docs: sync docs with rust/rayon migration, fill api.md gaps, add docs-audit gate#264
Merged
Conversation
e04f6ea to
eab9623
Compare
Collaborator
Author
|
Rebased onto |
Recent Phase 5 work (numba read-path deleted → Rust-only, awkward → _core migration, new rayon threading env vars) left user-facing docs stale. - faq: rewrite Ragged answer (seqpro _core.Ragged Rust backend, not Awkward/numba); document GVL_NUM_THREADS / GVL_FORCE_PARALLEL / RAYON_NUM_THREADS override - README: drop tbb/pyomp-for-numba note; parallelism is built-in Rust/rayon - SKILL: _core.Ragged is a Rust backend (rag layer is numba-free) - CLAUDE.md: require a docs audit before feature/breaking-change PRs Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
api.md drifted from __init__.__all__. Adds autodoc entries for every undocumented public symbol (verified: 39/39 __all__ names now present): - Insertion fill: InsertionFill + Constant/FlankSample/Interpolate/ Repeat5p/Repeat5pNormalized (new section) - Flat containers: FlatRagged/FlatAnnotatedHaps/FlatIntervals/ FlatVariants/FlatAlleles/FlatVariantWindows (new subsection) - Variant windows: VarWindowOpt, DummyVariant - Dataset maintenance: migrate, migrate_svar_link (new section) - Writing: update (sibling of write) - PyTorch interop: to_nested_tensor Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The api.md drift (18 missing symbols) motivates an explicit gate check: list api.md alongside the audited docs and add a one-liner that flags any __all__ export missing from the API reference. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The install-e task editable-installed sibling seqpro/genoray from hardcoded absolute cluster paths. It was never required: seqpro/genoray are transitive deps via pyproject.toml, and the locked seqpro 0.20.0 already provides every symbol gvl imports (e.g. seqpro.rag.reverse_complement). The docs build only failed on machines whose docs env had drifted stale from the lock (an ancient seqpro 0.11.0); `pixi install -e docs` reconciles it, no editable install needed. Verified: docs build cleanly on darwin (osx-arm64) with the plain PyPI wheels after reconciling the env. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
eab9623 to
e8ea5dd
Compare
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
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.
Summary
Audits and syncs user-facing docs against recent gvl changes (Phase 5: numba read-path deleted → Rust-only, awkward →
_coremigration, new rayon threading env vars), fills a largeapi.mdgap, adds a docs-audit gate toCLAUDE.md, and removes an obsolete docs task.Changes
Docs consistency
faq.md: rewrote the "Ragged objects" answer — dropped the false "subclass of Awkward Arrays / numba JIT'd functions" claim; describes theseqpro.rag.Ragged(_core.Ragged, Rust) backend and notesawkwardis no longer a dependency.faq.md: new entry documenting the threading env vars (GVL_NUM_THREADS,GVL_FORCE_PARALLEL,RAYON_NUM_THREADSoverride) — previously undocumented.README.md: replaced the staletbb/pyomp-for-numba install note with "parallelism is built-in Rust/rayon, tune viaGVL_NUM_THREADS".SKILL.md:_core.Ragged"Rust+numba backend" → "Rust backend" (rag layer is numba-free).api.md↔__all__drift (18 missing public symbols)InsertionFill,Constant,FlankSample,Interpolate,Repeat5p,Repeat5pNormalizedFlatRagged,FlatAnnotatedHaps,FlatIntervals,FlatVariants,FlatAlleles,FlatVariantWindowsVarWindowOpt,DummyVariantmigrate,migrate_svar_linkupdate; PyTorch interop:to_nested_tensor__all__symbols now present inapi.md.Process
CLAUDE.md: new "Docs audit before feature/breaking-change PRs" gate, including an explicitapi.md↔__all__sync check with a one-liner.Cleanup
pixi.toml: removed the obsoleteinstall-edocs task. It editable-installed sibling seqpro/genoray from hardcoded absolute cluster paths but was never required — they are transitive deps and the locked seqpro 0.20.0 already provides every symbol gvl imports. The docs build only broke when an env drifted stale from the lock (fixed bypixi install -e docs).Verification
api.htmlrenders and the new autodoc entries produce no new warnings/errors.__all__↔api.mddiff reports no missing symbols.docs/source/changelog.mdintentionally untouched (auto-generated from commits).🤖 Generated with Claude Code