"""
TradeSense AI — Canonical Market Data Normalizer
Transforms raw provider CanonicalBar collections into standardized Pandas DataFrames.
"""

from typing import List
import pandas as pd
from backend.data.providers.base import CanonicalBar


class CanonicalNormalizer:
    """Standardizes bar structures into canonical DataFrame representations."""

    @staticmethod
    def to_dataframe(bars: List[CanonicalBar]) -> pd.DataFrame:
        """Converts list of CanonicalBar objects to a sorted, cleaned DataFrame."""
        if not bars:
            return pd.DataFrame(columns=["session", "Open", "High", "Low", "Close", "Volume", "VWAP"])

        records = [
            {
                "session": b.session,
                "Open": b.open,
                "High": b.high,
                "Low": b.low,
                "Close": b.close,
                "Volume": b.volume,
                "VWAP": b.vwap if b.vwap is not None else b.close,
            }
            for b in bars
        ]

        df = pd.DataFrame(records)
        # Ensure correct datatypes
        df["Open"] = df["Open"].astype(float)
        df["High"] = df["High"].astype(float)
        df["Low"] = df["Low"].astype(float)
        df["Close"] = df["Close"].astype(float)
        df["Volume"] = df["Volume"].astype(float)
        df["VWAP"] = df["VWAP"].astype(float)

        # Drop duplicates by session and sort ascending
        df = df.drop_duplicates(subset=["session"]).sort_values("session").reset_index(drop=True)
        return df