ChessCalcNextTour/swiss_calc/pgn_fetcher.py

"""
PGN-партии игрока с Lichess-трансляций по FIDE ID.

Источник: https://lichess.org/api/broadcast
Формат: в PGN-тегах есть WhiteFideId / BlackFideId / TimeControl.
"""

import re
import time
import requests
from bs4 import BeautifulSoup
from typing import Optional, List, Dict

HEADERS = {
    'User-Agent': 'Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36'
}

RATE_DELAY = 1.5
MAX_BROADCASTS = 200
MIN_YEAR = 2025


def get_player_name(fide_id: str) -> str:
    """Получить имя игрока с ratings.fide.com."""
    url = f'https://ratings.fide.com/profile/{fide_id}'
    resp = _fetch_with_retry(url)
    if not resp:
        return ''

    soup = BeautifulSoup(resp.text, 'html.parser')

    for tag in soup.find_all(['h1', 'h2', 'span']):
        text = tag.get_text(strip=True)
        if text and 3 < len(text) < 80 and re.search(r'[A-ZА-ЯЁ]', text):
            if not re.match(r'^(MAIN|NEWS|RATINGS|CHAMPIONSHIP|CALENDAR|FIDE|DIRECTORY|PARTNERS|CONTACTS|SHOP|HOME|ARCHIVE)', text, re.IGNORECASE):
                return text

    full_text = soup.get_text()
    m = re.search(r'FIDE\s*ID\s*\d+', full_text)
    if m:
        before = full_text[:m.start()].strip().split('\n')
        for line in reversed(before):
            line = line.strip()
            if line and 3 < len(line) < 80 and re.search(r'[A-ZА-ЯЁ]', line):
                return line

    return ''


def _fetch_with_retry(url: str, max_retries: int = 2):
    """GET-запрос с retry при таймауте."""
    for attempt in range(max_retries + 1):
        try:
            return requests.get(url, headers=HEADERS, timeout=15 + attempt * 10)
        except requests.exceptions.Timeout:
            if attempt < max_retries:
                time.sleep(1)
            else:
                return None
        except requests.exceptions.RequestException:
            return None
    return None


def search_broadcasts(name_part: str) -> List[Dict]:
    """Поиск Lichess-трансляций по имени игрока."""
    url = f'https://lichess.org/api/broadcast?q={name_part}'
    resp = requests.get(url, headers=HEADERS, timeout=30)
    resp.raise_for_status()

    broadcasts = []
    for line in resp.text.strip().split('\n'):
        line = line.strip()
        if not line:
            continue
        try:
            import json
            data = json.loads(line)
            if 'tour' in data:
                broadcasts.append(data)
        except json.JSONDecodeError:
            continue
        if len(broadcasts) >= MAX_BROADCASTS:
            break

    return broadcasts


def fetch_round_pgn(round_id: str, delay: float = RATE_DELAY) -> Optional[str]:
    """Скачать PGN одного раунда Lichess-трансляции."""
    time.sleep(delay)
    url = f'https://lichess.org/api/broadcast/round/{round_id}.pgn'
    resp = requests.get(url, headers=HEADERS, timeout=30)
    if resp.status_code == 429:
        time.sleep(5)
        resp = requests.get(url, headers=HEADERS, timeout=30)
    if resp.status_code != 200:
        return None
    text = resp.text.strip()
    return text if text else None


def filter_by_fide_id(pgn_text: str, fide_id: str) -> List[str]:
    """Выкусить из PGN только партии игрока с данным FIDE ID."""
    games = []
    raw_games = re.split(r'\n(?=\[Event )', pgn_text)
    for game in raw_games:
        game = game.strip()
        if not game:
            continue
        if f'WhiteFideId "{fide_id}"' in game or f'BlackFideId "{fide_id}"' in game:
            games.append(game)
    return games


def classify_time_control(game_pgn: str) -> str:
    """Определить категорию: standard / rapid / blitz."""
    m = re.search(r'\[TimeControl\s+"([^"]+)"\]', game_pgn)
    if not m:
        return 'unknown'
    tc_str = m.group(1)
    parts = tc_str.split('+')
    try:
        initial = int(parts[0])
    except ValueError:
        return 'unknown'
    # initial time in seconds
    if initial >= 3600:
        return 'standard'
    elif initial >= 600:
        return 'rapid'
    else:
        return 'blitz'


def get_player_games(fide_id: str) -> Dict[str, str]:
    """Оркестратор: найти все PGN-партии игрока, сгруппировать по контролю."""
    name = get_player_name(fide_id)
    if not name:
        return {}

    parts = name.lower().split()
    last_name = parts[0].rstrip(',') if parts else name.lower()

    broadcasts = search_broadcasts(last_name)

    # Приоритезируем трансляции, где имя игрока в названии
    relevant_bcs = []
    others = []
    for bc in broadcasts:
        bc_name = bc.get('tour', {}).get('name', '').lower()
        if any(part in bc_name for part in parts if len(part) > 2):
            relevant_bcs.append(bc)
        else:
            others.append(bc)

    # Собираем finished-раунды: сначала из релевантных, потом остальные
    rounds_to_fetch = []
    MAX_ROUNDS = 30

    for bc_list in (relevant_bcs, others):
        for bc in bc_list:
            for rd in bc.get('rounds', []):
                if rd.get('finished') and len(rounds_to_fetch) < MAX_ROUNDS:
                    rounds_to_fetch.append(rd['id'])
            if len(rounds_to_fetch) >= MAX_ROUNDS:
                break
        if len(rounds_to_fetch) >= MAX_ROUNDS:
            break

    result: Dict[str, list] = {'standard': [], 'rapid': [], 'blitz': []}
    seen_games = set()
    total_found = 0

    for round_id in rounds_to_fetch:
        pgn = fetch_round_pgn(round_id)
        if not pgn:
            continue
        for game in filter_by_fide_id(pgn, fide_id):
            key = game[:200]
            if key in seen_games:
                continue
            seen_games.add(key)
            tc = classify_time_control(game)
            if tc in result:
                result[tc].append(game)
            else:
                result['standard'].append(game)
            total_found += 1

    return {
        'standard': '\n\n'.join(result['standard']),
        'rapid': '\n\n'.join(result['rapid']),
        'blitz': '\n\n'.join(result['blitz']),
    }


def count_games(pgn_text: str) -> int:
    """Количество партий в PGN-строке."""
    if not pgn_text.strip():
        return 0
    return len([g for g in pgn_text.split('\n\n[Event ') if '[Event ' in g or g.startswith('[Event ')])


def extract_fide_id(text: str) -> Optional[str]:
    """Вытащить FIDE ID из текста (ссылка ratings.fide.com или просто ID)."""
    m = re.search(r'(\d{5,8})', text)
    return m.group(1) if m else None