Move Python SDK to Objects

[NolanTrem]

Important

Transition Python SDK to object-oriented responses, updating methods and tests to handle new response objects across documents, users, and retrieval operations.

• SDK Changes:
  • Transition to object-oriented responses across the SDK, affecting documents, users, retrieval, indices, graphs, and conversations.
  • Update methods to return specific response objects like WrappedDocumentResponse, WrappedUserResponse, etc.
  • Modify ingestion, retrieval, and user management functions to align with new response handling.
• Test Updates:
  • Update integration tests to use new response objects and verify correct behavior.
  • Ensure tests cover all major functionalities including document ingestion, user operations, and retrieval queries.
  • Add tests for error handling and edge cases, such as invalid inputs and unauthorized access.
• Miscellaneous:
  • Remove deprecated or unused code paths.
  • Improve error messages and logging for better debugging.

^{This description was created by for b370f9d. It will automatically update as commits are pushed.}

[ellipsis-dev]

👍 Looks good to me! Reviewed everything up to b370f9d in 4 minutes and 57 seconds

More details

• Looked at 10476 lines of code in 58 files
• Skipped 0 files when reviewing.
• Skipped posting 10 drafted comments based on config settings.

1. tests/integration/test_chunks.py:80

• Draft comment:
  Using a fixed sleep (5s) to wait for ingestion might lead to flaky tests under load. Consider polling for status until the document is ingested.
• Reason this comment was not posted:
  Comment was not on a location in the diff, so it can't be submitted as a review comment.

2. tests/integration/test_chunks.py:173

• Draft comment:
  Good practice: the unauthorized access test cleanly instantiates a new client & checks for 403. Make sure your backend consistently enforces permissions.
• Reason this comment was not posted:
  Confidence changes required: 0% <= threshold 50%
  None

3. tests/integration/test_chunks.py:239

• Draft comment:
  The pagination tests are straightforward. Consider adding edge-case tests (e.g., offset=0 with fewer results) to further validate your pagination logic.
• Reason this comment was not posted:
  Confidence changes required: 50% <= threshold 50%
  None

4. py/tests/unit/test_chunks.py:81

• Draft comment:
  Consider increasing the sleep time during document ingestion (currently 5 seconds) only if necessary. Using async polling (or a retry mechanism) might reduce test flakiness.
• Reason this comment was not posted:
  Confidence changes required: 50% <= threshold 50%
  None

5. py/tests/integration/test_retrieval.py:143

• Draft comment:
  For consistency, use the '.results' attribute rather than dictionary key access (e.g. using resp.results instead of resp["results"]).
• Reason this comment was not posted:
  Comment was not on a location in the diff, so it can't be submitted as a review comment.

6. py/tests/integration/test_ingestion.py:89

• Draft comment:
  The polling loop for ingestion status works correctly, but adding a brief log (or comment) about breaking on failure may improve debuggability in case of failures.
• Reason this comment was not posted:
  Confidence changes required: 50% <= threshold 50%
  None

7. py/tests/integration/test_users.py:142

• Draft comment:
  Ensure that after user registration and login, the tests properly log out the user to avoid side effects between tests.
• Reason this comment was not posted:
  Confidence changes required: 30% <= threshold 50%
  None

8. py/tests/integration/test_filters.py:115

• Draft comment:
  The filter tests correctly cast collection IDs to strings ensuring type consistency when comparing UUIDs. This is good practice for cross-system compatibility.
• Reason this comment was not posted:
  Confidence changes required: 0% <= threshold 50%
  None

9. py/tests/integration/test_graphs.py:150

• Draft comment:
  Tests in the graphs module appropriately test functionality (e.g., create, update, list entities) and consistently compare string representations of IDs.
• Reason this comment was not posted:
  Confidence changes required: 0% <= threshold 50%
  None

10. py/tests/integration/test_ingestion.py:310

• Draft comment:
  Tests for document ingestion (file, raw text, chunks, etc.) are comprehensive. Consider adding more comments to guide future maintainers through the ingestion process logic.
• Reason this comment was not posted:
  Confidence changes required: 30% <= threshold 50%
  None

Workflow ID: wflow_YqNh0do0omnoOtIi

---

You can customize Ellipsis with 👍 / 👎 feedback, review rules, user-specific overrides, quiet mode, and more.