arazzo: 1.0.1 info: title: Power BI Import a PBIX and Poll Until Published summary: Upload a PBIX file into a workspace, wait for the import to finish publishing, and list the resulting reports and datasets. description: >- The deployment pattern for shipping Power BI Desktop content into a workspace from CI. The import is a multipart upload that returns either 200 or 202 with an import record in the Publishing state, and publishing then continues asynchronously. This API surface exposes no get-import-by-id operation, so completion is observed by listing the workspace's imports and waiting until none are still Publishing, after which the reports and datasets the import produced are enumerated. 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: powerBiApi url: ../openapi/power-bi-rest-api-openapi.yml type: openapi workflows: - workflowId: pbix-import-publish summary: Publish a PBIX into a workspace and wait for it to finish importing. description: >- Uploads a PBIX file to a workspace, polls the workspace import list until publishing settles, and enumerates the reports and datasets that resulted. inputs: type: object required: - groupId - datasetDisplayName - file properties: groupId: type: string description: The identifier of the workspace to import into. datasetDisplayName: type: string description: The display name for the imported file, including the .pbix extension. file: type: string format: binary description: The PBIX file content to upload. nameConflict: type: string description: >- How to handle an existing item with the same name. One of Ignore, Abort, Overwrite, or CreateOrOverwrite. default: Abort steps: - stepId: startImport description: >- Upload the PBIX as multipart form data. The service accepts the file and returns an import record that begins life in the Publishing state. operationId: createImportInGroup parameters: - name: groupId in: path value: $inputs.groupId - name: datasetDisplayName in: query value: $inputs.datasetDisplayName - name: nameConflict in: query value: $inputs.nameConflict requestBody: contentType: multipart/form-data payload: file: $inputs.file successCriteria: - condition: $statusCode == 200 || $statusCode == 202 outputs: importId: $response.body#/id importName: $response.body#/name importState: $response.body#/importState - stepId: pollImport description: >- List the workspace's imports and wait until none remain in the Publishing state. This API exposes no get-import-by-id operation, so the workspace import list is the only available completion signal. operationId: getImportsInGroup parameters: - name: groupId in: path value: $inputs.groupId successCriteria: - condition: $statusCode == 200 - context: $response.body condition: $.value[?(@.importState == 'Publishing')].length == 0 type: jsonpath onFailure: - name: importStillPublishing type: retry stepId: pollImport retryAfter: 15 retryLimit: 40 outputs: imports: $response.body#/value - stepId: listPublishedDatasets description: >- Enumerate the workspace datasets so the dataset the import created can be matched by name and handed to a refresh or rebind flow. operationId: getDatasetsInGroup parameters: - name: groupId in: path value: $inputs.groupId successCriteria: - condition: $statusCode == 200 outputs: datasets: $response.body#/value - stepId: listPublishedReports description: >- Enumerate the workspace reports so the report the import created can be matched by name and its embed URL captured. operationId: getReportsInGroup parameters: - name: groupId in: path value: $inputs.groupId successCriteria: - condition: $statusCode == 200 outputs: reports: $response.body#/value outputs: importId: $steps.startImport.outputs.importId imports: $steps.pollImport.outputs.imports datasets: $steps.listPublishedDatasets.outputs.datasets reports: $steps.listPublishedReports.outputs.reports