File size: 2,163 Bytes
1a94b98 226b7a0 1a94b98 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 | #!/usr/bin/env python3
"""Build HuggingFace-compatible dataset from SEC-bench-Pro instances."""
import json
from pathlib import Path
import datasets
BASE_DIR = Path(__file__).resolve().parent.parent / "SEC-bench-Pro"
OUTPUT_DIR = Path(__file__).resolve().parent
def load_instances(engine: str) -> list[dict]:
engine_dir = BASE_DIR / engine
rows = []
for meta_path in sorted(engine_dir.glob("*/meta.json")):
with open(meta_path) as f:
meta = json.load(f)
instance_id = meta_path.parent.name
fix_patches = []
for fix in meta.get("fixes", []):
patch_path = meta_path.parent / fix["path"]
if patch_path.exists():
fix_patches.append(patch_path.read_text())
else:
fix_patches.append("")
row = {
"instance_id": f"{engine}__{instance_id}",
"project": engine,
"bug_id": instance_id,
"image_name": meta["image_name"],
"work_dir": meta["work_dir"],
"verification_binary": meta["verification_binary"],
"command_options": meta.get("command_options", ""),
"target_source_files": meta["target_source_files"],
"target_subdir": meta["target_subdir"],
"target_vulnerability_type": meta["target_vulnerability_type"],
"error_type": meta["error_type"],
"description": meta["description"],
"vrp": str(meta.get("vrp") or ""),
"fix_patches": fix_patches,
}
rows.append(row)
return rows
def main():
data_dir = OUTPUT_DIR / "data"
data_dir.mkdir(exist_ok=True)
v8_rows = load_instances("v8")
sm_rows = load_instances("sm")
all_rows = v8_rows + sm_rows
datasets.Dataset.from_list(all_rows).to_parquet(data_dir / "all" / "test.parquet")
datasets.Dataset.from_list(v8_rows).to_parquet(data_dir / "v8" / "test.parquet")
datasets.Dataset.from_list(sm_rows).to_parquet(data_dir / "sm" / "test.parquet")
print(f"Built dataset: all={len(all_rows)}, v8={len(v8_rows)}, sm={len(sm_rows)}")
if __name__ == "__main__":
main()
|