Add pydocstyle to the ruff rules (#52136)

Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
Signed-off-by: Harry Mellor <hej.mellor@gmail.com>
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
This commit is contained in:
Harry Mellor
2026-09-16 16:44:29 +00:00
committed by GitHub
co-authored by Claude Opus 5
parent 804e5377aa
commit 8c1557a79c
2002 changed files with 15024 additions and 14269 deletions
+21 -22
View File
@@ -84,8 +84,7 @@ class StreamState:
class Parser:
"""
Abstract Parser class that unifies ReasoningParser and ToolParser into
"""Abstract Parser class that unifies ReasoningParser and ToolParser into
a single interface for parsing model output.
This class provides a unified way to handle both reasoning extraction
@@ -191,8 +190,7 @@ class Parser:
@abstractmethod
def is_reasoning_end(self, input_ids: list[int]) -> bool:
"""
Check if the reasoning content ends in the input_ids.
"""Check if the reasoning content ends in the input_ids.
Used by structured engines like `xgrammar` to check if the
reasoning content ends in the model output.
@@ -204,13 +202,13 @@ class Parser:
Returns:
True if the reasoning content ends in the input_ids.
"""
def is_reasoning_end_streaming(
self, input_ids: list[int], delta_ids: list[int]
) -> bool:
"""
Check if the reasoning content ends during a decode step.
"""Check if the reasoning content ends during a decode step.
Must be a pure function of the input_ids and delta_ids.
@@ -220,12 +218,12 @@ class Parser:
Returns:
True if the reasoning content ends in the delta_ids.
"""
return self.is_reasoning_end(input_ids)
def find_reasoning_end_offset(self, token_ids: Sequence[int]) -> int | None:
"""
Locate the token that ends reasoning within one window of tokens.
"""Locate the token that ends reasoning within one window of tokens.
Args:
token_ids: The tokens to examine.
@@ -235,13 +233,13 @@ class Parser:
``len(token_ids)`` when reasoning does not end in this window. Parsers
that cannot answer from a window alone always return ``None``; callers
fall back to :meth:`is_reasoning_end_streaming`.
"""
return None
@abstractmethod
def extract_content_ids(self, input_ids: list[int]) -> list[int]:
"""
Extract content token IDs from the input_ids.
"""Extract content token IDs from the input_ids.
This extracts the non-reasoning content (e.g., everything after
the </think> tag).
@@ -251,6 +249,7 @@ class Parser:
Returns:
The extracted content token IDs.
"""
@abstractmethod
@@ -259,8 +258,7 @@ class Parser:
model_output: str,
request: ChatCompletionRequest | ResponsesRequest,
) -> tuple[str | None, str | None]:
"""
Extract reasoning content from a complete model-generated string.
"""Extract reasoning content from a complete model-generated string.
Used for non-streaming responses where we have the entire model
response available before sending to the client.
@@ -271,6 +269,7 @@ class Parser:
Returns:
A tuple of (reasoning, response_content).
"""
@abstractmethod
@@ -283,8 +282,7 @@ class Parser:
current_token_ids: Sequence[int],
delta_token_ids: Sequence[int],
) -> DeltaMessage | None:
"""
Extract reasoning content from a streaming delta message.
"""Extract reasoning content from a streaming delta message.
Args:
previous_text: Text from all previous tokens.
@@ -296,6 +294,7 @@ class Parser:
Returns:
A DeltaMessage with reasoning and/or content fields, or None.
"""
# ========== Tool Parser Methods ==========
@@ -303,8 +302,7 @@ class Parser:
def adjust_request(
self, request: ChatCompletionRequest | ResponsesRequest
) -> ChatCompletionRequest | ResponsesRequest:
"""
Adjust the request parameters for tool calling.
"""Adjust the request parameters for tool calling.
Can be overridden by subclasses to modify request parameters
(e.g., setting structured output schemas for tool calling).
@@ -314,6 +312,7 @@ class Parser:
Returns:
The adjusted request.
"""
return request
@@ -323,8 +322,7 @@ class Parser:
model_output: str,
request: ChatCompletionRequest | ResponsesRequest,
) -> ExtractedToolCallInformation:
"""
Extract tool calls from a complete model-generated string.
"""Extract tool calls from a complete model-generated string.
Used for non-streaming responses.
@@ -334,6 +332,7 @@ class Parser:
Returns:
ExtractedToolCallInformation containing the tool calls.
"""
@abstractmethod
@@ -347,8 +346,7 @@ class Parser:
delta_token_ids: Sequence[int],
request: ChatCompletionRequest | ResponsesRequest,
) -> DeltaMessage | None:
"""
Extract tool calls from a streaming delta message.
"""Extract tool calls from a streaming delta message.
Args:
previous_text: Text from all previous tokens.
@@ -361,6 +359,7 @@ class Parser:
Returns:
A DeltaMessage with tool_calls field, or None.
"""
@abstractmethod
@@ -381,6 +380,7 @@ class Parser:
Returns:
A tuple of (reasoning, content, tool_calls).
"""
@abstractmethod
@@ -403,8 +403,7 @@ class Parser:
class DelegatingParser(Parser):
"""
A Parser implementation that delegates to separate ReasoningParser and
"""A Parser implementation that delegates to separate ReasoningParser and
ToolParser instances.
This is the recommended base class for creating model-specific parsers
+4 -8
View File
@@ -169,8 +169,7 @@ def collect_tool_schema(tool_schema: list[CohereNormalizedTool]) -> str:
def _tool_definitions_to_schema_list(
tools: str | list[Any],
) -> list[CohereNormalizedTool]:
"""
Build the list of ``CohereNormalizedTool`` dicts expected by
"""Build the list of ``CohereNormalizedTool`` dicts expected by
``collect_tool_schema``.
Accepts:
@@ -214,8 +213,7 @@ def _tool_definitions_to_schema_list(
def _has_effective_tools(
tools: str | list[Any] | None,
) -> TypeGuard[str | list[Any]]:
"""
True when ``tools`` contains at least one tool definition to convert.
"""True when ``tools`` contains at least one tool definition to convert.
``ResponsesRequest`` defaults ``tools`` to ``[]``; ``ChatCompletionRequest``
uses ``None``. Both mean "no tools" here. Strings (e.g. a JSON blob) are
@@ -236,8 +234,7 @@ def convert_schema_to_structural_tags(
tools: str | list[Any] | None = None,
model_architecture: str | None = None,
) -> str | None:
"""
Returns a response_format string accepted by xgrammar's structural tag format.
"""Returns a response_format string accepted by xgrammar's structural tag format.
Uses the canonical shape: {"type": "structural_tag", "format": {...}} with
format.type "triggered_tags" and tag content type "json_schema" or "grammar".
@@ -325,8 +322,7 @@ def _unwrap_nested_schema(candidate: Any) -> dict | None:
def _schema_from_json_schema_field(js_wr: Any) -> dict | None:
"""
Extract the JSON Schema object from Chat Completions ``json_schema`` payload.
"""Extract the JSON Schema object from Chat Completions ``json_schema`` payload.
Accepts:
- ``JsonSchemaResponseFormat`` (Pydantic) with ``schema`` / ``json_schema`` field
+1 -2
View File
@@ -1,7 +1,6 @@
# SPDX-License-Identifier: Apache-2.0
# SPDX-FileCopyrightText: Copyright contributors to the vLLM project
"""
Streaming parser engine framework for tool call and reasoning extraction.
"""Streaming parser engine framework for tool call and reasoning extraction.
Instead of hand-rolling a parser for every model's tool-call / reasoning
format, each format is declared as a ParserEngineConfig (terminals,
+2 -1
View File
@@ -1,7 +1,8 @@
# SPDX-License-Identifier: Apache-2.0
# SPDX-FileCopyrightText: Copyright contributors to the vLLM project
"""Incremental text lexer that converts text chunks into terminal
tokens, with prefix-match buffering for ambiguous boundaries."""
tokens, with prefix-match buffering for ambiguous boundaries.
"""
from __future__ import annotations
@@ -1,7 +1,8 @@
# SPDX-License-Identifier: Apache-2.0
# SPDX-FileCopyrightText: Copyright contributors to the vLLM project
"""Streaming parser engine that orchestrates token ID scanning,
incremental lexing, and state-machine-driven semantic event emission."""
incremental lexing, and state-machine-driven semantic event emission.
"""
from __future__ import annotations
+2 -1
View File
@@ -1,7 +1,8 @@
# SPDX-License-Identifier: Apache-2.0
# SPDX-FileCopyrightText: Copyright contributors to the vLLM project
"""Scan delta token IDs for special tokens and split the stream into
pre-lexed terminals and plain text chunks."""
pre-lexed terminals and plain text chunks.
"""
from __future__ import annotations
+1
View File
@@ -83,6 +83,7 @@ def _parse_gemma4_args(args_str: str, *, partial: bool = False) -> dict:
(e.g. partial boolean parsed as bare string).
Returns a dict ready for ``json.dumps()``.
"""
if not args_str or not args_str.strip():
return {}
+2 -1
View File
@@ -67,7 +67,8 @@ _WS = " \t\r\n"
def _scan_json_value(raw: str, start: int) -> int | None:
"""Return the end index (exclusive) of the JSON object starting at
``raw[start]``, or ``None`` when the object is still unterminated."""
``raw[start]``, or ``None`` when the object is still unterminated.
"""
depth = 0
in_string = False
escape = False
+2
View File
@@ -136,6 +136,7 @@ def mistral_config(
Returns:
A frozen :class:`ParserEngineConfig` with ``initial_state=CONTENT``.
"""
if reasoning_encoding == "special_token":
think_start = _THINK_START_SPECIAL
@@ -467,6 +468,7 @@ class MistralParser(ParserEngine):
Returns:
A JSON Schema dict, or ``None`` if the named tool is not found.
"""
tool_choice = request.tool_choice
tools = request.tools or []
+3 -5
View File
@@ -16,9 +16,7 @@ logger = init_logger(__name__)
class ParserManager:
"""
Provides a unified Parser by composing reasoning and tool parser adapters.
"""
"""Provides a unified Parser by composing reasoning and tool parser adapters."""
@classmethod
def get_tool_parser(
@@ -80,8 +78,7 @@ class ParserManager:
model_name: str | None = None,
is_harmony: bool = False,
) -> type[Parser] | None:
"""
Get a Parser that handles both reasoning and tool parsing.
"""Get a Parser that handles both reasoning and tool parsing.
Composes the individual parsers into a ``DelegatingParser`` subclass.
@@ -95,6 +92,7 @@ class ParserManager:
Returns:
A Parser class, or None if neither parser is specified.
"""
if not tool_parser_name and not reasoning_parser_name:
return None