context: $response.body
        condition: $.folder
        type: jsonpath
  - stepId: downloadContent
    description: >-
      Download the file's content. Graph either redirects to a short-lived
      pre-authenticated download URL or streams the bytes directly, so both a
      302 and a 200 are accepted as success.
    operationId: downloadMyDriveItemContent
    parameters:
    - name: item_id
      in: path
      value: $steps.inspectItem.outputs.itemId
    successCriteria:
    - condition: $statusCode == 200 || $statusCode == 302
    outputs:
      downloadStatus: $statusCode
  outputs:
    driveId: $steps.resolveDrive.outputs.driveId
    downloadedItemId: $steps.inspectItem.outputs.itemId
    downloadedItemName: $steps.inspectItem.outputs.itemName
    downloadedItemSize: $steps.inspectItem.outputs.itemSize