"""pubsalesacc/utils/io.py — DataFrame and file I/O helpers."""

import logging
import math
import os
from pathlib import Path

import pandas as pd

from pubsalesacc.utils.dates import get_timestamp

logger = logging.getLogger(__name__)


def df_excel_output(
    df: pd.DataFrame,
    filename: str,
    interval: int = 1_000_000,
    timestamp: bool = True,
) -> Path:
    """Write a DataFrame to an XLSX file, splitting into sheets if rows > interval.

    Args:
        df:        DataFrame to write.
        filename:  Output filename (extension optional; .xlsx appended if absent).
        interval:  Max rows per sheet. Default 1,000,000 (Excel limit).
        timestamp: Append a YYYYMMDD-HHMMSS suffix to the filename.

    Returns:
        Path to the written file.
    """
    base = Path(os.path.splitext(filename)[0])
    stem = str(base)
    out = Path(f"{stem}-{get_timestamp()}.xlsx" if timestamp else f"{stem}.xlsx")

    if df.empty:
        logger.warning("Cannot write %s — DataFrame is empty.", out.name)
        return out

    logger.info("Writing %s (%d rows)...", out.name, len(df))
    sheet_count = math.ceil(len(df) / interval)

    with pd.ExcelWriter(out) as writer:
        for sheet_idx in range(sheet_count):
            start = sheet_idx * interval
            end = min(start + interval, len(df))
            logger.info(
                "  Sheet %d/%d: rows %d–%d", sheet_idx + 1, sheet_count, start, end
            )
            df.iloc[start:end].to_excel(
                writer, sheet_name=f"DataSheet{sheet_idx + 1}", index=False
            )

    logger.info("Excel output complete: %s", out.name)
    return out


def df_csv_output(
    df: pd.DataFrame,
    filename: str,
    encoding: str = "utf-8",
) -> Path:
    """Write a DataFrame to a timestamped CSV file.

    Returns:
        Path to the written file.
    """
    base = Path(os.path.splitext(filename)[0])
    out = Path(f"{base}-{get_timestamp()}.csv")

    if df.empty:
        logger.warning("Cannot write %s — DataFrame is empty.", out.name)
        return out

    logger.info("Writing %s (%d rows)...", out.name, len(df))
    df.to_csv(out, encoding=encoding, index=False)
    logger.info("CSV output complete: %s", out.name)
    return out


def read_sql_file(filepath: str, sep: str = ";") -> list[str]:
    """Read a SQL file and return a list of non-empty statements split on sep."""
    with open(filepath, "r") as f:
        content = f.read()
    return [s.strip() for s in content.split(sep) if s.strip()]


def make_dir(path: str | Path, quiet: bool = False) -> None:
    """Create directory (and parents) if it does not already exist."""
    p = Path(path)
    if not p.exists():
        p.mkdir(parents=True, exist_ok=True)
        logger.info("Created directory: %s", p)
    elif not quiet:
        logger.debug("Directory already exists: %s", p)