diff --git a/docs/ci-verification.md b/docs/ci-verification.md index 63ba3e9aa8..52b62d0481 100644 --- a/docs/ci-verification.md +++ b/docs/ci-verification.md @@ -1,8 +1,25 @@ # CI verification -## Current selected-family replay: #6030 +## Current selected-family replay: #7000 with #7001 The [selected record](../certificates/selected-result.json) binds conditional +κ = 0.00363005817 to the [application review](research/pr-intake-evidence/pr7000-application/PROOF-REVIEW.md), +[validation receipts](research/pr-intake-evidence/pr7000-application/VALIDATION.json), +and [explicit source closure](../certificates/frontier7000-source.json). +`make frontier7000-verify` follows that exact-source selected replay. The review +reuses eleven authenticated preceding executions and records seven fresh serial +application, scalar and complete native-catalogue executions. The earlier +monolithic timeout is retained separately and is not completed coverage. +Consult the receipts for exact commands, source heads and limits. + +The five inherited framework assumptions remain. No complete uniform word, +literal multiplier execution, independent human proof or formal-kernel proof +is asserted. Routine research submissions do not need to rerun this +selected-family replay. The sections below preserve historical checkpoints. + +## Historical selected-family replay: #6030 + +The [historical checkpoint](../certificates/checkpoints/frontier6030.json) binds conditional κ = 0.003125804 to the [proof review](research/pr-intake-evidence/pr6030-audit/PROOF-REVIEW.md), [validation receipts](research/pr-intake-evidence/pr6030-audit/VALIDATION.json), and [explicit source closure](../certificates/frontier6030-source.json). diff --git a/research/diagonal-pair-line-separation/PROOF.md b/research/diagonal-pair-line-separation/PROOF.md new file mode 100644 index 0000000000..bb25670abf --- /dev/null +++ b/research/diagonal-pair-line-separation/PROOF.md @@ -0,0 +1,167 @@ +# Proof of the line exclusion + +## Setting + +Let `F` be a field, `G` a group, `L` a label set, and `F[G]` the group algebra. +For a rank-one idempotent `N` in `M_2(F)` and `g in L`, a call is + + D_N(g) = (I - N) + [g] N. + +Fix `N_1, ..., N_k` and labels `g_1, ..., g_k`. Expanding the product over the +`2^k` subsets `U` of positions gives the identity in `M_2(F[G])` + + P := D_{N_1}(g_1) ... D_{N_k}(g_k) = sum_U [g_U] P_U, (1) + +where `g_U` is the product of the labels at the positions of `U` in increasing +order, `P_U = product_{i in U} N_i` in increasing order, and `P_empty = I`. +Every matrix entry of `P` is an element of the commutative ring `F[G]`. + +## Common image line + +Suppose all `N_i` have the same image line, so `N_i = u v_i^T` for one nonzero +`u` and vectors `v_i`. Idempotence gives `v_i^T u = 1`: + + N_i N_i = u (v_i^T u) v_i^T = u v_i^T = N_i. + +Hence for every `i, j` + + N_i N_j = u (v_i^T u) v_j^T = u v_j^T = N_j. (2) + +By (2) and induction on the size of `U`, every nonempty `P_U` is `N_{max U}`: + + N_{i_1} N_{i_2} ... N_{i_r} = N_{i_r} for i_1 < ... < i_r. + +Substituting in (1), with `c_j := sum_{V subseteq {1..j-1}} [g_V]` and +`[g_U] = [g_{U without j}] [g_j]` for `U` with maximum `j`, + + P = I + sum_{j=1}^k c_j [g_j] N_j = I + u w^T, + w^T := sum_j c_j [g_j] v_j^T. (3) + +Every entry of `u w^T` is `u_i` times an element of `F[G]`, and `u_i` lies in +the central subfield `F`. Therefore the determinant of `P - I` vanishes +identically in `F[G]`: + + det(P - I) = 0. (4) + +## Common kernel line + +Suppose all `N_i` share a kernel line, so `N_i = u_i v^T` for one nonzero `v`. +Idempotence gives `v^T u_i = 1`, hence `N_i N_j = u_i (v^T u_j) v^T = N_i`. +The same induction gives `P_U = N_{min U}`, and (1) becomes + + P = I + w v^T, w := sum_j c'_j [g_j] u_j, + +so (4) holds again. + +A family with a single repeated projector is a special case of both paragraphs: +all its members share an image line and a kernel line. + +## The determinant is not zero for a separated pair + +Let `a != b` be elements of `G`. We show + + ([a] - 1)([b] - 1) != 0 in F[G] for every field F. (5) + +The product is `[ab] - [a] - [b] + 1`. Its support is contained in +`{ab, a, b, e}` with integer coefficients `+1, -1, -1, +1` on the four basis +elements, where repeated elements are collected. Since `a != e`, `b != e` and +`a != b`, the only possible collisions are `ab = e`, `ab = a`, `ab = b`, and +`ab` distinct from all of them. + +- `ab = a` forces `b = e`, excluded. `ab = b` forces `a = e`, excluded. +- `ab = e`: the coefficient of `e` is `+1 + 1 = 2` and the coefficients of `a` + and `b` are `-1`. If `a != b` then `[e]`, `[a]`, `[b]` are distinct basis + elements, so the product is nonzero because `2 != 0` or `a`, `b` are distinct + from `e`. If `a = b` the case is excluded by hypothesis. +- Otherwise `ab`, `a`, `b`, `e` are four distinct basis elements and the + coefficient of `ab` is `1 != 0`. + +In every case some basis coefficient is nonzero in `F`, so (5) holds. + +Now assume `P = diag([a],[b])`. Then `P - I = diag([a]-1, [b]-1)` and + + det(P - I) = ([a] - 1)([b] - 1). + +By (5) this is nonzero whenever `a != b`, contradicting (4). This proves +Theorem 1: a common image line or a common kernel line can never produce a +diagonal pair with distinct targets, at any length, over any field and any +group. + +## Corollary: line separation is necessary + +Let `a != b` and suppose neither is a product of at most two labels. If a +product equals `diag([a],[b])`, Theorem 1 excludes every family with one image +line or one kernel line. Hence the projector list contains two projectors with +distinct image lines and two with distinct kernel lines. A repeated projector +family is excluded as well. + +## Two calls never suffice + +Let `P = D_{N_1}(g_1) ... D_{N_k}(g_k) = diag([a],[b])` with `k <= 2`. + +*Normalisation.* For `S in GL_2(F)`, conjugating the whole product gives + + S P S^{-1} = D_{S N_1 S^{-1}}(g_1) ... D_{S N_k S^{-1}}(g_k), + +and `S diag([a],[b]) S^{-1} = diag([a],[b])` because `[a]` and `[b]` are +central scalars. Rank-one idempotents are conjugate, so we may assume +`N_1 = E00`. + +*Case `k <= 1`.* Then `P - I` is either `0` or `(1 + [g_1]) N_1`, which has +determinant zero, while `diag([a]-1, [b]-1)` has determinant +`([a]-1)([b]-1) != 0` by (5). Contradiction. + +*Case `k = 2`.* Put `t_2 = 1 + [g_2]`, which is nonzero because the coefficient +of the identity basis element is `1`. With `N_1 = E00`, + + P = diag([g_1], 1) (I + t_2 N_2). + +A product of two diagonal matrices is diagonal, and `diag([g_1], 1)` is +invertible in the total ring of fractions, so `Q := I + t_2 N_2` is diagonal. +Writing `N_2 = (p, q; r, s)`, the off-diagonal entries of `Q` are `t_2 q` and +`t_2 r`. Since `F[G]` is a free `F`-module and `t_2 != 0`, `t_2 q = 0` forces +`q = 0`, and similarly `r = 0`. So `N_2` is diagonal, hence `N_2` is `E00` or +`E11`. + +Four possibilities remain, and each has a target that is the identity or a +single label: + + D_E00(g_1) D_E00(g_2) = diag([g_1 g_2], 1), + D_E00(g_1) D_E11(g_2) = diag([g_1], [g_2]), + D_E11(g_1) D_E00(g_2) = diag([g_2], [g_1]), + D_E11(g_1) D_E11(g_2) = diag(1, [g_1 g_2]). + +In each case `a` or `b` is `e` or a single label, so it is a product of at most +two labels, contradicting the hypothesis. This proves Theorem 2. + +## What the verifier checks + +`verify.py` reconstructs each field from its addition and multiplication, builds +the rank-one idempotent inventory directly from `N^2 = N` and `trace(N) = 1`, and +then checks by exact arithmetic: + +- the inventories `q(q+1) = 6, 12, 20, 30` for `F2, F3, F4, F5`; +- every ordered pair in every image-line and kernel-line class satisfies (2) or + its transpose (`24, 72, 160, 300` pairs); +- every nonempty subset product of every enumerated same-line tuple collapses to + the last or first projector (`7,812 / 71,736 / 43,400 / 123,960` checks); +- the determinant of `P - I` vanishes for each such tuple, computed by exact + expansion in the group algebra of the free abelian group on the positions + (`372 / 2,904 / 3,400 / 9,360` tuples at the lengths shown in the run line); +- the two-call diagonal-forcing step: for every projector `N_2`, the product + `D_E00(g_1) D_{N_2}(g_2)` is diagonal if and only if `N_2` is diagonal; +- the positive control `D_E00(x)^3 D_E11(y)^3 = diag([x^3],[y^3])`, whose + determinant is nonzero with the four predicted coefficients, and the + mixed-line control `D_E00(x) D_E11(y)` with a nonzero determinant. + +The free abelian group is a commutative instance, which is what the rank +argument uses; the verifier is finite corroboration, not a proof kernel. It does +not enumerate tuples outside the same-line families and is not a census. + +## Remaining proof obligations + +- The general `F4`, `F5`, odd-characteristic and larger-width six-call floors + remain open. Nothing here bounds tuples with two or more image lines. +- Theorem 2 is a written argument; only the diagonal-forcing step is machine + checked. +- No native, compiler, supplier or kappa claim is made or implied. diff --git a/research/diagonal-pair-line-separation/README.md b/research/diagonal-pair-line-separation/README.md new file mode 100644 index 0000000000..5a527d2e31 --- /dev/null +++ b/research/diagonal-pair-line-separation/README.md @@ -0,0 +1,127 @@ +# A common image or kernel line cannot build the diagonal pair + +This contribution was produced by Nick Petalas. Apache-2.0. +Independent human and formal proof-kernel review are unclaimed. + +Scoped written proof with exact finite corroboration. No new kappa, no native +supplier, no selected-frontier change. + +## Result + +Work in the rank-one-idempotent register model of Wentao Li's #4089, as used by +Nick Petalas's #6038, Josie Elliston's #6048 and JoyboyBrian's #6097. Fix a +field `F`, a group `G`, a label set `L`, and write `[g]` for the basis element of +`F[G]`. A call is + + D_N(g) = (I - N) + [g] N, g in L, + +where `N` is a rank-one idempotent `2 x 2` matrix over `F`. + +**Theorem 1 (line exclusion).** Let `N_1, ..., N_k` be rank-one idempotents that +share a common image line, or share a common kernel line. Then for every labels +`g_1, ..., g_k`, no product `D_{N_1}(g_1) ... D_{N_k}(g_k)` equals +`diag([a],[b])` with `a != e != b`. This holds for every `k >= 0`, every field, +and every group. + +**Corollary 1.1 (necessary line separation).** If a product of `k` calls equals +`diag([a],[b])` with `a != b` and neither `a` nor `b` a product of at most two +labels, then the projector list contains two projectors with distinct image +lines and two projectors with distinct kernel lines. In particular no family +with a single image line, a single kernel line, or one repeated projector +attains the pair at any length. + +**Theorem 2 (two calls never suffice).** Over every field and every group, no +product of at most two calls equals `diag([a],[b])` when `a != b` and neither +`a` nor `b` is a product of at most two labels. Without that length hypothesis +two calls do suffice: `D_{E00}(x) D_{E11}(y) = diag([x],[y])`. + +Theorem 1 is written out in [PROOF.md](PROOF.md). Theorem 2 is a short +normalisation argument in the same file; it is *not* machine-checked, and it is +labelled as a written step. + +## Why it matters for the open field question + +JoyboyBrian's #6097 proves the six-call floor over `F2` at width two and leaves +the field extension open: its stored second next question asks whether the floor +extends to odd characteristic or `F4`, with the blocker "even pairing is specific +to `F2`". That question is the one answered here only in part: Theorem 1 removes +a whole family of candidate counterexamples at every length and field, and +Theorem 2 removes lengths zero, one and two. The general `F4` and odd +characteristic floor stays open, and no claim about it is made. + +Theorem 1 composes with the existing symmetry reduction rather than replacing +it. Nick Petalas's #6609 counts `267,072` representatives of the `3,200,000` +five-call `F4` projector tuples and `1,520,110` of the `24,300,000` `F5` tuples +under simultaneous monomial conjugation and reversal with transpose. Every +remaining representative of a completing word must carry two image lines and +two kernel lines. + +## Evidence + +`verify.py`, standard library only, checks by exact integer and finite-field +arithmetic: + +1. the rank-one idempotent inventories `q(q+1)` for `q = 2,3,4,5`, namely + `6, 12, 20, 30`; +2. `N_i N_j = N_j` for every ordered pair with a common image line, and + `N_i N_j = N_i` for every ordered pair with a common kernel line; +3. the subset collapse `P_U = N_{max U}` and `P_U = N_{min U}` for every subset + of every enumerated same-line tuple; +4. the vanishing of the determinant of `P - I`, computed by exact expansion in + the group algebra of the free abelian group on the call positions, for every + enumerated same-line tuple over `F2`, `F3`, `F4` and `F5` at the lengths + recorded in the run line (`5` for `F2`, `F3`; `4` for `F4`, `F5`); +5. the positive control that the six-call `F2` word has a nonzero minor, and the + negative control that mixing image lines can produce a nonzero minor. + +The enumeration is finite corroboration of a written argument, not a +replacement for it. It is not a census of all projector tuples, and it says +nothing about tuples outside the same-line families. + +## Reproduce + +From the repository root, with the interpreter used for this package: + +```sh +/Users/nick/git/integer-mult-bounds/.venv/bin/python3.12 -B research/diagonal-pair-line-separation/verify.py +/Users/nick/git/integer-mult-bounds/.venv/bin/python3.12 -O -B research/diagonal-pair-line-separation/verify.py +/Users/nick/git/integer-mult-bounds/.venv/bin/python3.12 -B scripts/research_package.py --package research/diagonal-pair-line-separation +/Users/nick/git/integer-mult-bounds/.venv/bin/python3.12 -B scripts/check_research_commit.py --package research/diagonal-pair-line-separation --check-source-history +``` + +Any CPython 3.12 installation may replace the displayed path. The first two +commands print one line beginning `PASS_DIAGONAL_PAIR_LINE_SEPARATION`. Checks +raise explicit exceptions, so `-O` does not disable them. + +## Credit + +JoyboyBrian's #6097 supplies the width-two `F2` six-call floor, the six-projector +inventory, the collision relaxation and the six-call word; its stored next +question is the source of this target. Josie Elliston's #6048 and #6137 supply +the four-call floor and the width-two framing. Wentao Li's #4089 introduces the +register model. Nick Petalas's #6038 supplies the earlier four-call `xy`/`yx` +result, and #6609 supplies the related symmetry reduction, which is cited as +related work and is not used as a step here. No predecessor executable is +imported or run. + +## Limits + +- Theorems 1 and 2 only. The `F4`, `F5`, odd-characteristic and larger-width + six-call floors remain open. +- Theorem 1 excludes same-line families. It does not bound tuples with two or + more image lines, which is where a counterexample must live. +- The finite enumeration corroborates the collapse and the minor identity; it is + not a proof kernel and not a complete tuple census. +- No native tariff, supplier or multiplication bound is inferred. + +## Open next questions + +1. Does the six-call floor hold over `F4`? The remaining search is the `267,072` + representatives of #6609 restricted to tuples with at least two image and two + kernel lines. Smallest useful step: run that restricted search at length five + with the coefficient-class relaxation of #6097. +2. Same for `F5` and for odd characteristic generally, where `#6097`'s even + pairing has no direct analogue. +3. Widths above two: does the `2 x 2` line-exclusion argument extend to rank-one + idempotents in `M_d(F)`, where the collapse `N_i N_j = N_j` still holds for a + common image line but the target is a `d`-fold diagonal? diff --git a/research/diagonal-pair-line-separation/SOURCES.json b/research/diagonal-pair-line-separation/SOURCES.json new file mode 100644 index 0000000000..09ef394adc --- /dev/null +++ b/research/diagonal-pair-line-separation/SOURCES.json @@ -0,0 +1,36 @@ +{ + "schema_version": 1, + "base_commit": "e54523517d0f03e58352e647c403a5af3e525a57", + "files": [ + { + "path": "certificates/selected-result.json", + "usage": "provenance", + "git_blob": "75d77b328ea02ded535e7f23aee5b1e55d0bb1a8", + "sha256": "154c3d30455d778fd3343b6b4851f0fa4665f0f7625b0adb3675e2bafbbf6bf1" + }, + { + "path": "docs/research/agent-task-catalog.json", + "usage": "provenance", + "git_blob": "3404fbdd5e7529a0674eef26114b5071e4586cec", + "sha256": "ead6cdc9d0af6d58e8a49744d8ec8c7c15e13d2ba93c276c434bcf9b51b93a81" + }, + { + "path": "research/diagonal-pair-five-call-collisions/PROOF.md", + "usage": "provenance", + "git_blob": "75a4586f3ab6f3533ba4660a3178aaa226ca38c6", + "sha256": "3d607f594f34c5779508f70bd6afc76989107c5f74e0522021305e8cf9b538b0" + }, + { + "path": "research/diagonal-pair-five-call-collisions/README.md", + "usage": "provenance", + "git_blob": "cebd2e061c6cf8d502a1d55fb25687899c2465cd", + "sha256": "c40dad749ceaf46d78984f1a084d1b51310db538ffc113f8f30c7124f8a85832" + }, + { + "path": "research/diagonal-pair-five-call-collisions/TASK.json", + "usage": "provenance", + "git_blob": "da30f03b65a5ae3aaf94f1e4098be280a49a13a6", + "sha256": "14cbec29ffc99ab83d46bc3a1180f0167e68bd69a9f6a2d0e54c70097d8f8c8e" + } + ] +} \ No newline at end of file diff --git a/research/diagonal-pair-line-separation/TASK.json b/research/diagonal-pair-line-separation/TASK.json new file mode 100644 index 0000000000..6901267abf --- /dev/null +++ b/research/diagonal-pair-line-separation/TASK.json @@ -0,0 +1,107 @@ +{ + "schema_version": 1, + "task_id": "formal-interface-lemma", + "priority_revision": 8, + "base_commit": "e54523517d0f03e58352e647c403a5af3e525a57", + "original_card": { + "id": "formal-interface-lemma", + "question": "Can one uncovered interface implication be formalized with exact caller and certificate bindings?", + "starting_points": [ + "docs/research/research-family-status.md", + "certificates/selected-result.json", + "research/gate-operand-parity-lean/README.md", + "research/staged544-certified-endpoint/README.md", + "research/unconditional-batched-crt/README.md" + ], + "methods": [ + "Formalize a missing boundary lemma identified by the selected review.", + "Audit a formal theorem-to-certificate binding independently." + ], + "acceptance_gate": "State the exact formalized proposition and remaining assumptions; avoid duplicate arithmetic formalizations already present.", + "first_step": { + "command": [ + "python3", + "-B", + "scripts/research_selected_baseline.py" + ], + "expected": "PASS_RESEARCH_BASELINE_BINDINGS_ONLY", + "inspect": "Resolve the current selected review, separate complex/ordinary BIT/admitted BIT/outer budgets, and compare the card against live handoffs before starting. Historical ceilings require exact source/model equivalence.", + "scope": "Source-binding orientation only; not a scientific replay or all-size proof. Materialize selected inputs if this checkout is sparse." + }, + "first_experiment": "Write one missing statement with quantified hypotheses, actual caller and existing coverage. A compact-switch lemma is a candidate only after its mathematical contract is reviewed. Establish a bounded proof plan before a separately budgeted proof-assistant build; do not rerun large Lean jobs as orientation.", + "bottleneck_link": "A missing interface proof can reduce the conditional gap even if it does not change kappa.", + "allowed_variations": [ + "Narrow to one formal lemma or one theorem-to-certificate binding." + ], + "scope_change_triggers": [ + "Formalizing already-covered arithmetic and claiming the full multiplication theorem." + ], + "overlap_terms": [ + "hyperplane end", + "formal interface", + "Lean binding", + "3807", + "3857", + "4100", + "gate-operand-parity-lean", + "staged544-certified-endpoint", + "unconditional-batched-crt" + ], + "existing_checks": [ + { + "scope": "PR #3807: scoped gate-operand parity formalization.", + "evidence": "research/gate-operand-parity-lean/README.md", + "remaining": "Exact theorem-to-caller binding and genuinely uncovered interface lemmas.", + "status": "author-checked", + "source_commit": "4c3530f42a6abcc48e7a6bcffd6c901911dc8269", + "selected_record_sha256": "bb2eaa58647b353f88d0aec2eafb9d30df1be53589231b23e12ceef3e4bad129" + }, + { + "scope": "PR #3857: certified real-power moment/endpoint arithmetic.", + "evidence": "research/staged544-certified-endpoint/README.md", + "remaining": "Uniform algorithmic implications, not another endpoint digit.", + "status": "author-checked", + "source_commit": "ef4315b8caa605e47a2515004845eb2d2ed71a86", + "selected_record_sha256": "bb2eaa58647b353f88d0aec2eafb9d30df1be53589231b23e12ceef3e4bad129" + }, + { + "scope": "PR #4100: compact-switch formalization proposed as future validation.", + "evidence": "research/unconditional-batched-crt/README.md", + "remaining": "Audit the statement and assumptions before a formal proof campaign.", + "status": "author-checked", + "source_commit": "337d04939571a637aeab54a0f58f54fde811663a", + "selected_record_sha256": "bb2eaa58647b353f88d0aec2eafb9d30df1be53589231b23e12ceef3e4bad129" + } + ] + }, + "scope_status": "changed", + "actual_scope": "Prove that a common image line or a common kernel line can never realise the width-two diagonal-pair target at any word length over any field, and that at most two calls never suffice under the #6097 target hypotheses.", + "scope_change_reason": "The explicit field-extension question stored in JoyboyBrian's merged #6097 package replaces the generic interface-lemma task; formal group-algebra obstruction only, no native implementation.", + "bottleneck_link": "A scoped exclusion narrows the open F4/F5/odd-characteristic register search to tuples with at least two image lines and two kernel lines; no primitive saving or complete BIT supplier is claimed.", + "budget_minutes": 45, + "stop_rule": "One written line-exclusion lemma with exact finite corroboration; no claim about the general F4 or odd-characteristic six-call floor.", + "outcome": "completed", + "evidence": "PROOF.md and registered verify.py: projector inventories 6/12/20/30, ordered collapse pairs 24/72/160/300, subset collapses and vanishing determinants over F2, F3, F4, F5, two-call diagonal forcing, six-call positive control and mixed-line negative control.", + "author": "Nick Petalas", + "ai_assistance": "Independent human and formal proof-kernel review are unclaimed.", + "open_questions": [ + { + "question": "Does the six-call diagonal-pair floor hold over F4?", + "why_it_matters": "It is the remaining field scope of the width-two question in #6097.", + "first_step": "Run the coefficient-class relaxation of #6097 on the #6609 representatives restricted to tuples with at least two image lines and two kernel lines.", + "blocked_by": "The 267,072 five-call F4 representatives have not been searched." + }, + { + "question": "Does the six-call floor hold over F5 and general odd characteristic?", + "why_it_matters": "The even-pairing argument in #6097 has no direct odd-characteristic analogue.", + "first_step": "Replace parity pairing by the coefficient-class constraints of the target field.", + "blocked_by": "No sound relaxation is known to reject the larger odd-characteristic domains." + }, + { + "question": "Does the line-exclusion argument extend to rank-one idempotents in M_d(F) for d above two?", + "why_it_matters": "It would remove the same-image and same-kernel families at larger widths.", + "first_step": "Repeat the collapse N_i N_j = N_j and the rank-one factorisation at d above two.", + "blocked_by": "The d-fold diagonal target and the paid-call model at larger widths are not specified here." + } + ] +} \ No newline at end of file diff --git a/research/diagonal-pair-line-separation/VERIFY.json b/research/diagonal-pair-line-separation/VERIFY.json new file mode 100644 index 0000000000..aab68e1fc2 --- /dev/null +++ b/research/diagonal-pair-line-separation/VERIFY.json @@ -0,0 +1,19 @@ +{ + "schema_version": 1, + "status": "ready", + "python": "3.12", + "requirements": [], + "inputs": [], + "commands": [ + { + "argv": ["{python}", "-B", "-W", "error", "{package}/verify.py"], + "timeout_seconds": 60, + "checks": "Projector inventories 6/12/20/30, ordered same-image and same-kernel collapse pairs, subset collapse, vanishing determinant of P-I over F2/F3/F4/F5, two-call diagonal forcing, six-call positive control and mixed-line negative control." + }, + { + "argv": ["{python}", "-O", "-B", "-W", "error", "{package}/verify.py"], + "timeout_seconds": 60, + "checks": "Same exact checks with optimisations enabled; all assertions are explicit raises." + } + ] +} diff --git a/research/diagonal-pair-line-separation/verify.py b/research/diagonal-pair-line-separation/verify.py new file mode 100644 index 0000000000..c6c4cb532f --- /dev/null +++ b/research/diagonal-pair-line-separation/verify.py @@ -0,0 +1,321 @@ +#!/usr/bin/env python3 +"""Same-image-line and same-kernel-line exclusion for the diagonal-pair word. + +Standard library only. Exact finite-field arithmetic; the group algebra is the +free group on the call positions, with words as multisets of call positions (the free abelian group), so a +vanishing group-algebra element is a coefficient-wise vanishing identity. +""" +import itertools +import json + +# ---------------------------------------------------------------- fields + +class PrimeField: + def __init__(self, p): + self.q = p + self.elements = tuple(range(p)) + self.zero = 0 + self.one = 1 + + def add(self, a, b): + return (a + b) % self.q + + def mul(self, a, b): + return (a * b) % self.q + + def neg(self, a): + return (-a) % self.q + + def inv(self, a): + for b in self.elements: + if self.mul(a, b) == self.one: + return b + raise ValueError('not invertible') + + +class GF4: + q = 4 + elements = (0, 1, 2, 3) + zero = 0 + one = 1 + MUL = ((0, 0, 0, 0), (0, 1, 2, 3), (0, 2, 3, 1), (0, 3, 1, 2)) + + def add(self, a, b): + return a ^ b + + def mul(self, a, b): + return self.MUL[a][b] + + def neg(self, a): + return a + + def inv(self, a): + for b in self.elements: + if self.mul(a, b) == self.one: + return b + raise ValueError('not invertible') + + +FIELDS = (PrimeField(2), PrimeField(3), GF4(), PrimeField(5)) + +# ---------------------------------------------------------------- matrices + +def mmul(F, A, B): + return (F.add(F.mul(A[0], B[0]), F.mul(A[1], B[2])), + F.add(F.mul(A[0], B[1]), F.mul(A[1], B[3])), + F.add(F.mul(A[2], B[0]), F.mul(A[3], B[2])), + F.add(F.mul(A[2], B[1]), F.mul(A[3], B[3]))) + + +def madd(F, A, B): + return tuple(F.add(x, y) for x, y in zip(A, B)) + + +def mscale(F, c, A): + return tuple(F.mul(c, x) for x in A) + + +I2 = (1, 0, 0, 1) + + +def projectors(F): + out = [] + for a in F.elements: + for b in F.elements: + for c in F.elements: + for d in F.elements: + N = (a, b, c, d) + if mmul(F, N, N) != N: + continue + if F.add(a, d) != F.one: + continue + out.append(N) + return tuple(out) + + +def normalise(F, v): + for x in v: + if x != F.zero: + inv = F.inv(x) + return tuple(F.mul(inv, y) for y in v) + return None + + +def image_line(F, N): + a, b, c, d = N + for v in ((a, c), (b, d)): + w = normalise(F, v) + if w is not None: + return w + raise ValueError('zero image') + + +def kernel_line(F, N): + a, b, c, d = N + for v in ((b, F.neg(a)), (d, F.neg(c))): + w = normalise(F, v) + if w is not None: + # keep only vectors that really lie in the kernel + if (F.add(F.mul(a, w[0]), F.mul(b, w[1])) == F.zero and + F.add(F.mul(c, w[0]), F.mul(d, w[1])) == F.zero): + return w + raise ValueError('no kernel vector found') + +# ---------------------------------------------------------------- group algebra + +def gadd(A, B, F): + out = dict(A) + for w, c in B.items(): + v = F.add(out.get(w, F.zero), c) + if v == F.zero: + out.pop(w, None) + else: + out[w] = v + return out + + +def gmul(A, B, F): + # commutative group algebra: words are multisets of call positions + out = {} + for w1, c1 in A.items(): + for w2, c2 in B.items(): + w = tuple(sorted(w1 + w2)) + v = F.add(out.get(w, F.zero), F.mul(c1, c2)) + if v == F.zero: + out.pop(w, None) + else: + out[w] = v + return out + + +def gscale(A, c, F): + if c == F.zero: + return {} + return {w: F.mul(c, v) for w, v in A.items() if F.mul(c, v) != F.zero} + + +def gneg(A, F): + return {w: F.neg(v) for w, v in A.items() if F.neg(v) != F.zero} + + +def expand(F, Ns): + """Return the four entries of the product as group-algebra elements. + + Ward 0 of the dict is the empty word (the identity of the free group). + """ + terms = {(): I2} + for i, N in enumerate(Ns): + nxt = {} + for mask, C in terms.items(): + left = mmul(F, C, madd(F, I2, mscale(F, F.neg(F.one), N))) + right = mmul(F, C, N) + nxt[mask] = left + nxt[mask + (i,)] = right + terms = nxt + entries = [{}, {}, {}, {}] + for mask, C in terms.items(): + for e in range(4): + if C[e] != F.zero: + entries[e][mask] = C[e] + return entries + + +def minor(F, entries): + A = [dict(entries[e]) for e in range(4)] + A[0] = gadd(A[0], {(): F.neg(F.one)}, F) + A[3] = gadd(A[3], {(): F.neg(F.one)}, F) + return gadd(gmul(A[0], A[3], F), gneg(gmul(A[1], A[2], F), F), F) + +# ---------------------------------------------------------------- main + +def require(cond, msg): + if not cond: + raise SystemExit('FAIL ' + msg) + + +def same_class(projector_list, key): + classes = {} + for N in projector_list: + classes.setdefault(key(N), []).append(N) + return classes + + +def main(): + report = {} + for F in FIELDS: + q = F.q + inv = projectors(F) + require(len(inv) == q * (q + 1), + 'projector inventory %d for q=%d' % (len(inv), q)) + require(all(mmul(F, N, N) == N for N in inv), 'idempotence') + require(all(F.add(N[0], N[3]) == F.one for N in inv), 'trace one') + + by_image = same_class(inv, lambda N: image_line(F, N)) + by_kernel = same_class(inv, lambda N: kernel_line(F, N)) + require(len(by_image) == q + 1, 'image line count') + require(len(by_kernel) == q + 1, 'kernel line count') + require(all(len(v) == q for v in by_image.values()), 'image class size') + require(all(len(v) == q for v in by_kernel.values()), 'kernel class size') + + pairs = 0 + for cls in by_image.values(): + for X in cls: + for Y in cls: + require(mmul(F, X, Y) == Y, 'image collapse') + pairs += 1 + for cls in by_kernel.values(): + for X in cls: + for Y in cls: + require(mmul(F, X, Y) == X, 'kernel collapse') + pairs += 1 + + # subset collapse and vanishing minors, bounded by tuple length + max_len = 5 if q <= 3 else 4 + tuples = 0 + subs = 0 + minors = 0 + for cls in by_image.values(): + for k in range(1, max_len + 1): + for Ns in _product(cls, k): + tuples += 1 + for mask in range(1, 1 << k): + prod = I2 + last = None + for i in range(k): + if mask >> i & 1: + prod = mmul(F, prod, Ns[i]) + last = Ns[i] + require(prod == last, 'subset collapse to last') + subs += 1 + M = minor(F, expand(F, Ns)) + require(M == {}, 'same-image minor did not vanish') + minors += 1 + for cls in by_kernel.values(): + for k in range(1, max_len + 1): + for Ns in _product(cls, k): + tuples += 1 + for mask in range(1, 1 << k): + prod = I2 + first = None + for i in range(k): + if mask >> i & 1: + prod = mmul(F, prod, Ns[i]) + if first is None: + first = Ns[i] + require(prod == first, 'subset collapse to first') + subs += 1 + M = minor(F, expand(F, Ns)) + require(M == {}, 'same-kernel minor did not vanish') + minors += 1 + + # two-call normal form: with N1 = E00, diag(diag([g1],1) (I+t2 N2)) + # forces N2 diagonal, checked in the free group on two letters. + E00 = (1, 0, 0, 0) + t2 = {(): F.one, (1,): F.one} + two_call = 0 + for N2 in inv: + entries = expand(F, (E00, N2)) # free group on two letters + diag_q = (entries[1] == {} and entries[2] == {}) + diag_n = (N2[1] == F.zero and N2[2] == F.zero) + require(diag_q == diag_n, 'two-call diagonal forcing') + two_call += 1 + + report[q] = { + 'projectors': len(inv), + 'ordered_collapse_pairs': pairs, + 'same_line_tuples': tuples, + 'subset_collapses': subs, + 'vanishing_minors': minors, + 'two_call_projectors': two_call, + 'max_tuple_length': max_len, + } + + _controls() + print('PASS_DIAGONAL_PAIR_LINE_SEPARATION ' + + json.dumps(report, sort_keys=True)) + + +def _product(cls, k): + return itertools.product(cls, repeat=k) + + +def _controls(): + F = PrimeField(2) + E00 = (1, 0, 0, 0) + E11 = (0, 0, 0, 1) + # six-call F2 word: diag([g1 g2 g3], [g4 g5 g6]); the minor is nonzero + w = (E00, E00, E00, E11, E11, E11) + M = minor(F, expand(F, w)) + require(M != {}, 'six-call positive control minor vanished') + expected = {(0, 1, 2): 1, (3, 4, 5): 1, (0, 1, 2, 3, 4, 5): 1, (): 1} + require(M == expected, 'six-call minor coefficients') + # mixing image lines can give a nonzero minor: E00 then E11 is diagonal + # with two distinct targets, so (P-I) has rank two there. + require(image_line(F, E11) != image_line(F, E00), 'control lines coincide') + require(minor(F, expand(F, (E00, E11))) != {}, 'mixed-line control vanished') + # a same-image pair does vanish (already covered above, repeat explicitly) + require(minor(F, expand(F, (E00, E00))) == {}, 'same-image control') + + +if __name__ == '__main__': + main() diff --git a/research/historical-selected-replay-gap/README.md b/research/historical-selected-replay-gap/README.md index 55d8cc3b49..36d9a873ca 100644 --- a/research/historical-selected-replay-gap/README.md +++ b/research/historical-selected-replay-gap/README.md @@ -69,7 +69,11 @@ python3 -B scripts/research_package.py --package research/historical-selected-re The direct command prints `PASS_HISTORICAL_SELECTED_REPLAY_GAP`. A registered replay prints `PASS_REGISTERED_RESEARCH_PACKAGE`. Both require Git history -containing the two inspected commits. +containing the two inspected commits and the source and manifest objects at +`2485805a6d47f36ea83ddde6327d453ac82d44dc`. The verifier uses a temporary +sparse checkout of that commit, checks the declared source digests there, and +uses its planner and manifest inventory. The recorded rejections and selection +therefore remain reproducible after later changes to the live Makefile or CI guide. ## Open next questions diff --git a/research/historical-selected-replay-gap/SOURCES.json b/research/historical-selected-replay-gap/SOURCES.json index c1010eec95..62bb8b3950 100644 --- a/research/historical-selected-replay-gap/SOURCES.json +++ b/research/historical-selected-replay-gap/SOURCES.json @@ -4,13 +4,13 @@ "files": [ { "path": "Makefile", - "usage": "input", + "usage": "provenance", "git_blob": "0d683a06b9e7ee162496108ce9a7c5d2ee94fccf", "sha256": "4f2d4f2dcd15139ec82658a975f85ed300897385f85799218539f4c8d60567da" }, { "path": "certificates/partial-swap-network.json", - "usage": "input", + "usage": "provenance", "git_blob": "eccafa830c16170c36faf93045ddf1f638f7bcd4", "sha256": "4c0b9d37e1034b2076e30a70f8e514416298c462d9ee98bc87fe6556b14df786" }, @@ -22,7 +22,7 @@ }, { "path": "docs/ci-verification.md", - "usage": "input", + "usage": "provenance", "git_blob": "63ba3e9aa8bdc5e942b56e34244b92b16f854085", "sha256": "c7d3027e9ad8bcbff68f07aa9576b30fb2f3c76c44d876d40c7fdbf4a3e2228a" }, @@ -34,163 +34,163 @@ }, { "path": "research/clifford-state-performance/SOURCE.json", - "usage": "input", + "usage": "provenance", "git_blob": "6f98a0061a5104dfab82c9e806f819b41811f684", "sha256": "396d0f08df24952248140d5c65c51b8c902ef1c580892ff7ea182890530f67ee" }, { "path": "research/community-batch1047-audit/CONTENT-COVERAGE.json", - "usage": "input", + "usage": "provenance", "git_blob": "5e0f92af463ec12c5e1dcc3effd44081a88af647", "sha256": "07a9daeaf2bda308a92e9a4684f13d5fc766c7772d74801fceb7f4599fdab226" }, { "path": "research/community-batch1099-audit/CONTENT-COVERAGE.json", - "usage": "input", + "usage": "provenance", "git_blob": "9e13266bf85e627f1b7dc225e9530ea69fd2c7f0", "sha256": "421e1211fb1897c45621dc6801c276e53f2e6f335583c9c9049e535f87c0a3b1" }, { "path": "research/community-batch932-audit/CONTENT-COVERAGE.json", - "usage": "input", + "usage": "provenance", "git_blob": "cc4d03eb59b9ab3f952735f7e88ff18c2d15d7af", "sha256": "184629cac91a5cfaca50330a88596008b824221433867dcf8117539c913c284a" }, { "path": "research/community-batch970-audit/CONTENT-COVERAGE.json", - "usage": "input", + "usage": "provenance", "git_blob": "8b9f4111badf7de5d97684799560b55110e9da79", "sha256": "23188ed920e38a54006cb883e181585620821c63bf7e8b5e05d32719417cc37e" }, { "path": "research/community-frontier741-audit/CONTENT-COVERAGE.json", - "usage": "input", + "usage": "provenance", "git_blob": "a5300d4e7a5381a3f443fc54c768ca6f0f236d68", "sha256": "3c820fb30e1b4d9bcfc6a7ab524375e5b2abd66b940af2c509146104f0bcd79c" }, { "path": "research/community-frontier890-audit/CONTENT-COVERAGE.json", - "usage": "input", + "usage": "provenance", "git_blob": "d7ff6d777b01d69b6ce84caf25d3b7e4a4d2f211", "sha256": "6b185a9902652f893f3ebd8ada92d2f11b6de6533a9a4d11718a8a4496101128" }, { "path": "research/community-hyperplane599-audit/SOURCE.json", - "usage": "input", + "usage": "provenance", "git_blob": "546f37d326d10a99910ae8f908e71ef768524d98", "sha256": "65099d764bfcc39bda5e4d9966783611b094c3522dbdd1435cd498d657a8ef8e" }, { "path": "research/community-k492-audit/SOURCE.json", - "usage": "input", + "usage": "provenance", "git_blob": "e538481e4cb16ed936d3195f2dfb5b5a9ca0a76b", "sha256": "2c0f4fa5bc150d05658af2e7c54c4d2c337dd52b7efd3d69530ae73bcbdf452a" }, { "path": "research/community-reconciliation92-audit/MANIFEST.json", - "usage": "input", + "usage": "provenance", "git_blob": "2a2e93685173d5f13cbb5abe740f5d90352e0228", "sha256": "5f501f3c118131e7d069f30ad77186eb2ad6961f9680a0ef105f4514ed41808a" }, { "path": "research/partial-swap-digest/baseline.py", - "usage": "input", + "usage": "provenance", "git_blob": "a2f0190704db83bd8ef48e46e723c49b925804fe", "sha256": "81e9af92cdf38243a540cdcff2062778e449467848d444da37a9553c12ca4c34" }, { "path": "research/r420-two-sided-basis/SOURCES.json", - "usage": "input", + "usage": "provenance", "git_blob": "5d718e11e62f4bcadecf1e02d6b61f87dbe190b0", "sha256": "d9810b0d139c5a1211fd2c1f0d142187cc1fca30026e1e7bc4f9ecfacf1596d1" }, { "path": "research/skip-clones/SOURCE.json", - "usage": "input", + "usage": "provenance", "git_blob": "8d1bb3b9c8189e35e5d8bfdfa632d7473e8be0d5", "sha256": "0b68e63e35a97ce6b0cdb6e291b88ebaa8f2dd4afb70b2498f10ee89428258dc" }, { "path": "research/split-skip/SOURCE.json", - "usage": "input", + "usage": "provenance", "git_blob": "2ab610b6ea2edd2865fda793ef2168b0c76afb4f", "sha256": "7c2b4646a3ecfef6f881da5c43e4bd0fa9a4d1b00548b8ab458d897e16735134" }, { "path": "scripts/historical-ci-inputs/Makefile", - "usage": "input", + "usage": "provenance", "git_blob": "aff68409021c9d401a6f8013b9dfd7dbcc30f2a8", "sha256": "ea8830dd79da493e2d75baf753d733ff7efbcf2b4794afe41fae0ac8f3055646" }, { "path": "scripts/historical-ci-inputs/ledger-batch932-run_isolated_tests.py", - "usage": "input", + "usage": "provenance", "git_blob": "65db599ab0b1dea0312c4ca048e0bd0f44c3736f", "sha256": "96ea543e6c707424afe247e77bd86c82152f90fc76a0773c4232145da5cdf8d1" }, { "path": "scripts/historical-ci-inputs/ledger-batch932-test_isolated_test_runner.py", - "usage": "input", + "usage": "provenance", "git_blob": "34820a8f24678f918f4d0e7a35380d8013fc3495", "sha256": "628391b6abfbc8674ceec1751867a4392f2cc0b5b66886243e139af86cf355ac" }, { "path": "scripts/historical-ci-inputs/ledger-batch970-verify_selected_result.py", - "usage": "input", + "usage": "provenance", "git_blob": "c755b76e0d44ba4917aa94440ebe54daefce5b2a", "sha256": "d7fd929726b326aeeaf0a3996475adec88eee9aac09ae0bebeeaa0bc1fc8f199" }, { "path": "scripts/historical-ci-inputs/ledger-frontier741-ci-verification.md", - "usage": "input", + "usage": "provenance", "git_blob": "7f3c012504187917a728b9fe87043f0fc8cd006a", "sha256": "a854022e0e1d5d4db4d60e586be3ea56005013ef14a1fc06d02bc06e8529c4b0" }, { "path": "scripts/historical-ci-inputs/partial-swap-context.json", - "usage": "input", + "usage": "provenance", "git_blob": "ad65114886bf34a4fd27e79f3be502389e689faa", "sha256": "833a9642f33aeab2ffea17ec9c6b82e0a953f90192b2c82e10bef21538dccba9" }, { "path": "scripts/historical-ci-inputs/partial-swap-network.json", - "usage": "input", + "usage": "provenance", "git_blob": "52075f2463f60af5e16571347f24274fce73a1ea", "sha256": "f3c4be597b2f2136da4ae2cc80c7454f63c90aa01f84e5b8131cee4dd55de815" }, { "path": "scripts/partial_swap/shared.py", - "usage": "input", + "usage": "provenance", "git_blob": "44a0e454c70f9c4b98e258b8443612e6a119904b", "sha256": "5a0646a50b9f3e8ea55168620a78e8853b0ab810517cefc8cfb8696aedee278a" }, { "path": "scripts/replay_historical_check.py", - "usage": "input", + "usage": "provenance", "git_blob": "1e92fff2782970ef71378f594f2350d245a4db26", "sha256": "5467619c78183d3adc82b0cb55af28e5c1a01fca2b0588677816244b53e2bafd" }, { "path": "scripts/research_package.py", - "usage": "input", + "usage": "provenance", "git_blob": "091b4c8f58bac9a9580d76025b887dc5c873d05f", "sha256": "e1256016e7f51827d04fa2179255989f639ae1d97afaa9c16265f85a0c6abd70" }, { "path": "scripts/run_isolated_tests.py", - "usage": "input", + "usage": "provenance", "git_blob": "a96ab35b315ce492e7fbbfd6c5ae211aa066b541", "sha256": "49eb56dd0fc0d8d8ffdb209970a7adf2176223580646f575d5c22e59543b381c" }, { "path": "scripts/verify_selected_result.py", - "usage": "input", + "usage": "provenance", "git_blob": "6882a4700e410dec569b51ae2eaf1a19cc019195", "sha256": "d6b305eac5b9bdb8172c469ad30a1655a60fa0aceac8b1bf610e1b73f51babd8" }, { "path": "tests/test_isolated_test_runner.py", - "usage": "input", + "usage": "provenance", "git_blob": "343b24abf6719b81805eba5a84c74e57634c8288", "sha256": "3a5470d0b139828b90e19bf9583ddeafcd69f8e34f2824664eb3a2958a52e34d" } diff --git a/research/historical-selected-replay-gap/VERIFY.json b/research/historical-selected-replay-gap/VERIFY.json index 98553e72ed..ab7efa3471 100644 --- a/research/historical-selected-replay-gap/VERIFY.json +++ b/research/historical-selected-replay-gap/VERIFY.json @@ -4,36 +4,7 @@ "python": "3.12", "requirements": [], "inputs": [ - "Makefile", - "certificates/partial-swap-network.json", - "docs/ci-verification.md", - "research/clifford-state-performance/SOURCE.json", - "research/community-batch1047-audit/CONTENT-COVERAGE.json", - "research/community-batch1099-audit/CONTENT-COVERAGE.json", - "research/community-batch932-audit/CONTENT-COVERAGE.json", - "research/community-batch970-audit/CONTENT-COVERAGE.json", - "research/community-frontier741-audit/CONTENT-COVERAGE.json", - "research/community-frontier890-audit/CONTENT-COVERAGE.json", - "research/community-hyperplane599-audit/SOURCE.json", - "research/community-k492-audit/SOURCE.json", - "research/community-reconciliation92-audit/MANIFEST.json", - "research/partial-swap-digest/baseline.py", - "research/r420-two-sided-basis/SOURCES.json", - "research/skip-clones/SOURCE.json", - "research/split-skip/SOURCE.json", - "scripts/historical-ci-inputs/Makefile", - "scripts/historical-ci-inputs/ledger-batch932-run_isolated_tests.py", - "scripts/historical-ci-inputs/ledger-batch932-test_isolated_test_runner.py", - "scripts/historical-ci-inputs/ledger-batch970-verify_selected_result.py", - "scripts/historical-ci-inputs/ledger-frontier741-ci-verification.md", - "scripts/historical-ci-inputs/partial-swap-context.json", - "scripts/historical-ci-inputs/partial-swap-network.json", - "scripts/partial_swap/shared.py", - "scripts/replay_historical_check.py", - "scripts/research_package.py", - "scripts/run_isolated_tests.py", - "scripts/verify_selected_result.py", - "tests/test_isolated_test_runner.py" + "scripts/research_package.py" ], "commands": [ { diff --git a/research/historical-selected-replay-gap/verify.py b/research/historical-selected-replay-gap/verify.py index 83b995941a..9d266c5a4e 100644 --- a/research/historical-selected-replay-gap/verify.py +++ b/research/historical-selected-replay-gap/verify.py @@ -6,8 +6,10 @@ """ import hashlib import importlib.util +import json import subprocess import sys +import tempfile from pathlib import Path ROOT = Path(__file__).resolve().parents[2] @@ -61,31 +63,47 @@ def require(ok, message): raise SystemExit(message) -def load_checker(): - path = ROOT / "scripts/replay_historical_check.py" - spec = importlib.util.spec_from_file_location("historical_selected_replay_gap_checker", path) +def load_module(path): + spec = importlib.util.spec_from_file_location(path.stem, path) module = importlib.util.module_from_spec(spec) spec.loader.exec_module(module) return module +def checkout_snapshot(snapshot, sources, runner, env): + paths = [runner.canonical_relative_path(row["path"]) for row in sources["files"]] + git = ["git", "-c", "maintenance.auto=false", "-c", "gc.auto=0", + "-c", "core.autocrlf=false"] + subprocess.run([*git, "clone", "--quiet", "--shared", "--no-tags", "--no-checkout", + "--", str(ROOT), str(snapshot)], cwd=ROOT, env=env, check=True) + subprocess.run([*git, "update-ref", "--no-deref", "HEAD", INSPECTED], + cwd=snapshot, env=env, check=True) + # The pinned planner reads omitted manifests from the sparse HEAD/index. + patterns = "".join(f"/{path.as_posix()}\n" for path in paths) + subprocess.run([*git, "sparse-checkout", "set", "--no-cone", "--stdin"], + cwd=snapshot, env=env, input=patterns.encode("utf-8"), check=True) + subprocess.run([*git, "read-tree", "-mu", "HEAD"], cwd=snapshot, env=env, check=True) + for path, row in zip(paths, sources["files"], strict=True): + require(hashlib.sha256((snapshot / path).read_bytes()).hexdigest() == row["sha256"], + "historical checkout source digest changed: " + path.as_posix()) + + def entries(checker): for name, config in checker.CONFIG.items(): for entry in [config, *config.get("additional_inputs", [])]: yield name, entry -def reject(checker, entry): +def reject(root, checker, entry): try: - checker.authenticated_input(ROOT, entry) + checker.authenticated_input(root, entry) except ValueError as exc: return str(exc) raise SystemExit("historical input was admitted: " + entry["path"]) -def main(): - require(not sys.flags.optimize, "assertions must stay enabled") - checker = load_checker() +def check_snapshot(root, env): + checker = load_module(root / "scripts/replay_historical_check.py") selected = [entry for name, entry in entries(checker) if name == "tooling-ledger" and entry["path"] == SELECTED_PATH] require(len(selected) == 1, "selected-result checker overlay is not unique") @@ -94,7 +112,7 @@ def main(): require(entry["accepted_live_sha256"] == [SELECTED_ACCEPTED], "accepted selected-checker successors changed") frozen = (checker.INPUTS / entry["input"]).read_bytes() - live = (ROOT / SELECTED_PATH).read_bytes() + live = (root / SELECTED_PATH).read_bytes() require(hashlib.sha256(frozen).hexdigest() == SELECTED_PIN, "frozen selected checker bytes changed") require(hashlib.sha256(live).hexdigest() == SELECTED_LIVE, "live selected checker digest changed") require(SELECTED_LIVE not in (SELECTED_PIN, SELECTED_ACCEPTED), "live checker was already admitted") @@ -102,7 +120,7 @@ def main(): # live successor, not a broken comparison. require(SELECTED_PIN in (SELECTED_PIN, *entry["accepted_live_sha256"]), "pin membership control failed") - require(reject(checker, entry) == REJECTIONS[SELECTED_PATH][1], + require(reject(root, checker, entry) == REJECTIONS[SELECTED_PATH][1], "selected checker rejection text changed") counts = {path: 0 for path in REJECTIONS} @@ -110,7 +128,7 @@ def main(): for _name, item in entries(checker): if item["path"] not in REJECTIONS: continue - message = reject(checker, item) + message = reject(root, checker, item) counts[item["path"]] += 1 messages.setdefault(item["path"], message) for path, (count, message) in REJECTIONS.items(): @@ -118,20 +136,31 @@ def main(): require(messages[path] == message, "rejection text changed for " + path) require(sum(counts.values()) == 19, "rejection total changed") ci = [item for _name, item in entries(checker) if item["path"] == CI_PATH] - require(len(ci) == 1 and hashlib.sha256((ROOT / CI_PATH).read_bytes()).hexdigest() == CI_LIVE, + require(len(ci) == 1 and hashlib.sha256((root / CI_PATH).read_bytes()).hexdigest() == CI_LIVE, "selected CI note digest changed") diff = subprocess.check_output( - ["git", "diff", "--name-only", PARENT, INSPECTED], cwd=ROOT, text=True) + ["git", "diff", "--name-only", PARENT, INSPECTED], cwd=root, env=env, text=True) changed = sorted(name for name in diff.splitlines() if name) require(changed == CHANGED, "inspected commit diff changed") - sys.path.insert(0, str(ROOT / "scripts")) - from research_package import plan_details - details = plan_details(ROOT, set(changed)) + planner = load_module(root / "scripts/research_package.py") + details = planner.plan_details(root, set(changed)) require([row["package"] for row in details] == PLAN_PACKAGES, "helper plan selection changed") matched = {name for row in details for names in row["causes"].values() for name in names} require(sorted(set(changed) - matched) == UNMATCHED, "unmatched plan paths changed") - require(plan_details(ROOT, set()) == [], "empty change set selected a package") + require(planner.plan_details(root, set()) == [], "empty change set selected a package") + + +def main(): + require(not sys.flags.optimize, "assertions must stay enabled") + sources = json.loads(Path(__file__).with_name("SOURCES.json").read_bytes()) + require(sources["base_commit"] == INSPECTED, "historical checkout base changed") + runner = load_module(ROOT / "scripts/research_package.py") + env = runner.git_source_env() + with tempfile.TemporaryDirectory(prefix="historical-selected-replay-") as temporary: + snapshot = Path(temporary) / "snapshot" + checkout_snapshot(snapshot, sources, runner, env) + check_snapshot(snapshot, env) print("PASS_HISTORICAL_SELECTED_REPLAY_GAP") print("selected_checker_live=" + SELECTED_LIVE) print("rejections=19 distinct_paths=7 plan_packages=9")