Inclusion of function to move an item accross the stages of a deployment pipeline

[ecampuss]

I have been following the new functions under the sempy_labs.deployment_pipeline and I am taking the advantage of some of them already, but it would be really helpful if we have a function to move an item (semantic model, report, etc) from one stage to the next stage in the deployment pipeline.

[m-kovalsky]

I believe a function for this API would help with this. Please confirm.

[ecampuss]

Yes, that would help a lot. We are planning to automate the deployment pipeline and this API will help us to achieve this goal.

[m-kovalsky]

Available in 0.12.7.

[ecampuss]

Hi @m-kovalsky , I have tested the new function, but I am keeping getting the same error message as the following. I have added the dict (with a masked code) that I am passing to the function for your reference, which is in the same format as the documentation:

masked dict:
{'itemId': 'XXXXXXXX-XXXX-XXXX-XXXX-5041ebf64bb4', 'itemType': 'SemanticModel'}

TypeError Traceback (most recent call last)
Cell In[145], line 4
2 item = deploy_dict[key]
3 print(item)
----> 4 labs_deploy.deploy_stage_content(
5 deployment_pipeline=pipeline_name,
6 source_stage_id=source_stage,
7 target_stage_id=target_stage,
8 items=item
9 )

File ~/cluster-env/trident_env/lib/python3.11/site-packages/sempy/_utils/_log.py:370, in mds_log..get_wrapper..log_decorator_wrapper(*args, **kwargs)
367 start_time = time.perf_counter()
369 try:
--> 370 result = func(*args, **kwargs)
372 # The invocation for get_message_dict moves after the function
373 # so it can access the state after the method call
374 message.update(extractor.get_completion_message_dict(result, arg_dict))

File ~/cluster-env/trident_env/lib/python3.11/site-packages/sempy_labs/deployment_pipeline/_items.py:533, in deploy_stage_content(deployment_pipeline, source_stage_id, target_stage_id, items, note, allow_cross_region_deployment, capacity, workspace_name)
490 @log
491 def deploy_stage_content(
492 deployment_pipeline: str | UUID,
(...)
499 workspace_name: Optional[str] = None,
500 ):
501 """
502 Deploys items from the specified stage of the specified deployment pipeline.
503
(...)
530 The workspace name to use for the deployment operation if creating a new workspace. Required when deploying to a stage that has no assigned workspaces, otherwise it is ignored. The deployment will fail if the new workspace configuration details aren't provided when required.
531 """
--> 533 deployment_pipeline_id = resolve_deployment_pipeline_id(
534 deployment_pipeline=deployment_pipeline
535 )
537 payload = {
538 "sourceStageId": source_stage_id,
539 "targetStageId": target_stage_id,
540 }
541 if note:

File ~/cluster-env/trident_env/lib/python3.11/site-packages/sempy/_utils/_log.py:370, in mds_log..get_wrapper..log_decorator_wrapper(*args, **kwargs)
367 start_time = time.perf_counter()
369 try:
--> 370 result = func(*args, **kwargs)
372 # The invocation for get_message_dict moves after the function
373 # so it can access the state after the method call
374 message.update(extractor.get_completion_message_dict(result, arg_dict))

File ~/cluster-env/trident_env/lib/python3.11/site-packages/sempy_labs/deployment_pipeline/_items.py:36, in resolve_deployment_pipeline_id(deployment_pipeline)
34 else:
35 dfP = list()
---> 36 dfP_filt = dfP[dfP["Deployment Pipeline Name"] == deployment_pipeline]
37 if len(dfP_filt) == 0:
38 raise ValueError(
39 f"{icons.red_dot} The '{deployment_pipeline}' deployment pipeline is not valid."
40 )

TypeError: list indices must be integers or slices, not str

[ecampuss]

I found where was the issue with the function.

Related to the error above, the rootcause was that I was passing the deployment pipeline's name as a value for deployment_pipeline parameter. Once I changed its value to the pipeline's ID, I did not get the error message again
After that I got another different error message about my dictionary. The documentation is saying to create a list with the key "itemId", but the API is expecting the key "sourceItemId".

After doing these two changes, the function worked successfully.

[m-kovalsky]

Would you please try PR #999? There was an issue in the dependent function resolve_deployment_pipeline_id so it should now be able to resolve names or IDs.