collect_feedback.py

# collect_feedback.py
import os, json, pandas as pd, fsspec

def collect_feedback():
    REPO = os.environ.get("HF_DATASET_REPO", "ai-tomoni/julia-dialog-logs")
    HF_TOKEN = os.environ.get("HF_TOKEN")  # set in Space secrets if private

    fs = fsspec.filesystem("hf", token=HF_TOKEN)
    paths = fs.glob(f"datasets/{REPO}/flat_logs/**/*.ndjson")
    print("paths",paths)

    rows = []
    for p in paths:
        with fs.open(p, "rb") as f:
            for line in f:
                try:
                    rec = json.loads(line)
                except Exception:
                    continue
                # keep full dict + path for later filtering
                rec["_file"] = p
                rows.append(rec)

    if not rows:
        print("⚠️ No records found.")
        return None

    df = pd.DataFrame(rows)

    # optional: sort by timestamp if available
    if "ts" in df.columns:
        df = df.sort_values("ts", na_position="last")

    df.to_csv("all_logs.csv", index=False)
    print(f"✅ all_logs.csv written with {len(df)} records and {len(df.columns)} columns.")
    return df


from huggingface_hub import HfApi, CommitOperationAdd

def push_csv_to_space():
    HF_TOKEN = os.environ.get("HF_TOKEN")  # <- uses the secret
    api = HfApi(token=HF_TOKEN)
    repo_id = "ai-tomoni/tomoni.helps"  # your Space repo id
    api.create_commit(
        repo_id=repo_id,
        repo_type="space",
        operations=[CommitOperationAdd(
            path_in_repo="all_logs.csv",
            path_or_fileobj="all_logs.csv"
        )],
        commit_message="Add/update all_logs.csv",
    )
    print("✅ all_logs.csv pushed to repo")



if __name__ == "__main__":
    collect_feedback()