Operationalize RC10 data retention

2026-07-18 06:58:10 +02:00
parent bf867f8f4f
commit 7b96037853
22 changed files with 1306 additions and 2 deletions
@@ -0,0 +1,95 @@
+#!/usr/bin/env python3
+"""Dry-run-first cleanup for old unreferenced derived/cache artifacts."""
+
+from __future__ import annotations
+
+import argparse
+import json
+from pathlib import Path
+
+from audit_data_operations import build_report
+from release_backup_guard import require_confirmation, verify_current_backup
+
+from app.core.config import get_settings
+from app.db.session import SessionLocal
+
+
+CONFIRMATION = "DELETE_STORAGE_ARTIFACTS"
+
+
+def parse_args() -> argparse.Namespace:
+    parser = argparse.ArgumentParser(description=__doc__)
+    parser.add_argument("--storage-root", type=Path)
+    parser.add_argument("--minimum-age-days", type=int, default=7)
+    parser.add_argument("--max-delete", type=int, default=25)
+    parser.add_argument("--apply", action="store_true")
+    parser.add_argument("--confirm")
+    parser.add_argument("--backup-dir", type=Path)
+    parser.add_argument("--backup-max-age-hours", type=float, default=24.0)
+    return parser.parse_args()
+
+
+def main() -> int:
+    args = parse_args()
+    if args.max_delete < 0:
+        raise SystemExit("--max-delete must be greater than or equal to zero")
+    storage_root = (args.storage_root or Path(get_settings().storage_root)).resolve()
+    with SessionLocal() as db:
+        report, candidates = build_report(
+            storage_root,
+            db,
+            minimum_age_days=args.minimum_age_days,
+            max_candidate_records=max(args.max_delete, 500),
+        )
+
+    blocked_reason = None
+    backup = None
+    deleted: list[str] = []
+    if args.apply:
+        require_confirmation(args.confirm, CONFIRMATION)
+        if args.backup_dir is None:
+            raise RuntimeError("--backup-dir is required with --apply")
+        backup = verify_current_backup(
+            args.backup_dir,
+            max_age_hours=args.backup_max_age_hours,
+        )
+        if len(candidates) > args.max_delete:
+            blocked_reason = (
+                f"candidate_count {len(candidates)} exceeds --max-delete {args.max_delete}; "
+                "review the dry run and raise the explicit limit"
+            )
+        else:
+            for candidate in candidates:
+                candidate.path.unlink()
+                deleted.append(candidate.relative_path)
+
+    payload = {
+        "schema_version": 1,
+        "mode": "apply" if args.apply else "dry-run",
+        "storage_root": str(storage_root),
+        "minimum_age_days": args.minimum_age_days,
+        "max_delete": args.max_delete,
+        "candidate_count": len(candidates),
+        "candidate_bytes": sum(item.size_bytes for item in candidates),
+        "candidates": [item.relative_path for item in candidates],
+        "deleted_count": len(deleted),
+        "deleted": deleted,
+        "blocked_reason": blocked_reason,
+        "protected_prefixes": report["cleanup"]["protected_prefixes"],
+        "backup": (
+            {
+                "release_id": backup.release_id,
+                "created_at": backup.created_at.isoformat(),
+                "age_hours": round(backup.age_hours, 3),
+                "git_commit": backup.git_commit,
+            }
+            if backup
+            else None
+        ),
+    }
+    print(json.dumps(payload, indent=2, sort_keys=True))
+    return 1 if blocked_reason else 0
+
+
+if __name__ == "__main__":
+    raise SystemExit(main())