Updated code quality tooling (#29)

* Ran 'pre-commit autoupdate'

* Ran 'uv lock --upgrade'

* Updated pytest disables

* Regenerated cassettes so tests pass

* Removed dead test cassettes

* Allowing latest Python 3.13 in CI since pylint fixed their issue

* Propagated fetch-depth comment in CI

* Modernized codeflash CI and fixed pull_request path

* Fixed codeflash formatter-cmds

* Added Renovate config with validation in CI

* Removed pytest-mock dependency

* Added mutable-override to mypy

* Added doctesting to pytest config

* Modernized refurb config for ruff port

* Added blacken-docs and removed extra exclude for codespell

* Updated ruff config and all fixes for latest ruff

.github/renovate.json5

@@ -0,0 +1,28 @@
+{
+  $schema: "https://docs.renovatebot.com/renovate-schema.json",
+  extends: ["config:recommended"],
+  schedule: ["* 2 1-7 * 1"],
+  prHourlyLimit: 4,
+  timezone: "America/Los_Angeles",
+  rangeStrategy: "widen",
+  lockFileMaintenance: {
+    enabled: true,
+    schedule: ["* 2 1-7 * 1"], // Work around https://github.com/renovatebot/renovate/discussions/33152
+  },
+  minimumReleaseAge: "2 weeks",
+  "pre-commit": { enabled: true },
+  packageRules: [
+    {
+      matchUpdateTypes: ["lockFileMaintenance"],
+      automerge: true,
+    },
+    {
+      // group:allNonMajor, with automerge
+      groupName: "all non-major dependencies",
+      groupSlug: "all-minor-patch",
+      matchPackageNames: ["*"],
+      matchUpdateTypes: ["minor", "patch"],
+      automerge: true,
+    },
+  ],
+}

.github/workflows/codeflash.yaml

@@ -3,7 +3,7 @@ name: CodeFlash
 on:
   pull_request:
     paths:
-      - "ldp/**"
+      - "llmclient/**"
   workflow_dispatch:
 
 concurrency: # Cancel prior if new push, SEE: https://stackoverflow.com/a/72408109
@@ -13,31 +13,16 @@ concurrency: # Cancel prior if new push, SEE: https://stackoverflow.com/a/724081
 jobs:
   optimize: # SEE: https://docs.codeflash.ai/getting-started/codeflash-github-actions
     runs-on: ubuntu-latest
+    if: ${{ github.actor != 'codeflash-ai[bot]' }}
     env:
       CODEFLASH_API_KEY: ${{ secrets.CODEFLASH_API_KEY }}
       CODEFLASH_PR_NUMBER: ${{ github.event.number }}
     steps:
-      - name: Check if PR is from CodeFlash bot
-        id: bot_check
-        working-directory: .
-        run: |
-          echo "Checking if this PR is created by CodeFlash bot..."
-          if [ "${{ github.event.pull_request.user.login }}" == "codeflash-ai[bot]" ]; then
-            echo "PR created by Codeflash bot. Skipping optimization."
-            echo "skip_remaining_steps=yes" >> $GITHUB_OUTPUT
-          else
-            echo "skip_remaining_steps=no" >> $GITHUB_OUTPUT
-            echo "It's not. Proceeding with the optimization."
-          fi
-      - if: steps.bot_check.outputs.skip_remaining_steps == 'no'
-        uses: actions/checkout@v4
+      - uses: actions/checkout@v4
         with:
           fetch-depth: 0
-      - if: steps.bot_check.outputs.skip_remaining_steps == 'no'
-        uses: astral-sh/setup-uv@v3
+      - uses: astral-sh/setup-uv@v4
         with:
           enable-cache: true
-      - if: steps.bot_check.outputs.skip_remaining_steps == 'no'
-        run: uv sync
-      - if: steps.bot_check.outputs.skip_remaining_steps == 'no'
-        run: uv run codeflash
+      - run: uv sync --group=codeflash
+      - run: uv run codeflash

.github/workflows/test.yaml

@@ -12,7 +12,7 @@ jobs:
     if: github.event_name == 'pull_request' # pre-commit-ci/lite-action only runs here
     strategy:
       matrix:
-        python-version: [3.11, 3.13.0] # Our min and max supported Python versions
+        python-version: [3.11, 3.13] # Our min and max supported Python versions
     steps:
       - uses: actions/checkout@v4
         with:
@@ -27,26 +27,28 @@ jobs:
     runs-on: ubuntu-latest
     strategy:
       matrix:
-        python-version: [3.11, 3.13.0] # Our min and max supported Python versions
+        python-version: [3.11, 3.13] # Our min and max supported Python versions
     steps:
       - uses: actions/checkout@v4
         with:
-          fetch-depth: 0
+          fetch-depth: 0 # For setuptools-scm, replace with fetch-tags after https://github.com/actions/checkout/issues/1471
       - uses: astral-sh/setup-uv@v4
         with:
           enable-cache: true
       - run: uv python pin ${{ matrix.python-version }}
       - name: Check build
-        if: matrix.python-version == '3.11'
+        if: matrix.python-version == '3.11' # Only need to run this on one version
         uses: hynek/build-and-inspect-python-package@v2
       - run: uv sync --python-preference=only-managed
       - run: uv run refurb llmclient tests
       - run: uv run pylint llmclient
+      - if: matrix.python-version == '3.11' # Only need to run this on one version
+        uses: suzuki-shunsuke/[email protected]
   test:
     runs-on: ubuntu-latest
     strategy:
       matrix:
-        python-version: [3.11, 3.13.0] # Our min and max supported Python versions
+        python-version: [3.11, 3.13] # Our min and max supported Python versions
     steps:
       - uses: actions/checkout@v4
       - uses: astral-sh/setup-uv@v3

.pre-commit-config.yaml

@@ -1,3 +1,4 @@
+---
 default_language_version:
   python: python3
 repos:
@@ -18,18 +19,22 @@ repos:
       - id: mixed-line-ending
       - id: trailing-whitespace
   - repo: https://github.com/astral-sh/ruff-pre-commit
-    rev: v0.8.0
+    rev: v0.8.6
     hooks:
       - id: ruff
         args: [--fix, --exit-non-zero-on-fix]
   - repo: https://github.com/rbubley/mirrors-prettier
-    rev: v3.4.0
+    rev: v3.4.2
     hooks:
       - id: prettier
   - repo: https://github.com/psf/black-pre-commit-mirror
     rev: 24.10.0
     hooks:
       - id: black
+  - repo: https://github.com/adamchainz/blacken-docs
+    rev: 1.19.1
+    hooks:
+      - id: blacken-docs
   - repo: https://github.com/jumanjihouse/pre-commit-hooks
     rev: 3.0.0
     hooks:
@@ -40,10 +45,6 @@ repos:
       - id: codespell
         additional_dependencies: [".[toml]"]
         exclude_types: [jupyter]
-        exclude: |
-          (?x)^(
-            tests/cassettes.*
-          )$
   - repo: https://github.com/pappasam/toml-sort
     rev: v0.24.2
     hooks:
@@ -62,11 +63,11 @@ repos:
     hooks:
       - id: uv-lock
   - 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]

llmclient/cost_tracker.py

@@ -1,7 +1,7 @@
 import contextvars
 import logging
 from collections.abc import Awaitable, Callable
-from contextlib import asynccontextmanager
+from contextlib import contextmanager
 from functools import wraps
 from typing import ParamSpec, TypeVar
 
@@ -44,8 +44,8 @@ def enable_cost_tracking(enabled: bool = True) -> None:
     GLOBAL_COST_TRACKER.enabled.set(enabled)
 
 
-@asynccontextmanager
-async def cost_tracking_ctx(enabled: bool = True):
+@contextmanager
+def cost_tracking_ctx(enabled: bool = True):
     prev = GLOBAL_COST_TRACKER.enabled.get()
     GLOBAL_COST_TRACKER.enabled.set(enabled)
     try:

llmclient/embeddings.py

@@ -28,7 +28,8 @@ class EmbeddingModel(ABC, BaseModel):
     config: dict[str, Any] = Field(
         default_factory=dict,
         description=(
-            "Optional `rate_limit` key, value must be a RateLimitItem or RateLimitItem string for parsing"
+            "Optional `rate_limit` key, value must be a RateLimitItem or RateLimitItem"
+            " string for parsing"
         ),
     )
 
@@ -162,7 +163,7 @@ class SparseEmbeddingModel(EmbeddingModel):
     model_config = ConfigDict(arbitrary_types_allowed=True)
 
     name: str = "sparse"
-    ndim: int = 256
+    ndim: int = 256  # type: ignore[mutable-override]
     enc: tiktoken.Encoding = Field(
         default_factory=lambda: tiktoken.get_encoding("cl100k_base")
     )
@@ -221,7 +222,8 @@ def __init__(self, **kwargs):
             from sentence_transformers import SentenceTransformer
         except ImportError as exc:
             raise ImportError(
-                "Please install fh-llm-client[local] to use SentenceTransformerEmbeddingModel."
+                "Please install fh-llm-client[local] to use"
+                " SentenceTransformerEmbeddingModel."
             ) from exc
 
         self._model = SentenceTransformer(self.name)

llmclient/rate_limiter.py

@@ -82,7 +82,7 @@ class GlobalRateLimiter:
     def __init__(
         self,
         rate_config: (
-            None | dict[tuple[str, str | MatchAllInputs], RateLimitItem]
+            dict[tuple[str, str | MatchAllInputs], RateLimitItem] | None
         ) = None,
         use_in_memory: bool = False,
     ):
@@ -194,25 +194,25 @@ def parse_rate_limits_and_namespace(
         # this needs to be checked first, since it's more specific than the stub machine id
         if (namespace_w_machine_id_stripped, primary_key) in self.rate_config:
             return (
-                self.rate_config[(namespace_w_machine_id_stripped, primary_key)],
+                self.rate_config[namespace_w_machine_id_stripped, primary_key],
                 namespace_w_machine_id_stripped,
             )
         # we keep the old namespace if we match on the namespace_w_stub_machine_id
         if (namespace_w_stub_machine_id, primary_key) in self.rate_config:
             return (
-                self.rate_config[(namespace_w_stub_machine_id, primary_key)],
+                self.rate_config[namespace_w_stub_machine_id, primary_key],
                 namespace,
             )
         # again we only want the original namespace, keep the old namespace
         if (namespace_w_stub_machine_id, MATCH_ALL) in self.rate_config:
             return (
-                self.rate_config[(namespace_w_stub_machine_id, MATCH_ALL)],
+                self.rate_config[namespace_w_stub_machine_id, MATCH_ALL],
                 namespace,
             )
         # again we want to use the stripped namespace if it matches
         if (namespace_w_machine_id_stripped, MATCH_ALL) in self.rate_config:
             return (
-                self.rate_config[(namespace_w_machine_id_stripped, MATCH_ALL)],
+                self.rate_config[namespace_w_machine_id_stripped, MATCH_ALL],
                 namespace_w_machine_id_stripped,
             )
         return FALLBACK_RATE_LIMIT, namespace
@@ -294,7 +294,7 @@ async def rate_limit_status(self):
                 rate_limit.amount,
                 rate_limit.get_expiry(),
             )
-            limit_status[(namespace, primary_key)] = {
+            limit_status[namespace, primary_key] = {
                 "period_start": period_start,
                 "n_items_in_period": n_items_in_period,
                 "period_seconds": rate_limit.GRANULARITY.seconds,
@@ -343,14 +343,14 @@ async def try_acquire(
             namespace_and_key, machine_id=machine_id
         )
 
-        _rate_limit, new_namespace = self.parse_rate_limits_and_namespace(
+        rate_limit_, new_namespace = self.parse_rate_limits_and_namespace(
             namespace, primary_key
         )
 
         if isinstance(rate_limit, str):
             rate_limit = limit_parse(rate_limit)
 
-        rate_limit = rate_limit or _rate_limit
+        rate_limit = rate_limit or rate_limit_
 
         if rate_limit.amount < weight and raise_impossible_limits:
             raise ValueError(