feat: Add sync streaming support for Anthropic instrumentation#4155
feat: Add sync streaming support for Anthropic instrumentation#4155vasantteja wants to merge 45 commits intoopen-telemetry:mainfrom
Conversation
- Add support for Messages.create(stream=True) with StreamWrapper - Add support for Messages.stream() with MessageStreamManagerWrapper - Add MessageWrapper for non-streaming response telemetry - Rename MessageCreateParams to MessageRequestParams - Add comprehensive tests for sync streaming functionality
- Add type: ignore[arg-type] for Union type narrowing in messages_create - Add type: ignore[return-value] for wrapper return types - Add type: ignore[return-value] for __exit__ returning None
99a2596 to
504d0df
Compare
...opentelemetry-instrumentation-anthropic/src/opentelemetry/instrumentation/anthropic/patch.py
Outdated
Show resolved
Hide resolved
...opentelemetry-instrumentation-anthropic/src/opentelemetry/instrumentation/anthropic/patch.py
Outdated
Show resolved
Hide resolved
...opentelemetry-instrumentation-anthropic/src/opentelemetry/instrumentation/anthropic/utils.py
Outdated
Show resolved
Hide resolved
|
tagging @anirudha who was interested to review the PR :) |
|
Thanks. Taking a look today |
…r handling - Introduce constants for provider name and cache token attributes. - Normalize stop reasons and aggregate cache token fields in MessageWrapper and StreamWrapper. - Enhance tests to validate input token aggregation and stop reason normalization. - Update cassettes for new request and response structures in streaming scenarios.
…d consistency - Simplify constant definitions and normalize function calls in utils.py. - Enhance test cases by removing unnecessary line breaks and improving formatting. - Ensure consistent usage of type hints and comments in test functions.
- Update the pylint directive to disable too-many-arguments warning for better clarity. - Maintain consistency in function signature and improve code readability.
anirudha
left a comment
There was a problem hiding this comment.
Tests all pass locally. Nice work overall — the wrapper separation is clean. One bug to fix (double finalize), rest are suggestions.
Note: conftest.py isn't in this diff so I can't leave a line comment, but scrub_response_headers is a no-op and all new cassettes leak anthropic-organization-id: 455ea6be-bd92-4199-83ec-0c6b39c5c169. Worth scrubbing that or adding it to filter_headers.
Also, the PR description says Fixes #3949 but async streaming isn't covered. Totally fine to scope this to sync only, but Fixes will auto-close the issue on merge. Maybe Partially addresses #3949 instead?
...opentelemetry-instrumentation-anthropic/src/opentelemetry/instrumentation/anthropic/utils.py
Outdated
Show resolved
Hide resolved
...opentelemetry-instrumentation-anthropic/src/opentelemetry/instrumentation/anthropic/utils.py
Outdated
Show resolved
Hide resolved
...opentelemetry-instrumentation-anthropic/src/opentelemetry/instrumentation/anthropic/utils.py
Outdated
Show resolved
Hide resolved
...opentelemetry-instrumentation-anthropic/src/opentelemetry/instrumentation/anthropic/patch.py
Show resolved
Hide resolved
…tion - Update test cases to validate streaming behavior with various parameters, including token usage and stop reasons. - Introduce new cassettes for different scenarios, ensuring comprehensive coverage of streaming interactions. - Refactor existing tests for clarity and consistency in structure and assertions.
…ocals in test_stream_wrapper_finalize_idempotent function
…eja/opentelemetry-python-contrib into anthropic-sync-streaming
…Ds, timestamps, and token usage across various test cases. Refine content capture logic and ensure consistency in message formats, including adjustments to event data and headers for improved clarity and accuracy.
...ntelemetry-instrumentation-anthropic/src/opentelemetry/instrumentation/anthropic/wrappers.py
Outdated
Show resolved
Hide resolved
...ntelemetry-instrumentation-anthropic/src/opentelemetry/instrumentation/anthropic/wrappers.py
Outdated
Show resolved
Hide resolved
...ntelemetry-instrumentation-anthropic/src/opentelemetry/instrumentation/anthropic/wrappers.py
Outdated
Show resolved
Hide resolved
...ntelemetry-instrumentation-anthropic/src/opentelemetry/instrumentation/anthropic/wrappers.py
Outdated
Show resolved
Hide resolved
...opentelemetry-instrumentation-anthropic/src/opentelemetry/instrumentation/anthropic/patch.py
Outdated
Show resolved
Hide resolved
...opentelemetry-instrumentation-anthropic/src/opentelemetry/instrumentation/anthropic/patch.py
Outdated
Show resolved
Hide resolved
...opentelemetry-instrumentation-anthropic/src/opentelemetry/instrumentation/anthropic/patch.py
Outdated
Show resolved
Hide resolved
...instrumentation-anthropic/src/opentelemetry/instrumentation/anthropic/messages_extractors.py
Outdated
Show resolved
Hide resolved
…oved type safety. Replace 'Any' with 'object' in several function signatures and class attributes. Introduce logging for error handling in MessagesStreamWrapper to enhance instrumentation reliability.
…eja/opentelemetry-python-contrib into anthropic-sync-streaming
… clarity and safety. Update function signatures to use specific types instead of 'object', including changes to parameters in extract_params, get_input_messages, and get_system_instruction. Refactor messages_create to ensure correct type handling for streaming and non-streaming responses. Additionally, streamline message handling in MessagesStreamWrapper for better performance and reliability.
…function signatures in `messages_extractors.py` and `wrappers.py` to include specific types, improving clarity and reliability. Introduce handling for `None` values in `get_input_messages` and `get_system_instruction`. Refactor `MessagesStreamWrapper` to better manage usage updates and ensure correct type handling for streaming responses. Add new test cases for aggregating cache tokens and handling streaming errors.
…. Simplify assertion statements by removing unnecessary parentheses, enhancing code clarity in cache token tests.
| return False | ||
| if mode == ContentCapturingMode.EVENT_ONLY and not should_emit_event(): | ||
| return False | ||
| return True |
There was a problem hiding this comment.
@aabmass We don't have this in util. This is specifically used for understanding whether we should we do the work of converting messages/content blocks into structured parts. I couldn't see any function which is having same functionality as this one in util.
...instrumentation-anthropic/src/opentelemetry/instrumentation/anthropic/messages_extractors.py
Outdated
Show resolved
Hide resolved
...instrumentation-anthropic/src/opentelemetry/instrumentation/anthropic/messages_extractors.py
Outdated
Show resolved
Hide resolved
...opentelemetry-instrumentation-anthropic/src/opentelemetry/instrumentation/anthropic/utils.py
Show resolved
Hide resolved
…equirements.oldest.txt for compatibility improvements.
…s for improved clarity and type safety. Update extract_usage_tokens function to return UsageTokens instead of a tuple, and adjust related invocations in MessageWrapper and MessagesStreamWrapper accordingly.
…ility improvements.
Description
This PR adds sync streaming support for the Anthropic instrumentation. It enables telemetry capture for:
Messages.create(stream=True)- Streaming responses via the create method with stream parameterMessages.stream()- The dedicated streaming method that returns aMessageStreamManagerKey changes:
StreamWrapperclass to wrapStream[RawMessageStreamEvent]and extract telemetry from streaming chunksMessageStreamManagerWrapperto wrapMessageStreamManagercontext managerMessageWrapperfor non-streaming response telemetry extractionMessageCreateParamstoMessageRequestParamsto reflect broader API coveragemessages_createto use manual lifecycle management (start_llm/stop_llm) instead of context manager to support both streaming and non-streamingFixes #3949 partially.
Type of change
Please delete options that are not relevant.
How Has This Been Tested?
Added comprehensive tests for sync streaming functionality:
test_sync_messages_create_streaming- Tests streaming with context managertest_sync_messages_create_streaming_iteration- Tests direct iteration without context managertest_sync_messages_create_streaming_connection_error- Tests error handling for streamingtest_sync_messages_stream_basic- TestsMessages.stream()methodtest_sync_messages_stream_with_params- Tests stream with additional parameters (temperature, top_p, top_k)test_sync_messages_stream_token_usage- Tests token usage capture in streamingtest_sync_messages_stream_connection_error- Tests error handling for stream methodAll tests use VCR cassettes for reproducible HTTP interaction replay.
Does This PR Require a Core Repo Change?
Checklist:
See contributing.md for styleguide, changelog guidelines, and more.