Python azure-search-backup-and-restore: appending the list failed_keys appears to fail

[marcindulak]

I'm not sure this is an artifact of my use of the notebook functions, or some type mismatch in the notebook implementation.

You can try to trigger the problem I'm describing by negating this condition

azure-search-vector-samples/demo-python/code/index-backup-restore/azure-search-backup-and-restore.ipynb

Line 171 in fcfd950

" if item.succeeded is not True: \n",

As a consequence, the azure-search-backup-and-restore notebook appears to fail on appending the list failed_keys, since page appears to be a list of dicts

azure-search-vector-samples/demo-python/code/index-backup-restore/azure-search-backup-and-restore.ipynb

Line 173 in fcfd950

" failed_keys.append(page[result.index_of(item)].id) \n",

    failed_keys.append(page[result.index_of(item)].id)  
                            ^^^^^^^^^^^^^^^
AttributeError: 'list' object has no attribute 'index_of'

Trying to use the list.index fails on dict.id attribute access

    failed_keys.append(page[result.index(item)].id)  
                       ^^^^^^^^^^^^^^^^^^^^^^^^^^^
AttributeError: 'dict' object has no attribute 'id'

The following failed_keys.append(page[result.index(item)]["id"]) allows to advance further, but then fails at

    print(f"Document upload error: {item.error.message}")
                                    ^^^^^^^^^^
AttributeError: 'IndexingResult' object has no attribute 'error'

which should probably be print(f"Document upload error: {item.error_message}"), since item appears to be an instance of https://learn.microsoft.com/en-us/python/api/azure-search-documents/azure.search.documents.models.indexingresult

An extra question: are there any news about Azure AI Search getting a functionality of automatic index snapshots/backups and restores? The lack of this may create a friction in adoption of Azure AI search in enterprise setups.