feat: instrument the openai responses API

[satyadevai]

New Changes:

• Added support for tracing and instrumenting OpenAI responses using the new Responses API integration.
• Included sample code and usage examples to simplify testing and validation of the integration.
• Added unit tests to ensure functionality and maintain reliability of the newly added features.

Resolves #1351

[Copilot]

Copilot reviewed 14 out of 16 changed files in this pull request and generated 1 comment.

Files not reviewed (2)

• python/instrumentation/openinference-instrumentation-openai/examples/requirements.txt: Language not supported
• python/instrumentation/openinference-instrumentation-openai/test-requirements.txt: Language not supported

Comments suppressed due to low confidence (1)

python/instrumentation/openinference-instrumentation-openai/src/openinference/instrumentation/openai/_response_accumulator.py:64

• [nitpick] Consider renaming 'chat_completion_type' to 'responses_type' for clarity, as it represents a Response type in the context of responses.

chat_completion_type: Type["Response"],

[Copilot]

Copilot reviewed 14 out of 16 changed files in this pull request and generated no comments.

Files not reviewed (2)

• python/instrumentation/openinference-instrumentation-openai/examples/requirements.txt: Language not supported
• python/instrumentation/openinference-instrumentation-openai/test-requirements.txt: Language not supported

Comments suppressed due to low confidence (3)

python/instrumentation/openinference-instrumentation-openai/tests/openinference/instrumentation/openai/test_instrumentor.py:541

• The delta field value '{' may result in malformed JSON parsing; please verify if this is intentional or adjust the mock data to represent a properly concatenated delta message.

b'event: response.function_call_arguments.delta\ndata: {"type":"response.function_call_arguments.delta","item_id":"call_caea3f3962f34170b06d77f8f69da8e1","output_index":0,"delta":"{"}\n\n'

python/instrumentation/openinference-instrumentation-openai/src/openinference/instrumentation/openai/_utils.py:40

• [nitpick] Consider providing more descriptive documentation or renaming the 'cast_to' parameter to better clarify its purpose in transforming the input object.

def _io_value_and_type(obj: Any, cast_to: Optional[type] = None) -> _ValueAndType:

python/instrumentation/openinference-instrumentation-openai/src/openinference/instrumentation/openai/_response_accumulator.py:75

• Consider using isinstance() for type checks instead of comparing the name string to improve type safety and handle subclassing properly.

if type(chunk).__name__ == "ResponseCompletedEvent":

[Copilot]

Copilot reviewed 16 out of 18 changed files in this pull request and generated no comments.

Files not reviewed (2)

• python/instrumentation/openinference-instrumentation-openai/examples/requirements.txt: Language not supported
• python/instrumentation/openinference-instrumentation-openai/test-requirements.txt: Language not supported

Comments suppressed due to low confidence (2)

python/instrumentation/openinference-instrumentation-openai/tests/openinference/instrumentation/openai/test_instrumentor.py:541

• The delta JSON fragment appears incomplete, with the delta value being just an opening brace. Please verify that the split fragments (lines 541–542) will be concatenated correctly in the stream so that the overall JSON is valid.

b'event: response.function_call_arguments.delta\ndata: {"type":"response.function_call_arguments.delta","item_id":"call_caea3f3962f34170b06d77f8f69da8e1","output_index":0,"delta":"{"}\n\n',  # noqa: E501

python/instrumentation/openinference-instrumentation-openai/src/openinference/instrumentation/openai/_response_accumulator.py:75

• Instead of comparing the type name as a string, consider using isinstance() with the actual ResponseCompletedEvent type (if available) to improve robustness and maintainability.

if type(chunk).__name__ == "ResponseCompletedEvent":

[RogerHYang]

transferred to #1521