DY个人视频列表demo.7z

import base64 import os import random import json import re import time import datetime import execjs import requests from retrying import retry from fake_useragent import UserAgent from urllib.parse import urlencode from loguru import logger # 忽略requests证书警告 from requests.packages.urllib3.exceptions import InsecureRequestWarning requests.packages.urllib3.disable_warnings(InsecureRequestWarning) PATH = os.path.dirname(os.path.abspath(__file__)) # 获取绝对路径 from dy_slide_verify import Verify COUNT = 1 def run_time(func): def new_func(*args, **kwargs): start_time = datetime.datetime.now() logger.info("程序开始时间：{}".format(start_time)) res = func(*args, **kwargs) end_time = datetime.datetime.now() logger.info("程序结束时间：{}".format(end_time)) logger.info("程序执行用时：{}s".format((end_time - start_time))) return res return new_func class sessions(requests.Session): def request(self, *args, **kwargs): kwargs.setdefault('timeout', 30) kwargs.setdefault('cookies', '') return super(sessions, self).request(*args, **kwargs) class DouYinComment(object): def __init__(self): logger.add("file_X.log", retention="1 days") self.session = sessions() @retry(stop_max_attempt_number=3) def _parse_url(self, url, headers=None, params=None): """url请求""" count = 1 while count < 20: try: if params: response = requests.get(url, headers=headers, params=params, verify=False, allow_redirects=False, timeout=20) else: response = requests.get(url, headers=headers, verify=False, allow_redirects=False, timeout=20) if count % 5 == 0: self.get_cookie() if not response.text: count += 1 logger.info('数据获取失败,重试...') time.sleep(1) continue return response except Exception as e: count += 1 continue def get_passport_csrf_token(self): headers = { "authority": "sso.douyin.com", "accept": "application/json, text/plain, */*", "accept-language": "zh-CN,zh;q=0.9", "cache-control": "no-cache", "origin": "https://www.douyin.com", "pragma": "no-cache", "referer": "https://www.douyin.com/", "sec-ch-ua": "^\\^Chromium^^;v=^\\^110^^, ^\\^Not", "sec-ch-ua-mobile": "?0", "sec-ch-ua-platform": "^\\^Windows^^", "sec-fetch-dest": "empty", "sec-fetch-mode": "cors", "sec-fetch-site": "same-site", "user-agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/110.0.0.0 Safari/537.36" } url = "https://sso.douyin.com/check_login/" params = { "service": "https:^%^2F^%^2Fwww.douyin.com", "aid": "6383", "account_sdk_source": "sso", "language": "zh", "sdk_version": "2.1.3" } response = self.session.get(url, headers=headers, params=params) Set_Cookie = response.headers.get('Set-Cookie') passport_csrf_token_default = re.findall(r'passport_csrf_token_default=(.*?);', Set_Cookie, re.S | re.M) if not passport_csrf_token_default: logger.error('token: 提取失败') raise self.cookie += f'passport_csrf_token={passport_csrf_token_default[0]};' @retry(stop_max_attempt_number=3) def get_cookie(self): count = 10 while count: try: session = requests.session() url = 'https://www.douyin.com/aweme/v1/web/aweme/post/?device_platform=webapp&aid=6383&channel=channel_pc_web&sec_user_id=MS4wLjABAAAA_py8TGmFe6t8KDY04LU0JH9Yr9ml54dCjRFi0mc1lwI&max_cursor=1644497505000&locate_item_id=7064150578586193188&locate_query=false&show_live_replay_strategy=1&count=10&publish_video_strategy_type=2&pc_client_type=1&version_code=170400&version_name=17.4.0&cookie_enabled=true&screen_width=1680&screen_height=1050&browser_language=zh-CN&browser_platform=Win32&browser_name=Chrome&browser_version=108.0.0.0&browser_online=true&engine_name=Blink&engine_version=108.0.0.0&os_name=Windows&os_version=10&cpu_core_num=4&device_memory=8&platform=PC&downlink=10&effective_type=4g&round_trip_time=100&webid=7180319539912607284&msToken=mgDDRa5pg4ZXS8pPtZSYMAQSePr3lEy0rZ_jCzpOcY2c7opOKqNwSSR1cz991481ddf0AQAl2S4aTTBs19-O4VwM9K3JfMVB6-Q-tn2k2LqJVlgOl-2sdA==&X-Bogus=DFSzswVYA9sANryXSkh5PRXAIQR0' h = session.get(url, verify=False, allow_redirects=False, timeout=20).headers x_vc_bdturing_parameters = h.get('x-vc-bdturing-parameters') if not x_vc_bdturing_parameters: count -= 1 logger.info(f'提取：x_vc_bdturing_parameters 失败，重试！') time.sleep(random.randint(3, 5)) continue verify_data = json.loads(base64.b64decode(h.get('x-vc-bdturing-parameters')).decode("utf-8")) fp = verify_data.get("fp") detail = verify_data.get("detail") logger.info(f"成功提取：{fp}, 开始验证") try: # 有几率报错，报错重试 msg = Verify().verify(fp, detail) logger.info(msg) except Exception as e: logger.info(f"{e}") continue if msg.get('code') != 200: logger.info(f"{msg.get('message')}，重试") continue logger.info(f"ck s_v_web_id：{fp}, {msg.get('message')}") s_v_web_id = f's_v_web_id={verify_data.get("fp")};' self.cookie = s_v_web_id return except Exception as e: logger.info(f'提取：x_vc_bdturing_parameters 出错：{e}') time.sleep(random.randint(3, 5)) continue input('没有提取到fp') def _get_xb(self, params): """ 计算X-Bogus """ # 读取xb算法js with open('x-b.js', encoding='utf-8') as f: js = f.read() ctx = execjs.compile(js) pathanme = urlencode(params) res = ctx.call('Xb', pathanme) return res def get_aweme_list(self, sec_user_id): """ 根据作者ID获取所有视频ID """ headers = { "authority": "www.douyin.com", "accept": "application/json, text/plain, */*", "accept-language": "zh-CN,zh;q=0.9", "bd-ticket-guard-client-csr": "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", "bd-ticket-guard-version": "2", "cache-control": "no-cache", "pragma": "no-cache", "referer": "https://www.douyin.com/", "sec-ch-ua": "^\\^Chromium^^;v=^\\^110^^, ^\\^Not",