scripts/upload_traces.py

#!/usr/bin/env python3
"""Upload agent_traces.jsonl to a Hugging Face Dataset (CodeFlow-style)."""

from __future__ import annotations

import argparse
import sys
from pathlib import Path

from huggingface_hub import HfApi

from trace_store import TRACE_JSONL_PATH, trace_file_exists


def main() -> int:
    parser = argparse.ArgumentParser(
        description="Upload agent_traces.jsonl to a Hugging Face Dataset.",
    )
    parser.add_argument(
        "--repo",
        required=True,
        help="Target dataset repo id, e.g. your-username/car-diagnostic-agent-traces",
    )
    parser.add_argument(
        "--path",
        type=Path,
        default=TRACE_JSONL_PATH,
        help=f"Local JSONL file (default: {TRACE_JSONL_PATH})",
    )
    parser.add_argument(
        "--remote-path",
        default="agent_traces.jsonl",
        help="Path inside the dataset repo (default: agent_traces.jsonl)",
    )
    parser.add_argument(
        "--private",
        action="store_true",
        help="Create the dataset as private if it does not exist",
    )
    args = parser.parse_args()

    if not args.path.is_file() or args.path.stat().st_size == 0:
        print(f"No trace file at {args.path}", file=sys.stderr)
        return 1

    api = HfApi()
    try:
        api.repo_info(args.repo, repo_type="dataset")
    except Exception:
        api.create_repo(
            args.repo,
            repo_type="dataset",
            private=args.private,
            exist_ok=True,
        )

    api.upload_file(
        path_or_fileobj=str(args.path),
        path_in_repo=args.remote_path,
        repo_id=args.repo,
        repo_type="dataset",
        commit_message="Update agent traces JSONL",
    )
    print(f"Uploaded {args.path} → {args.repo}/{args.remote_path}")
    return 0


if __name__ == "__main__":
    raise SystemExit(main())