|
| 1 | +# -*- encoding: utf-8 -*- |
| 2 | +# Auth: Zok Email: [email protected] |
| 3 | +# Date: 2020/3/6 |
| 4 | + |
| 5 | +import re |
| 6 | +import requests |
| 7 | +import json |
| 8 | + |
| 9 | + |
| 10 | +class ParseVideo: |
| 11 | + |
| 12 | + def __init__(self, share): |
| 13 | + path = self.get_url(share) |
| 14 | + self.url = 'https://v.douyin.com/' + path + '/' |
| 15 | + self.headers = { |
| 16 | + 'user-agent': 'Mozilla/5.0 (iPhone; CPU iPhone OS 11_0 like Mac OS X) AppleWebKit/604.1.38 (KHTML, like Gecko) Version/11.0 Mobile/15A372 Safari/604.1', |
| 17 | + } |
| 18 | + self.session = requests.session() |
| 19 | + self.first_url = None |
| 20 | + |
| 21 | + @staticmethod |
| 22 | + def get_url(share_url): |
| 23 | + return re.search(r'https://v\.douyin\.com/(.*?)/', share_url).group(1) |
| 24 | + |
| 25 | + def go_location(self): |
| 26 | + response = self.session.get(self.url, headers=self.headers) |
| 27 | + self.first_url = response.url |
| 28 | + result = re.search(r'itemId: "(.*?)",[\s\S]*?uid: "(.*?)",[\s\S]*?authorName: "(.*?)",[\s\S]*?dytk: "(.*?)"', |
| 29 | + response.text) |
| 30 | + return result |
| 31 | + |
| 32 | + def go_message(self, ret): |
| 33 | + url = 'https://www.iesdouyin.com/web/api/v2/aweme/iteminfo/?item_ids=' + ret.group(1) + '&dytk=' + ret.group(4) |
| 34 | + response = self.session.get(url, headers=self.headers) |
| 35 | + json_data = json.loads(response.text) |
| 36 | + user_id = ret.group(2) |
| 37 | + user_name = ret.group(3).encode('utf-8').decode('unicode_escape') |
| 38 | + |
| 39 | + if json_data.get('status_code') != 0: |
| 40 | + print('解析失败') |
| 41 | + exit() |
| 42 | + item_list = json_data.get('item_list')[0] |
| 43 | + aweme_id = item_list.get('aweme_id') |
| 44 | + desc = item_list.get('desc') |
| 45 | + comment_count = item_list.get('statistics').get('comment_count') |
| 46 | + digg_count = item_list.get('statistics').get('digg_count') |
| 47 | + |
| 48 | + video = item_list.get('video') |
| 49 | + cover = video.get('origin_cover').get('url_list')[0] |
| 50 | + play_addr = video.get('play_addr_lowbr').get('url_list')[0] |
| 51 | + |
| 52 | + play_addr_response = self.session.get(play_addr, headers=self.headers, allow_redirects=False) |
| 53 | + msg = """ |
| 54 | + 用户id：{user_id} |
| 55 | + 用户名：{user_name} |
| 56 | + 作品id：{aweme_id} |
| 57 | + 标题： {desc} |
| 58 | + 评论数： {comment_count} |
| 59 | + 点赞数： {digg_count} |
| 60 | + 封面地址：{cover} |
| 61 | + 无水印视频：{addr} |
| 62 | + """.format( |
| 63 | + user_id=user_id, |
| 64 | + user_name=user_name, |
| 65 | + aweme_id=aweme_id, |
| 66 | + desc=desc, |
| 67 | + comment_count=comment_count, |
| 68 | + digg_count=digg_count, |
| 69 | + cover=cover, |
| 70 | + addr=play_addr_response.headers['location'] |
| 71 | + ) |
| 72 | + print(msg) |
| 73 | + |
| 74 | + def start(self): |
| 75 | + result = self.go_location() |
| 76 | + self.go_message(result) |
| 77 | + |
| 78 | + |
| 79 | +if __name__ == '__main__': |
| 80 | + # text = '#在抖音，记录美好生活#要逆天！北京地坛医院证实新冠病毒攻击中枢神经系统 https://v.douyin.com/tW7qrw/ 复制此链接，打开【抖音短视频】，直接观看视频！' |
| 81 | + text = input('请输入分享链接>>>') |
| 82 | + pv = ParseVideo(text) |
| 83 | + pv.start() |
0 commit comments