Update Optional[x] -> x | None and Union[x, y] to x | y (#26633)

Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
This commit is contained in:
Harry Mellor
2025-10-12 09:51:31 -07:00
committed by GitHub
parent 9bb38130cb
commit 8fcaaf6a16
944 changed files with 9490 additions and 10121 deletions
+36 -36
View File
@@ -6,11 +6,11 @@ import json
import time
import uuid
from collections import deque
from collections.abc import AsyncGenerator, AsyncIterator, Sequence
from collections.abc import AsyncGenerator, AsyncIterator, Callable, Sequence
from contextlib import AsyncExitStack
from copy import copy
from http import HTTPStatus
from typing import Callable, Final, Optional, Union
from typing import Final
import jinja2
from fastapi import Request
@@ -109,14 +109,14 @@ class OpenAIServingResponses(OpenAIServing):
engine_client: EngineClient,
models: OpenAIServingModels,
*,
request_logger: Optional[RequestLogger],
chat_template: Optional[str],
request_logger: RequestLogger | None,
chat_template: str | None,
chat_template_content_format: ChatTemplateContentFormatOption,
return_tokens_as_token_ids: bool = False,
reasoning_parser: str = "",
enable_auto_tools: bool = False,
tool_parser: Optional[str] = None,
tool_server: Optional[ToolServer] = None,
tool_parser: str | None = None,
tool_server: ToolServer | None = None,
enable_prompt_tokens_details: bool = False,
enable_force_include_usage: bool = False,
enable_log_outputs: bool = False,
@@ -210,7 +210,7 @@ class OpenAIServingResponses(OpenAIServing):
def _validate_generator_input(
self, engine_prompt: EngineTokensPrompt
) -> Optional[ErrorResponse]:
) -> ErrorResponse | None:
"""Add validations to the input to the generator here."""
if self.max_model_len <= len(engine_prompt["prompt_token_ids"]):
error_message = (
@@ -229,12 +229,12 @@ class OpenAIServingResponses(OpenAIServing):
async def create_responses(
self,
request: ResponsesRequest,
raw_request: Optional[Request] = None,
) -> Union[
AsyncGenerator[StreamingResponsesResponse, None],
ResponsesResponse,
ErrorResponse,
]:
raw_request: Request | None = None,
) -> (
AsyncGenerator[StreamingResponsesResponse, None]
| ResponsesResponse
| ErrorResponse
):
error_check_ret = await self._check_model(request)
if error_check_ret is not None:
logger.error("Error with model %s", error_check_ret)
@@ -460,7 +460,7 @@ class OpenAIServingResponses(OpenAIServing):
async def _make_request(
self,
request: ResponsesRequest,
prev_response: Optional[ResponsesResponse],
prev_response: ResponsesResponse | None,
tokenizer: AnyTokenizer,
):
if len(request.tools) > 0:
@@ -481,7 +481,7 @@ class OpenAIServingResponses(OpenAIServing):
def _make_request_with_harmony(
self,
request: ResponsesRequest,
prev_response: Optional[ResponsesResponse],
prev_response: ResponsesResponse | None,
):
if request.tool_choice != "auto":
raise NotImplementedError(
@@ -522,8 +522,8 @@ class OpenAIServingResponses(OpenAIServing):
model_name: str,
tokenizer: AnyTokenizer,
request_metadata: RequestResponseMetadata,
created_time: Optional[int] = None,
) -> Union[ErrorResponse, ResponsesResponse]:
created_time: int | None = None,
) -> ErrorResponse | ResponsesResponse:
if created_time is None:
created_time = int(time.time())
@@ -642,9 +642,9 @@ class OpenAIServingResponses(OpenAIServing):
def _create_response_logprobs(
self,
token_ids: Sequence[int],
logprobs: Optional[SampleLogprobs],
logprobs: SampleLogprobs | None,
tokenizer: AnyTokenizer,
top_logprobs: Optional[int] = None,
top_logprobs: int | None = None,
) -> list[Logprob]:
assert logprobs is not None, "logprobs must be provided"
assert len(token_ids) == len(logprobs), (
@@ -676,9 +676,9 @@ class OpenAIServingResponses(OpenAIServing):
def _create_stream_response_logprobs(
self,
token_ids: Sequence[int],
logprobs: Optional[SampleLogprobs],
logprobs: SampleLogprobs | None,
tokenizer: AnyTokenizer,
top_logprobs: Optional[int] = None,
top_logprobs: int | None = None,
) -> list[response_text_delta_event.Logprob]:
lgs = self._create_response_logprobs(
token_ids=token_ids,
@@ -793,7 +793,7 @@ class OpenAIServingResponses(OpenAIServing):
def _construct_input_messages(
self,
request: ResponsesRequest,
prev_response: Optional[ResponsesResponse] = None,
prev_response: ResponsesResponse | None = None,
) -> list[ChatCompletionMessageParam]:
messages: list[ChatCompletionMessageParam] = []
if request.instructions:
@@ -833,7 +833,7 @@ class OpenAIServingResponses(OpenAIServing):
def _construct_input_messages_with_harmony(
self,
request: ResponsesRequest,
prev_response: Optional[ResponsesResponse],
prev_response: ResponsesResponse | None,
) -> list[OpenAIHarmonyMessage]:
messages: list[OpenAIHarmonyMessage] = []
if prev_response is None:
@@ -986,7 +986,7 @@ class OpenAIServingResponses(OpenAIServing):
async def responses_background_stream_generator(
self,
response_id: str,
starting_after: Optional[int] = None,
starting_after: int | None = None,
) -> AsyncGenerator[StreamingResponsesResponse, None]:
if response_id not in self.event_store:
raise ValueError(f"Unknown response_id: {response_id}")
@@ -1011,13 +1011,13 @@ class OpenAIServingResponses(OpenAIServing):
async def retrieve_responses(
self,
response_id: str,
starting_after: Optional[int],
stream: Optional[bool],
) -> Union[
ErrorResponse,
ResponsesResponse,
AsyncGenerator[StreamingResponsesResponse, None],
]:
starting_after: int | None,
stream: bool | None,
) -> (
ErrorResponse
| ResponsesResponse
| AsyncGenerator[StreamingResponsesResponse, None]
):
async with self.response_store_lock:
response = self.response_store.get(response_id)
@@ -1034,7 +1034,7 @@ class OpenAIServingResponses(OpenAIServing):
async def cancel_responses(
self,
response_id: str,
) -> Union[ErrorResponse, ResponsesResponse]:
) -> ErrorResponse | ResponsesResponse:
async with self.response_store_lock:
response = self.response_store.get(response_id)
if response is None:
@@ -1082,7 +1082,7 @@ class OpenAIServingResponses(OpenAIServing):
self,
request: ResponsesRequest,
sampling_params: SamplingParams,
result_generator: AsyncIterator[Optional[ConversationContext]],
result_generator: AsyncIterator[ConversationContext | None],
context: ConversationContext,
model_name: str,
tokenizer: AnyTokenizer,
@@ -1385,7 +1385,7 @@ class OpenAIServingResponses(OpenAIServing):
self,
request: ResponsesRequest,
sampling_params: SamplingParams,
result_generator: AsyncIterator[Optional[ConversationContext]],
result_generator: AsyncIterator[ConversationContext | None],
context: ConversationContext,
model_name: str,
tokenizer: AnyTokenizer,
@@ -1772,12 +1772,12 @@ class OpenAIServingResponses(OpenAIServing):
self,
request: ResponsesRequest,
sampling_params: SamplingParams,
result_generator: AsyncIterator[Optional[ConversationContext]],
result_generator: AsyncIterator[ConversationContext | None],
context: ConversationContext,
model_name: str,
tokenizer: AnyTokenizer,
request_metadata: RequestResponseMetadata,
created_time: Optional[int] = None,
created_time: int | None = None,
) -> AsyncGenerator[StreamingResponsesResponse, None]:
# TODO:
# 1. Handle disconnect