arazzo: 1.0.1
info:
  title: SharePoint Upload a Document and Verify It
  summary: Survey a library folder, upload a file with overwrite, then read the bytes back.
  description: >-
    The document management flow that matters most in practice. The upload
    endpoint hardcodes overwrite=true, so the workflow deliberately surveys the
    folder first — the caller sees exactly what is about to be replaced instead
    of discovering it afterwards. After the upload it downloads the file back
    from its server-relative URL, which is the only real proof the bytes are
    retrievable rather than merely accepted. 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: sharepointRestApi
  url: ../openapi/sharepoint-rest-api.yaml
  type: openapi
workflows:
- workflowId: upload-document
  summary: Upload a file into a SharePoint document library and verify it downloads.
  description: >-
    Lists the target folder to reveal any file that would be overwritten,
    uploads the document, and downloads it back to confirm the round trip.
  inputs:
    type: object
    required:
    - folderUrl
    - fileName
    - fileContent
    properties:
      folderUrl:
        type: string
        description: >-
          Server-relative folder URL, e.g. "/sites/mysite/Shared Documents".
      fileName:
        type: string
        description: File name to write, e.g. "quarterly-report.docx".
      fileContent:
        type: string
        format: binary
        description: Raw bytes of the file to upload.
  steps:
  - stepId: surveyFolder
    description: >-
      List the folder before writing. The upload path carries overwrite=true, so
      this is the caller's only chance to see what an existing file of the same
      name looks like before it is replaced.
    operationId: getFilesInFolder
    parameters:
    - name: folder_url
      in: path
      value: $inputs.folderUrl
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      existingFiles: $response.body#/value
      firstExistingName: $response.body#/value/0/Name
  - stepId: uploadDocument
    description: >-
      Upload the document as a raw octet-stream. The operation overwrites any
      file already at this name and returns the created file's metadata,
      including the server-relative URL used to read it back.
    operationId: uploadFile
    parameters:
    - name: folder_url
      in: path
      value: $inputs.folderUrl
    - name: file_name
      in: path
      value: $inputs.fileName
    requestBody:
      contentType: application/octet-stream
      payload: $inputs.fileContent
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      serverRelativeUrl: $response.body#/ServerRelativeUrl
      uniqueId: $response.body#/UniqueId
      length: $response.body#/Length
      majorVersion: $response.body#/MajorVersion
      timeLastModified: $response.body#/TimeLastModified
  - stepId: verifyDownload
    description: >-
      Download the file back from the server-relative URL the upload returned.
      A 200 here is the proof that the document is stored and retrievable, not
      merely that the write was accepted.
    operationId: downloadFile
    parameters:
    - name: file_url
      in: path
      value: $steps.uploadDocument.outputs.serverRelativeUrl
    successCriteria:
    - condition: $statusCode == 200
  outputs:
    serverRelativeUrl: $steps.uploadDocument.outputs.serverRelativeUrl
    uniqueId: $steps.uploadDocument.outputs.uniqueId
    length: $steps.uploadDocument.outputs.length
    majorVersion: $steps.uploadDocument.outputs.majorVersion