piaoquan_crawler/douyin/douyin_recommend/recommend_dy.py

# -*- coding: utf-8 -*-
# @Author: lierqiang
# @Time: 2023/4/06
import json
import os
import random
import shutil
import sys
import time
import requests
from hashlib import md5

from douyin.douyin_recommend import get_xb

sys.path.append(os.getcwd())
from common.common import Common
from common.feishu import Feishu
from common.db import MysqlHelper
from common.publish import Publish
from common.public import get_config_from_mysql
from common.public import random_title
from common.userAgent import get_random_user_agent


class DyRecommend(object):
    # 个人主页视频翻页参数
    # offset = 0
    platform = "抖音"

    # tag = "西瓜视频爬虫,定向爬虫策略"

    @classmethod
    def get_rule(cls, log_type, crawler):
        try:
            while True:
                rule_sheet = Feishu.get_values_batch(log_type, crawler, "a6L9Kb")
                if rule_sheet is None:
                    Common.logger(log_type, crawler).warning("rule_sheet is None! 10秒后重新获取")
                    time.sleep(10)
                    continue
                rule_dict = {
                    # "play_cnt": int(rule_sheet[0][2]),
                    # "comment_cnt": int(rule_sheet[1][2]),
                    "video_width": int(rule_sheet[0][2]),
                    "video_height": int(rule_sheet[1][2]),
                    "like_cnt": int(rule_sheet[2][2]),
                    "duration": int(rule_sheet[3][2]),
                    "publish_time": int(rule_sheet[4][2]),

                }
                return rule_dict
        except Exception as e:
            Common.logger(log_type, crawler).error(f"get_rule:{e}\n")

    # 下载规则
    @classmethod
    def download_rule(cls, video_info_dict, rule_dict):
        # if video_info_dict['play_cnt'] >= rule_dict['play_cnt']:
        #     if video_info_dict['comment_cnt'] >= rule_dict['comment_cnt']:
        if video_info_dict['like_cnt'] >= rule_dict['like_cnt']:
            if video_info_dict['duration'] >= rule_dict['duration']:
                if video_info_dict['video_width'] >= rule_dict['video_width'] \
                        or video_info_dict['video_height'] >= rule_dict['video_height']:
                    return True
                else:
                    return False
            else:
                return False
        else:
            return False

    # else:
    #     return False
    # else:
    #     return False

    # 过滤词库
    @classmethod
    def filter_words(cls, log_type, crawler):
        try:
            while True:
                filter_words_sheet = Feishu.get_values_batch(log_type, crawler, '6BS2RR')
                if filter_words_sheet is None:
                    Common.logger(log_type, crawler).warning(f"filter_words_sheet:{filter_words_sheet} 10秒钟后重试")
                    continue
                filter_words_list = []
                for x in filter_words_sheet:
                    for y in x:
                        if y is None:
                            pass
                        else:
                            filter_words_list.append(y)
                return filter_words_list
        except Exception as e:
            Common.logger(log_type, crawler).error(f'filter_words异常:{e}\n')

    @classmethod
    def video_title(cls, log_type, crawler, env, title):
        title_split1 = title.split(" #")
        if title_split1[0] != "":
            title1 = title_split1[0]
        else:
            title1 = title_split1[-1]

        title_split2 = title1.split(" #")
        if title_split2[0] != "":
            title2 = title_split2[0]
        else:
            title2 = title_split2[-1]

        title_split3 = title2.split("@")
        if title_split3[0] != "":
            title3 = title_split3[0]
        else:
            title3 = title_split3[-1]

        video_title = title3.strip().replace("\n", "") \
                          .replace("/", "").replace("抖音", "").replace(" ", "") \
                          .replace(" ", "").replace("&NBSP", "").replace("\r", "") \
                          .replace("#", "").replace(".", "。").replace("\\", "") \
                          .replace(":", "").replace("*", "").replace("？", "") \
                          .replace("?", "").replace('"', "").replace("<", "") \
                          .replace(">", "").replace("|", "").replace("@", "").replace('"', '').replace("'", '')[:40]
        if video_title.replace(" ", "") == "" or video_title == "。。。" or video_title == "...":
            return random_title(log_type, crawler, env, text='title')
        else:
            return video_title

    @classmethod
    def get_videolist(cls, log_type, crawler, strategy, our_id, oss_endpoint, env, machine):
        rule_dict = cls.get_rule(log_type, crawler)
        for page in range(1, 101):
            aweme_pc_rec_raw_data = '%7B%22videoPrefer%22:%7B%22fsn%22:%5B%5D,%22like%22:%5B%5D,%22halfMin%22:%5B%227188684310696742200%22,%224380080926896941%22%5D,%22min%22:%5B%5D%7D,%22seo_info%22:%22https:%2F%2Fwww.douyin.com%2F%22,%22is_client%22:false,%22ff_danmaku_status%22:1,%22danmaku_switch_status%22:1%7D'
            f_url = 'https://www.douyin.com/aweme/v1/web/tab/feed/?device_platform=webapp&aid=6383&channel=channel_pc_web&tag_id=&share_aweme_id=&count=10&refresh_index={page}&video_type_select=1&aweme_pc_rec_raw_data={aweme_pc_rec_raw_data}&globalwid=&pull_type=2&min_window=0&ug_source=&creative_id=&pc_client_type=1&version_code=170400&version_name=17.4.0&cookie_enabled=true&screen_width=1440&screen_height=900&browser_language=zh-CN&browser_platform=MacIntel&browser_name=Chrome&browser_version=109.0.0.0&browser_online=true&engine_name=Blink&engine_version=109.0.0.0&os_name=Mac+OS&os_version=10.15.7&cpu_core_num=8&device_memory=8&platform=PC&downlink=10&effective_type=4g&round_trip_time=50&webid=7219223873342260736&msToken=Sh6bVLWZUEZ3ruIHq1L3iUXnr1GT5yklyo-XZRO7lNgsFvpYq0C7tcu5Z4Jv9DrMESZ9kGVhwKT4ftFDkBL11ZGPUxth2ToA4M4q-qs2MK9ctR7GhwFkGw=='.format(
                page=page, aweme_pc_rec_raw_data=aweme_pc_rec_raw_data)
            headers = {
                # 'cookie': 'ttwid=1%7CzpiG_VTvd1xpRFvfHUjEHiaq3qkfUqPElZUu0wbTSr8%7C1680507728%7Cc61697d37b4d4d49d42b466a3bbe8ecd5c06ae6e9a751d9e410102d2c52a185d; douyin.com; passport_csrf_token=208a829b0156a2feaa0fa24ad026ea91; passport_csrf_token_default=208a829b0156a2feaa0fa24ad026ea91; s_v_web_id=verify_lg0iwv1g_BwfztkmU_azbL_4Gua_9Fb9_KWfGPVXCyWua; bd_ticket_guard_client_data=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; csrf_session_id=a5355d954d3c63ed1ba35faada452b4d; ttcid=60304e8eb309434784f02372ef36387b41; xgplayer_user_id=446878319428; __ac_nonce=0642bcb0a001596f3fe5b; __ac_signature=_02B4Z6wo00f012HU-wAAAIDA9QKgDGYSlVNh9P-AALxqUQdGOEO.l3IAhdmUh4D-Y9rXLut3p7moXUuAUmo7rUOUJzpnB9nLx0YdZcvdMNeUgQOjsGIHh9LTN38BOVtrElZBeXDLjuVVC5Hh81; strategyABtestKey=%221680591628.189%22; download_guide=%223%2F20230404%22; VIDEO_FILTER_MEMO_SELECT=%7B%22expireTime%22%3A1681196521206%2C%22type%22%3A0%7D; home_can_add_dy_2_desktop=%221%22; msToken=v_pzGLfpXwl4PugynDwIb5DeepUms68tZLZFNLHl8WQnEeNQZtaawWYVu4Y3TLWxpqbvgqkOFULGmld2BLBZydbzrMJgkx5q5GjqetkVI4GoxLX1QdJQ0CP607uEVw==; tt_scid=NZGTg99heu5lHFfAvBht7p3Qxl0TGP.TyfxOQ7cWIvZjEnOcZERFaJxQ.HnKY-UT18cb; msToken=JFY-VD2YHS-6IO6lNhTj4AcLVtHqjein_5giYzIUh_VRsXMPFXy9QOg-RKDDwQgW5TTbgQB_BLzpfQhNdNEQCv5sGXatzGei9yppG1eSLLkbI9fjhTdBWtdkAJpLIg==',
                'cookie': 'ttwid=1%7CI2Xp275XabSiVJ9GAmfLtqbtqturVSIS2yLbXVkAHnQ%7C1680856567%7Cfd36579475157e2303e36e5fd75cdea4ebad78c20da989be0590305f169242ca; douyin.com; strategyABtestKey=%221680856567.817%22; passport_csrf_token=6a28a7b1e1ad38570cc5ee39deaf587a; passport_csrf_token_default=6a28a7b1e1ad38570cc5ee39deaf587a; s_v_web_id=verify_lg6aloex_WG2LmlFk_kbA6_4tMb_BZlD_iuKxRvsMvrg7; bd_ticket_guard_client_data=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; csrf_session_id=a5355d954d3c63ed1ba35faada452b4d; ttcid=585ca9110e2345a09b4499dc543ec39959; odin_tt=d5fbc530c34bf8d4ea8cf8030c85a3f70202ee34dad7fb5b9ef7ff299d20e38ed01c852f02f2f3f864f2a45060480717f72083b58cdbc204c988edbf997fda7c; xgplayer_user_id=64984931555; SEARCH_RESULT_LIST_TYPE=%22single%22; pwa2=%222%7C1%22; download_guide=%223%2F20230407%22; __ac_nonce=0642fff310028297fec16; __ac_signature=_02B4Z6wo00f01M3AeJAAAIDDWRYjnPVAtVzN4HwAAFdlurQp2aR1Npvb7RYeaCY4fZs3DkMFlu7-Obn7zsvc34whBQesvTIc2p8nV-1crQtTacaxSYqP8nwNk3WqH.tkNdKQaUMw6sCC2.xZ15; VIDEO_FILTER_MEMO_SELECT=%7B%22expireTime%22%3A1681472712797%2C%22type%22%3A1%7D; home_can_add_dy_2_desktop=%221%22; tt_scid=PNmY2BL-Q9E9lAfdecenRGOzq64XgFOc0CJGFMN.JIE-QJO51S3Zvw56-Z6O12QVf4d9; msToken=uFCSX87jL9sTq6ScVYJucYgv9Hd5gCbTPvKIGBRMgVLuo7Pp9zLRrutBYzq4BmnCr83WnJAwZb8H78lBr3s3eyJLnySxYO5FgClQRXW1i_mAu7fLfBj3gA==; msToken=Sh6bVLWZUEZ3ruIHq1L3iUXnr1GT5yklyo-XZRO7lNgsFvpYq0C7tcu5Z4Jv9DrMESZ9kGVhwKT4ftFDkBL11ZGPUxth2ToA4M4q-qs2MK9ctR7GhwFkGw==',
                'referer': 'https://www.douyin.com/',
                'user-agent': get_random_user_agent('pc')
            }
            try:
                x_bogus = get_xb(f_url, headers['user-agent'])
                if not x_bogus:
                    continue
                url = f_url + '&X-Bogus={}'.format(x_bogus)
                res = requests.get(url=url, headers=headers, proxies=Common.tunnel_proxies()).json()
                aweme_list = res.get('aweme_list', [])
            except Exception as e:
                Common.logger(log_type, crawler).error(f"获取抖音推荐失败:{e}")
                continue
            if not aweme_list:
                Common.logger(log_type, crawler).warning(f"抖音推荐没有获取到更多数据,页数:{page}")
                continue
            for info in aweme_list:
                if info.get('is_ads'):
                    continue
                publish_time = info.get('create_time')
                if not publish_time:
                    continue
                publish_time_str = time.strftime("%Y-%m-%d %H:%M:%S", time.localtime(publish_time))
                publish_day = int((int(time.time()) - publish_time) / (3600 * 24))
                if not info['desc']:
                    video_title = random_title(log_type, crawler, env, text='title')
                else:
                    video_title = cls.video_title(log_type, crawler, env, info['desc'])

                video_dict = {'video_title': video_title,
                              'video_id': info['aweme_id'],
                              'play_cnt': info['statistics']['play_count'],
                              'comment_cnt': info['statistics']['comment_count'],
                              'like_cnt': info['statistics']['digg_count'],
                              'share_cnt': info['statistics']['share_count'],
                              'video_width': info['video']['width'],
                              'video_height': info['video']['height'],
                              'duration': round(info['video']['duration'] / 1000),
                              'publish_time': publish_day,
                              'publish_time_stamp': publish_time * 1000,
                              'publish_time_str': publish_time_str,
                              'user_name': info['author']['nickname'],
                              'user_id': info['author_user_id'],
                              'user_sec_id': info['author']['sec_uid'],
                              'avatar_url': info['author']['avatar_thumb']['url_list'][0],
                              'cover_url': info['video']['origin_cover']['url_list'][0].replace('\u0026', '&'),
                              'video_url': info['video']['play_addr']['url_list'][0],
                              'session': f"douyin{int(time.time())}"
                              }

                for k, v in video_dict.items():
                    Common.logger(log_type, crawler).info(f"{k}:{v}")
                cls.download_publish(log_type=log_type,
                                     crawler=crawler,
                                     video_dict=video_dict,
                                     rule_dict=rule_dict,
                                     strategy=strategy,
                                     our_uid=our_id,
                                     oss_endpoint=oss_endpoint,
                                     env=env,
                                     machine=machine)

    @classmethod
    def repeat_video(cls, log_type, crawler, video_id, env, machine):
        sql = f""" select * from crawler_video where platform="{cls.platform}" and out_video_id="{video_id}"; """
        repeat_video = MysqlHelper.get_values(log_type, crawler, sql, env, machine)
        return len(repeat_video)

    # 下载 / 上传
    @classmethod
    def download_publish(cls, log_type, crawler, strategy, video_dict, rule_dict, our_uid, oss_endpoint, env, machine):
        try:
            filter_words = get_config_from_mysql(log_type, crawler, env, text='filter')
            for filter_word in filter_words:
                if filter_word in video_dict['video_title']:
                    Common.logger(log_type, crawler).info('标题已中过滤词:{}\n', video_dict['video_title'])
                    return
            if cls.download_rule(video_dict, rule_dict) is False:
                Common.logger(log_type, crawler).info('不满足抓取规则\n')
            elif any(word if word in video_dict['video_title'] else False for word in
                     cls.filter_words(log_type, crawler)) is True:
                Common.logger(log_type, crawler).info('标题已中过滤词:{}\n', video_dict['video_title'])
            elif cls.repeat_video(log_type, crawler, video_dict['video_id'], env, machine) != 0:
                Common.logger(log_type, crawler).info('视频已下载\n')
            else:
                # 下载视频
                Common.download_method(log_type=log_type, crawler=crawler, text='video',
                                       title=video_dict['video_title'], url=video_dict['video_url'])
                md_title = md5(video_dict['video_title'].encode('utf8')).hexdigest()
                if os.path.getsize(f"./{crawler}/videos/{md_title}/video.mp4") == 0:
                    # 删除视频文件夹
                    shutil.rmtree(f"./{crawler}/videos/{md_title}")
                    Common.logger(log_type, crawler).info("视频size=0，删除成功\n")
                    return
                # ffmpeg_dict = Common.ffmpeg(log_type, crawler, f"./{crawler}/videos/{video_dict['video_title']}/video.mp4")
                # if ffmpeg_dict is None or ffmpeg_dict['size'] == 0:
                #     Common.logger(log_type, crawler).warning(f"下载的视频无效，已删除\n")
                #     # 删除视频文件夹
                #     shutil.rmtree(f"./{crawler}/videos/{video_dict['video_title']}")
                #     return
                # 下载封面
                Common.download_method(log_type=log_type, crawler=crawler, text='cover',
                                       title=video_dict['video_title'], url=video_dict['cover_url'])
                # 保存视频信息至txt
                Common.save_video_info(log_type=log_type, crawler=crawler, video_dict=video_dict)

                # 上传视频
                Common.logger(log_type, crawler).info("开始上传视频...")
                our_video_id = Publish.upload_and_publish(log_type=log_type,
                                                          crawler=crawler,
                                                          strategy=strategy,
                                                          our_uid=our_uid,
                                                          env=env,
                                                          oss_endpoint=oss_endpoint)
                if env == 'dev':
                    our_video_link = f"https://testadmin.piaoquantv.com/cms/post-detail/{our_video_id}/info"
                else:
                    our_video_link = f"https://admin.piaoquantv.com/cms/post-detail/{our_video_id}/info"
                Common.logger(log_type, crawler).info("视频上传完成")

                if our_video_id is None:
                    # 删除视频文件夹
                    shutil.rmtree(f"./{crawler}/videos/{video_dict['video_title']}")
                    return

                # 视频写入飞书
                Feishu.insert_columns(log_type, 'douyin', "82c8d9", "ROWS", 1, 2)
                upload_time = int(time.time())
                values = [[
                    our_video_id,
                    time.strftime("%Y-%m-%d %H:%M:%S", time.localtime(upload_time)),
                    "推荐",
                    str(video_dict['video_id']),
                    video_dict['video_title'],
                    our_video_link,
                    # video_dict['gid'],
                    video_dict['play_cnt'],
                    video_dict['comment_cnt'],
                    video_dict['like_cnt'],
                    video_dict['share_cnt'],
                    video_dict['duration'],
                    str(video_dict['video_width']) + '*' + str(video_dict['video_height']),
                    video_dict['publish_time_str'],
                    video_dict['user_name'],
                    video_dict['user_id'],
                    video_dict['avatar_url'],
                    video_dict['cover_url'],
                    video_dict['video_url']
                ]]
                # time.sleep(1)
                Feishu.update_values(log_type, 'douyin', "82c8d9", "A2:Z2", values)
                Common.logger(log_type, crawler).info(f"视频已保存至云文档\n")

                # 视频信息保存数据库
                insert_sql = f""" insert into crawler_video(video_id,
                                user_id,
                                out_user_id,
                                platform,
                                strategy,
                                out_video_id,
                                video_title,
                                cover_url,
                                video_url,
                                duration,
                                publish_time,
                                play_cnt,
                                comment_cnt,
                                like_cnt,
                                share_cnt,
                                crawler_rule,
                                width,
                                height)
                                values({our_video_id},
                                {our_uid},
                                "{video_dict['user_id']}",
                                "{cls.platform}",
                                "推荐爬虫策略",
                                "{video_dict['video_id']}",
                                "{video_dict['video_title']}",
                                "{video_dict['cover_url']}",
                                "{video_dict['video_url']}",
                                {int(video_dict['duration'])},
                                "{video_dict['publish_time_str']}",
                                {int(video_dict['play_cnt'])},
                                {int(video_dict['comment_cnt'])},
                                {int(video_dict['like_cnt'])},
                                {int(video_dict['share_cnt'])},
                                '{json.dumps(rule_dict)}',
                                {int(video_dict['video_width'])},
                                {int(video_dict['video_height'])}) """
                Common.logger(log_type, crawler).info(f"insert_sql:{insert_sql}")
                MysqlHelper.update_values(log_type, crawler, insert_sql, env, machine)
                Common.logger(log_type, crawler).info('视频信息插入数据库成功！\n')
        except Exception as e:
            Common.logger(log_type, crawler).error(f'download_publish异常:{e}\n')

    # @classmethod
    # def get_recommend(cls, log_type, crawler, strategy, oss_endpoint, env, machine):
    #     try:
    #
    #         user_list = get_user_from_mysql(log_type, crawler, crawler, env)
    #         for user in user_list:
    #             spider_link = user["spider_link"]
    #             out_uid = spider_link.split('/')[-1]
    #             user_name = user["nick_name"]
    #             our_uid = user["media_id"]
    #
    #             Common.logger(log_type, crawler).info(f"开始抓取 {user_name} 用户主页视频\n")
    #             cls.get_videolist(log_type=log_type,
    #                               crawler=crawler,
    #                               strategy=strategy,
    #                               our_uid=our_uid,
    #                               oss_endpoint=oss_endpoint,
    #                               env=env,
    #                               machine=machine)
    #             time.sleep(1)
    #     except Exception as e:
    #         Common.logger(log_type, crawler).error(f"get_follow_videos:{e}\n")


if __name__ == '__main__':
    # DyRecommend.get_recommend('recommend','douyin','推荐抓取策略', 'inner','prod', 'aliyun')
    DyRecommend.get_videolist('recommend', 'douyin', '推荐抓取策略', 6282709, 'outer', 'dev', 'aliyun')