Repository navigation
D-RA-2: relalg symbolic answers from contract::set_rules + #1651 closeout - #1654
Conversation
|
Important
This repository does not receive automatic reviews because it has fewer than 10 stars. ⚙️ Run configuration
Warning Billing warning: we have not been able to collect payment for this subscription for more than 72 hours. Please update the payment method or pay any pending invoices in Billing to avoid service interruption. Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: a0075b5c53
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
Codex review of #1654: the NoUniverse guard in range_bounds_ok had no negative control, so the 'every guard disable-verified' claim did not hold for it. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013z2e1YXkm6o6HNrDn5BPWA
|
The crepe-plan-probe suite passes unchanged: 46 tests, 0 failures. That run built against the repointed relalg and the extended Generated by Claude Code |
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013z2e1YXkm6o6HNrDn5BPWA
Codex review of #1654: the NoUniverse guard in range_bounds_ok had no negative control, so the 'every guard disable-verified' claim did not hold for it. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013z2e1YXkm6o6HNrDn5BPWA
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013z2e1YXkm6o6HNrDn5BPWA
relalg's `symbolic` decided Count(∅), Sum(∅), Min/Max(∅) and Count(All) locally. It now rewrites the measure with `set_rules::rewrite` under relalg's semantics and answers only a constant. The contract gains the two rules it needs: EmptyMeasure (Count/Sum of ∅ → 0, refused for _sym sums) and CountAll (Count(All) → N, refused without a universe, for bags and under IEEE). crepe's symbolic answers come through relalg::run, so they follow; crepe's join-body rewrites stay. Tests: set_rules 25 (3 new, 2 re-pinned because the new rules fire on their fixtures); relalg 26 unchanged; disable runs show relalg's answers depend on the contract rules. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013z2e1YXkm6o6HNrDn5BPWA
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013z2e1YXkm6o6HNrDn5BPWA
7c375fa to
ac147d9
Compare
|
I rebased onto main
Generated by Claude Code |
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013z2e1YXkm6o6HNrDn5BPWA
board: #1654 closeout (D-RA-2 repoint shipped)
This PR records the merge of #1651 (D-CMR-6), adds one test from Codex's review of this PR, and repoints relalg onto
contract::set_rules.Closeout of #1651
a0075b5(board files): STATUS marks D-CMR-6 Shipped (contract: D-CMR-6 guarded rule catalogue (set_rules) + #1650 closeout #1651,6bf4f86). Also a PR_ARC entry for contract: D-CMR-6 guarded rule catalogue (set_rules) + #1650 closeout #1651 and a "main after contract: D-CMR-6 guarded rule catalogue (set_rules) + #1650 closeout #1651" section in LATEST_STATE.b3927d6(test only): addsexperiment_e_no_universe_refuses_the_range_rules, after Codex's review of this PR. Without it, removing theNoUniverserefusal on the range rules failed no test.4b8b57aupdates the test counts on the board.184cadb: repoint ontoset_rules(D-RA-2, after D-CMR-6)normalize::symbolichandledCount(∅) = 0,Sum(∅) = 0,Min/Max(∅) = NoneandCount(All) = nitself, outside the contract.set_rulesgains the two rules it needs.EmptyMeasure:Count(∅) → 0andSum(∅) → 0. It refuses_symsums, because their empty sum is a placeholder value.CountAll:Count(All) → N. It refuses when the population has no fixed size, when it is a bag, and under IEEE.symbolicrewrites the measure withset_rules::rewriteunderSemantics::exact(n_rows)and answers only when the result is a constant.Min/MaxanswerNoneonly when theirCountfolds to 0. The public signature is unchanged.lance_graph_relalg::run, so they now come from the contract. crepe's own rewrites inoptimize.rs(binding propagation, duplicate atom, duplicate rule) work across several relations, so they stay where they are.set_rules: 25. Three are new. Two existing tests were re-pinned because the new rules now fire on their inputs; the PR's board entry explains both.symbolicforced to returnNone, 3 relalg tests fail.Count(∅)rule off, 1 relalg test fails; with itsCount(All)rule off, 2 fail.entries/2026-10-21-d-ra-2-relalg-symbolic-repoint.md.Coordination: claimed on EXO-50 as
session_013z2e1YXkm6o6HNrDn5BPWA/019. crepe, including U3'scse.rs, and relalg'sprogram.rs(U2) are untouched.🤖 Generated with Claude Code
https://claude.ai/code/session_013z2e1YXkm6o6HNrDn5BPWA