fix(watsonx): emit response events with logger

[pragnyanramtha]

Summary

• pass the Watsonx event logger through response event emission paths
• normalize dict and list-shaped Watsonx responses before emitting gen_ai.choice events
• add regression coverage for dict, batch/list, and stream response event logs

Validation

• uv run pytest tests/test_events.py -q
• uv run pytest tests/ -q
• uv run ruff check .

Summary by CodeRabbit

• Bug Fixes

  • Consistent emission of choice events for single, batched, and streaming responses; assistant role and default finish reason are applied uniformly.

• Tests

  • Added tests covering single, indexed batched, streamed responses and skipping invalid batched items.

• Chores

  • Redacted real upload URLs in test cassettes, replacing them with placeholder URLs.

[CLAassistant]

All committers have signed the CLA.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro

Run ID: 4e977813-9a1f-4590-b895-76f915d490f4

📥 Commits

Reviewing files that changed from the base of the PR and between f34d62d and d5ba8b6.

📒 Files selected for processing (1)

• packages/opentelemetry-instrumentation-watsonx/tests/test_events.py

---

📝 Walkthrough

Walkthrough

Refactors _emit_response_events to accept an event_logger, normalize single/list responses (flatten results), filter non-dict items, and emit gen_ai.choice/ChoiceEvent logs. Updates _handle_stream_response to pass event_logger. Adds tests and redacts pre-signed upload URLs in several test cassettes.

Changes

Response Event Emission and Testing

Layer / File(s)	Summary
Response event emission refactoring packages/opentelemetry-instrumentation-watsonx/opentelemetry/instrumentation/watsonx/__init__.py	_emit_response_events is refactored to accept event_logger, normalize single dict or list responses into a messages list, filter non-dict items, flatten results when present, and emit ChoiceEvent / gen_ai.choice log records. _handle_stream_response updated to pass event_logger to the refactored function.
Test infrastructure and validation packages/opentelemetry-instrumentation-watsonx/tests/test_events.py	New test module adding RecordingSpan test-double, event_logger fixture (uses LoggerProvider + InMemoryLogExporter), and tests asserting emitted gen_ai.choice log records for single response, list responses (indices 0/1), invalid batched inputs, and stream responses (mapping generated_text and finish reason).

Test Cassette Redactions

Layer / File(s)

Summary

Redact pre-signed upload URLs in cassettes packages/traceloop-sdk/tests/datasets/cassettes/test_dataset_with_attachments/test_create_dataset_with_file_attachments_mocked.yaml, packages/traceloop-sdk/tests/datasets/cassettes/test_dataset_with_attachments/test_create_dataset_with_in_memory_attachment.yaml, packages/traceloop-sdk/tests/datasets/cassettes/test_dataset_with_attachments/test_create_dataset_with_mixed_attachments.yaml

Replaced captured upload_url pre-signed S3 URLs with https://example.invalid/redacted-upload-url placeholders in multiple test cassettes; preserved storage_key, expires_at, and method fields.

🎯 3 (Moderate) | ⏱️ ~20 minutes

🐰 I hopped through code with eager paws,
Normalized responses, smoothed out the laws,
Lists or single, each choice now logs,
Streams whisper events through tidy logs,
Tests confirm the meadow's cause.

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 0.00% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title 'fix(watsonx): emit response events with logger' directly describes the main change: passing the event logger through Watsonx response event emission and normalizing response handling.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[Copilot]

Copilot encountered an error and was unable to review this pull request. You can try again by re-requesting a review.

[coderabbitai]

🧹 Nitpick comments (1)

packages/opentelemetry-instrumentation-watsonx/tests/test_events.py (1)

58-83: ⚡ Quick win

Add a regression test for mixed invalid entries in batched responses.

Please add one case where the list includes non-dict items and results contains non-dict elements, then assert only valid dict messages emit logs. This locks in the new filtering behavior.

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@packages/opentelemetry-instrumentation-watsonx/tests/test_events.py` around
lines 58 - 83, Add a regression test in tests/test_events.py alongside
test_emit_response_events_writes_choice_log_for_list_response that calls
_emit_response_events with a mixed list containing non-dict items (e.g., a
string or None) and entries whose "results" list includes non-dict elements;
then use the event_logger fixture/exporter to collect finished logs and assert
that only the valid dict messages produced gen_ai.choice logs (with correct
index/message/finish_reason) and that invalid entries produced no logs.
Reference _emit_response_events, event_logger, and exporter when locating where
to add the new assertions.

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Nitpick comments:
In `@packages/opentelemetry-instrumentation-watsonx/tests/test_events.py`:
- Around line 58-83: Add a regression test in tests/test_events.py alongside
test_emit_response_events_writes_choice_log_for_list_response that calls
_emit_response_events with a mixed list containing non-dict items (e.g., a
string or None) and entries whose "results" list includes non-dict elements;
then use the event_logger fixture/exporter to collect finished logs and assert
that only the valid dict messages produced gen_ai.choice logs (with correct
index/message/finish_reason) and that invalid entries produced no logs.
Reference _emit_response_events, event_logger, and exporter when locating where
to add the new assertions.

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro

Run ID: b3f80768-4106-44e0-9a69-2b0c3e8e0066

📥 Commits

Reviewing files that changed from the base of the PR and between 8ab3c08 and f34d62d.

📒 Files selected for processing (5)

• packages/opentelemetry-instrumentation-watsonx/opentelemetry/instrumentation/watsonx/__init__.py
• packages/opentelemetry-instrumentation-watsonx/tests/test_events.py
• packages/traceloop-sdk/tests/datasets/cassettes/test_dataset_with_attachments/test_create_dataset_with_file_attachments_mocked.yaml
• packages/traceloop-sdk/tests/datasets/cassettes/test_dataset_with_attachments/test_create_dataset_with_in_memory_attachment.yaml
• packages/traceloop-sdk/tests/datasets/cassettes/test_dataset_with_attachments/test_create_dataset_with_mixed_attachments.yaml

✅ Files skipped from review due to trivial changes (2)

• packages/traceloop-sdk/tests/datasets/cassettes/test_dataset_with_attachments/test_create_dataset_with_in_memory_attachment.yaml
• packages/traceloop-sdk/tests/datasets/cassettes/test_dataset_with_attachments/test_create_dataset_with_mixed_attachments.yaml

[pragnyanramtha]

@coderabbitai resume

[coderabbitai]

✅ Actions performed

Reviews resumed.