├── .gitignore ├── requirements.txt ├── app ├── nico │ ├── config.py │ ├── __init__.py │ └── crawler.py ├── gelbooru │ ├── config.py │ ├── __init__.py │ └── crawler.py ├── twitter │ ├── config.py │ ├── model.py │ ├── __init__.py │ └── crawler.py ├── pixiv │ ├── config.py │ ├── __init__.py │ ├── pixiv_auth.py │ ├── crawler.py │ └── model.py └── models.py ├── config.py ├── run.sh ├── run_crawlers.py ├── README.md └── main.py /.gitignore: -------------------------------------------------------------------------------- 1 | **/__pycache__ 2 | **/*.log 3 | test*.py 4 | **/*.sqlite3 5 | *.sqlite3-shm 6 | *.sqlite3-wal 7 | *.out 8 | # config.py 9 | # **/config.py 10 | # run.sh 11 | -------------------------------------------------------------------------------- /requirements.txt: -------------------------------------------------------------------------------- 1 | # pre 2 | python-telegram-bot==20.0a2 3 | aiohttp 4 | python-dotenv 5 | pytz 6 | lxml 7 | xmltodict 8 | tortoise-orm 9 | pydantic 10 | tweepy 11 | pixivpy-async -------------------------------------------------------------------------------- /app/nico/config.py: -------------------------------------------------------------------------------- 1 | # 要推送的图片 tags 2 | NICO_TAGS = ['a', 'b'] 3 | 4 | # user_sess 同名 cookie，推送原图时需要 5 | NICO_USER_SESS = '' 6 | 7 | # 爬取页数 8 | NICO_PAGE_NUM = 1 9 | 10 | # 用户黑名单，该名单内的用户的作品不会被推送 11 | NICO_USER_BLACKLIST = [] 12 | -------------------------------------------------------------------------------- /config.py: -------------------------------------------------------------------------------- 1 | # bot token 2 | TOKEN = 'xxxx' 3 | 4 | # 消息发送失败的最大重试次数 5 | MESSAGE_MAX_RETRY = 12 6 | 7 | # group/user/channel id 列表 8 | CHAD_ID_LIST = ['1234567'] 9 | 10 | # 代理地址（本地使用需要设置） 11 | PROXY = None # 'http://127.0.0.1:1081' 12 | -------------------------------------------------------------------------------- /run.sh: -------------------------------------------------------------------------------- 1 | cd /root/telegram-bot/th-telegram-image-bot 2 | LOCK="LOCK" 3 | 4 | if [ ! -f "$LOCK" ]; then 5 | touch $LOCK 6 | python run_crawlers.py 7 | sleep 1 8 | python main.py 9 | rm -rf $LOCK 10 | else 11 | ts=`stat -c %Y LOCK` 12 | now=`date +%s` 13 | if [ $[ $now - $ts ] -gt 1800 ]; then 14 | rm -rf $LOCK 15 | echo "Lock expired, deleted" 16 | fi 17 | fi 18 | -------------------------------------------------------------------------------- /app/gelbooru/config.py: -------------------------------------------------------------------------------- 1 | # 推送图片的 tags 2 | GELBOORU_TAGS = ['a', 'b'] 3 | 4 | # 爬取页数 5 | GELBOORU_PAGE_NUM = 1 6 | 7 | # 图片质量 8 | # file_url: 原图 9 | # sample_url: 压缩图 10 | GELBOORU_IMAGE_QUALITY = 'file_url' # or sample_url 11 | 12 | # 因为 Gelbooru 限制了 API 请求次数（目测是 110 次每天），所以需要减少请求次数 13 | # 每隔多少分钟请求 Gelbooru，取值为 1 - 59 14 | REQ_INTERVAL = 20 15 | 16 | # API KEY 17 | # 在 https://gelbooru.com/index.php?page=account&s=options 中 18 | API_KEY = '' 19 | USER_ID = '' -------------------------------------------------------------------------------- /app/twitter/config.py: -------------------------------------------------------------------------------- 1 | # 目标列表Id 2 | # 关于 LIST 参考：https://help.twitter.com/en/using-twitter/twitter-lists 3 | TWITTER_LIST_ID = ['123456789'] 4 | # 爬取页数 5 | TWITTER_PAGE_NUM = 5 6 | # 只推送含图片的动态 7 | TWITTER_ONLY_IMAGE = True 8 | 9 | # KEY 相关，生成 Authentication Tokens 10 | # https://developer.twitter.com/en/portal/dashboard 11 | # 获取方法： 12 | # https://developer.twitter.com/en/docs/authentication/oauth-2-0/bearer-tokens 13 | TWITTER_API_KEY = '' 14 | TWITTER_API_KEY_SECRET = '' 15 | TWITTER_ACCESS_TOKEN = '' 16 | TWITTER_ACCESS_SECRET = '' 17 | -------------------------------------------------------------------------------- /app/twitter/model.py: -------------------------------------------------------------------------------- 1 | from pydantic import BaseModel 2 | from datetime import datetime 3 | 4 | from typing import List 5 | 6 | 7 | class TimeLine(BaseModel): 8 | author_name: str 9 | created_at: datetime 10 | text: str 11 | id: str 12 | photos: List[str] 13 | 14 | def __str__(self): 15 | author = self.author_name 16 | created_at = self.created_at 17 | text = self.text 18 | id = self.id 19 | 20 | return id + '\n\n' + 'author: ' + author + '\n' + 'created_at(utc+8):\n' + str(created_at) + '\n\n' + text 21 | -------------------------------------------------------------------------------- /app/pixiv/config.py: -------------------------------------------------------------------------------- 1 | # 必须配置 PIXIV_TAGS 和 PIXIV_REFRESH_TOKEN，其他配置可默认 2 | 3 | # 推送图片的 tags 4 | PIXIV_TAGS = [ 5 | 'a', 6 | { 7 | 'tag': 'b', 8 | 'score': 1000 9 | } 10 | ] 11 | 12 | # refresh token 13 | # 获取方法：https://gist.github.com/ZipFile/c9ebedb224406f4f11845ab700124362 14 | PIXIV_REFRESH_TOKEN = 'dF8jpliqu0UlM519k09ICNLQQYnbLL6MBO9XuqDGEBPFAC' 15 | 16 | # 用户白名单，填入 uid，在该名单内的用户的作品会被立即推送（无视分数） 17 | PIXIV_USER_WHITELIST = [] 18 | 19 | # 每个 tag 爬取的最大页数 20 | PIXIV_MAX_PAGE = 1 21 | 22 | # 推送的投稿的最小分数，用于排除低质量的投稿 23 | # 计算公式 view + bookmark * 5 24 | PIXIV_MIN_POST_SCORE = 300 25 | 26 | # 是否推送关注的作者 27 | PIXIV_FOLLOW = False 28 | 29 | # 反代 i.pixiv.re，目前不需要 30 | PIXIV_REVERSE_PROXY = '' 31 | -------------------------------------------------------------------------------- /app/twitter/__init__.py: -------------------------------------------------------------------------------- 1 | from app.models import ImageDB, ImagePD 2 | from config import PROXY 3 | from .config import * 4 | 5 | from .crawler import TwitterListCrawler 6 | 7 | api_config = { 8 | 'api_key': TWITTER_API_KEY, 9 | 'api_key_secret': TWITTER_API_KEY_SECRET, 10 | 'access_token': TWITTER_ACCESS_TOKEN, 11 | 'access_secret': TWITTER_ACCESS_SECRET 12 | } 13 | 14 | 15 | async def run(): 16 | crawler = TwitterListCrawler(proxy=PROXY, **api_config) 17 | for list_id in TWITTER_LIST_ID: 18 | for timeline in crawler.get_timeline(list_id, pages=TWITTER_PAGE_NUM, only_img=TWITTER_ONLY_IMAGE): 19 | dct = { 20 | 'original_site': 'twitter', 21 | 'original_id': timeline.id, 22 | 'pic_hash_list': timeline.photos 23 | } 24 | 25 | if not await ImageDB.filter(**dct): 26 | pd = ImagePD(content=str(timeline), **dct) 27 | await ImageDB.create(**pd.dict()) 28 | 29 | yield pd 30 | -------------------------------------------------------------------------------- /run_crawlers.py: -------------------------------------------------------------------------------- 1 | """ 2 | 运行全部爬虫，入库数据 3 | """ 4 | import asyncio 5 | 6 | from tortoise import Tortoise, run_async 7 | 8 | from app.gelbooru import run as run_gel 9 | from app.nico import run as run_nico 10 | from app.pixiv import run as run_pixiv 11 | from app.twitter import run as run_twitter 12 | 13 | 14 | async def init_db(): 15 | db_url = 'sqlite://db.sqlite3' 16 | 17 | await Tortoise.init( 18 | db_url=db_url, 19 | modules={'models': ['app.models']} 20 | ) 21 | 22 | await Tortoise.generate_schemas() 23 | 24 | 25 | funcs = { 26 | 'Gelbooru': run_gel, 27 | 'Nico': run_nico, 28 | 'Twitter': run_twitter, 29 | 'Pixiv': run_pixiv 30 | } 31 | 32 | 33 | async def main(): 34 | await init_db() 35 | 36 | for name, func in funcs.items(): 37 | print('=' * 30) 38 | print(f'开始执行 {name} 爬虫') 39 | try: 40 | async for img in func(): 41 | print(img) 42 | except Exception as e: 43 | print(f'{name} 爬虫失败:', e) 44 | print('=' * 30) 45 | print() 46 | 47 | 48 | if __name__ == '__main__': 49 | run_async(main()) 50 | -------------------------------------------------------------------------------- /app/gelbooru/__init__.py: -------------------------------------------------------------------------------- 1 | from app.models import ImageDB, ImagePD 2 | from config import PROXY 3 | from datetime import datetime 4 | 5 | from .config import * 6 | from .crawler import GelbooruCrawler 7 | 8 | 9 | def parse_item(item, tag): 10 | content = f'[Gel #{tag}]\nid={item["id"]}' 11 | pics = [item[GELBOORU_IMAGE_QUALITY]] 12 | return ImagePD(content=content, pic_hash_list=pics, original_site='gelbooru', original_id=item['id']) 13 | 14 | 15 | async def run(): 16 | minute = datetime.now().minute 17 | 18 | if minute % REQ_INTERVAL != 0: 19 | return 20 | 21 | async with GelbooruCrawler(proxy=PROXY, api_key=API_KEY, user_id=USER_ID) as crawler: 22 | for tag in GELBOORU_TAGS: 23 | async for data in crawler.get_many_pages(tag, begin=0, end=GELBOORU_PAGE_NUM): 24 | for item in data['post']: 25 | pd = parse_item(item, tag) 26 | dct = pd.dict() 27 | if not await ImageDB.filter(original_site=pd.original_site, original_id=pd.original_id): 28 | await ImageDB.create(**dct) 29 | yield pd 30 | -------------------------------------------------------------------------------- /app/models.py: -------------------------------------------------------------------------------- 1 | from typing import List, Optional 2 | 3 | from pydantic import BaseModel 4 | from tortoise import fields, models 5 | 6 | 7 | class ImageDB(models.Model): 8 | id = fields.IntField(pk=True) 9 | content = fields.CharField(255) 10 | pic_hash_list = fields.JSONField() 11 | original_site = fields.CharField(max_length=10, index=True) 12 | original_id = fields.CharField(max_length=25, index=True) 13 | send_at = fields.DatetimeField(auto_now_add=True) 14 | send_successed = fields.BooleanField(default=False) 15 | retry = fields.IntField(default=0) 16 | reason = fields.CharField(max_length=50, default='') 17 | 18 | def __str__(self): 19 | return f'{self.content}' 20 | 21 | def __repr__(self): 22 | return f'{self.content}\n\n{self.original_site} - {self.original_id}\n{ self.send_at}\n{self.send_successed}' 23 | 24 | # pydantic creator doesn't work on pyright lint, so just written a pydantic model of ImageDB 25 | 26 | 27 | class ImagePD(BaseModel): 28 | content: str 29 | pic_hash_list: List[str] 30 | original_site: str 31 | original_id: str 32 | 33 | def __str__(self): 34 | pic_number = len(self.pic_hash_list) 35 | return self.content + f'\npic number: {pic_number}\n' 36 | 37 | # TODO: 添加一个新表/字段，实现记录推送成功的人/群的ID，以避免多消息推送其中部分失败无法重发的问题 38 | -------------------------------------------------------------------------------- /app/pixiv/__init__.py: -------------------------------------------------------------------------------- 1 | from typing import List 2 | from app.models import ImageDB, ImagePD 3 | 4 | from config import PROXY 5 | 6 | from .config import * 7 | from .crawler import PixivCrawler 8 | from .model import Illust 9 | 10 | 11 | async def run(): 12 | illust_list: List[Illust] = [] 13 | 14 | async with PixivCrawler(PIXIV_REFRESH_TOKEN, PROXY) as crawler: 15 | for tag_info in PIXIV_TAGS: 16 | min_score = PIXIV_MIN_POST_SCORE 17 | tag = tag_info 18 | 19 | if isinstance(tag_info, dict): 20 | tag, min_score = tag_info['tag'], tag_info['score'] 21 | 22 | async for illust in crawler.search_illust(tag, PIXIV_MAX_PAGE): 23 | if illust.score >= min_score or (illust.user_id in PIXIV_USER_WHITELIST): 24 | illust_list.append(illust) 25 | 26 | if PIXIV_FOLLOW: 27 | async for illust in crawler.craw_follow(PIXIV_MAX_PAGE): 28 | illust_list.append(illust) 29 | 30 | for illust in illust_list: 31 | dct = { 32 | 'original_site': 'pixiv', 33 | 'original_id': illust.illust_id, 34 | 'pic_hash_list': illust.images 35 | } 36 | 37 | if not await ImageDB.filter(**dct): 38 | pd = ImagePD(content=str(illust)[:254], **dct) 39 | await ImageDB.create(**pd.dict()) 40 | yield pd 41 | -------------------------------------------------------------------------------- /app/pixiv/pixiv_auth.py: -------------------------------------------------------------------------------- 1 | from pprint import pprint 2 | from typing import Optional 3 | 4 | 5 | import requests 6 | 7 | # Latest app version can be found using GET /v1/application-info/android 8 | USER_AGENT = "PixivAndroidApp/5.0.234 (Android 11; Pixel 5)" 9 | REDIRECT_URI = "https://app-api.pixiv.net/web/v1/users/auth/pixiv/callback" 10 | LOGIN_URL = "https://app-api.pixiv.net/web/v1/login" 11 | AUTH_TOKEN_URL = "https://oauth.secure.pixiv.net/auth/token" 12 | CLIENT_ID = "MOBrBDS8blbauoSck0ZfDbtuzpyT" 13 | CLIENT_SECRET = "lsACyCD94FhDUtGTXi3QzcFE2uU1hqtDaKeqrdwj" 14 | 15 | 16 | def auth_token_response(response): 17 | data = response.json() 18 | 19 | try: 20 | access_token = data["access_token"] 21 | refresh_token = data["refresh_token"] 22 | return access_token, refresh_token 23 | except KeyError: 24 | print("error:") 25 | pprint(data) 26 | exit(1) 27 | 28 | 29 | def refresh(refresh_token: str, proxy = ''): 30 | response = requests.post( 31 | AUTH_TOKEN_URL, 32 | data={ 33 | "client_id": CLIENT_ID, 34 | "client_secret": CLIENT_SECRET, 35 | "grant_type": "refresh_token", 36 | "include_policy": "true", 37 | "refresh_token": refresh_token, 38 | }, 39 | headers={"User-Agent": USER_AGENT}, 40 | proxies={ 41 | 'http': proxy, 42 | 'https': proxy 43 | } 44 | ) 45 | 46 | return auth_token_response(response) 47 | -------------------------------------------------------------------------------- /app/gelbooru/crawler.py: -------------------------------------------------------------------------------- 1 | from typing import Optional 2 | 3 | import aiohttp 4 | from aiohttp import TCPConnector 5 | 6 | 7 | class GelbooruCrawler(object): 8 | def __init__(self, proxy: Optional[str] = '', api_key: Optional[str] = '', user_id: Optional[str] = ''): 9 | self.proxy = proxy 10 | self.api_key = api_key 11 | self.user_id = user_id 12 | self.session = aiohttp.ClientSession(connector=TCPConnector(ssl=False)) 13 | 14 | async def __aenter__(self): 15 | return self 16 | 17 | async def close(self): 18 | await self.session.close() 19 | 20 | async def __aexit__(self, ex1, e2, e3): 21 | await self.close() 22 | 23 | async def get_one_page(self, tags: str, page: int, limit: int = 42): 24 | api = 'https://gelbooru.com/index.php?page=dapi&s=post&q=index&json=1' 25 | 26 | params = { 27 | 'tags': tags, 28 | 'pid': page, 29 | 'limit': limit, 30 | 'api_key': self.api_key, 31 | 'user_id': self.user_id 32 | } 33 | 34 | headers = { 35 | 'user-agent': 'wasp', 36 | 'cookie': 'fringeBenefits=yup' 37 | } 38 | 39 | async with self.session.get(api, params=params, headers=headers, proxy=self.proxy) as resp: 40 | return await resp.json() 41 | 42 | async def get_many_pages(self, tags: str, begin: int, end: int, limit: int = 42): 43 | for page in range(begin, end): 44 | yield await self.get_one_page(tags, page, limit) 45 | -------------------------------------------------------------------------------- /app/nico/__init__.py: -------------------------------------------------------------------------------- 1 | from app.models import ImageDB, ImagePD 2 | from config import PROXY 3 | 4 | from .config import * 5 | from .crawler import NSCrawler 6 | 7 | 8 | def parse_message(info): 9 | username = info['nickname'] 10 | title = info['title'] 11 | description = info['description'][:150] 12 | created = '投稿时间' + info['created'] 13 | sid = info['id'] 14 | tags = ['#' + tag['name'] for tag in info['tag_list']['tag']] 15 | 16 | content = 'im' + sid + '\n' + title + '\n' 17 | content += '投稿者: ' + username + '\n' 18 | content += description + '\n\n' 19 | content += ' | '.join(tags) + '\n\n' + created 20 | 21 | return content 22 | 23 | 24 | async def run(): 25 | async with NSCrawler(NICO_USER_SESS, proxy=PROXY) as crawler: 26 | for tag in NICO_TAGS: 27 | async for imlist in crawler.get_many_pages(tag, 1, NICO_PAGE_NUM): 28 | for im in imlist: 29 | dct = { 30 | 'original_site': 'nico', 31 | 'original_id': im, 32 | } 33 | 34 | if not await ImageDB.filter(**dct): 35 | info = await crawler.get_info(im) 36 | 37 | if info['user_id'] in NICO_USER_BLACKLIST: # type: ignore 38 | continue 39 | 40 | content = parse_message(info)[:250] 41 | pd = ImagePD(content=content, pic_hash_list=[], **dct) 42 | await ImageDB.create(**pd.dict()) 43 | 44 | yield pd 45 | -------------------------------------------------------------------------------- /app/pixiv/crawler.py: -------------------------------------------------------------------------------- 1 | import asyncio 2 | from random import randint 3 | 4 | from pixivpy_async import AppPixivAPI, PixivClient 5 | 6 | from .model import parse_illust 7 | from .pixiv_auth import refresh 8 | 9 | 10 | class PixivCrawler(object): 11 | def __init__(self, refresh_token, proxy='') -> None: 12 | self.client = PixivClient(proxy=proxy) 13 | self.aapi = AppPixivAPI(client=self.client.start()) 14 | 15 | ACCESS_TOKEN, _ = refresh(refresh_token, proxy) 16 | self.aapi.set_auth(ACCESS_TOKEN, refresh_token) 17 | 18 | async def __aenter__(self): 19 | return self 20 | 21 | async def __aexit__(self, ex, e1, e2): 22 | await self.client.close() 23 | 24 | async def search_illust(self, tag: str, max_page=3): 25 | """抓取指定 Tag 的内容并返回一个生成器 26 | :param max_page: 抓取的最大页数 27 | """ 28 | offset = 0 29 | page_idx = 1 30 | 31 | while page_idx <= max_page: 32 | detail = await self.aapi.search_illust(tag, offset=offset) 33 | 34 | offset += len(detail['illusts']) 35 | page_idx += 1 36 | 37 | for item in detail['illusts']: 38 | illust = parse_illust(item) 39 | yield illust 40 | 41 | # if page_idx <= max_page: 42 | # await asyncio.sleep(randint(1, 3)) 43 | 44 | async def craw_follow(self, max_page=3): 45 | for page in range(max_page): 46 | for illust in (await self.aapi.illust_follow(offset=page * 30))['illusts']: 47 | yield parse_illust(illust) 48 | 49 | # if page != max_page - 1: 50 | # await asyncio.sleep(randint(1, 3)) 51 | -------------------------------------------------------------------------------- /app/pixiv/model.py: -------------------------------------------------------------------------------- 1 | from typing import List 2 | 3 | from pydantic import BaseModel 4 | 5 | 6 | class Illust(BaseModel): 7 | title: str 8 | description: str 9 | illust_id: str 10 | tags: List[str] 11 | create_date: str 12 | user_id: str 13 | username: str 14 | images: List[str] 15 | score: int 16 | 17 | def __str__(self): 18 | pid = f'Pixiv ID: {self.illust_id}\n\n' 19 | title = f'{self.title}\n' if self.title else '(无题)' 20 | description = f'{self.description}\n\n' if self.description else '' 21 | poster = f'投稿者(id={self.user_id}): {self.username}\n\n' 22 | tags = f'{" | ".join(self.tags)}\n\n' 23 | post_at = f'投稿时间: {self.create_date}' 24 | 25 | return ''.join([pid, title, description, poster, tags, post_at]) 26 | 27 | 28 | def parse_illust(illust: dict) -> Illust: 29 | title = illust['title'] 30 | description = illust['caption'][:50] 31 | illust_id = illust['id'] 32 | tags = ['#' + tag['name'] for tag in illust['tags']] 33 | create_date = illust['create_date'] 34 | 35 | view = illust['total_view'] 36 | bookmark = illust['total_bookmarks'] 37 | 38 | user = illust['user'] 39 | user_id = user['id'] 40 | name = user['name'] 41 | 42 | meta_single_page = illust['meta_single_page'] 43 | meta_pages = illust['meta_pages'] 44 | 45 | if 'original_image_url' in meta_single_page: 46 | images = [meta_single_page['original_image_url']] 47 | else: 48 | images = [image['image_urls']['original'] for image in meta_pages] 49 | 50 | illust_model = Illust(**{ 51 | 'title': title, 52 | 'description': description, 53 | 'illust_id': illust_id, 54 | 'tags': tags, 55 | 'create_date': create_date, 56 | 'user_id': user_id, 57 | 'username': name, 58 | 'images': images, 59 | 'score': view + bookmark * 5 60 | }) 61 | 62 | return illust_model 63 | -------------------------------------------------------------------------------- /app/twitter/crawler.py: -------------------------------------------------------------------------------- 1 | from datetime import timedelta 2 | import os 3 | 4 | import tweepy 5 | 6 | from .model import TimeLine 7 | 8 | """ 9 | from datetime import timedelta, datetime 10 | import pytz 11 | utc8now = datetime.now(tz=pytz.timezone( 12 | 'Asia/Shanghai')).replace(tzinfo=None) 13 | 14 | date_interval = utc8now - \ 15 | timedelta(minutes=config.INTERVAL) 16 | 17 | date_interval = date_interval.replace(tzinfo=None) 18 | 19 | if date_interval > created_at: 20 | continue 21 | 22 | for media in entities['media']: 23 | if media['type'] == 'photo': 24 | photos.append(media['media_url_https'] + ':orig') 25 | 26 | 27 | async def parse_message(): 28 | timeline = await get_formatted_timeline() 29 | 30 | for item in timeline: 31 | author_name = item.author_name 32 | created_at = item.created_at 33 | text = item.text 34 | id = item.id 35 | photos = item.photos 36 | 37 | content = ('author: ' + author_name + '\n' + 38 | 'created_at(utc+8):\n' + str(created_at) + '\n\n' + 39 | text + '\n' + 'id=' + id + '\n' + 40 | ''.join([f'[CQ:image,file={photo}]' for photo in photos]) 41 | ) 42 | 43 | yield content, item 44 | """ 45 | 46 | 47 | def parse_timeline_item(item, only_img): 48 | author_name = item.author.name 49 | # 转为 utc+8 50 | created_at = item.created_at + timedelta(hours=8) 51 | created_at = created_at.replace(tzinfo=None) 52 | text = item.text 53 | id = item.id_str 54 | photos = [] 55 | 56 | has_entities = hasattr(item, 'extended_entities') 57 | 58 | # 如果没有 entities 59 | if not has_entities: 60 | # 且指定只返回包含图片的 twitter，直接结束函数 61 | if only_img: 62 | return 63 | # 否则赋值一个空 entities 64 | else: 65 | entities = {'media': []} 66 | else: 67 | entities = item.extended_entities 68 | 69 | for media in entities['media']: 70 | if media['type'] == 'photo': 71 | photos.append(media['media_url_https']) 72 | 73 | if not photos: 74 | return 75 | 76 | return TimeLine(author_name=author_name, 77 | created_at=created_at, text=text, id=id, photos=photos) 78 | 79 | 80 | class TwitterListCrawler(object): 81 | def __init__(self, api_key, api_key_secret, access_token, access_secret, proxy='') -> None: 82 | self.api_key = api_key 83 | self.api_key_secret = api_key_secret 84 | self.access_token = access_token 85 | self.access_secret = access_secret 86 | self.proxy = proxy 87 | 88 | def get_timeline(self, list_id: str, count=200, pages=1, include_rts=False, only_img=True): 89 | """获取 count 条列表时间线数据，支持翻页""" 90 | if self.proxy: 91 | os.environ['http_proxy'] = self.proxy 92 | os.environ['https_proxy'] = self.proxy 93 | 94 | auth = tweepy.OAuthHandler(self.api_key, self.api_key_secret) 95 | auth.set_access_token(self.access_token, self.access_secret) 96 | 97 | api = tweepy.API(auth) 98 | 99 | for data in tweepy.Cursor(api.list_timeline, list_id=list_id, count=count, include_rts=include_rts, include_entities={'extended_entities': True}).pages(pages): 100 | for item in data: 101 | twitter = parse_timeline_item(item, only_img) 102 | 103 | if twitter: 104 | yield twitter 105 | -------------------------------------------------------------------------------- /app/nico/crawler.py: -------------------------------------------------------------------------------- 1 | # import logging 2 | from typing import List, Optional 3 | 4 | from aiohttp import ClientSession, TCPConnector 5 | from lxml.html import fromstring 6 | 7 | # logging.Logger = 8 | 9 | def get_id(sid: str): 10 | if sid.startswith('im'): 11 | sid = sid[2:] 12 | 13 | return sid 14 | 15 | 16 | class NSCrawler(): 17 | def __init__(self, user_sess: Optional[str] = '', session: Optional[ClientSession] = None, proxy: Optional[str] = None): 18 | self.session = session or ClientSession( 19 | connector=TCPConnector(ssl=False)) 20 | self.proxy = proxy 21 | self.headers = { 22 | # 'Host': 'seiga.nicovideo.jp', 23 | 'User-Agent': ('Mozilla/5.0 (Windows NT 10.0; Win64; x64)' 24 | ' AppleWebKit/537.36 (KHTML, like Gecko) ' 25 | 'Chrome/87.0.4280.141 Safari/537.36'), 26 | # image_search_sort item 指定按投稿时间排序 27 | 'Cookie': f'user_session={user_sess}; sort_search_image_keyword=image_created;image_search_sort=image_created;' 28 | } 29 | 30 | async def __aenter__(self): 31 | return self 32 | 33 | async def close(self): 34 | await self.session.close() 35 | 36 | async def __aexit__(self, ex1, e2, e3): 37 | await self.close() 38 | 39 | async def request(self, url): 40 | async with self.session.get(url, proxy=self.proxy, headers=self.headers) as resp: 41 | if 'content-type' in resp.headers: 42 | ctype = resp.headers['content-type'] 43 | else: 44 | # print(resp.headers) 45 | ctype = '' 46 | 47 | if 'application/json' in ctype: 48 | return await resp.json() 49 | 50 | return await resp.text() 51 | 52 | async def get_one_page(self, tags: str, pn: int) -> List[str]: 53 | """ 54 | :return: 55 | [1, 2, ...] 56 | """ 57 | 58 | url = f'https://seiga.nicovideo.jp/tag/{tags}?page={pn}' 59 | 60 | try: 61 | html = await self.request(url) 62 | es = fromstring(html) 63 | im_xpath = '//li[@class="list_item list_no_trim2"]/a/@href' 64 | 65 | return [item.replace('/seiga/im', '') for item in es.xpath(im_xpath)] 66 | except Exception as e: 67 | # logging.error(f'[NICO] {tags} 请求第 {pn} 页数据出错:', e) 68 | return [] 69 | 70 | async def get_many_pages(self, tags: str, begin: int, end: int): 71 | if end == begin: 72 | end += 1 73 | 74 | for pn in range(begin, end): 75 | im_list = await self.get_one_page(tags, pn) 76 | 77 | if not im_list: 78 | return 79 | 80 | yield im_list 81 | 82 | async def get_info(self, sid: str): 83 | id = get_id(sid) 84 | 85 | api = f'https://sp.seiga.nicovideo.jp/ajax/seiga?id={id}' 86 | 87 | json_data = await self.request(api) 88 | 89 | if 'errors' in json_data: 90 | # logging.error('[NICO]%s 请求图片信息出错: %s', id, json_data['errors']) 91 | return [] 92 | 93 | return json_data['target_image'] 94 | 95 | async def get_source_url(self, sid: str): 96 | id = get_id(sid) 97 | # 如果用户未登录默认使用缩略图 98 | source_url = f'https://lohas.nicoseiga.jp/thumb/{id}i' 99 | api = f'https://seiga.nicovideo.jp/image/source?id={id}' 100 | 101 | async with self.session.get(api, headers=self.headers, proxy=self.proxy, allow_redirects=False) as resp: 102 | if 'location' in resp.headers: 103 | location = resp.headers['location'] 104 | if 'lohas.nicoseiga.jp' in location: 105 | source_url = location.replace('/o/', '/priv/') 106 | 107 | return source_url 108 | 109 | async def get_tag_list(self, sid: str): 110 | id = get_id(sid) 111 | api = f'https://seiga.nicovideo.jp/ajax/illust/tag/list?id={id}' 112 | json_data = await self.request(api) 113 | 114 | if 'errors' in json_data: 115 | # logging.error('[NICO]%s 请求 tag list 失败: %s', id, json_data['errors']) 116 | return [] 117 | 118 | return json_data['tag_list'] 119 | -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | ## Telegram 图片推送 bot 2 | 3 | **支持多种网站推送** 4 | - [Pixiv](https://www.pixiv.net/) 5 | - [Nico Seiga](http://seiga.nicovideo.jp/) 6 | - [Twitter](https://twitter.com/home) 7 | - [Gelbooru](https://gelbooru.com/) 8 | 9 | **其它功能** 10 | * 消息自动重发 11 | * 当图片数量过多时支持图片分段发送 12 | * 判断图片大小以选择不同的消息类型 13 | 14 | ## 开始 15 | 16 | ### 0. 依赖 17 | 18 | 1. 需要 Python3.7+ 19 | 2. 运行 `pip install -r requirements.txt` 安装依赖 20 | 3. 一个能用的 TG 账号 21 | 22 | ### 1. 创建bot 23 | 24 | 1. 前往 [BotFather](https://core.telegram.org/bots#3-how-do-i-create-a-bot) 创建自己的bot 25 | 2. 获取 BotFather 给出的 token 26 | 27 | ### 2. 推送配置 28 | 29 | **获取 chat id** 30 | 31 | 参考：[Get the Telegram channel ID (github.com)](https://gist.github.com/mraaroncruz/e76d19f7d61d59419002db54030ebe35) 32 | 33 | 如果运行 bot 时提示 chat not found，可以试试不加 `-100` 前缀。 34 | 35 | **推送前需要把bot加入到你的频道/群，也可以私聊推送，具体取决于传入的 chat id 是哪一种。** 36 | 37 | 修改项目根目录的全局配置 `vim config.py` 38 | 39 | ```python 40 | # bot token 41 | TOKEN = 'xxxx' 42 | 43 | # 消息发送失败时的重试次数 44 | MESSAGE_MAX_RETRY = 12 45 | 46 | # group/user/channel id 列表 47 | CHAD_ID_LIST = ['1234567'] 48 | 49 | # 代理地址（本地使用需要设置） 50 | PROXY = None # 'http://127.0.0.1:1081' 51 | ``` 52 | 53 | > 推送网站均为可选，可根据实际需求配置自己需要推送的网站 54 | 55 | **Gelbooru 配置** 56 | 57 | `vim app/gelbooru/config.py` 58 | 59 | ```python 60 | # 推送图片的 tags 61 | GELBOORU_TAGS = ['a', 'b'] 62 | 63 | # 爬取页数 64 | GELBOORU_PAGE_NUM = 1 65 | 66 | # 图片质量 67 | # file_url: 原图 68 | # sample_url: 压缩图 69 | GELBOORU_IMAGE_QUALITY = 'file_url' # or sample_url 70 | 71 | # 因为 Gelbooru 限制了 API 请求次数（目测是 24 次每天），所以需要减少请求次数 72 | # 当前时间的分钟数 % REQ_INTERVAL == 0 的时候执行爬虫 73 | # e.g.: REQ_INTERVAL = 20，表示每个小时的 0、20、40 分钟会执行爬虫 74 | # 具体能不能执行取决于 cron 的执行时间 75 | REQ_INTERVAL = 20 76 | 77 | # API KEY（可选；如果被限制请求频率使用 API_KEY 可解除） 78 | # 在 https://gelbooru.com/index.php?page=account&s=options 中 79 | API_KEY = '' 80 | USER_ID = '' 81 | ``` 82 | 83 | **Pixiv 配置** 84 | 85 | `vim app/pixiv/config.py` 86 | 87 | ```python 88 | # 必须配置 PIXIV_TAGS 和 PIXIV_REFRESH_TOKEN，其他配置可默认 89 | 90 | # 推送图片的 tags 91 | PIXIV_TAGS = ['a', 'b'] 92 | 93 | # refresh token 94 | # 获取方法：https://gist.github.com/ZipFile/c9ebedb224406f4f11845ab700124362 95 | PIXIV_REFRESH_TOKEN = 'dF8jpliqu0UlM519k09ICNLQQYnbLL6MBO9XuqDGEBPFAC' 96 | 97 | # 用户白名单，填入 uid，在该名单内的用户的作品会被立即推送（无视分数） 98 | PIXIV_USER_WHITELIST = [] 99 | 100 | # 每个 tag 爬取的最大页数 101 | PIXIV_MAX_PAGE = 1 102 | 103 | # 推送的投稿的最小分数，用于排除低质量的投稿 104 | # 计算公式 view + bookmark * 5 105 | PIXIV_MIN_POST_SCORE = 300 106 | 107 | # 是否推送关注的作者 108 | PIXIV_FOLLOW = False 109 | 110 | # 反代 i.pixiv.re，目前不需要 111 | PIXIV_REVERSE_PROXY = '' 112 | ``` 113 | 114 | **Twitter 配置** 115 | 116 | `vim app/twitter/config.py` 117 | 118 | ```python 119 | # 目标列表Id 120 | # 关于 LIST 参考：https://help.twitter.com/en/using-twitter/twitter-lists 121 | TWITTER_LIST_ID = ['123456789'] 122 | # 爬取页数 123 | TWITTER_PAGE_NUM = 5 124 | # 只推送含图片的动态 125 | TWITTER_ONLY_IMAGE = True 126 | 127 | # KEY 相关，生成 Authentication Tokens 128 | # https://developer.twitter.com/en/portal/dashboard 129 | # 获取方法： 130 | # https://developer.twitter.com/en/docs/authentication/oauth-2-0/bearer-tokens 131 | TWITTER_API_KEY = '' 132 | TWITTER_API_KEY_SECRET = '' 133 | TWITTER_ACCESS_TOKEN = '' 134 | TWITTER_ACCESS_SECRET = '' 135 | ``` 136 | 137 | **N静配置** 138 | 139 | `vim app/nico/config.py` 140 | 141 | ```python 142 | # 要推送的图片 tags 143 | NICO_TAGS = ['a', 'b'] 144 | 145 | # user_sess 同名 cookie，推送原图时需要 146 | NICO_USER_SESS = '' 147 | 148 | # 爬取页数 149 | NICO_PAGE_NUM = 1 150 | 151 | # 用户黑名单，该名单内的用户的作品不会被推送 152 | NICO_USER_BLACKLIST = [] 153 | ``` 154 | 155 | ### 3. 运行 156 | 157 | 1. `python run_crawlers.py` 运行爬虫 158 | 2. `python main.py` 发送消息 159 | 160 | 服务器端可设置 cron 定时运行，只需编辑 `run.sh` 第一行 cd 的目录为项目在服务器的目录，然后 `sh 项目目录/run.sh` 即可。 161 | 162 | **run.sh** 163 | 164 | 提供一个运行脚本，该脚本先执行图片爬取功能，然后执行推送，并且实现脚本单例运行。 165 | 166 | ```bash 167 | # 修改这行为你项目的根目录地址 168 | cd /root/telegram-bot/th-telegram-image-bot 169 | 170 | LOCK="LOCK" 171 | 172 | if [ ! -f "$LOCK" ]; then 173 | touch $LOCK 174 | python run_crawlers.py 175 | sleep 1 176 | python main.py 177 | rm -rf $LOCK 178 | else 179 | ts=`stat -c %Y LOCK` 180 | now=`date +%s` 181 | if [ $[ $now - $ts ] -gt 1800 ]; then 182 | rm -rf $LOCK 183 | echo "Lock expired, deleted" 184 | fi 185 | fi 186 | 187 | ``` 188 | 189 | ## 4. 配置本地 API server（可选） 190 | 191 | 配置 TG 的本地 API server，可以把上传图片的大小提升至 2000MB，因此发送图片大于 10 MB 的消息时不会发送文件形式的图片。 192 | 193 | [官方地址](https://github.com/tdlib/telegram-bot-api) 中提供了安装方法，不过这里推荐用 docker 安装，更加方便且不会污染服务器环境。 194 | 195 | 如果你还没安装 docker，去 [官方文档](https://docs.docker.com/) 找自己系统的安装方式，之后使用 [docker 镜像安装](https://hub.docker.com/r/aiogram/telegram-bot-api)，可以看到容器需要两个环境变量，具体的获取方式在文档中有写，需要注意申请 API Key 的时候关闭 AdBlock 之类的插件，以避免申请失败。 196 | 197 | ## TODOs 198 | 199 | 1. 只发送发送失败的消息 - group message 发送失败视为整体发送失败 200 | 2. 只有一张表 201 | 3. 爬虫提供一个 yield 方法获取来获取数据 202 | 4. 爬虫 yield 的 model 应该符合数据表的 model 203 | 5. 发送成功 204 | 6. 可在爬虫 insert 部分添加时间判断 205 | 206 | 数据库表结构 - 同 qq 推送bot -------------------------------------------------------------------------------- /main.py: -------------------------------------------------------------------------------- 1 | import asyncio 2 | import logging 3 | import os 4 | from typing import List, Optional, Union 5 | import imghdr 6 | 7 | import httpx 8 | from telegram import Bot, InputMediaDocument, InputMediaPhoto, InputFile 9 | from telegram.error import BadRequest, RetryAfter, TimedOut 10 | from tortoise import run_async 11 | from tortoise.expressions import F 12 | 13 | from app.models import ImageDB 14 | from app.nico import config as nico_config 15 | from app.nico.crawler import NSCrawler 16 | from app.pixiv.config import PIXIV_REVERSE_PROXY 17 | from config import * 18 | from run_crawlers import init_db 19 | 20 | 21 | logging.basicConfig(format='%(asctime)s - %(name)s - %(levelname)s - %(message)s', 22 | level=logging.INFO) 23 | 24 | 25 | logger = logging.getLogger(__name__) 26 | logger.setLevel(logging.INFO) 27 | 28 | 29 | # 发现 tg 有提供常量... 30 | FILE_MAX_SIZE = 50 31 | HTTP_FILE_MAXSIZE = 5 32 | DONWLOAD_FILE_MAXSIZE = 10 33 | PER_MESSAGE_MAX_IMAGE_COUNT = 10 34 | 35 | 36 | if PROXY: 37 | os.environ['HTTP_PROXY'] = PROXY 38 | os.environ['HTTPS_PROXY'] = PROXY 39 | 40 | 41 | def get_headers(url: str): 42 | headers = {} 43 | 44 | if url.find('i.pximg.net') != -1: 45 | headers['Referer'] = 'https://app-api.pixiv.net/' 46 | 47 | return headers 48 | 49 | 50 | def get_filesizeMB(url: str): 51 | # 可以添加返回文件类型功能 52 | proxies = PROXY or None 53 | headers = get_headers(url) 54 | 55 | try: 56 | resp = httpx.head(url, headers=headers, proxies=proxies) 57 | except TimeoutError: 58 | logger.error('获取文件大小请求超时') 59 | return -1 60 | 61 | if 'content-length' in resp.headers: 62 | length = int(resp.headers['content-length']) 63 | return round(length / 1024 / 1024, 2) 64 | else: 65 | return -1 66 | 67 | 68 | def get_file_size_type(url: str): 69 | filesize = get_filesizeMB(url) 70 | 71 | if filesize < HTTP_FILE_MAXSIZE: 72 | return 'photo' 73 | elif filesize < DONWLOAD_FILE_MAXSIZE: 74 | return 'download' 75 | elif filesize > HTTP_FILE_MAXSIZE and filesize < FILE_MAX_SIZE: 76 | return 'document' 77 | elif filesize >= FILE_MAX_SIZE: 78 | return 'exceed' 79 | 80 | 81 | def download_media(url: str): 82 | headers = get_headers(url) 83 | 84 | return httpx.get(url, headers=headers, timeout=30).read() 85 | 86 | 87 | def get_media_list(urls: List[str], caption): 88 | """获取 PhotoMediaList 或者 DocumentMediaList 89 | 当所有文件大小 < 5MB 时，全为 PhotoMedia 90 | 当有文件大小 >= 5MB 且 < 50 MB 时，全为 DocumentMedia 91 | 当有文件大小 >= 50MB 时，跳过该文件 92 | """ 93 | media_list = [] 94 | section = 1 95 | document = False 96 | 97 | for idx, url in enumerate(urls): 98 | # 仅一个 caption 或者 11n 个 caption 设置标题 99 | title = None 100 | # idx = 0, true; idx = 10, true, idx = 20, true; ... 101 | # 第 idx + 1 张图片 102 | if idx % PER_MESSAGE_MAX_IMAGE_COUNT == 0: 103 | title = caption 104 | if len(urls) > PER_MESSAGE_MAX_IMAGE_COUNT: 105 | title = title + '\n\n' + f'SECTION: {section}' 106 | section += 1 107 | 108 | ft = get_file_size_type(url) 109 | 110 | if ft == 'exceed': 111 | continue 112 | elif ft == 'document': 113 | document = True 114 | 115 | filename = None 116 | 117 | if document: 118 | filename = url[url.rfind('/') + 1:] 119 | 120 | media = { 121 | 'media': url, 122 | 'caption': title, 123 | 'filename': filename 124 | } 125 | 126 | media_list.append(media) 127 | 128 | media_method = InputMediaDocument if document else InputMediaPhoto 129 | 130 | def List2InputMedia(li): 131 | return media_method(**li) 132 | 133 | media_list = list(map(List2InputMedia, media_list)) 134 | 135 | for idx in range(0, len(media_list), PER_MESSAGE_MAX_IMAGE_COUNT): 136 | yield media_list[idx:idx+PER_MESSAGE_MAX_IMAGE_COUNT] 137 | 138 | 139 | async def do_send_message(bot: Bot, chat_id: str, photos, reply_message_id: int = None, retry=1): 140 | timeout = { 141 | 'read_timeout': 30, 142 | 'write_timeout': 30, 143 | 'connect_timeout': 30, 144 | 'pool_timeout': 30 145 | } 146 | 147 | if retry > 6: 148 | return 149 | 150 | try: 151 | # type: ignore 152 | return await bot.send_media_group(chat_id, photos, reply_to_message_id=reply_message_id, **timeout) 153 | except RetryAfter as e: 154 | after = e.retry_after 155 | logger.info('发送消息过于频繁，将于 %s 秒后进行第 %s 尝试' % (after, retry)) 156 | await asyncio.sleep(after) 157 | return await do_send_message(bot, chat_id, photos, reply_message_id, retry + 1) 158 | except BadRequest as e: 159 | exstr = str(e) 160 | errors = ['wrong file', 'wrong type', 161 | 'photo_invalid_dimensions', 162 | 'failed to get http url content', 163 | 'image_process_failed'] 164 | 165 | logger.info('Bad Request %s', e) 166 | 167 | if list(filter(lambda e: exstr.find(e) != -1, errors)): 168 | downloaded_photos = [] 169 | logger.info('发送图片失败，尝试下载后发送 %s', retry) 170 | 171 | for photo in photos: 172 | obj = photo.media 173 | caption = photo.caption 174 | media_obj = None 175 | filename = None 176 | 177 | if isinstance(photo.media, str): 178 | url = photo.media 179 | obj = download_media(photo.media) 180 | filename = url[url.rfind('/') + 1:] 181 | elif isinstance(photo.media, bytes): 182 | filename = 'wth.' + str(imghdr.what('', photo.media)) 183 | 184 | if isinstance(photo, InputMediaDocument) or exstr.find('photo_invalid_dimensions') != -1: 185 | # filename arg doesn't work 186 | media_obj = InputMediaDocument(obj, caption=caption) 187 | elif isinstance(photo, InputMediaPhoto): 188 | media_obj = InputMediaPhoto(obj, caption=caption) 189 | 190 | if media_obj: 191 | downloaded_photos.append(media_obj) 192 | 193 | return await do_send_message(bot, chat_id, downloaded_photos, reply_message_id, retry + 1) 194 | else: 195 | raise 196 | 197 | 198 | async def send_message(bot: Bot, chat_id, message: str, urls: Optional[List[str]] = None, document=False, download=False): 199 | """发送消息，多块消息只要有一个被发送成功则视整个消息发送成功""" 200 | 201 | if not urls: 202 | await bot.send_message(chat_id, message) 203 | return 204 | 205 | reply_message_id = None 206 | for data in get_media_list(urls, message): 207 | try: 208 | msg_objs = await do_send_message(bot, chat_id, data, reply_message_id) 209 | if msg_objs: 210 | reply_message_id = msg_objs[0].id 211 | except Exception as e: 212 | logger.error('下载失败 %s' % e) 213 | message_with_error = str( 214 | message) + '\n\n发送图片失败: TG 无法处理图片 URL，请点击下面的链接访问原图。\n' + '\n'.join(urls) + '\n\n' + str(e) 215 | 216 | msg_objs = await bot.send_message(chat_id, message_with_error) 217 | 218 | if msg_objs: 219 | reply_message_id = msg_objs.id 220 | 221 | 222 | async def preprocess_message(message: ImageDB) -> List[str]: 223 | """预处理数据库数据 224 | :return: img list""" 225 | if message.original_site == 'nico': 226 | async with NSCrawler(nico_config.NICO_USER_SESS, proxy=PROXY) as crw: 227 | img_list = [await crw.get_source_url(message.original_id)] 228 | elif message.original_site == 'pixiv': 229 | if PIXIV_REVERSE_PROXY: 230 | img_list = list(map(lambda url: url.replace( 231 | 'i.pximg.net', PIXIV_REVERSE_PROXY), message.pic_hash_list)) 232 | else: 233 | img_list = message.pic_hash_list 234 | else: 235 | img_list = message.pic_hash_list 236 | 237 | return img_list 238 | 239 | 240 | async def send_message_and_update_db(bot: Bot, chat_id: str, message: ImageDB): 241 | """包装发送消息方法并更新数据库""" 242 | img_list = await preprocess_message(message) 243 | 244 | # 图片超过 60 张直接退出（避免消息太长出错） 245 | if len(img_list) > 60: 246 | return 247 | 248 | original_site = message.original_site 249 | original_id = message.original_id 250 | orm = ImageDB.filter( 251 | original_site=original_site, original_id=original_id) 252 | 253 | # 发送时标记该消息已被发送 254 | await orm.update(send_successed=True) 255 | 256 | logger.info(f'{original_site} {original_id} 开始发送') 257 | 258 | reason = '' 259 | 260 | try: 261 | await send_message(bot, chat_id, str(message), img_list) 262 | except TimedOut: 263 | reason = 'time out' 264 | 265 | if reason: 266 | await orm.update(retry=F('retry') + 1, send_successed=False, reason=reason) 267 | logger.info('发送失败: %s', reason) 268 | else: 269 | await orm.update(retry=F('retry') + 1, send_successed=True, reason='') 270 | logger.info('发送成功') 271 | 272 | print() 273 | 274 | 275 | async def main(): 276 | logger.info('开始启动推送程序...') 277 | 278 | await init_db() 279 | bot = Bot(TOKEN) 280 | 281 | async with bot: 282 | logger.info('启动完成！') 283 | 284 | async for message in ImageDB.filter(send_successed=False, retry__lt=MESSAGE_MAX_RETRY): 285 | for chat_id in CHAD_ID_LIST: 286 | await send_message_and_update_db(bot, chat_id, message) 287 | 288 | 289 | if __name__ == '__main__': 290 | run_async(main()) 291 | --------------------------------------------------------------------------------