feat(accuracy): complete phase 3 data scan and quarantine

2026-08-02 01:06:35 +02:00
parent 41f8a54236
commit 7b92e29e49
13 changed files with 37368 additions and 6 deletions
@@ -0,0 +1,70 @@
+from __future__ import annotations
+
+import json
+import subprocess
+import sys
+from pathlib import Path
+
+
+SCRIPT = Path(__file__).resolve().parents[2] / "scripts" / "run_accuracy_phase3_full_data_scan.py"
+
+
+def run_scan(repo: Path, output: Path, *, resume: bool = False) -> dict:
+    command = [
+        sys.executable,
+        str(SCRIPT),
+        "--repo-root",
+        str(repo),
+        "--output-dir",
+        str(output),
+        "--roots",
+        "data",
+        "--batch-size",
+        "2",
+    ]
+    if resume:
+        command.append("--resume")
+    completed = subprocess.run(command, check=True, capture_output=True, text=True)
+    return json.loads(completed.stdout)
+
+
+def test_phase3_scan_reconciles_and_resumes_deterministically(tmp_path: Path) -> None:
+    data = tmp_path / "data"
+    data.mkdir()
+    (data / "valid.geojson").write_text(
+        json.dumps(
+            {
+                "type": "FeatureCollection",
+                "features": [
+                    {
+                        "type": "Feature",
+                        "properties": {"source": "GRB"},
+                        "geometry": {"type": "Point", "coordinates": [4.4, 50.8]},
+                    }
+                ],
+            }
+        ),
+        encoding="utf-8",
+    )
+    invalid = data / "invalid.geojson"
+    invalid.write_text(
+        '{"type":"FeatureCollection","features":[{"type":"Feature","geometry":{"type":"Polygon","coordinates":[[[0,0],[1,1],[1,0],[0,1],[0,0]]]}}]}',
+        encoding="utf-8",
+    )
+    duplicate_payload = '{"schema_version":1,"value":"same"}'
+    (data / "one.json").write_text(duplicate_payload, encoding="utf-8")
+    (data / "two.json").write_text(duplicate_payload, encoding="utf-8")
+    (data / "broken.tif").write_bytes(b"not a geotiff")
+
+    output = tmp_path / "evidence"
+    first = run_scan(tmp_path, output)
+    second = run_scan(tmp_path, output, resume=True)
+    manifest = json.loads((output / "full-scan-manifest.json").read_text(encoding="utf-8"))
+    quarantine = json.loads((output / "quarantine-manifest.json").read_text(encoding="utf-8"))
+
+    assert first["reconciliation"] == {"examined": 5, "skipped": 0, "unreachable": 3, "inventory_total": 8, "reconciles": True}
+    assert second["content_hash"] == first["content_hash"]
+    assert manifest["determinism"]["content_hash"] == first["content_hash"]
+    assert any(item["path"] == "data/broken.tif" for item in quarantine["items"])
+    assert any(item["path"] == "data/invalid.geojson" for item in quarantine["items"])
+    assert len(manifest["duplicates"]["exact_duplicate_groups"]) == 1