[pre-commit.ci] pre-commit autoupdate - #1693
Closed
pre-commit-ci[bot] wants to merge 3 commits into
Closed
pre-commit-ci[bot] wants to merge 3 commits into
pre-commit-ci[bot] wants to merge 3 commits into
Conversation
Codecov Report✅ All modified and coverable lines are covered by tests. Additional details and impacted files@@ Coverage Diff @@
## master #1693 +/- ##
==========================================
+ Coverage 86.87% 89.43% +2.55%
==========================================
Files 56 73 +17
Lines 6532 10618 +4086
==========================================
+ Hits 5675 9496 +3821
- Misses 857 1122 +265
Flags with carried forward coverage won't be shown. Click here to find out more. ☔ View full report in Codecov by Harness. 🚀 New features to boost your workflow:
|
pre-commit-ci
Bot
force-pushed
the
pre-commit-ci-update-config
branch
2 times, most recently
from
April 20, 2026 18:23
6eec861 to
fe16fdb
Compare
pre-commit-ci
Bot
force-pushed
the
pre-commit-ci-update-config
branch
from
April 27, 2026 18:47
fe16fdb to
eb33c7e
Compare
pre-commit-ci
Bot
force-pushed
the
pre-commit-ci-update-config
branch
from
May 11, 2026 18:27
eb33c7e to
606ded9
Compare
pre-commit-ci
Bot
force-pushed
the
pre-commit-ci-update-config
branch
2 times, most recently
from
May 25, 2026 18:43
ddf979a to
64a17f9
Compare
pre-commit-ci
Bot
force-pushed
the
pre-commit-ci-update-config
branch
2 times, most recently
from
June 8, 2026 18:32
b797e47 to
0e66088
Compare
pre-commit-ci
Bot
force-pushed
the
pre-commit-ci-update-config
branch
from
June 15, 2026 18:38
0e66088 to
ee5896b
Compare
pre-commit-ci
Bot
force-pushed
the
pre-commit-ci-update-config
branch
2 times, most recently
from
June 29, 2026 18:34
e2adae1 to
d285d87
Compare
pre-commit-ci
Bot
force-pushed
the
pre-commit-ci-update-config
branch
2 times, most recently
from
July 20, 2026 18:40
53dc551 to
38d4185
Compare
pre-commit-ci
Bot
force-pushed
the
pre-commit-ci-update-config
branch
from
July 27, 2026 18:35
38d4185 to
ff594d8
Compare
pre-commit-ci
Bot
force-pushed
the
pre-commit-ci-update-config
branch
from
August 3, 2026 18:45
8dbefe3 to
0697abb
Compare
pre-commit-ci
Bot
force-pushed
the
pre-commit-ci-update-config
branch
from
August 10, 2026 18:50
a71a6bd to
8053ada
Compare
pre-commit-ci
Bot
force-pushed
the
pre-commit-ci-update-config
branch
2 times, most recently
from
August 24, 2026 18:50
0d3b5b8 to
ff60780
Compare
updates: - [github.com/astral-sh/ruff-pre-commit: v0.15.8 → v0.16.5](astral-sh/ruff-pre-commit@v0.15.8...v0.16.5) - [github.com/pre-commit/mirrors-mypy: v1.19.1 → v2.3.1](pre-commit/mirrors-mypy@v1.19.1...v2.3.1)
pre-commit-ci
Bot
force-pushed
the
pre-commit-ci-update-config
branch
from
August 31, 2026 19:01
e65841c to
bca7af9
Compare
chrisjsewell
added a commit
that referenced
this pull request
Sep 2, 2026
…hem (#1805) `ruff` was pinned in two places that disagreed — `v0.15.8` in `.pre-commit-config.yaml`, `0.14.11` in the `ruff` dependency group — so `uv run prek run` and `tox -e ruff-check` have been running different linters. `mypy` is pinned in two places too. This moves ruff to **0.16.5** and mypy to **2.3.1** in both places each, and adds the one config line that makes a ruff version bump mean "a new ruff", not "a new rule set". **ruff 0.16 replaced its default rule set.** In an empty project, `ruff check --isolated --show-settings` reports **59** default rules on 0.15.8 and **413** on 0.16.5. This project only ever set `extend-select`, never `select`, so its base set has always been whatever the installed ruff happened to default to. On today's `master` that means: | ruff | `ruff check` findings | | --- | --- | | 0.14.11 (dependency group) | 0 | | 0.15.8 (hook) | 0 | | 0.16.5, unpinned base set | **298** | | 0.16.5, base set pinned | 148 | | 0.16.5, base set pinned, as merged here | **0** | The bump is only a version bump if `select` is pinned, so the first commit does that: ```toml select = ["E4", "E7", "E9", "F"] # ruff's pre-0.16 default ``` At the versions in force at that commit it is a byte-for-byte no-op (same 259 enabled rules, same "All checks passed!", same 326 files already formatted). What it buys is that the enabled set is now a property of this file rather than of the pinned version. Enabled-rule set diff, old config on 0.15.8 vs the config in this PR on 0.16.5 (`ruff check --show-settings`, counting `linter.rules.enabled`): **259 -> 262, gained exactly `RUF036`, `RUF063`, `RUF068`, lost none** — all three were preview-only in 0.15.8 and stabilised in 0.16. `ISC004` stabilised with them and is added to `extend-ignore` next to the existing `ISC001`: all 35 occurrences here are deliberately wrapped string literals and its only fix is unsafe. So the entire lint surface of this bump is `RUF036 none-not-at-end-of-union`, 113 findings, all safe autofixes, applied machine-wise with `ruff check --select RUF036 --fix` followed by `ruff format`. Every one is `None | X` -> `X | None`: identical at runtime and to every type checker, already the prevailing spelling here, and never evaluated at all, since these modules use `from __future__ import annotations`. `git diff -w` is the same as `git diff` — there is no reformatting hiding in it. **Why #1693 goes red in all 13 Core cells and this does not.** That same RUF036 autofix wave rewrote `constraints_results: None | Mapping[str, Mapping[str, bool]]` to `Mapping[...] | None` in `sphinx_needs/data.py`. `tests/test_data.py::test_consistent` cross-checks each `NeedsCoreFields` entry against its `TypedDict` annotation by string-matching the raw `ForwardRef` text, and while its scalar branches listed both union orders, its *container* branches only listed `None | dict[` / `None | Mapping[` — so the reordered annotation fell into the non-optional branch and asserted `"object"` against `["object", "null"]`. One brittle test, no runtime behaviour involved; this PR teaches the container branches both orders (optional case tested first, so it cannot be swallowed) and the suite is green. **Why #1693's Lint job stays red as well.** Of the 298 findings the unpinned bump produces, **151 are not autofixable** — `BLE001` 43, `PLW1510` 30, `TRY004` 20, `DTZ005`/`TRY002` 4 each, and so on. The bot force-pushes a partially-fixed tree, and no amount of re-running clears the rest. **And one thing #1693 does silently.** `E402` (module-import-not-at-top-of-file) is in ruff's pre-0.16 default set and *not* in 0.16's, along with 17 others (`E401 E402 E7xx E711-E714 E721 E731 E741-E743 F403 F405 F406 F722`). So on that branch E402 stopped being enforced and `RUF100 unused-noqa` then offered to delete all twelve `# noqa: E402` suppressions in `docs/conf.py` as unused — lint coverage lost, plus a twelve-site re-annotation for whoever wants it back. Pinning `select` keeps both the rule and the suppressions: ``` ruff check --isolated --select RUF100 docs/conf.py -> 12x "Unused `noqa` (non-enabled: E402)" ruff check --isolated --select RUF100,E402 docs/conf.py -> All checks passed! ``` **The ruleset expansion is deliberately deferred.** Those 151 findings are a real backlog worth reviewing — but rule by rule, in its own PR, not as a side effect of a version pin. The `select` pin is precisely what turns that into a clean opt-in: adopting 0.16's defaults will then be a legible one-line diff with its own reviewable consequences. `ruff format` in 0.16 also formats python code blocks embedded in markdown, which brings `AGENTS.md` and `tests/conformance/needflow/README.md` into its scope. `[tool.ruff.format] exclude = ["*.md"]` keeps it out — same reasoning as the existing yamlfmt `^tests/conformance/` exclude, since that corpus is shared byte-for-byte with ubCode and checksummed in its own manifest. Without the exclude, `ruff format` rewrites 25 lines of `AGENTS.md` on this branch. **mypy 2.3.1** is folded in because its two pins have to move together: 1.19.1 *requires* the `# type: ignore[literal-required]` in `compile_validator` and 2.3.1 rejects it as unused, so splitting the change would force a bridging `[literal-required, unused-ignore]` into the tree and a follow-up to take it out again. Moving both pins in one commit makes it a plain deletion. mypy 2.3.1 reports nothing else: no new error classes, no deprecation warnings, no config migration. Two follow-ups this PR deliberately does not do: - The mypy hook's `additional_dependencies` and the `mypy` dependency group are maintained by hand and have already drifted apart — the hook carries `minijinja~=2.15` and the dependency group does not, so `uv run prek run mypy` and `tox -e mypy` type-check against different environments. It is pre-existing and orthogonal to a version bump; the fix is one line (add `minijinja~=2.15` to the group) and belongs in its own PR. - Routine hook autoupdates should be owned by a scheduled `prek autoupdate` job rather than by pre-commit.ci reopening a PR every release; that is a separate change. With `select` pinned, such updates are finally safe to take. This supersedes #1693, which can be closed.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
updates: