Conversation
The jit audit still built its context with `Context(..., enable_checks=False)`. Validation policy moved off the context onto the bound object in this release, so that keyword now raises TypeError and the audit step failed. Drop it and lower the ambient default instead -- argument validation would otherwise add non-traced Python work that muddies the retrace counts the script measures. Separately, the lint configuration named no `select`, so `ruff check .` enforced whatever ruff's implicit default happened to be. That default has grown between releases -- 0.16 folded I, UP, B, S, SIM, PL and RUF into it -- so an unpinned `pip install ruff` in CI silently turned a green tree red without a single source edit (569 findings on master, which last ran CI against an older ruff). Name the rules explicitly; the result is now identical under 0.15.8 and 0.16.7, and widening the set becomes a deliberate commit rather than a release surprise. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Two CI gates were failing on documentation rather than behaviour. `ruff check --select D` found five regressions from the submodule refactors: the newly *public* `spacecore.contextual` package needed a module docstring (the old `_contextual` was private, so D104 never applied), two summaries were not in the imperative mood, one ran its summary into its description, and `functional/_realified.py` carried LaTeX in a non-raw docstring. `scripts/docstring_audit.py` reported 36 numpydoc issues. These predate the branch -- master reports the identical set -- but they only became a CI failure now that numpydoc 1.10 is what gets installed, and they block this PR either way. Most are the `check_level` constructor argument going undocumented across the spaces, operators and functionals that gained it; the rest are summary placement, documented parameter order that no longer matched the signature, section ordering, and missing Returns/Yields on the ambient-state helpers. Also fix two Sphinx warnings introduced by this branch: the `[Conway]` citation on `LinOp` was defined but never cited, and `SumFunctional`'s summary put `F_1 + ... + F_n` inside an inline literal, where autosummary cuts the sentence at the ellipsis and leaves the literal unterminated. Documented behaviour was verified against the runtime rather than the type hints -- notably, a boolean `check_level` is annotated but rejected, so the prose describes the four literals only. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
The version, the changelog heading and the public-API version assertion were already at 0.4.3, but an `[Unreleased]` section had since accumulated on top of the dated 0.4.3 section. Fold it in, redate to the actual release day, and keep the empty `[Unreleased]` placeholder the file has always carried. The docs still described the API this release removes, which would have shipped examples that raise: * README and the checking-policy design note both built contexts as `Context(ops, dtype=..., check_level=...)`. The level is now a property of the bound object, so both are rewritten around per-object levels and the ambient `set_check_level` / `use_check_level` helpers. The corrected README example was run; it prints the output the page claims. * `api/context.rst` pointed autodoc at `spacecore.backend.Context`, which no longer imports, and still described the context as carrying validation policy. It now also lists the scoping helpers. * README advertised `SpectralLpNormFunctional`, removed here in favour of `spectralize` / `SpectralFunctional`. Release notes had no 0.4.3 entry, and no 0.4.2 entry either -- the page jumped from 0.4.3's predecessor straight to 0.4.1 -- so both are written up. Sphinx now builds with no warnings, matching master. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
…atory-comments 63 review the code and add explanatory comments
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.
What this PR does
0.4.3 release