two version of R2R are here HEAD master

3 files changed, 615 insertions, 0 deletions

diff --git a/.venv/lib/python3.12/site-packages/litellm/responses/main.py b/.venv/lib/python3.12/site-packages/litellm/responses/main.py
new file mode 100644
index 00000000..aec2f8fe
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/litellm/responses/main.py
@@ -0,0 +1,248 @@
+import asyncio
+import contextvars
+from functools import partial
+from typing import Any, Dict, Iterable, List, Literal, Optional, Union
+
+import httpx
+
+import litellm
+from litellm.constants import request_timeout
+from litellm.litellm_core_utils.litellm_logging import Logging as LiteLLMLoggingObj
+from litellm.llms.base_llm.responses.transformation import BaseResponsesAPIConfig
+from litellm.llms.custom_httpx.llm_http_handler import BaseLLMHTTPHandler
+from litellm.responses.utils import ResponsesAPIRequestUtils
+from litellm.types.llms.openai import (
+    Reasoning,
+    ResponseIncludable,
+    ResponseInputParam,
+    ResponsesAPIOptionalRequestParams,
+    ResponsesAPIResponse,
+    ResponseTextConfigParam,
+    ToolChoice,
+    ToolParam,
+)
+from litellm.types.router import GenericLiteLLMParams
+from litellm.utils import ProviderConfigManager, client
+
+from .streaming_iterator import BaseResponsesAPIStreamingIterator
+
+####### ENVIRONMENT VARIABLES ###################
+# Initialize any necessary instances or variables here
+base_llm_http_handler = BaseLLMHTTPHandler()
+#################################################
+
+
+@client
+async def aresponses(
+    input: Union[str, ResponseInputParam],
+    model: str,
+    include: Optional[List[ResponseIncludable]] = None,
+    instructions: Optional[str] = None,
+    max_output_tokens: Optional[int] = None,
+    metadata: Optional[Dict[str, Any]] = None,
+    parallel_tool_calls: Optional[bool] = None,
+    previous_response_id: Optional[str] = None,
+    reasoning: Optional[Reasoning] = None,
+    store: Optional[bool] = None,
+    stream: Optional[bool] = None,
+    temperature: Optional[float] = None,
+    text: Optional[ResponseTextConfigParam] = None,
+    tool_choice: Optional[ToolChoice] = None,
+    tools: Optional[Iterable[ToolParam]] = None,
+    top_p: Optional[float] = None,
+    truncation: Optional[Literal["auto", "disabled"]] = None,
+    user: Optional[str] = None,
+    # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs.
+    # The extra values given here take precedence over values defined on the client or passed to this method.
+    extra_headers: Optional[Dict[str, Any]] = None,
+    extra_query: Optional[Dict[str, Any]] = None,
+    extra_body: Optional[Dict[str, Any]] = None,
+    timeout: Optional[Union[float, httpx.Timeout]] = None,
+    # LiteLLM specific params,
+    custom_llm_provider: Optional[str] = None,
+    **kwargs,
+) -> Union[ResponsesAPIResponse, BaseResponsesAPIStreamingIterator]:
+    """
+    Async: Handles responses API requests by reusing the synchronous function
+    """
+    local_vars = locals()
+    try:
+        loop = asyncio.get_event_loop()
+        kwargs["aresponses"] = True
+
+        # get custom llm provider so we can use this for mapping exceptions
+        if custom_llm_provider is None:
+            _, custom_llm_provider, _, _ = litellm.get_llm_provider(
+                model=model, api_base=local_vars.get("base_url", None)
+            )
+
+        func = partial(
+            responses,
+            input=input,
+            model=model,
+            include=include,
+            instructions=instructions,
+            max_output_tokens=max_output_tokens,
+            metadata=metadata,
+            parallel_tool_calls=parallel_tool_calls,
+            previous_response_id=previous_response_id,
+            reasoning=reasoning,
+            store=store,
+            stream=stream,
+            temperature=temperature,
+            text=text,
+            tool_choice=tool_choice,
+            tools=tools,
+            top_p=top_p,
+            truncation=truncation,
+            user=user,
+            extra_headers=extra_headers,
+            extra_query=extra_query,
+            extra_body=extra_body,
+            timeout=timeout,
+            custom_llm_provider=custom_llm_provider,
+            **kwargs,
+        )
+
+        ctx = contextvars.copy_context()
+        func_with_context = partial(ctx.run, func)
+        init_response = await loop.run_in_executor(None, func_with_context)
+
+        if asyncio.iscoroutine(init_response):
+            response = await init_response
+        else:
+            response = init_response
+        return response
+    except Exception as e:
+        raise litellm.exception_type(
+            model=model,
+            custom_llm_provider=custom_llm_provider,
+            original_exception=e,
+            completion_kwargs=local_vars,
+            extra_kwargs=kwargs,
+        )
+
+
+@client
+def responses(
+    input: Union[str, ResponseInputParam],
+    model: str,
+    include: Optional[List[ResponseIncludable]] = None,
+    instructions: Optional[str] = None,
+    max_output_tokens: Optional[int] = None,
+    metadata: Optional[Dict[str, Any]] = None,
+    parallel_tool_calls: Optional[bool] = None,
+    previous_response_id: Optional[str] = None,
+    reasoning: Optional[Reasoning] = None,
+    store: Optional[bool] = None,
+    stream: Optional[bool] = None,
+    temperature: Optional[float] = None,
+    text: Optional[ResponseTextConfigParam] = None,
+    tool_choice: Optional[ToolChoice] = None,
+    tools: Optional[Iterable[ToolParam]] = None,
+    top_p: Optional[float] = None,
+    truncation: Optional[Literal["auto", "disabled"]] = None,
+    user: Optional[str] = None,
+    # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs.
+    # The extra values given here take precedence over values defined on the client or passed to this method.
+    extra_headers: Optional[Dict[str, Any]] = None,
+    extra_query: Optional[Dict[str, Any]] = None,
+    extra_body: Optional[Dict[str, Any]] = None,
+    timeout: Optional[Union[float, httpx.Timeout]] = None,
+    # LiteLLM specific params,
+    custom_llm_provider: Optional[str] = None,
+    **kwargs,
+):
+    """
+    Synchronous version of the Responses API.
+    Uses the synchronous HTTP handler to make requests.
+    """
+    local_vars = locals()
+    try:
+        litellm_logging_obj: LiteLLMLoggingObj = kwargs.get("litellm_logging_obj")  # type: ignore
+        litellm_call_id: Optional[str] = kwargs.get("litellm_call_id", None)
+        _is_async = kwargs.pop("aresponses", False) is True
+
+        # get llm provider logic
+        litellm_params = GenericLiteLLMParams(**kwargs)
+        model, custom_llm_provider, dynamic_api_key, dynamic_api_base = (
+            litellm.get_llm_provider(
+                model=model,
+                custom_llm_provider=custom_llm_provider,
+                api_base=litellm_params.api_base,
+                api_key=litellm_params.api_key,
+            )
+        )
+
+        # get provider config
+        responses_api_provider_config: Optional[BaseResponsesAPIConfig] = (
+            ProviderConfigManager.get_provider_responses_api_config(
+                model=model,
+                provider=litellm.LlmProviders(custom_llm_provider),
+            )
+        )
+
+        if responses_api_provider_config is None:
+            raise litellm.BadRequestError(
+                model=model,
+                llm_provider=custom_llm_provider,
+                message=f"Responses API not available for custom_llm_provider={custom_llm_provider}, model: {model}",
+            )
+
+        local_vars.update(kwargs)
+        # Get ResponsesAPIOptionalRequestParams with only valid parameters
+        response_api_optional_params: ResponsesAPIOptionalRequestParams = (
+            ResponsesAPIRequestUtils.get_requested_response_api_optional_param(
+                local_vars
+            )
+        )
+
+        # Get optional parameters for the responses API
+        responses_api_request_params: Dict = (
+            ResponsesAPIRequestUtils.get_optional_params_responses_api(
+                model=model,
+                responses_api_provider_config=responses_api_provider_config,
+                response_api_optional_params=response_api_optional_params,
+            )
+        )
+
+        # Pre Call logging
+        litellm_logging_obj.update_environment_variables(
+            model=model,
+            user=user,
+            optional_params=dict(responses_api_request_params),
+            litellm_params={
+                "litellm_call_id": litellm_call_id,
+                **responses_api_request_params,
+            },
+            custom_llm_provider=custom_llm_provider,
+        )
+
+        # Call the handler with _is_async flag instead of directly calling the async handler
+        response = base_llm_http_handler.response_api_handler(
+            model=model,
+            input=input,
+            responses_api_provider_config=responses_api_provider_config,
+            response_api_optional_request_params=responses_api_request_params,
+            custom_llm_provider=custom_llm_provider,
+            litellm_params=litellm_params,
+            logging_obj=litellm_logging_obj,
+            extra_headers=extra_headers,
+            extra_body=extra_body,
+            timeout=timeout or request_timeout,
+            _is_async=_is_async,
+            client=kwargs.get("client"),
+            fake_stream=responses_api_provider_config.should_fake_stream(
+                model=model, stream=stream, custom_llm_provider=custom_llm_provider
+            ),
+        )
+
+        return response
+    except Exception as e:
+        raise litellm.exception_type(
+            model=model,
+            custom_llm_provider=custom_llm_provider,
+            original_exception=e,
+            completion_kwargs=local_vars,
+            extra_kwargs=kwargs,
+        )
diff --git a/.venv/lib/python3.12/site-packages/litellm/responses/streaming_iterator.py b/.venv/lib/python3.12/site-packages/litellm/responses/streaming_iterator.py
new file mode 100644
index 00000000..3039efb9
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/litellm/responses/streaming_iterator.py
@@ -0,0 +1,270 @@
+import asyncio
+import json
+from datetime import datetime
+from typing import Optional
+
+import httpx
+
+from litellm.constants import STREAM_SSE_DONE_STRING
+from litellm.litellm_core_utils.asyncify import run_async_function
+from litellm.litellm_core_utils.litellm_logging import Logging as LiteLLMLoggingObj
+from litellm.litellm_core_utils.thread_pool_executor import executor
+from litellm.llms.base_llm.responses.transformation import BaseResponsesAPIConfig
+from litellm.types.llms.openai import (
+    ResponseCompletedEvent,
+    ResponsesAPIStreamEvents,
+    ResponsesAPIStreamingResponse,
+)
+from litellm.utils import CustomStreamWrapper
+
+
+class BaseResponsesAPIStreamingIterator:
+    """
+    Base class for streaming iterators that process responses from the Responses API.
+
+    This class contains shared logic for both synchronous and asynchronous iterators.
+    """
+
+    def __init__(
+        self,
+        response: httpx.Response,
+        model: str,
+        responses_api_provider_config: BaseResponsesAPIConfig,
+        logging_obj: LiteLLMLoggingObj,
+    ):
+        self.response = response
+        self.model = model
+        self.logging_obj = logging_obj
+        self.finished = False
+        self.responses_api_provider_config = responses_api_provider_config
+        self.completed_response: Optional[ResponsesAPIStreamingResponse] = None
+        self.start_time = datetime.now()
+
+    def _process_chunk(self, chunk):
+        """Process a single chunk of data from the stream"""
+        if not chunk:
+            return None
+
+        # Handle SSE format (data: {...})
+        chunk = CustomStreamWrapper._strip_sse_data_from_chunk(chunk)
+        if chunk is None:
+            return None
+
+        # Handle "[DONE]" marker
+        if chunk == STREAM_SSE_DONE_STRING:
+            self.finished = True
+            return None
+
+        try:
+            # Parse the JSON chunk
+            parsed_chunk = json.loads(chunk)
+
+            # Format as ResponsesAPIStreamingResponse
+            if isinstance(parsed_chunk, dict):
+                openai_responses_api_chunk = (
+                    self.responses_api_provider_config.transform_streaming_response(
+                        model=self.model,
+                        parsed_chunk=parsed_chunk,
+                        logging_obj=self.logging_obj,
+                    )
+                )
+                # Store the completed response
+                if (
+                    openai_responses_api_chunk
+                    and openai_responses_api_chunk.type
+                    == ResponsesAPIStreamEvents.RESPONSE_COMPLETED
+                ):
+                    self.completed_response = openai_responses_api_chunk
+                    self._handle_logging_completed_response()
+
+                return openai_responses_api_chunk
+
+            return None
+        except json.JSONDecodeError:
+            # If we can't parse the chunk, continue
+            return None
+
+    def _handle_logging_completed_response(self):
+        """Base implementation - should be overridden by subclasses"""
+        pass
+
+
+class ResponsesAPIStreamingIterator(BaseResponsesAPIStreamingIterator):
+    """
+    Async iterator for processing streaming responses from the Responses API.
+    """
+
+    def __init__(
+        self,
+        response: httpx.Response,
+        model: str,
+        responses_api_provider_config: BaseResponsesAPIConfig,
+        logging_obj: LiteLLMLoggingObj,
+    ):
+        super().__init__(response, model, responses_api_provider_config, logging_obj)
+        self.stream_iterator = response.aiter_lines()
+
+    def __aiter__(self):
+        return self
+
+    async def __anext__(self) -> ResponsesAPIStreamingResponse:
+        try:
+            while True:
+                # Get the next chunk from the stream
+                try:
+                    chunk = await self.stream_iterator.__anext__()
+                except StopAsyncIteration:
+                    self.finished = True
+                    raise StopAsyncIteration
+
+                result = self._process_chunk(chunk)
+
+                if self.finished:
+                    raise StopAsyncIteration
+                elif result is not None:
+                    return result
+                # If result is None, continue the loop to get the next chunk
+
+        except httpx.HTTPError as e:
+            # Handle HTTP errors
+            self.finished = True
+            raise e
+
+    def _handle_logging_completed_response(self):
+        """Handle logging for completed responses in async context"""
+        asyncio.create_task(
+            self.logging_obj.async_success_handler(
+                result=self.completed_response,
+                start_time=self.start_time,
+                end_time=datetime.now(),
+                cache_hit=None,
+            )
+        )
+
+        executor.submit(
+            self.logging_obj.success_handler,
+            result=self.completed_response,
+            cache_hit=None,
+            start_time=self.start_time,
+            end_time=datetime.now(),
+        )
+
+
+class SyncResponsesAPIStreamingIterator(BaseResponsesAPIStreamingIterator):
+    """
+    Synchronous iterator for processing streaming responses from the Responses API.
+    """
+
+    def __init__(
+        self,
+        response: httpx.Response,
+        model: str,
+        responses_api_provider_config: BaseResponsesAPIConfig,
+        logging_obj: LiteLLMLoggingObj,
+    ):
+        super().__init__(response, model, responses_api_provider_config, logging_obj)
+        self.stream_iterator = response.iter_lines()
+
+    def __iter__(self):
+        return self
+
+    def __next__(self):
+        try:
+            while True:
+                # Get the next chunk from the stream
+                try:
+                    chunk = next(self.stream_iterator)
+                except StopIteration:
+                    self.finished = True
+                    raise StopIteration
+
+                result = self._process_chunk(chunk)
+
+                if self.finished:
+                    raise StopIteration
+                elif result is not None:
+                    return result
+                # If result is None, continue the loop to get the next chunk
+
+        except httpx.HTTPError as e:
+            # Handle HTTP errors
+            self.finished = True
+            raise e
+
+    def _handle_logging_completed_response(self):
+        """Handle logging for completed responses in sync context"""
+        run_async_function(
+            async_function=self.logging_obj.async_success_handler,
+            result=self.completed_response,
+            start_time=self.start_time,
+            end_time=datetime.now(),
+            cache_hit=None,
+        )
+
+        executor.submit(
+            self.logging_obj.success_handler,
+            result=self.completed_response,
+            cache_hit=None,
+            start_time=self.start_time,
+            end_time=datetime.now(),
+        )
+
+
+class MockResponsesAPIStreamingIterator(BaseResponsesAPIStreamingIterator):
+    """
+    mock iterator - some models like o1-pro do not support streaming, we need to fake a stream
+    """
+
+    def __init__(
+        self,
+        response: httpx.Response,
+        model: str,
+        responses_api_provider_config: BaseResponsesAPIConfig,
+        logging_obj: LiteLLMLoggingObj,
+    ):
+        self.raw_http_response = response
+        super().__init__(
+            response=response,
+            model=model,
+            responses_api_provider_config=responses_api_provider_config,
+            logging_obj=logging_obj,
+        )
+        self.is_done = False
+
+    def __aiter__(self):
+        return self
+
+    async def __anext__(self) -> ResponsesAPIStreamingResponse:
+        if self.is_done:
+            raise StopAsyncIteration
+        self.is_done = True
+        transformed_response = (
+            self.responses_api_provider_config.transform_response_api_response(
+                model=self.model,
+                raw_response=self.raw_http_response,
+                logging_obj=self.logging_obj,
+            )
+        )
+        return ResponseCompletedEvent(
+            type=ResponsesAPIStreamEvents.RESPONSE_COMPLETED,
+            response=transformed_response,
+        )
+
+    def __iter__(self):
+        return self
+
+    def __next__(self) -> ResponsesAPIStreamingResponse:
+        if self.is_done:
+            raise StopIteration
+        self.is_done = True
+        transformed_response = (
+            self.responses_api_provider_config.transform_response_api_response(
+                model=self.model,
+                raw_response=self.raw_http_response,
+                logging_obj=self.logging_obj,
+            )
+        )
+        return ResponseCompletedEvent(
+            type=ResponsesAPIStreamEvents.RESPONSE_COMPLETED,
+            response=transformed_response,
+        )
diff --git a/.venv/lib/python3.12/site-packages/litellm/responses/utils.py b/.venv/lib/python3.12/site-packages/litellm/responses/utils.py
new file mode 100644
index 00000000..49d850ec
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/litellm/responses/utils.py
@@ -0,0 +1,97 @@
+from typing import Any, Dict, cast, get_type_hints
+
+import litellm
+from litellm.llms.base_llm.responses.transformation import BaseResponsesAPIConfig
+from litellm.types.llms.openai import (
+    ResponseAPIUsage,
+    ResponsesAPIOptionalRequestParams,
+)
+from litellm.types.utils import Usage
+
+
+class ResponsesAPIRequestUtils:
+    """Helper utils for constructing ResponseAPI requests"""
+
+    @staticmethod
+    def get_optional_params_responses_api(
+        model: str,
+        responses_api_provider_config: BaseResponsesAPIConfig,
+        response_api_optional_params: ResponsesAPIOptionalRequestParams,
+    ) -> Dict:
+        """
+        Get optional parameters for the responses API.
+
+        Args:
+            params: Dictionary of all parameters
+            model: The model name
+            responses_api_provider_config: The provider configuration for responses API
+
+        Returns:
+            A dictionary of supported parameters for the responses API
+        """
+        # Remove None values and internal parameters
+
+        # Get supported parameters for the model
+        supported_params = responses_api_provider_config.get_supported_openai_params(
+            model
+        )
+
+        # Check for unsupported parameters
+        unsupported_params = [
+            param
+            for param in response_api_optional_params
+            if param not in supported_params
+        ]
+
+        if unsupported_params:
+            raise litellm.UnsupportedParamsError(
+                model=model,
+                message=f"The following parameters are not supported for model {model}: {', '.join(unsupported_params)}",
+            )
+
+        # Map parameters to provider-specific format
+        mapped_params = responses_api_provider_config.map_openai_params(
+            response_api_optional_params=response_api_optional_params,
+            model=model,
+            drop_params=litellm.drop_params,
+        )
+
+        return mapped_params
+
+    @staticmethod
+    def get_requested_response_api_optional_param(
+        params: Dict[str, Any]
+    ) -> ResponsesAPIOptionalRequestParams:
+        """
+        Filter parameters to only include those defined in ResponsesAPIOptionalRequestParams.
+
+        Args:
+            params: Dictionary of parameters to filter
+
+        Returns:
+            ResponsesAPIOptionalRequestParams instance with only the valid parameters
+        """
+        valid_keys = get_type_hints(ResponsesAPIOptionalRequestParams).keys()
+        filtered_params = {k: v for k, v in params.items() if k in valid_keys}
+        return cast(ResponsesAPIOptionalRequestParams, filtered_params)
+
+
+class ResponseAPILoggingUtils:
+    @staticmethod
+    def _is_response_api_usage(usage: dict) -> bool:
+        """returns True if usage is from OpenAI Response API"""
+        if "input_tokens" in usage and "output_tokens" in usage:
+            return True
+        return False
+
+    @staticmethod
+    def _transform_response_api_usage_to_chat_usage(usage: dict) -> Usage:
+        """Tranforms the ResponseAPIUsage object to a Usage object"""
+        response_api_usage: ResponseAPIUsage = ResponseAPIUsage(**usage)
+        prompt_tokens: int = response_api_usage.input_tokens or 0
+        completion_tokens: int = response_api_usage.output_tokens or 0
+        return Usage(
+            prompt_tokens=prompt_tokens,
+            completion_tokens=completion_tokens,
+            total_tokens=prompt_tokens + completion_tokens,
+        )