"""Clean Yahoo Finance OHLCV data and compute log returns."""

from __future__ import annotations

import numpy as np
import pandas as pd

from src.config import DATA_PROCESSED, DATA_RAW_YAHOO
from src.utils import setup_logging

logger = setup_logging("yahoo_cleaner")


def clean_yahoo_data() -> pd.DataFrame:
    """Load all Yahoo CSVs, compute log returns, return a combined DataFrame.

    For each ticker file (SPY.csv, BTC_USD.csv, etc.) we:
      1. Parse dates and set DatetimeIndex.
      2. Keep Close and Volume.
      3. Compute log return = ln(Close_t / Close_{t-1}).

    Returns a wide DataFrame with columns like ``SPY_close``, ``SPY_volume``,
    ``SPY_log_return``, etc.
    """
    combined = pd.DataFrame()

    for csv_path in sorted(DATA_RAW_YAHOO.glob("*.csv")):
        # Skip the combined all_tickers file -- we use per-ticker files
        if csv_path.stem == "all_tickers":
            continue

        ticker = csv_path.stem  # e.g. "SPY", "BTC_USD"
        df = pd.read_csv(csv_path, parse_dates=["date"], index_col="date")
        df.index = pd.to_datetime(df.index, utc=True).normalize().tz_localize(None)

        # Coerce numeric
        for col in ["Close", "Volume"]:
            if col in df.columns:
                df[col] = pd.to_numeric(df[col], errors="coerce")

        if "Close" not in df.columns:
            logger.warning("No 'Close' column in %s, skipping", csv_path.name)
            continue

        close = df["Close"].rename(f"{ticker}_close")
        log_ret = np.log(df["Close"] / df["Close"].shift(1)).rename(f"{ticker}_log_return")

        combined[f"{ticker}_close"] = close
        combined[f"{ticker}_log_return"] = log_ret

        if "Volume" in df.columns:
            volume = df["Volume"].rename(f"{ticker}_volume")
            combined[f"{ticker}_volume"] = volume

        logger.info("Loaded Yahoo ticker '%s': %d observations", ticker, len(df))

    combined = combined.sort_index()
    logger.info(
        "Combined Yahoo data: %d rows, %d columns, date range %s to %s",
        len(combined), len(combined.columns),
        combined.index.min(), combined.index.max(),
    )
    return combined


def clean_all() -> None:
    """Save ``financial_data.parquet`` to ``data/processed/``."""
    DATA_PROCESSED.mkdir(parents=True, exist_ok=True)

    financial = clean_yahoo_data()
    if not financial.empty:
        out = DATA_PROCESSED / "financial_data.parquet"
        financial.to_parquet(out)
        logger.info("Saved %s (%d rows)", out.name, len(financial))

    logger.info("=== Yahoo cleaning complete ===")