#!/usr/bin/env python # -*- encoding: utf-8 -*- # @Author: https://github.com/Evil0ctal/ # @Time: 2021/11/06 # @Update: 2022/04/05 # @Function: # 核心代码，估值1块(๑•̀ㅂ•́)و✧ # 用于爬取Douyin/TikTok数据并以字典形式返回。 import re import json import time import requests from retrying import retry class Scraper: """ Scraper.douyin():抖音视频/图集解析，返回字典。 Scraper.tiktok():TikTok视频解析，返回字典。 """ def __init__(self): self.headers = { 'user-agent': 'Mozilla/5.0 (Linux; Android 8.0; Pixel 2 Build/OPD3.170816.012) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/87.0.4280.88 Mobile Safari/537.36 Edg/87.0.664.66' } self.tiktok_headers = { "Accept": "text/html,application/xhtml+xml,application/xml;q=0.9,image/avif,image/webp,image/apng,*/*;q=0.8,application/signed-exchange;v=b3;q=0.9", "authority": "www.tiktok.com", "Accept-Encoding": "gzip, deflate", "Connection": "keep-alive", "Host": "www.tiktok.com", "User-Agent": "Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) coc_coc_browser/86.0.170 Chrome/80.0.3987.170 Safari/537.36", } @retry(stop_max_attempt_number=6) def douyin(self, original_url): """ 利用官方接口解析抖音链接信息 :param original_url: 抖音/TikTok链接(支持长/短链接) :return:包含信息的字典 """ headers = self.headers try: # 开始时间 start = time.time() # 原视频链接 r = requests.get(url=original_url, headers=headers, allow_redirects=False) try: # 2021/12/11 发现抖音做了限制，会自动重定向网址，但是可以从回执头中获取 long_url = r.headers['Location'] except: # 报错后判断为长链接，直接截取视频id long_url = original_url # 正则匹配出视频ID key = re.findall('video/(\d+)?', long_url)[0] # 构造抖音API链接 api_url = f'https://www.iesdouyin.com/web/api/v2/aweme/iteminfo/?item_ids={key}' print("正在请求抖音API链接: " + '\n' + api_url) # 将回执以JSON格式处理 js = json.loads(requests.get(url=api_url, headers=headers).text) # 判断是否为图集 if js['item_list'][0]['images'] is not None: print("类型 = 图集") # 类型为图集 url_type = 'album' # 图集标题 album_title = str(js['item_list'][0]['desc']) # 图集作者昵称 album_author = str(js['item_list'][0]['author']['nickname']) # 图集作者签名 album_author_signature = str(js['item_list'][0]['author']['signature']) # 图集作者UID album_author_uid = str(js['item_list'][0]['author']['uid']) # 图集作者抖音号 album_author_id = str(js['item_list'][0]['author']['unique_id']) if album_author_id == "": # 如果作者未修改过抖音号，应使用此值以避免无法获取其抖音ID album_author_id = str(js['item_list'][0]['author']['short_id']) # 图集BGM链接 if len(js['item_list'][0]['music']['play_url']['url_list']) > 0: album_music = str(js['item_list'][0]['music']['play_url']['url_list'][0]) else: # 部分视频的API数据中没有BGM链接，返回None album_music = "None" # 图集BGM标题 album_music_title = str(js['item_list'][0]['music']['title']) # 图集BGM作者 album_music_author = str(js['item_list'][0]['music']['author']) # 图集BGM ID album_music_id = str(js['item_list'][0]['music']['id']) # 图集BGM MID album_music_mid = str(js['item_list'][0]['music']['mid']) # 图集ID album_aweme_id = str(js['item_list'][0]['statistics']['aweme_id']) # 评论数量 album_comment_count = str(js['item_list'][0]['statistics']['comment_count']) # 获赞数量 album_digg_count = str(js['item_list'][0]['statistics']['digg_count']) # 播放次数 album_play_count = str(js['item_list'][0]['statistics']['play_count']) # 分享次数 album_share_count = str(js['item_list'][0]['statistics']['share_count']) # 上传时间戳 album_create_time = str(js['item_list'][0]['create_time']) # 将话题保存在列表中 album_hashtags = [] for tag in js['item_list'][0]['text_extra']: album_hashtags.append(tag['hashtag_name']) # 将无水印图片链接保存在列表中 images_list = [] for data in js['item_list'][0]['images']: images_list.append(data['url_list'][0]) # 结束时间 end = time.time() # 解析时间 analyze_time = format((end - start), '.4f') # 将信息储存在字典中 album_data = {'status': 'success', 'analyze_time': (analyze_time + 's'), 'url_type': url_type, 'platform': 'douyin', 'original_url': original_url, 'api_url': api_url, 'album_aweme_id': album_aweme_id, 'album_title': album_title, 'album_author': album_author, 'album_author_signature': album_author_signature, 'album_author_uid': album_author_uid, 'album_author_id': album_author_id, 'album_music': album_music, 'album_music_title': album_music_title, 'album_music_author': album_music_author, 'album_music_id': album_music_id, 'album_music_mid': album_music_mid, 'album_comment_count': album_comment_count, 'album_digg_count': album_digg_count, 'album_play_count': album_play_count, 'album_share_count': album_share_count, 'album_create_time': album_create_time, 'album_list': images_list, 'album_hashtags': album_hashtags} return album_data else: print("类型 = 视频") # 类型为视频 url_type = 'video' # 视频标题 video_title = str(js['item_list'][0]['desc']) # 视频作者昵称 video_author = str(js['item_list'][0]['author']['nickname']) # 视频作者抖音号 video_author_id = str(js['item_list'][0]['author']['unique_id']) if video_author_id == "": # 如果作者未修改过抖音号，应使用此值以避免无法获取其抖音ID video_author_id = str(js['item_list'][0]['author']['short_id']) # 有水印视频链接 wm_video_url = str(js['item_list'][0]['video']['play_addr']['url_list'][0]) # 无水印视频链接 (在回执JSON中将关键字'playwm'替换为'play'即可获得无水印地址) nwm_video_url = str(js['item_list'][0]['video']['play_addr']['url_list'][0]).replace('playwm', 'play') # 去水印后视频链接(2022年1月1日抖音APi获取到的URL会进行跳转，需要在Location中获取直链) r = requests.get(url=nwm_video_url, headers=headers, allow_redirects=False) video_url = r.headers['Location'] # 视频背景音频 if len(js['item_list'][0]['music']['play_url']['url_list']) > 0: video_music = str(js['item_list'][0]['music']['play_url']['url_list'][0]) else: # 部分视频的API数据中没有BGM链接，返回None video_music = "None" # 视频作者签名 video_author_signature = str(js['item_list'][0]['author']['signature']) # 视频作者UID video_author_uid = str(js['item_list'][0]['author']['uid']) # 视频BGM标题 video_music_title = str(js['item_list'][0]['music']['title']) # 视频BGM作者 video_music_author = str(js['item_list'][0]['music']['author']) # 视频BGM ID video_music_id = str(js['item_list'][0]['music']['id']) # 视频BGM MID video_music_mid = str(js['item_list'][0]['music']['mid']) # 视频ID video_aweme_id = str(js['item_list'][0]['statistics']['aweme_id']) # 评论数量 video_comment_count = str(js['item_list'][0]['statistics']['comment_count']) # 获赞数量 video_digg_count = str(js['item_list'][0]['statistics']['digg_count']) # 播放次数 video_play_count = str(js['item_list'][0]['statistics']['play_count']) # 分享次数 video_share_count = str(js['item_list'][0]['statistics']['share_count']) # 上传时间戳 video_create_time = str(js['item_list'][0]['create_time']) # 将话题保存在列表中 video_hashtags = [] for tag in js['item_list'][0]['text_extra']: video_hashtags.append(tag['hashtag_name']) # 结束时间 end = time.time() # 解析时间 analyze_time = format((end - start), '.4f') # 返回包含数据的字典 video_data = {'status': 'success', 'analyze_time': (analyze_time + 's'), 'url_type': url_type, 'platform': 'douyin', 'original_url': original_url, 'api_url': api_url, 'video_title': video_title, 'nwm_video_url': video_url, 'wm_video_url': wm_video_url, 'video_aweme_id': video_aweme_id, 'video_author': video_author, 'video_author_signature': video_author_signature, 'video_author_uid': video_author_uid, 'video_author_id': video_author_id, 'video_music': video_music, 'video_music_title': video_music_title, 'video_music_author': video_music_author, 'video_music_id': video_music_id, 'video_music_mid': video_music_mid, 'video_comment_count': video_comment_count, 'video_digg_count': video_digg_count, 'video_play_count': video_play_count, 'video_share_count': video_share_count, 'video_create_time': video_create_time, 'video_hashtags': video_hashtags} return video_data except Exception as e: # 返回异常 return {'status': 'failed', 'reason': e, 'function': 'Scraper.douyin()', 'value': original_url} @retry(stop_max_attempt_number=6) def tiktok(self, original_url): """ 解析TikTok链接 :param original_url:TikTok链接 :return:包含信息的字典 """ headers = self.headers # 开始时间 start = time.time() # 校验TikTok链接 if original_url[:12] == "https://www.": original_url = original_url print("目标链接: ", original_url) else: # 从请求头中获取原始链接 response = requests.get(url=original_url, headers=headers, allow_redirects=False) true_link = response.headers['Location'].split("?")[0] original_url = true_link # TikTok请求头返回的第二种链接类型 if '.html' in true_link: response = requests.get(url=true_link, headers=headers, allow_redirects=False) original_url = response.headers['Location'].split("?")[0] print("目标链接: ", original_url) try: # 开始获取TikTok数据 tiktok_headers = self.tiktok_headers html = requests.get(url=original_url, headers=tiktok_headers) res = re.search('