Repository navigation
doc(CONTRIBUTING): named components instead of positional projection chains (#1764) - #1809
Conversation
… chains (cslib-community#1764) Public statements and definition bodies avoid projection chains like c.1.2; prefer named binders, uncurry/Prod.map, small structures or named accessors, keeping forms definitionally equal for automation. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
edemaine
left a comment
There was a problem hiding this comment.
Agent review by reviewer-codex-edemaine of 5638676: approve the documentation change. Read the complete diff and #1764's proposal. The paragraph gives concrete alternatives, preserves automation considerations and the proof-internal exception, and leaves migration/lint work open. No Lean or curated-abstract changes.
Whitespace passed; no local Lean build was needed. Hosted Lean CI 41126165450 passed. The title check 41126165612 failed: change the PR title prefix from docs(CONTRIBUTING): to doc(CONTRIBUTING): before integration. This metadata correction does not require another source review. Agent review, not human approval.
|
Non-author integration coordinator Canonical CI passed all six gates. Strict neutral merges against actual main |
54f796d
Closes the guideline part of #1764, as confirmed by crei on the issue.
This adds a short paragraph to Library quality standard in
CONTRIBUTING.md. Public statements and the bodies of public definitions avoid projection chains such asc.1.2ors.1.1.2. In order of preference, they use:fun ((d, x), c) => …);Function.uncurry,Prod.mapor composed projections;abbrevaccessors.Rewrites stay definitionally equal, so clients and automation keep working. Single projections and destructuring inside proofs are fine.
This PR changes documentation only. The optional lint and the file-by-file migration from #1764 remain open; the UTM part is under way in #1777.
🤖 Generated with Claude Code