"""
===============================================================================
FILE: services/scud_etl/pipeline.py
===============================================================================
"""

import os
import logging
from typing import Optional, Dict, Any, Tuple
import pandas as pd

from core.connection import get_connection
from core.database import load_scud_from_db_by_snapshot
from config import DATA_DIR
from services.data_loader import load_staff_data, load_absent_data

logger = logging.getLogger("SCUD_PIPELINE")


def load_best_snapshot_for_date(date_str: str, prefer_final_y: bool = False) -> Optional[pd.DataFrame]:
    """
    Загружает наилучший срез СКУД за дату.
    Если prefer_final_y=True — отдает предпочтение финишному Y (23:59:59).
    """
    with get_connection() as conn:
        cursor = conn.cursor()
        target_snap_id = None

        if prefer_final_y:
            cursor.execute("""
                SELECT snapshot_id 
                FROM scud_logs 
                WHERE log_date = ? 
                  AND (snapshot_id LIKE 'Y%' OR snapshot_time LIKE '%23:59:59' OR snapshot_time LIKE '%22:00:00')
                ORDER BY id DESC LIMIT 1
            """, (date_str,))
            row = cursor.fetchone()
            if row:
                target_snap_id = row[0]

        if not target_snap_id:
            cursor.execute("""
                SELECT snapshot_id 
                FROM scud_logs 
                WHERE log_date = ? 
                ORDER BY id DESC LIMIT 1
            """, (date_str,))
            row = cursor.fetchone()
            if row:
                target_snap_id = row[0]

    if not target_snap_id:
        return None

    return load_scud_from_db_by_snapshot(date_str, snapshot_param=target_snap_id)


def load_1c_files_for_date(date_str: str) -> Tuple[Optional[pd.DataFrame], Optional[pd.DataFrame]]:
    """
    Загружает реестры штата и отсутствий 1С на указанную дату через data_loader.
    """
    df_staff = load_staff_data(date_str)
    df_abs = load_absent_data(date_str)
    return df_staff, df_abs