fix(sdk): warn when both exporter and processor are passed to Traceloop.init()

[dvirski]

Fixes #3046.
Problem: passing both exporter and processor is a mistake — the processor already wraps the exporter
internally. The exporter was silently ignored with no feedback to the user.

Fix: emit a UserWarning pointing at the caller's line. Also removes the redundant
exporter= from the exporter_with_custom_span_processor test fixture which had the same bug.

Summary by CodeRabbit

• New Features

  • Traceloop.init now emits a UserWarning when both an exporter and a processor are supplied, clarifying the exporter will be ignored and advising wrapping the exporter with a processor instead.

• Tests

  • Added a test confirming the warning is raised and that span export behavior follows the processor-wrapped exporter.

[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: 12b1f3b2-d411-41b2-a9b4-3712476f6500

📥 Commits

Reviewing files that changed from the base of the PR and between 0668e0c and ccd0832.

📒 Files selected for processing (3)

• packages/traceloop-sdk/tests/conftest.py
• packages/traceloop-sdk/tests/test_sdk_initialization.py
• packages/traceloop-sdk/traceloop/sdk/__init__.py

💤 Files with no reviewable changes (1)

• packages/traceloop-sdk/tests/conftest.py

🚧 Files skipped from review as they are similar to previous changes (1)

• packages/traceloop-sdk/tests/test_sdk_initialization.py

---

📝 Walkthrough

Walkthrough

Emits a UserWarning when both exporter and processor are passed to Traceloop.init, adds a test verifying the warning and span behavior, and updates a fixture to pass processor=CustomSpanProcessor(exporter) instead of exporter=.

Changes

Exporter and Processor Warning

Layer / File(s)	Summary
Warning implementation in Traceloop.init packages/traceloop-sdk/traceloop/sdk/__init__.py	Added import warnings and conditional logic in Traceloop.init that emits a UserWarning when both exporter and processor are provided, indicating the exporter will be ignored.
Test for exporter and processor warning packages/traceloop-sdk/tests/test_sdk_initialization.py	Expanded imports and added test_both_exporter_and_processor_warns which asserts a UserWarning matching exporter.*ignored, and verifies processor-wrapped exporter receives spans while a standalone exporter does not.
Fixture update to use processor pattern packages/traceloop-sdk/tests/conftest.py	Updated exporter_with_custom_span_processor fixture to pass processor=CustomSpanProcessor(exporter) instead of supplying exporter= directly.

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~8 minutes

Poem

🐰 I hop a warning, soft and clear,
When exporter and processor both appear.
Wrap the exporter, let spans flow true,
A gentle nudge from this rabbit to you. ✨

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 50.00% 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 accurately describes the primary change: adding a warning when both exporter and processor are passed to Traceloop.init().
Linked Issues check	✅ Passed	The PR addresses issue #3046 by implementing option (b): emitting a UserWarning when both exporter and processor are provided, preventing silent failures.
Out of Scope Changes check	✅ Passed	All changes are directly related to the issue: adding the warning mechanism, updating the test fixture that exhibited the bug, and adding a test to verify the warning behavior.

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

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch dr/new/fix(sdk)-warn-when-both-exporter-and-processor-are-passed-to-Traceloop.init()

---

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.}

[coderabbitai]

🧹 Nitpick comments (1)

packages/traceloop-sdk/tests/test_sdk_initialization.py (1)

235-237: ⚡ Quick win

Prefer a more robust pattern for instance cleanup.

The pattern if "_instance" in dir(): works but is fragile and non-idiomatic. A more explicit approach improves clarity and reliability.

♻️ Proposed refactor using try-except

-    if hasattr(TracerWrapper, "instance"):
-        _instance = TracerWrapper.instance
+    try:
+        _instance = TracerWrapper.instance
         del TracerWrapper.instance
+    except AttributeError:
+        _instance = None
 
     exporter = InMemorySpanExporter()
     processor = SimpleSpanProcessor(exporter)
@@ -245,5 +246,5 @@
             processor=processor,
         )
 
-    if "_instance" in dir():
+    if _instance is not None:
         TracerWrapper.instance = _instance

Also applies to: 248-249

🤖 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/traceloop-sdk/tests/test_sdk_initialization.py` around lines 235 -
237, Replace the fragile instance cleanup with a robust try/except/finally
pattern: use getattr(TracerWrapper, "instance", None) to grab the current
instance, then if it is not None call delattr(TracerWrapper, "instance") inside
a try block and in finally restore the saved instance with
setattr(TracerWrapper, "instance", saved) so the test cannot leak state; apply
the same change for the other cleanup at the second occurrence (lines
referencing TracerWrapper, _instance, instance).

🤖 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/traceloop-sdk/tests/test_sdk_initialization.py`:
- Around line 235-237: Replace the fragile instance cleanup with a robust
try/except/finally pattern: use getattr(TracerWrapper, "instance", None) to grab
the current instance, then if it is not None call delattr(TracerWrapper,
"instance") inside a try block and in finally restore the saved instance with
setattr(TracerWrapper, "instance", saved) so the test cannot leak state; apply
the same change for the other cleanup at the second occurrence (lines
referencing TracerWrapper, _instance, instance).

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro

Run ID: 3721c9b8-fc97-4609-a09c-6e92f952dffa

📥 Commits

Reviewing files that changed from the base of the PR and between 6d3e696 and 0668e0c.

⛔ Files ignored due to path filters (1)

• packages/traceloop-sdk/uv.lock is excluded by !**/*.lock

📒 Files selected for processing (3)

• packages/traceloop-sdk/tests/conftest.py
• packages/traceloop-sdk/tests/test_sdk_initialization.py
• packages/traceloop-sdk/traceloop/sdk/__init__.py

💤 Files with no reviewable changes (1)

• packages/traceloop-sdk/tests/conftest.py

[dvirski]

@doronkopit5

Fixed both your comments (in resolved state now)