Latest tooling (#804)

Author: jamesbraza

.pre-commit-config.yaml

@@ -23,7 +23,7 @@ repos:
       - id: mixed-line-ending
       - id: trailing-whitespace
   - repo: https://github.com/astral-sh/ruff-pre-commit
-    rev: v0.8.4
+    rev: v0.9.1
     hooks:
       - id: ruff
         args: [--fix, --exit-non-zero-on-fix]
@@ -55,15 +55,15 @@ repos:
     hooks:
       - id: check-mailmap
   - repo: https://github.com/henryiii/validate-pyproject-schema-store
-    rev: 2024.11.25
+    rev: 2025.01.10
     hooks:
       - id: validate-pyproject
   - repo: https://github.com/astral-sh/uv-pre-commit
     rev: 0.4.30
     hooks:
       - id: uv-lock
   - repo: https://github.com/renovatebot/pre-commit-hooks
-    rev: 39.75.1
+    rev: 39.100.1
     hooks:
       - id: renovate-config-validator
         args: [--strict]
@@ -72,11 +72,11 @@ repos:
     hooks:
       - id: blacken-docs
   - repo: https://github.com/jsh9/markdown-toc-creator
-    rev: 0.0.8
+    rev: 0.0.10
     hooks:
       - id: markdown-toc-creator
   - repo: https://github.com/pre-commit/mirrors-mypy
-    rev: v1.13.0
+    rev: v1.14.1
     hooks:
       - id: mypy
         args: [--pretty, --ignore-missing-imports]

paperqa/agents/search.py

@@ -9,7 +9,7 @@
 import pickle
 import warnings
 import zlib
-from collections.abc import Awaitable, Callable, Collection, Sequence
+from collections.abc import Callable, Collection, Sequence
 from enum import StrEnum, auto
 from typing import TYPE_CHECKING, Any, ClassVar
 from uuid import UUID
@@ -456,7 +456,7 @@ async def process_file(
     manifest: dict[str, Any],
     semaphore: anyio.Semaphore,
     settings: Settings,
-    progress_bar_update: Callable[[], Awaitable] | None = None,
+    progress_bar_update: Callable[[], Any] | None = None,
 ) -> None:
     abs_file_path = (
         pathlib.Path(settings.agent.index.paper_directory).absolute() / rel_file_path
@@ -500,7 +500,7 @@ async def process_file(
                 # separate process_file invocation leads to a segfault or crash
                 await search_index.save_index()
                 if progress_bar_update:
-                    await progress_bar_update()
+                    progress_bar_update()
                 return
 
             this_doc = next(iter(tmp_docs.docs.values()))
@@ -527,15 +527,15 @@ async def process_file(
 
         # Update progress bar for either a new or previously indexed file
         if progress_bar_update:
-            await progress_bar_update()
+            progress_bar_update()
 
 
 WARN_IF_INDEXING_MORE_THAN = 999
 
 
 def _make_progress_bar_update(
     sync_index_w_directory: bool, total: int
-) -> tuple[contextlib.AbstractContextManager, Callable[[], Awaitable] | None]:
+) -> tuple[contextlib.AbstractContextManager, Callable[[], Any] | None]:
     # Disable should override enable
     env_var_disable = (
         os.environ.get("PQA_INDEX_DISABLE_PROGRESS_BAR", "").lower() in ENV_VAR_MATCH
@@ -562,7 +562,7 @@ def _make_progress_bar_update(
         )
         task_id = progress.add_task("Indexing...", total=total)
 
-        async def progress_bar_update() -> None:
+        def progress_bar_update() -> None:
             progress.update(task_id, advance=1)
 
         return progress, progress_bar_update

paperqa/clients/openalex.py

@@ -33,7 +33,7 @@ def reformat_name(name: str) -> str:
     return f"{given_names} {family}"
 
 
-async def get_openalex_mailto() -> str | None:
+def get_openalex_mailto() -> str | None:
     """Get the OpenAlex mailto address.
 
     Returns:
@@ -71,7 +71,7 @@ async def get_doc_details_from_openalex(
         ValueError: If neither DOI nor title is provided.
         DOINotFoundError: If the paper cannot be found.
     """
-    mailto = await get_openalex_mailto()
+    mailto = get_openalex_mailto()
     params = {"mailto": mailto} if mailto else {}
 
     if doi is title is None:
@@ -126,10 +126,10 @@ async def get_doc_details_from_openalex(
         if doi and results_data.get("doi") != doi:
             raise DOINotFoundError(f"DOI {doi!r} not found in OpenAlex.")
 
-        return await parse_openalex_to_doc_details(results_data)
+        return parse_openalex_to_doc_details(results_data)
 
 
-async def parse_openalex_to_doc_details(message: dict[str, Any]) -> DocDetails:
+def parse_openalex_to_doc_details(message: dict[str, Any]) -> DocDetails:
     """Parse OpenAlex API response to DocDetails.
 
     Args:

pyproject.toml

@@ -287,12 +287,12 @@ enable_all = true
 ignore = [
     "FURB101",  # Rely on ruff FURB101 for this
     "FURB103",  # Rely on ruff FURB103 for this
-    "FURB141",  # FURB141, FURB144, FURB146, FURB147, FURB150, FURB155: don't care to enforce pathlib
-    "FURB144",
-    "FURB146",
-    "FURB147",
-    "FURB150",
-    "FURB155",
+    "FURB141",  # Rely on ruff PTH110 for this
+    "FURB144",  # Rely on ruff PTH107 for this
+    "FURB146",  # Rely on ruff PTH113 for this
+    "FURB147",  # Rely on ruff PTH118 for this
+    "FURB150",  # Rely on ruff PTH102 for this
+    "FURB155",  # Rely on ruff PTH202 for this
 ]
 
 [tool.ruff]
@@ -310,10 +310,13 @@ preview = true
 [tool.ruff.lint]
 explicit-preview-rules = true
 extend-select = [
-    "A006",
+    "AIR301",
+    "AIR302",
+    "AIR303",
     "ASYNC116",
     "B901",
     "B909",
+    "B911",
     "C420",
     "DOC202",
     "DOC403",
@@ -359,6 +362,7 @@ extend-select = [
     "E305",
     "E306",
     "E502",
+    "FAST003",
     "FURB101",
     "FURB103",
     "FURB110",
@@ -379,32 +383,51 @@ extend-select = [
     "FURB166",
     "FURB171",
     "FURB180",
-    "FURB181",
-    "FURB188",
     "FURB189",
     "FURB192",
     "LOG015",
+    "PLC1802",
     "PLC1901",
     "PLC2701",
     "PLC2801",
     "PLE0304",
     "PLE1141",
     "PLE4703",
+    "PLR0202",
+    "PLR0203",
+    "PLR0916",
+    "PLR1733",
+    "PLR6104",
     "PLR6201",
     "PLW0108",
+    "PLW0177",
+    "PLW1507",
+    "PLW3201",
+    "PTH210",
     "PYI059",
+    "PYI061",
     "RUF027",
     "RUF028",
+    "RUF029",
     "RUF031",
-    "RUF032",
-    "RUF033",
-    "RUF034",
     "RUF035",
     "RUF036",
+    "RUF037",
     "RUF038",
     "RUF039",
+    "RUF040",
+    "RUF041",
+    "RUF043",
     "RUF048",
+    "RUF049",
+    "RUF051",
+    "RUF052",
+    "RUF055",
+    "RUF056",
+    "RUF057",
     "SIM905",
+    "TC007",
+    "TC008",
     "UP042",
     "UP044",
     "W391",
@@ -416,6 +439,7 @@ external = [
     "FURB",  # refurb
 ]
 ignore = [
+    "A005",  # Overly pedantic
     "ANN",  # Don't care to enforce typing
     "BLE001",  # Don't care to enforce blind exception catching
     "C901",  # we can be complex
@@ -453,6 +477,7 @@ ignore = [
     "PLW2901",  # Allow modifying loop variables
     "PTH",  # Overly pedantic
     "RUF027",  # Prompt templates may not be f-strings
+    "RUF052",  # Previous code uses leading underscore to indicate throwaway
     "S311",  # Ok to use python random
     "SLF001",  # Overly pedantic
     "T201",  # Overly pedantic

tests/test_agents.py

@@ -331,7 +331,9 @@ async def test_successful_memory_agent(agent_test_settings: Settings) -> None:
     thoughts: list[str] = []
     orig_llm_model_call = MultipleCompletionLLMModel.call
 
-    async def on_agent_action(action: OpResult[ToolRequestMessage], *_) -> None:
+    async def on_agent_action(  # noqa: RUF029
+        action: OpResult[ToolRequestMessage], *_
+    ) -> None:
         thoughts.append(extract_thought(content=action.value.content))
 
     async def llm_model_call(*args, **kwargs):
@@ -415,7 +417,7 @@ async def test_gather_evidence_rejects_empty_docs(
 ) -> None:
 
     @wraps(GenerateAnswer.gen_answer)
-    async def gen_answer(self, state) -> str:  # noqa: ARG001
+    async def gen_answer(self, state) -> str:  # noqa: ARG001, RUF029
         return f"{CANNOT_ANSWER_PHRASE}."
 
     # Patch GenerateAnswerTool.gen_answer so that if this tool is chosen first,
@@ -919,7 +921,7 @@ async def test_unsure_answer(
     ) -> None:
         unsure_answer = "Based on the sources provided, it appears no one has done x."
 
-        async def emulate_answered_but_unsure(
+        async def emulate_answered_but_unsure(  # noqa: RUF029
             *_, query: PQASession, **__
         ) -> PQASession:
             query.answer = unsure_answer

tests/test_clients.py

@@ -529,10 +529,9 @@ async def test_ensure_robust_to_timeouts() -> None:
     assert details is None, "Should return None for timeout"
 
 
-@pytest.mark.asyncio
-async def test_bad_init() -> None:
+def test_bad_init() -> None:
     with pytest.raises(
-        ValueError, match="At least one MetadataProvider must be provided."
+        ValueError, match=r"At least one MetadataProvider must be provided."
     ):
         client = DocMetadataClient(clients=[])  # noqa: F841