single_video_crawler/single_video/douyin/douyin_video.py

# -*- coding: utf-8 -*-
# @Time: 2023/12/22
import datetime
import os
import random
import re
import sys
import requests
import json
import urllib3



sys.path.append(os.getcwd())
from requests.adapters import HTTPAdapter
from single_video.douyin.douyin_help import DouYinHelper
from common.url_manage import urlManage
from common import Oss, Common
from common.pq_utility import PQ
from common.userAgent import get_random_user_agent

class douyinVideo():



    # 获取抖音 标题+视频链接
    @classmethod
    def get_videoList(cls, vx_message, channel):
        data_link = vx_message[1]
        content_id = ''
        for i in range(3):
            content_id = urlManage.url_manage(data_link, channel)
            if content_id:
                break
        if content_id == '':
            return "无法获取到视频ID"
        Common.logger().info(f'渠道:{channel},视频ID:{content_id}')
        for i in range(3):
            try:
                url = f'https://www.douyin.com/aweme/v1/web/aweme/detail/'
                headers = {
                    'Accept': 'application/json, text/plain, */*',
                    'Accept-Language': 'zh-CN,zh;q=0.9',
                    'Cache-Control': 'no-cache',
                    'Cookie' : DouYinHelper.get_cookie(),
                    'Pragma': 'no-cache',
                    'Referer': f'https://www.douyin.com/video/{content_id}',
                    'User-Agent': get_random_user_agent()
                }
                query = DouYinHelper.get_full_query(ua=headers['User-Agent'], extra_data={'aweme_id': content_id})
                urllib3.disable_warnings()
                s = requests.session()
                s.mount('http://', HTTPAdapter(max_retries=3))
                s.mount('https://', HTTPAdapter(max_retries=3))
                response = requests.request(method='GET', url=url, headers=headers, params=query)
                status_code = response.status_code
                if status_code != 200:
                    return "cookie过期"
                body = response.content.decode()
                obj = json.loads(body).get('aweme_detail', {})
                if len(obj) == 0:
                    return "cookie过期"
                if obj.get('desc'):
                    title = obj.get('desc').split('\n')[0].strip()
                    title = f'{title[:20]}...' if len(title) > 20 else title
                    title = title if title else ' '
                else:
                    title = ' '
                if obj.get('video', {}).get('bit_rate', []):
                    video_uri = obj.get('video', {}).get('play_addr', {}).get('uri')
                    ratio = f'{obj.get("video", {}).get("height")}p'
                    video_url = f'https://www.iesdouyin.com/aweme/v1/play/?video_id={video_uri}&ratio={ratio}&line=0'
                    # 随机生成视频oss_id
                    video_id = urlManage.random_id()
                    oss_object_key = Oss.video_url_upload_oss(video_url, video_id)
                    oss_object_key = oss_object_key.get("oss_object_key")
                    Common.logger().info(f'准备发送站内参数:{oss_object_key},{title},{vx_message[3]}')
                    piaoquantv = PQ.insert_piaoquantv(oss_object_key, title, vx_message[3])
                    if piaoquantv == False:
                        return '视频发送到站内失败'
                else:
                    return "无法获取到视频链接"
            except Exception as e:
                Common.logger().info(f'报错信息:{e}')


if __name__ == '__main__':
    douyinVideo.get_videoList()