import os
import json
import socket
import platform
import numpy as np
import pandas as pd
import streamlit as st
import sys
import requests
from streamlit_autorefresh import st_autorefresh
import errno
import time
import re
from datetime import datetime, timedelta, timezone


st.set_page_config(
    page_title="Баскетбол",
    page_icon="🏀",
    layout="wide",
    initial_sidebar_state="expanded",
    menu_items={"About": "версия 3.2 от 23.10.2025"},
)
REMOVE_PADDING_FROM_SIDES = """
<style>
    .block-container {
        padding-top: 0rem;
        padding-bottom: 0rem;
    }   
</style>
"""

st.markdown(REMOVE_PADDING_FROM_SIDES, unsafe_allow_html=True)
st_autorefresh()


def schedule_daily_restart():
    """Перезапуск get_season_and_schedule каждый день в 00:05 (только не на Windows)."""
    if not sys.platform.startswith("win"):
        while True:
            now = datetime.now()
            next_run = (now + timedelta(days=1)).replace(
                hour=0, minute=5, second=0, microsecond=0
            )
            sleep_time = (next_run - now).total_seconds()
            time.sleep(sleep_time)

            st.cache_data.clear()


# Функции для стилизации
def highlight_max(data):
    # Преобразуем данные к числовому типу, заменяя некорректные значения на NaN
    numeric_data = pd.to_numeric(data, errors="coerce")
    max_value = numeric_data.max() if pd.notna(numeric_data.max()) else None
    return [
        "background-color: green" if pd.notna(v) and v == max_value and v > 0 else ""
        for v in numeric_data
    ]


def color_win(s):
    return [
        (
            "background-color: ForestGreen"
            if v == True
            else "background-color: #FF4B4B" if v == False else None
        )
        for v in s
    ]


def highlight_grey(s):
    return ["background-color: grey"] * len(s) if s.foul == 5 else [""] * len(s)


def highlight_foul(s):
    return [
        (
            "background-color: orange"
            if v == 4
            else "background-color: red" if v == 5 else ""
        )
        for v in s
    ]


def load_json_data(filepath):
    """
    Загружает данные из JSON файла и кэширует их.
    Возвращает None, если файл не удается прочитать.
    """
    try:
        with open(filepath, "r", encoding="utf-8") as file:
            return json.load(file)
    except (json.JSONDecodeError, FileNotFoundError):
        return None


# Функция для обработки данных одной команды
def process_team_data(team_json, columns_to_include):
    team_data = pd.json_normalize(team_json)

    # Оставляем только нужные колонки
    team_data = team_data[:12][columns_to_include]

    # Обработка height и weight
    for column in ["height", "weight"]:
        if column in team_data.columns:
            team_data[column] = team_data[column].apply(
                lambda value: "" if value == 0 else value
            )

    return team_data


def process_player_data(team_json, player_index):
    team_data = pd.json_normalize(team_json)
    player_data = team_data.iloc[player_index]
    season_total = {
        "name": "Season Total",
        "game_count": str(player_data["TGameCount"]),
        "start_count": str(player_data["TStartCount"]),
        "pts": str(player_data["TPoints"]),
        "pt-2": str(player_data["TShots2"]),
        "pt-3": str(player_data["TShots3"]),
        "pt-1": str(player_data["TShots1"]),
        "fg": str(player_data["TShots23"]),
        "ast": str(player_data["TAssist"]),
        "stl": str(player_data["TSteal"]),
        "blk": str(player_data["TBlocks"]),
        "dreb": str(player_data["TDefRebound"]),
        "oreb": str(player_data["TOffRebound"]),
        "reb": str(player_data["TRebound"]),
        # "to": str(player_data["TTurnover"]),
        # "foul": str(player_data["TFoul"]),
        "fouled": str(player_data["TOpponentFoul"]),
        "dunk": str(player_data["TDunk"]),
        "time": str(player_data["TPlayedTime"]),
    }
    season_avg = {
        "name": "Season Average",
        "game_count": "",
        "start_count": "",
        "pts": str(player_data["AvgPoints"]),
        "pt-2": str(player_data["Shot2Percent"]),
        "pt-3": str(player_data["Shot3Percent"]),
        "pt-1": str(player_data["Shot1Percent"]),
        "fg": str(player_data["Shot23Percent"]),
        "ast": str(player_data["AvgAssist"]),
        "stl": str(player_data["AvgSteal"]),
        "blk": str(player_data["AvgBlocks"]),
        "dreb": str(player_data["AvgDefRebound"]),
        "oreb": str(player_data["AvgOffRebound"]),
        "reb": str(player_data["AvgRebound"]),
        # "to": str(player_data["AvgTurnover"]),
        # "foul": str(player_data["AvgFoul"]),
        "fouled": str(player_data["AvgOpponentFoul"]),
        "dunk": str(player_data["AvgDunk"]),
        "time": str(player_data["AvgPlayedTime"]),
    }
    career_total = {
        "name": "Career Total",
        "game_count": str(player_data["CareerTGameCount"]),
        "start_count": str(player_data["CareerTStartCount"]),
        "pts": str(player_data["CareerTPoints"]),
        "pt-2": str(player_data["CareerTShots2"]),
        "pt-3": str(player_data["CareerTShots3"]),
        "pt-1": str(player_data["CareerTShots1"]),
        "fg": str(player_data["CareerTShots23"]),
        "ast": str(player_data["CareerTAssist"]),
        "stl": str(player_data["CareerTSteal"]),
        "blk": str(player_data["CareerTBlocks"]),
        "dreb": str(player_data["CareerTDefRebound"]),
        "oreb": str(player_data["CareerTOffRebound"]),
        "reb": str(player_data["CareerTRebound"]),
        # "to": str(player_data["CareerTTurnover"]),
        # "foul": str(player_data["CareerTFoul"]),
        "fouled": str(player_data["CareerTOpponentFoul"]),
        "dunk": str(player_data["CareerTDunk"]),
        "time": str(player_data["CareerTPlayedTime"]),
    }

    career_avg = {
        "name": "Career Average",
        "game_count": "",
        "start_count": "",
        "pts": str(player_data["AvgCarPoints"]),
        "pt-2": str(player_data["CareerTShot2Percent"]),
        "pt-3": str(player_data["CareerTShot3Percent"]),
        "pt-1": str(player_data["CareerTShot1Percent"]),
        "fg": str(player_data["CareerTShot23Percent"]),
        "ast": str(player_data["AvgCarAssist"]),
        "stl": str(player_data["AvgCarSteal"]),
        "blk": str(player_data["AvgCarBlocks"]),
        "dreb": str(player_data["AvgCarDefRebound"]),
        "oreb": str(player_data["AvgCarOffRebound"]),
        "reb": str(player_data["AvgCarRebound"]),
        # "to": str(player_data["AvgTurnover"]),
        # "foul": str(player_data["AvgFoul"]),
        "fouled": str(player_data["AvgCarOpponentFoul"]),
        "dunk": str(player_data["AvgCarDunk"]),
        "time": str(player_data["AvgCarPlayedTime"]),
    }

    return [season_total, season_avg, career_total, career_avg], player_data


config = {
    "flag": st.column_config.ImageColumn("flag"),
    "roleShort": st.column_config.TextColumn("R", width=27),
    "num": st.column_config.TextColumn("#", width=27),
    "NameGFX": st.column_config.TextColumn(width=170),
    "isOn": st.column_config.TextColumn("🏀", width=27),
    "pts": st.column_config.TextColumn("PTS", width="content", help="⭐ = Career High"),
    "pt-2": st.column_config.TextColumn("2-PT", width="content"),
    "pt-3": st.column_config.TextColumn("3-PT", width="content"),
    "pt-1": st.column_config.TextColumn("FT", width="content"),
    "fg": st.column_config.TextColumn("FG", width="content"),
    "ast": st.column_config.TextColumn("AS", width="content"),
    "stl": st.column_config.TextColumn("ST", width="content"),
    "blk": st.column_config.TextColumn("BL", width="content"),
    "blkVic": st.column_config.TextColumn("BV", width="content"),
    "dreb": st.column_config.TextColumn("DR", width="content"),
    "oreb": st.column_config.TextColumn("OR", width="content"),
    "reb": st.column_config.TextColumn("R", width="content"),
    "to": st.column_config.TextColumn("TO", width="content"),
    "foul": st.column_config.TextColumn("F", width="content"),
    "fouled": st.column_config.TextColumn("Fed", width="content"),
    "plusMinus": st.column_config.TextColumn("+/-", width="content"),
    "dunk": st.column_config.TextColumn("DUNK", width="content"),
    "kpi": st.column_config.TextColumn("KPI", width="content"),
    "time": st.column_config.TextColumn("TIME"),
    "game_count": st.column_config.TextColumn("G", width=27),
    "start_count": st.column_config.TextColumn("S", width=27),
    "q_pts": st.column_config.TextColumn("PTS", width=27),
    "q_ast": st.column_config.TextColumn("AS", width=27),
    "q_stl": st.column_config.TextColumn("ST", width=27),
    "q_blk": st.column_config.TextColumn("BL", width=27),
    "q_reb": st.column_config.TextColumn("R", width=27),
    "q_rnk": st.column_config.TextColumn("KPI", width=27),
    "q_f": st.column_config.TextColumn("F", width=27),
    "q_f_on": st.column_config.TextColumn("Fed", width=27),
    "q_to": st.column_config.TextColumn("TO", width=27),
    "q_time": st.column_config.TextColumn("TIME"),
    "q_pt2": st.column_config.TextColumn("2-PT", width=45),
    "q_pt3": st.column_config.TextColumn("3-PT", width=45),
    "q_pt23": st.column_config.TextColumn("FG", width=45),
    "q_ft": st.column_config.TextColumn("FT", width=45),
}
config_season = {
    "flag": st.column_config.ImageColumn("flag"),
    "roleShort": st.column_config.TextColumn("R", width=27),
    "num": st.column_config.TextColumn("#", width=27),
    "NameGFX": st.column_config.TextColumn(width=170),
    "isOn": st.column_config.TextColumn("🏀", width=27),
    "pts": st.column_config.TextColumn("PTS", width=40),
    "pt-2": st.column_config.TextColumn("2-PT", width=60),
    "pt-3": st.column_config.TextColumn("3-PT", width=60),
    "pt-1": st.column_config.TextColumn("FT", width=60),
    "fg": st.column_config.TextColumn("FG", width=60),
    "ast": st.column_config.TextColumn("AS", width=40),
    "stl": st.column_config.TextColumn("ST", width=40),
    "blk": st.column_config.TextColumn("BL", width=40),
    "blkVic": st.column_config.TextColumn("BV", width=40),
    "dreb": st.column_config.TextColumn("DR", width=40),
    "oreb": st.column_config.TextColumn("OR", width=40),
    "reb": st.column_config.TextColumn("R", width=40),
    "to": st.column_config.TextColumn("TO", width=40),
    "foul": st.column_config.TextColumn("F", width=40),
    "fouled": st.column_config.TextColumn("Fed", width=40),
    "plusMinus": st.column_config.TextColumn("+/-", width=40),
    "dunk": st.column_config.TextColumn("DUNK", width=40),
    "kpi": st.column_config.TextColumn("KPI", width=40),
    "time": st.column_config.TextColumn("TIME"),
    "game_count": st.column_config.TextColumn("G", width=40),
    "start_count": st.column_config.TextColumn("S", width=40),
}

if "player1" not in st.session_state:
    st.session_state.player1 = None
if "player2" not in st.session_state:
    st.session_state.player2 = None


FOLDER_JSON = "static"


def get_ip_address():
    try:
        # Попытка получить IP-адрес с использованием внешнего сервиса
        # Может потребоваться подключение к интернету
        s = socket.socket(socket.AF_INET, socket.SOCK_DGRAM)
        s.connect(("8.8.8.8", 80))
        ip_address = s.getsockname()[0]
    except socket.error:
        # Если не удалось получить IP-адрес через внешний сервис,
        # используем метод для локального получения IP
        ip_address = socket.gethostbyname(socket.gethostname())
    return ip_address


def _is_meaningful_payload(obj) -> bool:
    # Нулевые/битые/пустые — не считаем осмысленными
    if obj is None:
        return False
    # Пустая коллекция — не осмысленно
    if isinstance(obj, (list, dict, set, tuple)):
        return len(obj) > 0
    # Любой другой тип (str/числа) считаем осмысленным только если не пустая строка
    if isinstance(obj, str):
        return obj.strip() != ""
    return True


def load_data_from_json(filepath):
    """
    Читает JSON и обновляет session_state ТОЛЬКО если данные осмысленные.
    Иначе оставляет предыдущий кэш нетронутым (stale-if-error/stale-if-empty).
    """
    directory = FOLDER_JSON
    os.makedirs(directory, exist_ok=True)
    filepath_full = os.path.join(directory, f"{filepath}.json")
    print(filepath_full)
    print(filepath)
    # вычисление ключа
    # ip = get_ip_address()
    # host = ip_check.get(ip, {}).get("host") or ""
    host = _ipcheck()
    base = os.path.basename(filepath_full).replace(".json", "")
    key = base.replace(f"{host}", "", 1)
    new_payload = load_json_data(filepath_full)  # может быть None/битое/пустое
    # if "team1_" in filepath_full:
    #     print(filepath_full)
    #     print(new_payload)

    # Если новый payload осмысленный — обновляем кэш и "last_good_*"
    if _is_meaningful_payload(new_payload):
        st.session_state[key] = new_payload
        st.session_state.setdefault("_last_good", {})
        st.session_state["_last_good"][key] = new_payload
        return

    # Иначе: если раньше уже был хороший — восстанавливаем его и НЕ трогаем
    if "_last_good" in st.session_state and key in st.session_state["_last_good"]:
        # ничего не делаем — оставляем текущее значение как есть
        return

    # Иначе (нет ни нового, ни прошлого хорошего) — вообще не создаём ключ
    # чтобы ниже по коду .get(...) вернул None и UI ничего не нарисовал
    # (при желании можно явно "очищать": st.session_state.pop(key, None))


def _ipcheck() -> str:
    """Возвращает префикс для имени файла по IP.
    Если IP локальный или host не найден — возвращает пустую строку.
    """
    try:
        ip_str = get_ip_address()
    except Exception:
        ip_str = None

    ip_map = globals().get("ip_check") or {}
    if ip_str and isinstance(ip_map, dict):
        host = (ip_map.get(ip_str) or {}).get("host")
        if host:
            return f"{host}_"
    return ""


def read_match_id_json(path="match_id.json", attempts=10, delay=0.2):
    """Надёжное чтение match_id.json с ретраями при EBUSY/битом JSON."""
    d = delay
    for i in range(attempts):
        try:
            if not os.path.isfile(path):
                return {}
            with open(path, "r", encoding="utf-8") as f:
                return json.load(f)
        except json.JSONDecodeError:
            # файл переписывают — подождём и попробуем снова
            time.sleep(d)
            d = min(d * 1.6, 2.0)
        except OSError as e:
            # EBUSY (errno=16) — подождём и ещё раз
            if getattr(e, "errno", None) == errno.EBUSY:
                time.sleep(d)
                d = min(d * 1.6, 2.0)
                continue
            # иные ошибки — пробрасываем дальше
            raise
    print("Не удалось прочитать match_id.json после нескольких попыток; возвращаю {}")
    return {}


def rewrite_file(filename: str, data: dict, directory: str = "JSON") -> None:
    """
    Перезаписывает JSON-файл с заданными данными.
    Если запуск локальный (локальный IP), префикс host в имени файла не используется.
    Если IP не локальный и есть словарь ip_check с хостами — добавим префикс host_.
    """
    # Если глобальная константа задана — используем её
    try:
        directory = FOLDER_JSON  # type: ignore[name-defined]
    except NameError:
        # иначе используем аргумент по умолчанию/переданный
        pass

    os.makedirs(directory, exist_ok=True)

    filepath = os.path.join(directory, f"{filename}.json")
    # print(filepath)  # оставил как у тебя; можно заменить на logger.debug при желании

    try:
        with open(filepath, "w", encoding="utf-8") as f:
            json.dump(data, f, ensure_ascii=False, indent=2)
    except Exception as e:
        print(f"Ошибка при записи файла {filepath}: {e}")


ip_check = read_match_id_json("match_id.json") or {}


prefix = _ipcheck()

load_data_from_json("game")
cached_game_online = st.session_state.get("game")

load_data_from_json("team1")
cached_team1 = st.session_state.get("team1")

load_data_from_json("team2")
cached_team2 = st.session_state.get("team2")

load_data_from_json("referee")
cached_referee = st.session_state.get("referee")

# standings — может не быть тега/файла
league_tag = None
if isinstance(cached_game_online, dict):
    league_tag = (cached_game_online.get("league") or {}).get("tag")
    comp_name = (cached_game_online.get("comp") or {}).get("name").replace(" ", "_")
if league_tag:
    load_data_from_json(f"standings_{league_tag}_{comp_name}")
cached_standings = (
    st.session_state.get(f"standings_{league_tag}_{comp_name}") if league_tag else None
)

load_data_from_json("scores_quarter")
cached_scores_quarter = st.session_state.get("scores_quarter")

load_data_from_json("play_by_play")
cached_play_by_play = st.session_state.get("play_by_play")

load_data_from_json("team_stats")
cached_team_stats = st.session_state.get("team_stats")

load_data_from_json("scores")
cached_scores = st.session_state.get("scores") or []  # важно!

load_data_from_json("api_live-status")
cached_live_status = st.session_state.get("api_live-status")

load_data_from_json("schedule")
cached_schedule = st.session_state.get("schedule")

load_data_from_json("team_comparison")
cached_team_comparison = st.session_state.get("team_comparison")


def _is_empty_like(x) -> bool:
    if x is None:
        return True
    if isinstance(x, pd.DataFrame):
        return x.empty
    try:
        from pandas.io.formats.style import Styler

        if isinstance(x, Styler):
            return getattr(x, "data", pd.DataFrame()).empty
    except Exception:
        pass
    if isinstance(x, (list, tuple, dict, set)):
        return len(x) == 0
    return False


def safe_show(func, *args, **kwargs):
    for a in args:
        if _is_empty_like(a):
            return None
    for k, v in kwargs.items():
        if k.lower() in (
            "height",
            "width",
            "use_container_width",
            "unsafe_allow_html",
            "on_select",
            "selection_mode",
            "column_config",
            "hide_index",
            "border",
            "delta_color",
            "key",
        ):
            continue
        if _is_empty_like(v):
            return None
    if "height" in kwargs:
        h = kwargs.get("height")
        if h is None:
            kwargs.pop("height")
        else:
            try:
                h = int(h)
                if h < 0:
                    kwargs.pop("height")
                else:
                    kwargs["height"] = h
            except Exception:
                kwargs.pop("height")
    try:
        return func(*args, **kwargs)
    except Exception as e:
        st.warning(f"⚠️ Ошибка при отображении: {e}")
        return None


# ======== /SAFE WRAPPER ========
def ensure_state(key: str, default=None):
    # Инициализирует ключ один раз и возвращает значение
    return st.session_state.setdefault(key, default)


period_max = 0
if isinstance(cached_play_by_play, list) and isinstance(cached_game_online, dict):
    plays = (cached_game_online.get("result") or {}).get("plays") or []
    if plays:
        df_data_pbp = pd.DataFrame(cached_play_by_play)
        if not df_data_pbp.empty and "period" in df_data_pbp.columns:
            period_max = int(df_data_pbp.iloc[0]["period"])
            count_quarter = [
                f"Четверть {i}" if i < 5 else f"Овертайм {i-4}"
                for i in range(1, period_max + 1)
            ]

            for i in range(1, period_max + 1):
                key_team1 = f"team1_{i}"
                key_team2 = f"team2_{i}"
                load_data_from_json(key_team1)
                load_data_from_json(key_team2)
                _q1 = st.session_state.get(key_team1)  # может быть None — это ок
                _q2 = st.session_state.get(key_team2)


timeout1 = []
timeout2 = []

if isinstance(cached_game_online, dict):
    result = cached_game_online
    plays = result.get("plays") or []

    timeout1, timeout2 = [], []
    for event in plays:
        if isinstance(event, dict) and event.get("play") == 23:
            if event.get("startNum") == 1:
                timeout1.append(event)
            elif event.get("startNum") == 2:
                timeout2.append(event)

    col1, col4, col2, col5, col3 = st.columns([1, 5, 3, 5, 1])

    t1 = result.get("team1") or {}
    t2 = result.get("team2") or {}

    if t1.get("logo"):
        col1.image(t1["logo"], width=100)
    team1_name = t1.get("name") or ""
    team2_name = t2.get("name") or ""
    if team1_name or team2_name:
        col2.markdown(
            f"<h2 style='text-align: center'>{team1_name} — {team2_name}</h2>",
            unsafe_allow_html=True,
        )
        col2.markdown(
            f"<h3 style='text-align: center'>{result['game']['localDate']} {result['game']['defaultZoneTime']}</h3>",
            unsafe_allow_html=True,
        )
    if t2.get("logo"):
        col3.image(t2["logo"], width=100)

    col4_1, col4_2, col4_3 = col4.columns((1, 1, 1))
    col5_1, col5_2, col5_3 = col5.columns((1, 1, 1))

    # Points метрики безопасно
    val1 = val2 = None
    if isinstance(cached_team_stats, list) and len(cached_team_stats) > 0:
        v1 = cached_team_stats[0].get("val1")
        v2 = cached_team_stats[0].get("val2")
        if v1 is not None and v2 is not None:
            val1, val2 = int(v1), int(v2)
            delta_color_1 = "off" if val1 == val2 else "normal"
            col4_1.metric("Points", v1, val1 - val2, delta_color_1)
            col5_3.metric("Points", v2, val2 - val1, delta_color_1)

    col4_3.metric("TimeOuts", len(timeout1))
    col5_1.metric("TimeOuts", len(timeout2))

    if isinstance([cached_live_status], list) and cached_live_status:
        foulsA = (cached_live_status["result"] or {}).get("foulsA")
        foulsB = (cached_live_status["result"] or {}).get("foulsB")
        if foulsA is not None:
            col4_2.metric("Fouls", foulsA)
        if foulsB is not None:
            col5_2.metric("Fouls", foulsB)


if isinstance(cached_game_online, dict) and (
    cached_game_online.get("plays") or []
):
    col_1_col = [f"col_1_{i}" for i in range(1, period_max + 1)]
    col_2_col = [f"col_2_{i}" for i in range(1, period_max + 1)]
    count_q = 0

    score_by_quarter_1 = [
        x.get("score1")
        for x in cached_scores
        if isinstance(x, dict) and x.get("score1") not in ("", None)
    ]
    score_by_quarter_2 = [
        x.get("score2")
        for x in cached_scores
        if isinstance(x, dict) and x.get("score2") not in ("", None)
    ]

    if score_by_quarter_1:
        col_1_col = col4.columns([1 for _ in range(len(score_by_quarter_1))])
        col_2_col = col5.columns([1 for _ in range(len(score_by_quarter_2))])

    for q1, q2, col1_i, col2_i in zip(
        score_by_quarter_1, score_by_quarter_2, col_1_col, col_2_col
    ):
        count_q += 1
        name_q = f"OT{count_q-4}" if count_q > 4 else f"Q{count_q}"
        try:
            delta_color = "off" if int(q1) == int(q2) else "normal"
            col1_i.metric(name_q, q1, int(q1) - int(q2), delta_color, border=True)
            col2_i.metric(name_q, q2, int(q2) - int(q1), delta_color, border=True)
        except (ValueError, TypeError):
            # если кривые данные в JSON, просто пропустим
            pass


(
    tab_temp_1,
    tab_temp_2,
    tab_temp_3,
    tab_temp_4,
    tab_temp_5,
    tab_temp_6,
    tab_pbp,
    tab_temp_7,
    tab_temp_8,
    tab_schedule,
    tab_online,
) = st.tabs(
    [
        "Игроки",
        "Команды",
        "Судьи",
        "Турнирная таблица",
        "Статистика четвертей",
        "Ход игры",
        "События игры",
        "Статистика по четвертям",
        "Milestones",
        "Прошедшие/будущие матчи",
        "Сегодня",
    ]
)


def check_milestone(value, milestone_type, name, num, where):
    milestone_checks = {
        "PTS": "point",
        "AST": "assist",
        "BLK": "block",
        "REB": "rebound",
        "DREB": "defensive rebound",
        "OREB": "offensive rebound",
        "STL": "steal",
        "GAMES": "game",
    }
    if milestone_type == "GAMES":
        list_data = [*range(50, 5100, 50)]
        if int(value) % 100 in [49, 99]:
            diff = [l - int(value) for l in list_data]
            positive_numbers = [num for num in diff if num > -1]
            count = 0
            for i in diff:
                count += 1
                if i == min(positive_numbers):
                    break
            full_word = [
                word
                for w, word in milestone_checks.items()
                if w.lower() == milestone_type.lower()
            ][0]
            # print(positive_numbers)
            if min(positive_numbers) != 0:
                word = full_word if min(positive_numbers) == 1 else f"{full_word}s"
                if where == "season":
                    where = "in this season"
                elif where == "league":
                    where = "in career VTB"
                string_value = f"{name} needs {min(positive_numbers)} {word} to reach {list_data[count-1]} {full_word}s {where}"
            else:
                string_value = ""
            return {
                "NameGFX": f"{name} ({num})",
                "type": milestone_type.upper(),
                "value": value,
                "string_value": string_value,
            }
    else:
        list_data = [*range(100, 5100, 100)]
        if (int(value) % 100) >= 90 or (int(value) % 1000) in list_data:
            diff = [l - int(value) for l in list_data]
            positive_numbers = [num for num in diff if num > -1]
            count = 0
            for i in diff:
                count += 1
                if i == min(positive_numbers):
                    break
            # print(positive_numbers)
            full_word = [
                word
                for w, word in milestone_checks.items()
                if w.lower() == milestone_type.lower()
            ][0]
            # print(positive_numbers)
            if min(positive_numbers) != 0:
                word = full_word if min(positive_numbers) == 1 else f"{full_word}s"
                if where == "season":
                    where = "in this season"
                elif where == "league":
                    where = "in career VTB"
                string_value = f"{name} needs {min(positive_numbers)} {word} to reach {list_data[count-1]} {full_word}s {where}"
            else:
                string_value = ""
            return {
                "NameGFX": f"{name} ({num})",
                "type": milestone_type.upper(),
                "value": value,
                "string_value": string_value,
            }
    return None


def milestones(data):
    new_data_season = []
    new_data_career = []
    for d in data:
        if d["startRole"] == "Player":
            milestone_checks = {
                "PTS": d["TPoints"],
                "AST": d["TAssist"],
                "BLK": d["TBlocks"],
                "REB": d["TRebound"],
                "DREB": d["TDefRebound"],
                "OREB": d["TOffRebound"],
                "STL": d["TSteal"],
                "GAMES": d["TGameCount"],
            }
            milestone_career_checks = {
                "PTS": d["CareerTPoints"],
                "AST": d["CareerTAssist"],
                "BLK": d["CareerTBlocks"],
                "REB": d["CareerTRebound"],
                "DREB": d["CareerTDefRebound"],
                "OREB": d["CareerTOffRebound"],
                "STL": d["CareerTSteal"],
                "GAMES": d["CareerTGameCount"],
            }
            for milestone_type, value in milestone_checks.items():
                milestone_data = check_milestone(
                    value, milestone_type, d["NameGFX"], d["num"], "season"
                )
                if milestone_data:
                    new_data_season.append(milestone_data)
            for milestone_type_car, value_car in milestone_career_checks.items():
                milestone_data_car = check_milestone(
                    value_car, milestone_type_car, d["NameGFX"], d["num"], "league"
                )
                if milestone_data_car:
                    new_data_career.append(milestone_data_car)
    return new_data_season, new_data_career


columns_game = [
    "num",
    # "roleShort",
    "NameGFX",
    "isOn",
    # "flag",
    "pts",
    "pt-2",
    "pt-3",
    "pt-1",
    "fg",
    "ast",
    "stl",
    "blk",
    # "blkVic",
    "dreb",
    "oreb",
    "reb",
    "to",
    "foul",
    # "fouled",
    "plusMinus",
    "dunk",
    "kpi",
    "time",
]
if cached_team1 and cached_team2:
    team1_data = process_team_data(cached_team1, columns_game)
    team2_data = process_team_data(cached_team2, columns_game)

    # Добавляем звездочку, если pts > PTS_Career_High
    def _get_first_number(x):
        """Безопасно вытащить число из строки/значения (например '12 (60%)' -> 12)."""
        try:
            if x is None:
                return None
            s = str(x)
            # заберём ведущие число/знак (поддержим +/-)
            import re

            m = re.search(r"[-+]?\d+(\.\d+)?", s)
            return float(m.group(0)) if m else None
        except Exception:
            return None

    CAREER_HIGH_KEYS = {
        "pts": ["PTS_Career_High", "CareerHighPoints", "career_high_pts"],
        "ast": ["AST_Career_High", "CareerHighAssist", "career_high_ast"],
        "stl": ["STL_Career_High", "CareerHighSteal", "career_high_stl"],
        "blk": ["BLK_Career_High", "CareerHighBlocks", "career_high_blk"],
        "reb": ["REB_Career_High", "CareerHighRebound", "career_high_reb"],
        # если нужно — добавь ещё пары "df_column": ["possible_key1","possible_key2"...]
    }

    def _build_career_high_map(cached_team_list):
        """Вернёт словарь: player_id -> {stat_key: value} для всех доступных максимумов."""
        out = {}
        if not isinstance(cached_team_list, list):
            return out
        for p in cached_team_list:
            if not isinstance(p, dict):
                continue
            pid = p.get("id")
            if pid is None:
                continue
            out[pid] = {}
            for stat_col, aliases in CAREER_HIGH_KEYS.items():
                for k in aliases:
                    if k in p and p[k] not in (None, ""):
                        out[pid][stat_col] = _get_first_number(p[k])
                        break
        return out

    def _ensure_id_column(df, cached_team_list):
        """Присвоить игрокам id в том же порядке, что и в списке cached_team."""
        try:
            ids = [
                p.get("id") if isinstance(p, dict) else None for p in cached_team_list
            ][: len(df)]
            if "id" not in df.columns:
                df["id"] = ids
            else:
                # не затираем, только заполняем пустые
                df["id"] = df["id"].fillna(pd.Series(ids, index=df.index))
        except Exception:
            pass

    def _mark_star_for_columns(df, cached_team_list, columns):
        """
        Для каждого col в columns: если текущее значение > career high — добавляем ' ⭐️'.
        Преобразуем колонки в текст (для отображения эмодзи).
        """
        _ensure_id_column(df, cached_team_list)
        ch_map = _build_career_high_map(cached_team_list)

        def format_with_star(val, career_max):
            v = _get_first_number(val)
            cm = _get_first_number(career_max)
            if v is not None and cm is not None and v >= cm and val > 0:
                # сохраняем исходное текстовое представление + ⭐️
                return f"{val} ⭐️"
            return f"{val}" if val is not None else ""

        for col in columns:
            if col not in df.columns:
                continue
            new_vals = []
            for idx, row in df.iterrows():
                pid = row.get("id")
                career_max = (ch_map.get(pid, {}) or {}).get(col)
                new_vals.append(format_with_star(row[col], career_max))
            df[col] = new_vals  # теперь это текст для отображения

    STAR_COLUMNS = [
        "pts",
        "ast",
        "stl",
        "blk",
        "reb",
    ]
    team1_data["_pts_num"] = pd.to_numeric(team1_data["pts"], errors="coerce")
    team1_data["_kpi_num"] = pd.to_numeric(team1_data["kpi"], errors="coerce")
    team2_data["_pts_num"] = pd.to_numeric(team2_data["pts"], errors="coerce")
    team2_data["_kpi_num"] = pd.to_numeric(team2_data["kpi"], errors="coerce")

    def highlight_max_by_refcol(df, view_col, ref_col):
        ref = pd.to_numeric(df[ref_col], errors="coerce")
        mx = ref.max()
        return [
            ("background-color: green" if (pd.notna(v) and v == mx and v > 0) else "")
            for v in ref
        ]

    _mark_star_for_columns(team1_data, cached_team1, STAR_COLUMNS)
    _mark_star_for_columns(team2_data, cached_team2, STAR_COLUMNS)

    # Стилизация данных
    # team1_styled = (
    #     team1_data.style.apply(highlight_grey, axis=1)
    #     .apply(highlight_foul, subset="foul")
    #     .apply(highlight_max, subset="pts")
    #     .apply(highlight_max, subset="kpi")
    # )
    # team2_styled = (
    #     team2_data.style.apply(highlight_grey, axis=1)
    #     .apply(highlight_foul, subset="foul")
    #     .apply(highlight_max, subset="pts")
    #     .apply(highlight_max, subset="kpi")
    # )
    team1_styled = (
        team1_data[columns_game]
        .style.apply(highlight_grey, axis=1)
        .apply(highlight_foul, subset="foul")
        .apply(
            lambda _: highlight_max_by_refcol(team1_data, "pts", "_pts_num"),
            axis=0,
            subset=["pts"],
        )
        .apply(
            lambda _: highlight_max_by_refcol(team1_data, "kpi", "_kpi_num"),
            axis=0,
            subset=["kpi"],
        )
    )
    team2_styled = (
        team2_data[columns_game]
        .style.apply(highlight_grey, axis=1)
        .apply(highlight_foul, subset="foul")
        .apply(
            lambda _: highlight_max_by_refcol(team2_data, "pts", "_pts_num"),
            axis=0,
            subset=["pts"],
        )
        .apply(
            lambda _: highlight_max_by_refcol(team2_data, "kpi", "_kpi_num"),
            axis=0,
            subset=["kpi"],
        )
    )

    def get_player_all_game(player_data_1):

        try:
            directory = FOLDER_JSON  # type: ignore[name-defined]
        except NameError:
            # иначе используем аргумент по умолчанию/переданный
            pass

        os.makedirs(directory, exist_ok=True)

        host_prefix = _ipcheck()
        filename = player_data_1["id"]
        filepath = os.path.join(directory, f"{host_prefix}{filename}.json")
        # print(filepath)  # оставил как у тебя; можно заменить на logger.debug при желании

        try:
            with open(filepath, "r", encoding="utf-8") as f:
                players_game = json.load(f)
        except Exception:
            pass
        df_player_game = pd.json_normalize(players_game)
        # Отфильтровать строки, где class == "Normal"
        # print(player_data_1)
        df_filtered = df_player_game[df_player_game["class"] == "Normal"].copy()

        # Преобразуем game.gameDate в datetime
        df_filtered["game.gameDate"] = pd.to_datetime(
            df_filtered["game.gameDate"], errors="coerce", dayfirst=True
        )

        # Удалим строки без корректной даты (если такие есть)
        # df_filtered = df_filtered.dropna(subset=["game.gameDate"])

        # Сортировка от последнего матча к первому
        df_filtered = df_filtered.sort_values(by="game.gameDate", ascending=False)
        # Указать нужные колонки для вывода
        columns_to_show = [
            "Сезон",
            "Дата",
            "Команда 1",
            "Команда 2",
            "Счёт",
            "PTS",
            "2-PTS%",
            "3-PTS%",
            "FG%",
            "FT%",
            "AST",
            "STL",
            "BLK",
            "DR",
            "OR",
            "REB",
            "TO",
            "F",
            "TIME",
            "+/-",
        ]
        numeric_cols = [
            "stats.points",
            "stats.assist",
            "stats.steal",
            "stats.blockShot",
            "stats.rebound",
        ]
        # df_filtered[numeric_cols] = df_filtered[numeric_cols].apply(
        #     pd.to_numeric, errors="coerce"
        # )

        # 🟢 Переименовываем колонки для отображения в Streamlit
        rename_map = {
            "season": "Сезон",
            "game.gameDate": "Дата",
            "game.team1Name": "Команда 1",
            "game.team2Name": "Команда 2",
            "game.score": "Счёт",
            "stats.points": "PTS",
            "stats.shot2Percent": "2-PTS%",
            "stats.shot3Percent": "3-PTS%",
            "stats.shot23Percent": "FG%",
            "stats.shot1Percent": "FT%",
            "stats.assist": "AST",
            "stats.steal": "STL",
            "stats.blockShot": "BLK",
            "stats.defRebound": "DR",
            "stats.offRebound": "OR",
            "stats.rebound": "REB",
            "stats.turnover": "TO",
            "stats.foul": "F",
            "stats.playedTime": "TIME",
            "stats.plusMinus": "+/-",
        }
        df_filtered[numeric_cols] = df_filtered[numeric_cols].apply(
            pd.to_numeric, errors="coerce"
        )
        df_filtered[numeric_cols] = df_filtered[numeric_cols].round(0).astype("Int64")
        df_filtered = df_filtered.rename(columns=rename_map)
        df_filtered["Дата"] = df_filtered["Дата"].dt.strftime("%d.%m.%Y")
        styled = (
            df_filtered[columns_to_show]
            .style.apply(highlight_max, subset=["PTS"])
            .apply(highlight_max, subset=["AST"])
            .apply(highlight_max, subset=["STL"])
            .apply(highlight_max, subset=["BLK"])
            .apply(highlight_max, subset=["REB"])
            .format(
                {
                    "Дата": lambda x: x,  # уже строка, просто оставляем как есть
                    "PTS": "{:,.0f}".format,
                    "AST": "{:,.0f}".format,
                    "STL": "{:,.0f}".format,
                    "BLK": "{:,.0f}".format,
                    "REB": "{:,.0f}".format,
                }
            )
        )
        return styled

    # Вывод данных
    col_player1, col_player2 = tab_temp_1.columns((5, 5))
    config_copy = config.copy()

    event1 = col_player1.dataframe(
        team1_styled,
        column_config=config_copy,
        hide_index=True,
        height=460,
        on_select="rerun",
        selection_mode=[
            "single-row",
        ],
    )
    event2 = col_player2.dataframe(
        team2_styled,
        column_config=config,
        hide_index=True,
        height=460,
        on_select="rerun",
        selection_mode=[
            "single-row",
        ],
    )

    if event1.selection and event1.selection.get("rows"):
        selected_index1 = event1.selection["rows"][0]
        st.session_state["player1"] = (
            selected_index1  # Сохранение состояния в session_state
        )
    else:
        st.session_state["player1"] = None
    if st.session_state["player1"] is not None:
        selected_player_1, player_data_1 = process_player_data(
            cached_team1, st.session_state["player1"]
        )
        if player_data_1["num"]:
            z, a, q, b, c, d, e = col_player1.columns((1, 6, 1, 1, 1, 1, 1))
            z.metric("Номер", player_data_1["num"], border=False)
            a.metric("Игрок", player_data_1["NameGFX"], border=False)
            q.image(player_data_1["flag"])
            b.metric("Амплуа", player_data_1["roleShort"], border=False)
            c.metric("Возраст", player_data_1["age"], border=False)
            d.metric("Рост", player_data_1["height"].split()[0], border=False)
            e.metric("Вес", player_data_1["weight"].split()[0], border=False)

            col_player1.dataframe(
                selected_player_1,
                column_config=config_season,
                hide_index=True,
            )

            col_player1.title("Статистика каждой игры")
            col_player1.dataframe(get_player_all_game(player_data_1), hide_index=True)

    if event2.selection and event2.selection.get("rows"):
        selected_index2 = event2.selection["rows"][0]
        st.session_state["player2"] = (
            selected_index2  # Сохранение состояния в session_state
        )
    else:
        st.session_state["player2"] = None
    if st.session_state["player2"] is not None:
        selected_player_2, player_data_2 = process_player_data(
            cached_team2, st.session_state["player2"]
        )
        if player_data_2["num"]:
            z, a, q, b, c, d, e = col_player2.columns((1, 6, 1, 1, 1, 1, 1))
            z.metric("Номер", player_data_2["num"], border=False)
            a.metric("Игрок", player_data_2["NameGFX"], border=False)
            q.image(player_data_2["flag"])
            b.metric("Амплуа", player_data_2["roleShort"], border=False)
            c.metric("Возраст", player_data_2["age"], border=False)
            d.metric("Рост", player_data_2["height"].split()[0], border=False)
            e.metric("Вес", player_data_2["weight"].split()[0], border=False)

            col_player2.dataframe(
                selected_player_2,
                column_config=config_season,
                hide_index=True,
            )
            col_player2.title("Статистика каждой игры")
            col_player2.dataframe(get_player_all_game(player_data_2), hide_index=True)

team_col1, team_col2 = tab_temp_2.columns((5, 5))
if isinstance(cached_team_stats, list) and len(cached_team_stats) >= 34:
    cached_team_stats_new = [
        cached_team_stats[0],
        *cached_team_stats[25:29],
        cached_team_stats[7],
        cached_team_stats[33],
        *cached_team_stats[9:11],
        *cached_team_stats[15:17],
    ]
    team_col1.title("Статистика за матч")
    team_col1.dataframe(cached_team_stats_new, height=500)

if isinstance(cached_team_comparison, list):
    team_col2.title("Статистика за сезон")

    # задаём желаемый порядок ключей
    order = [
        "points",
        "points_1",
        "points_2",
        "points_3",
        "points_23",
        "assists",
        "rebounds",
        "steals",
        "blocks",
        "turnovers",
        "fouls",
        "games",
        # "team",
    ]

    # формируем результат с учётом порядка
    result = [
        {
            "name": key,
            "val1": cached_team_comparison[0][key],
            "val2": cached_team_comparison[1][key],
        }
        for key in order
        if key in cached_team_comparison[0]
    ]
    team_col2.dataframe(result, width="stretch", height=500)

if isinstance(cached_referee, (list, pd.DataFrame)):
    tab_temp_3.dataframe(
        cached_referee,
        height=600,
        width="content",
        column_config={"flag": st.column_config.ImageColumn("flag")},
    )


column_config_ref = {
    "flag": st.column_config.ImageColumn(
        "flag",
    ),
}


def highlight_teams(s):
    try:
        if s.iloc[0] in (
            cached_game_online["result"]["team1"]["teamId"],
            cached_game_online["result"]["team2"]["teamId"],
        ):
            return ["background-color: #FF4B4B"] * len(s)
        else:
            return [""] * len(s)
    except NameError:
        return [""] * len(s)


if cached_standings:
    df_st = pd.json_normalize(cached_standings)

    def highlight_teams(s):
        try:
            t1 = (cached_game_online or {}).get("team1", {}).get("teamId")
            t2 = (cached_game_online or {}).get("team2", {}).get("teamId")
            if s.iloc[0] in (t1, t2):
                return ["background-color: #FF4B4B"] * len(s)
            if s.iloc[0] in (t1, t2):
                return ["background-color: #FF4B4B"] * len(s)
        except Exception:
            pass
        return [""] * len(s)

    styled = df_st[
        [
            "teamId",
            "start",
            "place",
            "name",
            "regionName",
            "totalGames",
            "totalWin",
            "totalDefeat",
            "totalPoints",
            "totalGoalPlus",
            "totalGoalMinus",
            "logo",
            "w_l",
            "procent",
            "plus_minus",
        ]
    ].style.apply(highlight_teams, axis=1)
    tab_temp_4.dataframe(
        styled,
        column_config={"logo": st.column_config.ImageColumn("logo")},
        hide_index=True,
        height=610,
        width="content",
    )


if isinstance(cached_scores_quarter, list) and len(cached_scores_quarter) >= 2:
    column_config = {}
    for quarter in ["Q1", "Q2", "Q3", "Q4", "OT1", "OT2", "OT3", "OT4"]:
        column_name = f"score_avg{quarter}"
        column_config[column_name] = st.column_config.NumberColumn(
            column_name, format="%.1f"
        )

    columns_quarters_name = ["Q1", "Q2", "Q3", "Q4"]
    columns_quarters_name_ot = ["OT1", "OT2", "OT3", "OT4"]
    columns_quarters = tab_temp_5.columns((1, 1, 1, 1))

    # Основные четверти
    for index, col in enumerate(columns_quarters):
        q = columns_quarters_name[index]
        df_col = [
            {
                "team": cached_scores_quarter[0].get("team"),
                "W": cached_scores_quarter[0].get(f"win{q}"),
                "L": cached_scores_quarter[0].get(f"lose{q}"),
                "D": cached_scores_quarter[0].get(f"draw{q}"),
                "PTS": cached_scores_quarter[0].get(f"score{q}"),
                "AVG": cached_scores_quarter[0].get(f"score_avg{q}"),
            },
            {
                "team": cached_scores_quarter[1].get("team"),
                "W": cached_scores_quarter[1].get(f"win{q}"),
                "L": cached_scores_quarter[1].get(f"lose{q}"),
                "D": cached_scores_quarter[1].get(f"draw{q}"),
                "PTS": cached_scores_quarter[1].get(f"score{q}"),
                "AVG": cached_scores_quarter[1].get(f"score_avg{q}"),
            },
        ]
        col.write(q)
        col.dataframe(df_col)

    # Овертаймы
    for index, col in enumerate(columns_quarters):
        q = columns_quarters_name_ot[index]
        df_col = [
            {
                "team": cached_scores_quarter[0].get("team"),
                "W": cached_scores_quarter[0].get(f"win{q}"),
                "L": cached_scores_quarter[0].get(f"lose{q}"),
                "D": cached_scores_quarter[0].get(f"draw{q}"),
                "PTS": cached_scores_quarter[0].get(f"score{q}"),
                "AVG": cached_scores_quarter[0].get(f"score_avg{q}"),
            },
            {
                "team": cached_scores_quarter[1].get("team"),
                "W": cached_scores_quarter[1].get(f"win{q}"),
                "L": cached_scores_quarter[1].get(f"lose{q}"),
                "D": cached_scores_quarter[1].get(f"draw{q}"),
                "PTS": cached_scores_quarter[1].get(f"score{q}"),
                "AVG": cached_scores_quarter[1].get(f"score_avg{q}"),
            },
        ]
        col.write(q)
        col.dataframe(df_col)


if isinstance(cached_play_by_play, list) and isinstance(cached_game_online, dict):
    plays = (cached_game_online.get("result") or {}).get("plays") or []
    if plays:
        tab_temp_6.table(cached_play_by_play)


def _ensure_columns(df: pd.DataFrame, cols: list[str]) -> pd.DataFrame:
    """Гарантирует наличие всех столбцов в df, отсутствующие заполняет NaN/0 по типу."""
    for c in cols:
        if c not in df.columns:
            # для числовых метрик по четверти логично ставить 0
            df[c] = 0
    # Переупорядочим столбцы
    return df[cols]


def _safe_dataframe(
    container, df: pd.DataFrame, cols_for_st: list[str], height_per_row: int = 38
):
    df = _ensure_columns(df, cols_for_st)
    rows = len(df)

    # Стилизуем безопасно (если вдруг highlight_max упадёт — показываем без стиля)
    try:
        styled = (
            df.style.apply(highlight_max, subset="q_pts") if not df.empty else df.style
        )
    except Exception:
        styled = df.style

    kwargs = dict(column_config=config, hide_index=True)

    # height только если он действительно нужен
    if rows > 10:
        kwargs["height"] = height_per_row * rows
        # или: kwargs["height"] = "stretch"  # если хотите растягивать блок

    container.dataframe(styled, **kwargs)


# 1) Безопасная проверка наличия плей-данных
has_plays = bool(
    cached_game_online and cached_game_online.get("result", {}).get("plays")
)
if has_plays:

    # 2) Готовим вкладки по количеству периодов
    # Если у вас уже есть count_quarter — используйте его. Иначе берём из period_max.

    # === REPLACE: безопасное определение количества периодов и создание вкладок ===
    def _safe_int(x, default=None):
        try:
            v = int(x)
            return v if v is not None else default
        except Exception:
            return default

    def _max_period_from_plays(cached_game_online) -> int:
        plays = (cached_game_online or {}).get("result", {}).get("plays") or []
        periods = []
        for item in plays:
            try:
                v = int(item.get("period"))
                periods.append(v)
            except Exception:
                # пропускаем пустые/неконвертируемые значения
                continue
        return max(periods) if periods else 0

    count_quarter = _max_period_from_plays(cached_game_online)

    # 3) попытка по ключам team1_/team2_ в session_state (например, team1_1..team1_4)
    if count_quarter is None or count_quarter <= 0:
        import re

        mx = 0
        for k in st.session_state.keys():
            m1 = re.fullmatch(r"team1_(\d+)", k)
            m2 = re.fullmatch(r"team2_(\d+)", k)
            if m1:
                mx = max(mx, int(m1.group(1)))
            if m2:
                mx = max(mx, int(m2.group(1)))
        count_quarter = mx if mx > 0 else None

    # 4) дефолт, если ничего не получилось
    if count_quarter is None or count_quarter <= 0:
        # если совсем нет данных — не создаём вкладки и показываем инфо
        tab_temp_7.info(
            "Нет корректного числа периодов для отображения таблиц по четвертям."
        )
        count_quarter = 0

    # 5) создаём вкладки, только если их хотя бы одна
    if count_quarter > 0:
        tab_labels = [f"{i+1}Q" for i in range(count_quarter)]
        columns_quarter = tab_temp_7.tabs(tab_labels)
    else:
        columns_quarter = []  # чтобы ниже цикл просто не выполнялся

    columns_quarter_for_st = [
        "num",
        "NameGFX",
        "q_pts",
        "q_pt2",
        "q_pt3",
        "q_ft",
        "q_pt23",
        "q_ast",
        "q_stl",
        "q_blk",
        "q_reb",
        "q_to",
        "q_f",
        "q_f_on",
        "q_rnk",
        "q_time",
    ]

    # 3) Рендер таблиц по четвертям
    for i in range(count_quarter):
        # Каждая вкладка -> 2 колонки
        try:
            col_quarter1, col_quarter2 = columns_quarter[i].columns((5, 5))
        except IndexError:
            # На случай несоответствия размеров
            st.warning(f"Недостаточно вкладок для периода #{i+1}. Пропуск.")
            continue

        # --- Team 1 ---
        key_team1 = f"team1_{i+1}"
        try:
            load_data_from_json(key_team1)
        except Exception as e:
            tab_temp_7.warning(f"Не удалось загрузить данные для {key_team1}: {e}")
            continue

        data_team1 = st.session_state.get(key_team1, [])
        # print(data_team1)
        df_team1 = (
            pd.DataFrame(data_team1)
            if isinstance(data_team1, (list, tuple, dict))
            else pd.DataFrame()
        )

        # _safe_dataframe(col_quarter1, df_team1, columns_quarter_for_st)

        # --- Team 2 ---
        key_team2 = f"team2_{i+1}"
        try:
            load_data_from_json(key_team2)
        except Exception as e:
            tab_temp_7.warning(f"Не удалось загрузить данные для {key_team2}: {e}")
            continue

        data_team2 = st.session_state.get(key_team2, [])
        df_team2 = (
            pd.DataFrame(data_team2)
            if isinstance(data_team2, (list, tuple, dict))
            else pd.DataFrame()
        )
        # _safe_dataframe(col_quarter2, df_team2, columns_quarter_for_st)
    tab_temp_7.warning("В процессе разработки!!!!!!")

    # 4) Блок с milestones по командам (если есть кэш)
    if cached_team1 and cached_team2:
        try:
            data_team_season_1, data_team_career_1 = milestones(cached_team1)
            data_team_season_2, data_team_career_2 = milestones(cached_team2)
        except Exception as e:
            st.warning(f"Не удалось получить milestones: {e}")
        else:
            # На случай, если функции вернули не DataFrame
            if not isinstance(data_team_season_1, pd.DataFrame):
                data_team_season_1 = pd.DataFrame(data_team_season_1 or {})
            if not isinstance(data_team_career_1, pd.DataFrame):
                data_team_career_1 = pd.DataFrame(data_team_career_1 or {})
            if not isinstance(data_team_season_2, pd.DataFrame):
                data_team_season_2 = pd.DataFrame(data_team_season_2 or {})
            if not isinstance(data_team_career_2, pd.DataFrame):
                data_team_career_2 = pd.DataFrame(data_team_career_2 or {})

            tab7_col1, tab7_col2 = tab_temp_8.columns((5, 5))

            tab7_col1.title("Сезон")
            tab7_col1.dataframe(data_team_season_1)
            tab7_col2.title("Сезон")
            tab7_col2.dataframe(data_team_season_2)
            tab7_col1.title("Карьера")
            tab7_col1.dataframe(data_team_career_1)
            tab7_col2.title("Карьера")
            tab7_col2.dataframe(data_team_career_2)
else:
    tab_temp_7.info("Нет онлайн-плей-данных для отображения.")


def schedule_selected_team(team_id, data, game_id, selected, away_team_id):
    columns = [
        "game.localDate",
        "team1.name",
        "team1.logo",
        "game.score",
        "team2.logo",
        "team2.name",
        "game.fullScore",
        "win",
        "team1.teamId",
        "team2.teamId",
    ]
    df_schedule_new = data.loc[
        # (data["game.id"] < game_id)
        # &
        (
            (data["team1.teamId"].isin([team_id]))
            | (data["team2.teamId"].isin([team_id]))
        )
    ]
    df_schedule_new.loc[:, "game.fullScore"] = df_schedule_new[
        "game.fullScore"
    ].str.split(",")
    conditions = [
        (df_schedule_new["team1.teamId"] == team_id)
        & (df_schedule_new["game.score1"] > df_schedule_new["game.score2"]),
        (df_schedule_new["team1.teamId"] == team_id)
        & (df_schedule_new["game.score1"] < df_schedule_new["game.score2"]),
        (df_schedule_new["team2.teamId"] == team_id)
        & (df_schedule_new["game.score2"] > df_schedule_new["game.score1"]),
        (df_schedule_new["team2.teamId"] == team_id)
        & (df_schedule_new["game.score2"] < df_schedule_new["game.score1"]),
    ]
    values = [True, False, True, False]

    df_schedule_new = df_schedule_new.copy()
    df_schedule_new.loc[:, "win"] = np.select(conditions, values, default=None)
    mask = pd.Series(True, index=df_schedule_new.index)

    # Проверяем каждое выбранное условие и объединяем с маской
    if selected:
        if "Дома" in selected:
            mask &= df_schedule_new["team1.teamId"] == team_id
        if "В гостях" in selected:
            mask &= df_schedule_new["team2.teamId"] == team_id
        if "Выигрыши" in selected:
            mask &= df_schedule_new["win"] == True
        if "Поражения" in selected:
            mask &= df_schedule_new["win"] == False
        if "Друг с другом" in selected:
            mask &= df_schedule_new["team1.teamId"].isin(
                [away_team_id, team_id]
            ) & df_schedule_new["team2.teamId"].isin([away_team_id, team_id])
    return df_schedule_new[columns].loc[mask]


def get_in(d, path, default=None):
    cur = d
    for key in path:
        if not isinstance(cur, dict):
            return default
        cur = cur.get(key, default)
        if cur is default:
            return default
    return cur


if tab_schedule:
    if cached_schedule and "items" in cached_schedule:
        cached_schedule = cached_schedule["items"]
        pd_schedule = pd.json_normalize(cached_schedule)

        game_online = st.session_state.get("game_online")
        # print(game_online)
        team1_id = get_in(game_online, ["result", "team1", "teamId"])
        team1_name = get_in(game_online, ["result", "team1", "name"])
        team2_id = get_in(game_online, ["result", "team2", "teamId"])
        team2_name = get_in(game_online, ["result", "team2", "name"])
        game_id = get_in(game_online, ["result", "game", "id"])

        col1_schedule, col2_schedule = tab_schedule.columns((5, 5))
        options = ["Дома", "В гостях", "Выигрыши", "Поражения", "Друг с другом"]
        selection1 = col1_schedule.segmented_control(
            "Фильтр", options, selection_mode="multi", key="1"
        )
        selection2 = col2_schedule.segmented_control(
            "Фильтр", options, selection_mode="multi", key="2"
        )

        team1_data = schedule_selected_team(
            team1_id, pd_schedule, game_id, selection1, team2_id
        )
        # print(team1_data)
        team2_data = schedule_selected_team(
            team2_id, pd_schedule, game_id, selection2, team1_id
        )

        def highlight_two_teams(s):
            # print(s)
            try:
                if s.loc["team1.teamId"] in (
                    team1_id,
                    team2_id,
                ) and s.loc["team2.teamId"] in (
                    team1_id,
                    team2_id,
                ):
                    return ["background-color: #FF4B4B"] * len(s)
                else:
                    return [""] * len(s)
            except NameError:
                return [""] * len(s)

        column_config = {
            "team1.name": st.column_config.TextColumn("Команда1", width=150),
            "team2.name": st.column_config.TextColumn("Команда2", width=150),
            "game.score": st.column_config.TextColumn(
                "Счёт",
            ),
            "game.localDate": st.column_config.TextColumn(
                "Дата",
            ),
            "game.fullScore": st.column_config.Column(
                "Счёт по четвертям", width="medium"
            ),
            "team1.logo": st.column_config.ImageColumn("Лого1", width=50),
            "team2.logo": st.column_config.ImageColumn("Лого2", width=50),
        }
        count_game_1 = len(team1_data)
        count_game_2 = len(team2_data)
        team1_data = team1_data.style.apply(highlight_two_teams, axis=1).apply(
            color_win, subset="win"
        )
        team2_data = team2_data.style.apply(highlight_two_teams, axis=1).apply(
            color_win, subset="win"
        )
        height1 = 38 * max(count_game_1, 10)
        height2 = 38 * max(count_game_2, 10)
        col1_schedule.dataframe(
            team1_data,
            hide_index=True,
            height=int(min(height1, 1200)),
            column_config=column_config,
        )
        col2_schedule.dataframe(
            team2_data,
            hide_index=True,
            height=int(min(height2, 1200)),
            column_config=column_config,
        )


with tab_online:
    if isinstance(cached_schedule, list):
        current_date = pd.to_datetime(datetime.now().date())
        df_online = pd.json_normalize(cached_schedule)
        df_online["game.localDate"] = pd.to_datetime(
            df_online["game.localDate"], format="%d.%m.%Y", errors="coerce"
        )
        df_filtered = df_online[df_online["game.localDate"] == current_date].copy()

        names = {
            "Avtodor": r"D:\Графика\БАСКЕТБОЛ\VTB League\TEAM LOGOS\LOGOS 2025-2026\AVTODOR.png",
            "BETCITY PARMA": r"D:\Графика\БАСКЕТБОЛ\VTB League\TEAM LOGOS\LOGOS 2025-2026\BETCITY_PARMA_ENG.png",
            "CSKA": r"D:\Графика\БАСКЕТБОЛ\VTB League\TEAM LOGOS\LOGOS 2025-2026\CSKA_BLACK_BACK_ENG.png",
            "Enisey": r"D:\Графика\БАСКЕТБОЛ\VTB League\TEAM LOGOS\LOGOS 2025-2026\ENISEY_ENG.png",
            "Lokomotiv Kuban": r"D:\Графика\БАСКЕТБОЛ\VTB League\TEAM LOGOS\LOGOS 2025-2026\LOKOMOTIV KUBAN.png",
            "MBA-MAI": r"D:\Графика\БАСКЕТБОЛ\VTB League\TEAM LOGOS\LOGOS 2025-2026\MBA-MAI.png",
            "Pari Nizhny Novgorod": r"D:\Графика\БАСКЕТБОЛ\VTB League\TEAM LOGOS\LOGOS 2025-2026\PARI_NN_ENG.png",
            "Samara": r"D:\Графика\БАСКЕТБОЛ\VTB League\TEAM LOGOS\LOGOS 2025-2026\SAMARA_BLACK_BACK.png",
            "UNICS": r"D:\Графика\БАСКЕТБОЛ\VTB League\TEAM LOGOS\LOGOS 2025-2026\UNICS_BLACK_BACK.png",
            "Uralmash": r"D:\Графика\БАСКЕТБОЛ\VTB League\TEAM LOGOS\LOGOS 2025-2026\URALMASH_ENG.png",
            "Zenit": r"D:\Графика\БАСКЕТБОЛ\VTB League\TEAM LOGOS\LOGOS 2025-2026\ZENIT.png",
        }

        df_filtered["logo1"] = df_filtered["team1.name"].map(names)
        df_filtered["logo2"] = df_filtered["team2.name"].map(names)

        # Словарь для быстрого поиска по gameId
        live_data_map = {}

        # Собираем live-данные по каждому game.id

        for _, row in df_filtered.iterrows():
            game_id = row["game.id"]
            try:
                json_data = requests.get(
                    f"https://vtb-league.org/api/abc/games/live-status?Id={game_id}&Lang=en",
                ).json()
            except Exception as ex:

                print(ex)

            # Берём содержимое result (словарь с gameId и данными)
            if json_data:
                result = json_data.get("result", {})
                if result and "gameId" in result:
                    live_data_map[result["gameId"]] = result

        # Создаём колонки для live-данных
        df_filtered["live_period"] = None
        df_filtered["live_timeToGo"] = None
        df_filtered["live_scoreA"] = None
        df_filtered["live_scoreB"] = None
        df_filtered["live_status"] = None
        df_filtered["period"] = None
        # Заполняем колонки, где game.id совпадает с gameId
        for idx, row in df_filtered.iterrows():
            game_id = row["game.id"]
            if game_id in live_data_map:
                live_info = live_data_map[game_id]
                df_filtered.at[idx, "live_period"] = live_info.get("period")
                df_filtered.at[idx, "live_timeToGo"] = live_info.get("timeToGo")
                df_filtered.at[idx, "live_scoreA"] = live_info.get("scoreA")
                df_filtered.at[idx, "live_scoreB"] = live_info.get("scoreB")
                df_filtered.at[idx, "live_status"] = live_info.get("gameStatus")
                try:
                    if live_info.get("timeToGo") != "":
                        if float(live_info.get("timeToGo")) == 0.0:
                            if live_info.get("period") == 2:
                                df_filtered.at[idx, "period"] = "HT"
                            elif live_info.get("period") in [1, 3]:
                                df_filtered.at[idx, "period"] = (
                                    f"END {live_info.get('period')}Q"
                                )
                        elif float(live_info.get("timeToGo")) > 0.0:
                            df_filtered.at[idx, "period"] = (
                                f"{live_info.get('period')}Q"
                            )
                        else:
                            df_filtered.at[idx, "period"] = ""
                except Exception as ex:
                    df_filtered.at[idx, "period"] = ""
                    print(ex)

        df_filtered["live_scoreA"] = pd.to_numeric(
            df_filtered["live_scoreA"], errors="coerce"
        ).fillna(pd.to_numeric(df_filtered["game.score1"], errors="coerce"))

        df_filtered["live_scoreB"] = pd.to_numeric(
            df_filtered["live_scoreB"], errors="coerce"
        ).fillna(pd.to_numeric(df_filtered["game.score2"], errors="coerce"))

        # Вывод нужных колонок
        columns = [
            "game.id",
            "game.gameStatus",
            "status.displayName",
            "game.defaultZoneTime",
            "team1.name",
            "team2.name",
            "game.score1",
            "game.score2",
            "live_scoreA",
            "live_scoreB",
            "live_period",
            "live_timeToGo",
            "live_status",
            "period",
            "team1.logo",
            "team2.logo",
            "logo1",
            "logo2",
        ]

        existing_columns = [c for c in columns if c in df_filtered.columns]
        df_sel = df_filtered.loc[:, existing_columns].copy()
        data = df_sel.to_dict(orient="records")  # список словарей
        rewrite_file("online", data)

        st.dataframe(
            df_filtered[existing_columns],
            column_config={
                "team1.logo": st.column_config.ImageColumn("team1.logo"),
                "team2.logo": st.column_config.ImageColumn("team2.logo"),
            },
            width="content",
        )


def get_play_info(play):
    # Ищем в списке play_type_id элемент, у которого PlayTypeID совпадает с play
    for item in play_type_id:
        if item["PlayTypeID"] == play:
            return item["PlayInfoSite"]
    return None  # Если совпадение не найдено


def get_player_name(start_num):
    # Ищем в списке teams_temp элемент, у которого startNum совпадает с temp_data_pbp["startNum"]
    for player in teams_temp:
        if player["startNum"] == start_num:
            return f"{player['firstName']} {player['lastName']}"
    return None  # Если совпадение не найдено


def get_event_time(row):
    if row != 0:
        time_str = 6000 - row
        if time_str == 0:
            time_str = "0:00"
        else:
            time_str = time_str // 10
            time_str = f"{time_str // 60}:{str(time_str % 60).zfill(2)}"
        return time_str


# Безопасная загрузка PlayTypeID
try:
    with open("PlayTypeID.json", "r", encoding="utf-8") as f:
        play_type_id = json.load(f)
except (FileNotFoundError, json.JSONDecodeError):
    play_type_id = []

teams_section = ((cached_game_online or {}).get("result") or {}).get("teams") or {}

# Если teams_section — список (например, [{"starts": [...]}, {...}])
if isinstance(teams_section, list):
    if len(teams_section) >= 2:
        starts1 = next(
            (t.get("starts") for t in teams_section if t["teamNumber"] == 1), None
        )
        starts2 = next(
            (t.get("starts") for t in teams_section if t["teamNumber"] == 2), None
        )
    else:
        starts1 = []
        starts2 = []
# Если teams_section — словарь (обычно {"1": {...}, "2": {...}})
elif isinstance(teams_section, dict):
    starts1 = (teams_section.get(1) or teams_section.get("1") or {}).get("starts") or []
    starts2 = (teams_section.get(2) or teams_section.get("2") or {}).get("starts") or []
else:
    starts1 = []
    starts2 = []

teams_temp = sorted(
    [x for x in starts1 if isinstance(x, dict)], key=lambda x: x.get("playerNumber", 0)
) + sorted(
    [x for x in starts2 if isinstance(x, dict)], key=lambda x: x.get("playerNumber", 0)
)

list_fullname = [None] + [
    f"({x.get('displayNumber')}) {x.get('firstName','')} {x.get('lastName','')}".strip()
    for x in teams_temp
    if x.get("startRole") == "Player"
]


def get_play_info(play):
    for item in play_type_id:
        if isinstance(item, dict) and item.get("PlayTypeID") == play:
            return item.get("PlayInfoSite")
    return None


def get_player_name(start_num):
    for player in teams_temp:
        if player.get("startNum") == start_num:
            return f"{player.get('firstName','')} {player.get('lastName','')}".strip()
    return None


def get_event_time(row):
    if isinstance(row, (int, float)) and row != 0:
        time_val = 6000 - int(row)
        if time_val <= 0:
            return "0:00"
        time_val //= 10
        return f"{time_val // 60}:{str(time_val % 60).zfill(2)}"
    return None


with tab_pbp:
    plays = ((cached_game_online or {}).get("result") or {}).get("plays") or []
    if plays:
        temp_data_pbp = pd.DataFrame(plays)
        col1_pbp, col2_pbp = tab_pbp.columns((3, 4))
        option_player = col1_pbp.selectbox("Выбрать игрока", list_fullname)

        options_pbp = ["1 очко", "2 очка", "3 очка"]
        selection_pbp = col1_pbp.segmented_control(
            "Фильтр", options_pbp, selection_mode="multi", key=3
        )

        if not temp_data_pbp.empty and "period" in temp_data_pbp.columns:
            options_quarter = [
                (f"{i+1} четверть" if i + 1 < 5 else f"{i-3} овертайм")
                for i in range(int(temp_data_pbp["period"].max()))
            ]
            selection_quarter = col1_pbp.segmented_control(
                "Выбор четверти", options_quarter, selection_mode="multi", key=4
            )

            temp_data_pbp["info"] = temp_data_pbp["play"].map(get_play_info)
            temp_data_pbp["who"] = temp_data_pbp["startNum"].map(get_player_name)
            temp_data_pbp["time"] = temp_data_pbp["sec"].map(get_event_time)

            mask1 = pd.Series(True, index=temp_data_pbp.index)

            if option_player:
                # безопасный поиск startNum
                for x in teams_temp:
                    display = f"({x.get('displayNumber')}) {x.get('firstName','')} {x.get('lastName','')}".strip()
                    if display == option_player:
                        mask1 &= temp_data_pbp["startNum"] == x.get("startNum")
                        break

            if selection_pbp:
                plays_mapping = {"1 очко": 1, "2 очка": 2, "3 очка": 3}
                selected_plays = [
                    plays_mapping[p] for p in selection_pbp if p in plays_mapping
                ]
                mask1 &= temp_data_pbp["play"].isin(selected_plays)

            if selection_quarter:
                select_quart = [
                    i + 1
                    for i, q in enumerate(options_quarter)
                    if q in selection_quarter
                ]
                mask1 &= temp_data_pbp["period"].isin(select_quart)

            filtered_data_pbp = temp_data_pbp[mask1]
            count_pbp = len(filtered_data_pbp)

            column_pbp = ["num", "info", "who", "period", "time"]
            column_config_pbp = {
                "info": st.column_config.TextColumn(width="medium"),
                "who": st.column_config.TextColumn(width="large"),
            }
            col2_pbp.dataframe(
                filtered_data_pbp[column_pbp],
                column_config=column_config_pbp,
                hide_index=True,
                height=(38 * count_pbp if count_pbp > 10 else "auto"),
            )
    else:
        tab_pbp.info("Данных play-by-play нет.")