Databricks Run Job, Wait, Then Export the Notebook

arazzo: 1.0.1
info:
  title: Databricks Run Job, Wait, Then Export the Notebook
  summary: Trigger a job run, wait for it to finish, then export the source notebook.
  description: >-
    Runs a Databricks job, polls the run until it is TERMINATED, and then exports
    the source notebook that backs the job so the exact executed source can be
    archived alongside the run. The run_id drives the poll loop and the supplied
    notebook path drives the export. 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: jobsApi
  url: ../openapi/databricks-jobs-api-openapi.yml
  type: openapi
- name: workspaceApi
  url: ../openapi/databricks-workspace-api-openapi.yml
  type: openapi
workflows:
- workflowId: run-job-and-export-notebook
  summary: Run a job to completion, then export the backing notebook.
  description: >-
    Triggers the job, polls until the run is TERMINATED, then exports the source
    notebook content for archival.
  inputs:
    type: object
    required:
    - job_id
    - notebook_path
    properties:
      job_id:
        type: integer
        description: The job to run.
      notebook_path:
        type: string
        description: The workspace path of the notebook to export afterward.
  steps:
  - stepId: runJobNow
    description: >-
      Trigger an immediate run of the job and capture the run_id.
    operationId: runJobNow
    requestBody:
      contentType: application/json
      payload:
        job_id: $inputs.job_id
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      runId: $response.body#/run_id
  - stepId: pollRun
    description: >-
      Read the run and inspect the life cycle state. Loop while RUNNING or
      PENDING; continue once TERMINATED.
    operationId: getJobRun
    parameters:
    - name: run_id
      in: query
      value: $steps.runJobNow.outputs.runId
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      lifeCycleState: $response.body#/state/life_cycle_state
      resultState: $response.body#/state/result_state
    onSuccess:
    - name: stillRunning
      type: goto
      stepId: pollRun
      criteria:
      - context: $response.body
        condition: $.state.life_cycle_state == "RUNNING"
        type: jsonpath
    - name: stillPending
      type: goto
      stepId: pollRun
      criteria:
      - context: $response.body
        condition: $.state.life_cycle_state == "PENDING"
        type: jsonpath
    - name: terminated
      type: goto
      stepId: exportNotebook
      criteria:
      - context: $response.body
        condition: $.state.life_cycle_state == "TERMINATED"
        type: jsonpath
  - stepId: exportNotebook
    description: >-
      Export the source notebook backing the job in SOURCE format for archival
      next to the completed run.
    operationId: exportWorkspaceObject
    parameters:
    - name: path
      in: query
      value: $inputs.notebook_path
    - name: format
      in: query
      value: SOURCE
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      content: $response.body#/content
  outputs:
    runId: $steps.runJobNow.outputs.runId
    resultState: $steps.pollRun.outputs.resultState
    notebookContent: $steps.exportNotebook.outputs.content