SCRIP-CORPUS-084: pin Darwin VAR revision SHA-1 identities

.github/workflows/darwin-var-revision-identities.yml

@@ -0,0 +1,115 @@
+name: Darwin VAR frozen revision identities
+
+on:
+  pull_request:
+    paths:
+      - ".github/workflows/darwin-var-revision-identities.yml"
+      - "scriptorium/single_page_revision.py"
+      - "scriptorium/darwin_var_revision_identities.py"
+      - "tests/test_darwin_var_revision_identities.py"
+      - "corpus/candidates/source-edition-traces/darwin-origin-species-rachinsky-1864-ru.var-revision-identities-evidence.json"
+  workflow_dispatch:
+
+permissions:
+  contents: read
+
+jobs:
+  capture:
+    name: Capture six exact Wikisource revision identities
+    runs-on: ubuntu-latest
+    timeout-minutes: 10
+    steps:
+      - name: Checkout pull-request head
+        uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1
+        with:
+          ref: ${{ github.event.pull_request.head.sha || github.sha }}
+          persist-credentials: false
+
+      - name: Set up Python 3.13
+        uses: actions/setup-python@5fda3b95a4ea91299a34e894583c3862153e4b97 # v7.0.0
+        with:
+          python-version: "3.13"
+
+      - name: Run standard-library test suite
+        run: python -m unittest discover -s tests -v
+
+      - name: Capture provider-returned source-free identities
+        run: |
+          set -euo pipefail
+          mkdir -p build/repro
+          python - <<'PY'
+          import json
+          from pathlib import Path
+          from scriptorium.single_page_revision import fetch_pinned_revision
+
+          # The dependency evidence uses MediaWiki's English namespace alias in Lua
+          # require() strings. The Russian Wikisource API canonicalizes those page
+          # titles to the local `Модуль:` namespace, so preserve both identities.
+          resources = [
+              ("Шаблон:ВАР", "Шаблон:ВАР", 3684602),
+              ("Модуль:Дореформенная орфография", "Модуль:Дореформенная орфография", 5721277),
+              ("Module:Header", "Модуль:Header", 5746249),
+              ("Module:BEED", "Модуль:BEED", 5746253),
+              ("Module:Util", "Модуль:Util", 5750249),
+              ("Module:RomanNumber", "Модуль:RomanNumber", 3684553),
+          ]
+          observed = []
+          for closure_title, provider_title, revision_id in resources:
+              item = fetch_pinned_revision(title=provider_title, revision_id=revision_id)
+              item["closure_title"] = closure_title
+              observed.append(item)
+          assert [
+              (item["closure_title"], item["title"], item["revision_id"])
+              for item in observed
+          ] == resources
+          assert all(len(item["mediawiki_sha1"]) == 40 for item in observed)
+          payload = {
+              "schema_version": "scriptorium-darwin-var-provider-revision-identities-v1",
+              "provider": "Russian Wikisource",
+              "source_text_included": False,
+              "resources": observed,
+          }
+          path = Path("build/repro/darwin-var-provider-revision-identities.json")
+          path.write_text(json.dumps(payload, ensure_ascii=False, indent=2, sort_keys=True) + "\n", encoding="utf-8")
+          print("SCRIPTORIUM_DARWIN_VAR_IDENTITIES=" + json.dumps(payload, ensure_ascii=False, sort_keys=True, separators=(",", ":")))
+          PY
+
+      - name: Compare capture to committed identity contract
+        run: |
+          set -euo pipefail
+          python - <<'PY'
+          import json
+          from pathlib import Path
+          from scriptorium.darwin_var_revision_identities import validate_evidence
+
+          observed = json.loads(Path("build/repro/darwin-var-provider-revision-identities.json").read_text(encoding="utf-8"))
+          evidence = json.loads(Path("corpus/candidates/source-edition-traces/darwin-origin-species-rachinsky-1864-ru.var-revision-identities-evidence.json").read_text(encoding="utf-8"))
+          validate_evidence(evidence)
+          expected = evidence["resources"]
+          assert observed["source_text_included"] is False
+          assert len(observed["resources"]) == len(expected) == 6
+          for got, want in zip(observed["resources"], expected, strict=True):
+              for key in (
+                  "closure_title",
+                  "title",
+                  "page_id",
+                  "revision_id",
+                  "revision_timestamp",
+                  "mediawiki_sha1",
+                  "wikitext_character_count",
+                  "wikitext_utf8_byte_count",
+                  "wikitext_sha256",
+              ):
+                  assert got[key] == want[key], (key, got[key], want[key])
+          assert evidence["identity_gate"]["mediawiki_sha1_complete"] is True
+          assert evidence["identity_gate"]["replay_ready"] is False
+          print("SCRIPTORIUM_DARWIN_VAR_EVIDENCE_SHA256=" + evidence["evidence_sha256"])
+          PY
+
+      - name: Upload source-free identity capture
+        uses: actions/upload-artifact@ea165f8d65b6e75b540449e92b4886f43607fa02 # v4.6.2
+        with:
+          name: darwin-var-provider-revision-identities
+          path: build/repro/darwin-var-provider-revision-identities.json
+          if-no-files-found: error
+          retention-days: 7

corpus/candidates/darwin-origin-species-rachinsky-1864-ru-var-revision-identities.md

@@ -0,0 +1,24 @@
+# Darwin/Rachinsky 1864 — `{{ВАР}}` provider revision identities
+
+This source-free companion advances the independently reviewed SCRIP-CORPUS-083 observed-live `{{ВАР}}` closure by binding all six exact revisions to official Russian Wikisource identity metadata. It is an identity-completion artifact, not a full `{{ВАР}}` replay, historical-transclusion claim, or renderer promotion.
+
+The frozen closure remains:
+
+`Шаблон:ВАР@3684602` → `Модуль:Дореформенная орфография@5721277` → `Module:Header@5746249` → `Module:BEED@5746253` + `Module:Util@5750249`; loading BEED also loads `Module:RomanNumber@3684553`.
+
+The official MediaWiki revisions API returned a 40-hex SHA-1 for every exact revision. Russian Wikisource canonicalizes the English `Module:` namespace aliases used by Lua dependency strings to local provider titles under `Модуль:`; the evidence therefore stores both the reviewed closure alias and the provider-returned title.
+
+| Closure title | Provider title | Revision | MediaWiki SHA-1 |
+| --- | --- | ---: | --- |
+| `Шаблон:ВАР` | `Шаблон:ВАР` | 3684602 | `fdb6fa7c0d4b08157bc30d44f5f77c5b9313167c` |
+| `Модуль:Дореформенная орфография` | `Модуль:Дореформенная орфография` | 5721277 | `64d28378f4f620c67e1823a0d4292445d9ab293d` |
+| `Module:Header` | `Модуль:Header` | 5746249 | `2ab2bfdd7e49c73c4ec9ef24e4625c1d228cfba0` |
+| `Module:BEED` | `Модуль:BEED` | 5746253 | `9e9531e9178e0333bd3a939ace91e6a71e114cc5` |
+| `Module:Util` | `Модуль:Util` | 5750249 | `945f8e6bf173bb5712385995255a8b6eccef38ed` |
+| `Module:RomanNumber` | `Модуль:RomanNumber` | 3684553 | `51a3956469ee3fbd4ed56ce3a1f72f76f231e1d3` |
+
+The machine-readable contract is [`source-edition-traces/darwin-origin-species-rachinsky-1864-ru.var-revision-identities-evidence.json`](source-edition-traces/darwin-origin-species-rachinsky-1864-ru.var-revision-identities-evidence.json), SHA-256 `98dc8aed94fd8b4e18f0ad3325673603cba1575c1e745e1175ab4cd3a8aa2a48`. It also retains provider page IDs, UTC revision timestamps, source-free wikitext counts, and SHA-256 digests captured without persisting source prose.
+
+`mediawiki_sha1_complete=true`, but `replay_ready=false`. `ВАР` remains unresolved at **388 invocations**, and the effective backlog remains **4 tag shapes / 130 tag tokens** plus **37 template shapes / 2,356 template invocations**. Full candidate invocation replay, historical transclusion, an explicit version-pinned Scribunto/runtime boundary, renderer promotion, literary-body identity, `>=300k` admission, FantLab analyzer-input identity, and M2 parity all remain open.
+
+The next bounded step is to replay the candidate `{{ВАР}}` branch against exactly these six identities under an explicit version-pinned runtime boundary; no later live descendants may silently substitute for the frozen revisions.

corpus/candidates/source-edition-traces/darwin-origin-species-rachinsky-1864-ru.var-revision-identities-evidence.json

@@ -0,0 +1,108 @@
+{
+  "candidate_id": "darwin-origin-species-rachinsky-1864-ru",
+  "capture_provenance": {
+    "api_surface": "MediaWiki revisions API: rvprop=ids|timestamp|sha1|content, rvslots=main",
+    "provider": "Russian Wikisource",
+    "source_text_persisted": false
+  },
+  "evidence_sha256": "98dc8aed94fd8b4e18f0ad3325673603cba1575c1e745e1175ab4cd3a8aa2a48",
+  "fantlab_source_edition_match": "unknown",
+  "historical_transclusion_proven": false,
+  "identity_gate": {
+    "mediawiki_sha1_complete": true,
+    "missing_mediawiki_sha1_titles": [],
+    "provider_title_aliases_explicit": true,
+    "reason": "all six exact observed-live revisions are bound to provider-returned identities, but a full candidate invocation replay and an explicit version-pinned Scribunto/runtime boundary are still missing",
+    "replay_ready": false,
+    "revision_ids_complete": true
+  },
+  "literary_body_count_and_digests_frozen": false,
+  "m2_parity_admissible": false,
+  "minimum_300k_proved": false,
+  "offline_version_pinned_runtime_proven": false,
+  "prior_closure": {
+    "evidence_sha256": "245f77bf3d99d6610ca5fbbe4fcdcdb4ef68bebca612c5bb81db6c2d77f26509",
+    "schema_version": "scriptorium-darwin-var-live-snapshot-closure-v1"
+  },
+  "promotion_decision": {
+    "backlog_item_removed": false,
+    "effective_backlog_changed": false,
+    "full_var_candidate_branch_replayed": false,
+    "next_evidence_required": "perform a separately bounded full candidate invocation replay against these six exact identities with an explicit version-pinned Scribunto/runtime boundary",
+    "render_profile_rule_promoted": false,
+    "var_invocations_unresolved": 388
+  },
+  "renderer_implementation_ready": false,
+  "renderer_semantics_complete": false,
+  "researched_on": "2026-09-24",
+  "resources": [
+    {
+      "closure_title": "Шаблон:ВАР",
+      "mediawiki_sha1": "fdb6fa7c0d4b08157bc30d44f5f77c5b9313167c",
+      "page_id": 122258,
+      "revision_id": 3684602,
+      "revision_timestamp": "2019-06-04T20:36:46Z",
+      "title": "Шаблон:ВАР",
+      "wikitext_character_count": 206,
+      "wikitext_sha256": "583f46a8f7fa2444456abede564817e134d313fb20d5cabfa2f2c48591c54707",
+      "wikitext_utf8_byte_count": 269
+    },
+    {
+      "closure_title": "Модуль:Дореформенная орфография",
+      "mediawiki_sha1": "64d28378f4f620c67e1823a0d4292445d9ab293d",
+      "page_id": 378940,
+      "revision_id": 5721277,
+      "revision_timestamp": "2026-06-08T13:06:13Z",
+      "title": "Модуль:Дореформенная орфография",
+      "wikitext_character_count": 1553,
+      "wikitext_sha256": "d0199f6df8f5f789e6a5aec0e450dae6d35c3448adca411481244e40aec3004d",
+      "wikitext_utf8_byte_count": 1588
+    },
+    {
+      "closure_title": "Module:Header",
+      "mediawiki_sha1": "2ab2bfdd7e49c73c4ec9ef24e4625c1d228cfba0",
+      "page_id": 361783,
+      "revision_id": 5746249,
+      "revision_timestamp": "2026-09-10T02:29:30Z",
+      "title": "Модуль:Header",
+      "wikitext_character_count": 54598,
+      "wikitext_sha256": "b3950aa5a04e8a8f3bec71c6c82843643dfdc1a1a82f3788f48830f21e434ecb",
+      "wikitext_utf8_byte_count": 67792
+    },
+    {
+      "closure_title": "Module:BEED",
+      "mediawiki_sha1": "9e9531e9178e0333bd3a939ace91e6a71e114cc5",
+      "page_id": 362284,
+      "revision_id": 5746253,
+      "revision_timestamp": "2026-09-10T03:41:25Z",
+      "title": "Модуль:BEED",
+      "wikitext_character_count": 10647,
+      "wikitext_sha256": "108fee006a43f146302aafda316f05f15198932c3389487ed7c760c83a7c4875",
+      "wikitext_utf8_byte_count": 11979
+    },
+    {
+      "closure_title": "Module:Util",
+      "mediawiki_sha1": "945f8e6bf173bb5712385995255a8b6eccef38ed",
+      "page_id": 980298,
+      "revision_id": 5750249,
+      "revision_timestamp": "2026-09-18T07:31:28Z",
+      "title": "Модуль:Util",
+      "wikitext_character_count": 1328,
+      "wikitext_sha256": "a69fdcf9cf4dbbfe34d3e41585f2bcd418eee4efbe6e5c42d82e450cc4045a90",
+      "wikitext_utf8_byte_count": 1585
+    },
+    {
+      "closure_title": "Module:RomanNumber",
+      "mediawiki_sha1": "51a3956469ee3fbd4ed56ce3a1f72f76f231e1d3",
+      "page_id": 378544,
+      "revision_id": 3684553,
+      "revision_timestamp": "2019-06-04T20:13:27Z",
+      "title": "Модуль:RomanNumber",
+      "wikitext_character_count": 1152,
+      "wikitext_sha256": "90edb9a1d3bf5d283112fd58e28ee60f2569ae0fe7850e936a17f4e23869cf45",
+      "wikitext_utf8_byte_count": 1212
+    }
+  ],
+  "schema_version": "scriptorium-darwin-var-provider-revision-identities-v1",
+  "source_text_included": false
+}

project/CHANGELOG.d/SCRIP-CORPUS-084.md

@@ -0,0 +1,10 @@
+## 2026-09-24 — Darwin `{{ВАР}}` provider revision identities
+
+- Selected SCRIP-CORPUS-084 / Issue #257 from the normal-flow P2 corpus queue after confirming no interrupted/open PR or issue recovery work.
+- Added a dedicated source-free GitHub Actions capture for exactly the six SCRIP-CORPUS-083 observed-live revisions using the existing fail-closed MediaWiki revision fetch primitive.
+- The first capture run correctly failed because Russian Wikisource canonicalizes `Module:` aliases to the local `Модуль:` namespace. The workflow was repaired to bind both the reviewed closure alias and provider-returned canonical title instead of weakening title validation.
+- Successful provider capture bound all six exact revision IDs to page IDs, UTC timestamps, 40-hex MediaWiki SHA-1 values, source-free wikitext counts, and SHA-256 digests without committing source prose.
+- Added deterministic evidence builder/validator, six focused regressions, and a public source-free companion. Evidence SHA-256: `98dc8aed94fd8b4e18f0ad3325673603cba1575c1e745e1175ab4cd3a8aa2a48`.
+- `mediawiki_sha1_complete=true`, but `replay_ready=false`: full `{{ВАР}}` candidate replay, historical transclusion, explicit version-pinned Scribunto/runtime identity, renderer promotion, backlog reduction, literary-body identity, >=300k admission, FantLab-input identity and M2 parity remain outside this unit and closed.
+- `ВАР` remains unresolved at 388 invocations; effective backlog remains 4 tag shapes / 130 tag tokens and 37 template shapes / 2,356 template invocations. M2 remains 0/5.
+- Draft PR #258 remains for a later independent exact-head review; this authoring run does not self-approve or merge its substantive change.