test: gate V1 side files and V3 legacy receipts; early output rejection

- analyze_jet_cv_continuation: samples are analyzed only if the sibling
  re*_s*.json receipt is the same case with run_status COMPLETE and
  source_recheck UNCHANGED; receipts without run_status (pre-recheck
  runners) are refused unless --legacy-receipts, and labelled
  LEGACY_NO_FINAL_SOURCE_RECHECK. Results carry receipt_gate.
- V3: receipts are schema 2 and need source_recheck UNCHANGED; a missing
  recheck is accepted only for frozen-mode schema 1 receipts whose
  run_case.py hash is one of the pre-recheck runners (9805383, 2cf9853,
  db3d098; hashes checked against git history in a test).
- V1/V3 refuse --output inside the checkout before source checks or CUDA
  (shared source_provenance.reject_output_inside).
- A failing recheck is recorded as RECHECK_FAILED (still invalid); the
  in-run V1 receipt write is tmp+replace.
Negative host tests for each gate. No threshold, solver or evidence edits.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
ClaudeandClaude Opus 5.5 committed 2026-10-20 19:13:59 +00:00
1 parent df0f586331
commit 36ffc8b1b2
9 files changed
+185 -33

No files matched your search

+15 -4
View File
@@ -209,15 +209,15 @@ class Host(unittest.TestCase):
done = v.finish_receipt(path, payload, recorded, lambda: dict(recorded))
self.assertEqual(json.loads(path.read_text()), done)
self.assertEqual((done["run_status"], done["source_recheck"]), ("COMPLETE", "UNCHANGED"))
for recheck, after in ((lambda: dict(recorded, checkout_dirty=True), True),
(lambda: v.verify(Path(d)), None)):
for recheck, state in ((lambda: dict(recorded, checkout_dirty=True), "CHANGED"),
(lambda: v.verify(Path(d)), "RECHECK_FAILED")):
path.write_text(json.dumps(payload))
with self.assertRaisesRegex(RuntimeError, "Source changed during the run"):
v.finish_receipt(path, payload, recorded, recheck)
bad = json.loads(path.read_text())
self.assertEqual((bad["run_status"], bad["source_recheck"]),
("INVALID_SOURCE_CHANGED", "CHANGED"))
self.assertEqual(bad["source_after"] is None, after is None)
("INVALID_SOURCE_CHANGED", state))
self.assertEqual(bad["source_after"] is None, state == "RECHECK_FAILED")
self.assertEqual(bad["windows"], [1, 2])
self.assertFalse(Path(str(path) + ".tmp").exists())
@@ -231,5 +231,16 @@ class Host(unittest.TestCase):
self.assertIn("finish_receipt(", inspect.getsource(v.main))
def test_output_inside_checkout_is_refused_before_source_or_cuda(self):
from unittest import mock
import run_jet_circle_cv as v
argv = ["run_jet_circle_cv.py", "--repo", str(REPO), "--output", str(REPO / "v1-out")]
with mock.patch("sys.argv", argv), mock.patch.dict("os.environ", CUDA_VISIBLE_DEVICES="0"), \
mock.patch.object(v, "source_identity", side_effect=AssertionError("work started")):
with self.assertRaisesRegex(SystemExit, "outside the source checkout"):
v.main()
self.assertFalse((REPO / "v1-out").exists())
if __name__ == "__main__":
unittest.main()