piaoquan_crawler/kuaishou/kuaishou_recommend/kuaishou_recommend_shceduling.py

# -*- coding: utf-8 -*-
# @Author: wangkun
# @Time: 2023/2/24
import os
import random
import shutil
import sys
import time
import string
from hashlib import md5

import requests
import json

import urllib3
from requests.adapters import HTTPAdapter

sys.path.append(os.getcwd())
from common.common import Common
from common.feishu import Feishu
from common.getuser import getUser
# from common.db import MysqlHelper
from common.scheduling_db import MysqlHelper
from common.publish import Publish
from common.public import get_user_from_mysql, random_title, get_config_from_mysql
from common.userAgent import get_random_user_agent


class KuaiShouRecommendScheduling:
    platform = "快手"
    tag = "快手爬虫,推荐爬虫策略"

    @classmethod
    def get_rule(cls, log_type, crawler):
        try:
            rule_sheet = Feishu.get_values_batch(log_type, crawler, "NQ6CZN")
            rule_dict = {
                "play_cnt": f"{rule_sheet[0][1]}{rule_sheet[0][2]}",
                "video_width": f"{rule_sheet[1][1]}{rule_sheet[1][2]}",
                "video_height": f"{rule_sheet[2][1]}{rule_sheet[2][2]}",
                "like_cnt": f"{rule_sheet[5][1]}{rule_sheet[5][2]}",
                "duration": f"{rule_sheet[3][1]}{rule_sheet[3][2]}",
                "publish_time": f"{rule_sheet[4][1]}{rule_sheet[4][2]}",
            }
            return rule_dict
        except Exception as e:
            Common.logger(log_type, crawler).error(f"get_rule:{e}\n")

    @classmethod
    def download_rule(cls, video_dict, rule_dict):
        if video_dict['like_cnt'] >= rule_dict['like_cnt']['min']:
            if video_dict['publish_time'] >= rule_dict['publish_time']['min']:
                if video_dict['duration'] >= rule_dict['duration']['min']:
                    if video_dict['video_width'] >= rule_dict['width']['min'] \
                            or video_dict['video_height'] >= rule_dict['height']['min']:
                        return True
                    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, 'HIKVvs')
                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')

    # 获取用户信息列表
    @classmethod
    def get_user_list(cls, log_type, crawler, sheetid, env):
        try:
            while True:
                user_sheet = Feishu.get_values_batch(log_type, crawler, sheetid)
                if user_sheet is None:
                    Common.logger(log_type, crawler).warning(f"user_sheet:{user_sheet} 10秒钟后重试")
                    continue
                our_user_list = []
                for i in range(1, len(user_sheet)):
                    # for i in range(1, 2):
                    out_uid = user_sheet[i][2]
                    user_name = user_sheet[i][3]
                    our_uid = user_sheet[i][6]
                    our_user_link = user_sheet[i][7]
                    if out_uid is None or user_name is None:
                        Common.logger(log_type, crawler).info("空行\n")
                    else:
                        Common.logger(log_type, crawler).info(f"正在更新 {user_name} 用户信息\n")
                        if our_uid is None:
                            out_user_info = cls.get_out_user_info(log_type, crawler, out_uid)
                            out_user_dict = {
                                "out_uid": out_uid,
                                "user_name": user_name,
                                "out_avatar_url": out_user_info["out_avatar_url"],
                                "out_create_time": '',
                                "out_tag": '',
                                "out_play_cnt": 0,
                                "out_fans": out_user_info["out_fans"],
                                "out_follow": out_user_info["out_follow"],
                                "out_friend": 0,
                                "out_like": 0,
                                "platform": cls.platform,
                                "tag": cls.tag,
                            }
                            our_user_dict = getUser.create_user(log_type=log_type, crawler=crawler,
                                                                out_user_dict=out_user_dict, env=env)
                            our_uid = our_user_dict['our_uid']
                            our_user_link = our_user_dict['our_user_link']
                            Feishu.update_values(log_type, crawler, sheetid, f'G{i + 1}:H{i + 1}',
                                                 [[our_uid, our_user_link]])
                            Common.logger(log_type, crawler).info(f'站内用户信息写入飞书成功！\n')
                            our_user_list.append(our_user_dict)
                        else:
                            our_user_dict = {
                                'out_uid': out_uid,
                                'user_name': user_name,
                                'our_uid': our_uid,
                                'our_user_link': our_user_link,
                            }
                            our_user_list.append(our_user_dict)
                return our_user_list
        except Exception as e:
            Common.logger(log_type, crawler).error(f'get_user_list:{e}\n')

    # 处理视频标题
    @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("快手", "").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, task, our_uid, oss_endpoint, env):
        rule_dict_1 = task['rule_dict']
        for i in range(100):
            url = "https://www.kuaishou.com/graphql"

            payload = json.dumps({
                "operationName": "visionNewRecoFeed",
                "variables": {
                    "dailyFirstPage": False
                },
                "query": "fragment photoContent on PhotoEntity {\n  id\n  duration\n  caption\n  originCaption\n  likeCount\n  viewCount\n  realLikeCount\n  coverUrl\n  photoUrl\n  photoH265Url\n  manifest\n  manifestH265\n  videoResource\n  coverUrls {\n    url\n    __typename\n  }\n  timestamp\n  expTag\n  animatedCoverUrl\n  distance\n  videoRatio\n  liked\n  stereoType\n  profileUserTopPhoto\n  musicBlocked\n  __typename\n}\n\nfragment feedContent on Feed {\n  type\n  author {\n    id\n    name\n    headerUrl\n    following\n    headerUrls {\n      url\n      __typename\n    }\n    __typename\n  }\n  photo {\n    ...photoContent\n    __typename\n  }\n  canAddComment\n  llsid\n  status\n  currentPcursor\n  tags {\n    type\n    name\n    __typename\n  }\n  __typename\n}\n\nfragment photoResult on PhotoResult {\n  result\n  llsid\n  expTag\n  serverExpTag\n  pcursor\n  feeds {\n    ...feedContent\n    __typename\n  }\n  webPageArea\n  __typename\n}\n\nquery visionNewRecoFeed($semKeyword: String, $semCrowd: String, $utmSource: String, $utmMedium: String, $utmCampaign: String, $dailyFirstPage: Boolean) {\n  visionNewRecoFeed(semKeyword: $semKeyword, semCrowd: $semCrowd, utmSource: $utmSource, utmMedium: $utmMedium, utmCampaign: $utmCampaign, dailyFirstPage: $dailyFirstPage) {\n    ...photoResult\n    __typename\n  }\n}\n"
            })
            s = string.ascii_lowercase
            r = random.choice(s)

            headers = {
                'Accept-Language': 'zh-CN,zh;q=0.9',
                'Connection': 'keep-alive',
                'Cookie': 'kpf=PC_WEB; clientid=3; did=web_aba004b1780f4d7174d0a2ff42da1f{r}7; kpn=KUAISHOU_VISION;'.format(
                    r=r),
                'Origin': 'https://www.kuaishou.com',
                'Referer': 'https://www.kuaishou.com/new-reco',
                'Sec-Fetch-Dest': 'empty',
                'Sec-Fetch-Mode': 'cors',
                'Sec-Fetch-Site': 'same-origin',
                'User-Agent': get_random_user_agent('pc'),
                'accept': '*/*',
                'content-type': 'application/json',
                'sec-ch-ua': '"Google Chrome";v="111", "Not(A:Brand";v="8", "Chromium";v="111"',
                'sec-ch-ua-mobile': '?0',
                'sec-ch-ua-platform': '"macOS"'
            }

            try:
                urllib3.disable_warnings()
                s = requests.session()
                # max_retries=3 重试3次
                s.mount('http://', HTTPAdapter(max_retries=3))
                s.mount('https://', HTTPAdapter(max_retries=3))
                response = s.post(url=url, headers=headers, data=payload, proxies=Common.tunnel_proxies(), verify=False,
                                  timeout=10)
                response.close()
            except Exception as e:
                Common.logger(log_type, crawler).error(f"get_videoList:{e}\n")
                continue
            # Common.logger(log_type, crawler).info(f"get_videoList:{response.text}\n")
            if response.status_code != 200:
                Common.logger(log_type, crawler).warning(f"get_videoList_response:{response.text}\n")
                continue
            elif 'data' not in response.json():
                Common.logger(log_type, crawler).warning(f"get_videoList_response:{response.json()}\n")
                continue
            elif 'visionNewRecoFeed' not in response.json()['data']:
                Common.logger(log_type, crawler).warning(f"get_videoList_response:{response.json()['data']}\n")
                continue
            elif 'feeds' not in response.json()['data']['visionNewRecoFeed']:
                Common.logger(log_type, crawler).warning(
                    f"get_videoList_response:{response.json()['data']['visionNewRecoFeed']}\n")
                continue
            elif len(response.json()['data']['visionNewRecoFeed']['feeds']) == 0:
                Common.logger(log_type, crawler).info("没有更多视频啦 ~\n")
                continue
            else:
                feeds = response.json()['data']['visionNewRecoFeed']['feeds']
                for i in range(len(feeds)):
                    if 'photo' not in feeds[i]:
                        Common.logger(log_type, crawler).warning(f"get_videoList:{feeds[i]}\n")
                        continue

                    # video_title
                    if 'caption' not in feeds[i]['photo']:
                        video_title = random_title(log_type, crawler, env, text='title')

                    elif feeds[i]['photo']['caption'].strip() == "":
                        video_title = random_title(log_type, crawler, env, text='title')
                    else:
                        video_title = cls.video_title(log_type, crawler, env, feeds[i]['photo']['caption'])

                    if 'videoResource' not in feeds[i]['photo'] \
                            and 'manifest' not in feeds[i]['photo'] \
                            and 'manifestH265' not in feeds[i]['photo']:
                        Common.logger(log_type, crawler).warning(f"get_videoList:{feeds[i]['photo']}\n")
                        continue
                    videoResource = feeds[i]['photo']['videoResource']

                    if 'h264' not in videoResource and 'hevc' not in videoResource:
                        Common.logger(log_type, crawler).warning(f"get_videoList:{videoResource}\n")
                        continue

                    # video_id
                    if 'h264' in videoResource and 'videoId' in videoResource['h264']:
                        video_id = videoResource['h264']['videoId']
                    elif 'hevc' in videoResource and 'videoId' in videoResource['hevc']:
                        video_id = videoResource['hevc']['videoId']
                    else:
                        video_id = ""

                    # play_cnt
                    if 'viewCount' not in feeds[i]['photo']:
                        play_cnt = 0
                    else:
                        play_cnt = int(feeds[i]['photo']['viewCount'])

                    # like_cnt
                    if 'realLikeCount' not in feeds[i]['photo']:
                        like_cnt = 0
                    else:
                        like_cnt = feeds[i]['photo']['realLikeCount']

                    # publish_time
                    if 'timestamp' not in feeds[i]['photo']:
                        publish_time_stamp = 0
                        publish_time_str = ''
                        publish_time = 0
                    else:
                        publish_time_stamp = int(int(feeds[i]['photo']['timestamp']) / 1000)
                        publish_time_str = time.strftime("%Y-%m-%d %H:%M:%S", time.localtime(publish_time_stamp))
                        publish_time = int((int(time.time()) - publish_time_stamp) / (3600 * 24))

                    # duration
                    if 'duration' not in feeds[i]['photo']:
                        duration = 0
                    else:
                        duration = int(int(feeds[i]['photo']['duration']) / 1000)

                    # video_width / video_height / video_url
                    mapping = {}
                    for item in ['width', 'height']:
                        try:
                            val = str(videoResource['h264']['adaptationSet'][0]['representation'][0][item])
                        except Exception:
                            val = str(videoResource['hevc']['adaptationSet'][0]['representation'][0][item])
                        except:
                            val = ''
                        mapping[item] = val
                    video_width = int(mapping['width']) if mapping['width'] != '' else 0
                    video_height = int(mapping['height']) if mapping['height'] != '' else 0
                    # cover_url
                    if 'coverUrl' not in feeds[i]['photo']:
                        cover_url = ""
                    else:
                        cover_url = feeds[i]['photo']['coverUrl']

                    # user_name / avatar_url
                    try:
                        user_name = feeds[i]['author']['name']
                        avatar_url = feeds[i]['author']['headerUrl']
                        user_id = feeds[i]['author']['id']
                    except Exception:
                        user_name = ''
                        avatar_url = ''
                        user_id = ''
                    video_url = feeds[i]['photo']['photoUrl']
                    video_dict = {'video_title': video_title,
                                  'video_id': video_id,
                                  'play_cnt': play_cnt,
                                  'comment_cnt': 0,
                                  'like_cnt': like_cnt,
                                  'share_cnt': 0,
                                  'video_width': video_width,
                                  'video_height': video_height,
                                  'duration': duration,
                                  'publish_time': publish_time,
                                  'publish_time_stamp': publish_time_stamp,
                                  'publish_time_str': publish_time_str,
                                  'user_name': user_name,
                                  'user_id': user_id,
                                  'avatar_url': avatar_url,
                                  'cover_url': cover_url,
                                  'video_url': video_url,
                                  'session': f"kuaishou{int(time.time())}"}

                    rule_1 = cls.download_rule(video_dict, rule_dict_1)

                    if rule_1 is True:
                        cls.download_publish(log_type=log_type,
                                             crawler=crawler,
                                             strategy=strategy,
                                             video_dict=video_dict,
                                             rule_dict=rule_dict_1,
                                             our_uid=our_uid,
                                             oss_endpoint=oss_endpoint,
                                             env=env,
                                             )

                    else:
                        Common.logger(log_type, crawler).info("不满足下载规则\n")

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

    @classmethod
    def download_publish(cls, log_type, crawler, strategy, video_dict, rule_dict, our_uid, oss_endpoint, env):
        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
            download_finished = False
            if cls.repeat_video(log_type, crawler, video_dict['video_id'], video_dict['video_title'],
                                video_dict['publish_time_str'], env) != 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
                # 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:
                    Common.logger(log_type, crawler).warning(f"our_video_id:{our_video_id} 删除该视频文件夹")
                    # 删除视频文件夹
                    shutil.rmtree(f"./{crawler}/videos/{video_dict['video_title']}")
                    return download_finished

                # 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,
                                                        crawler_rule,
                                                        width,
                                                        height)
                                                        values({our_video_id},
                                                        {our_uid},
                                                        "{video_dict['user_id']}",
                                                        "{cls.platform}",
                                                        "{strategy}",
                                                        "{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'])},
                                                        '{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)
                Common.logger(log_type, crawler).info('视频信息插入数据库成功！\n')

                # 视频写入飞书
                Feishu.insert_columns(log_type, 'kuaishou', "Aps2BI", "ROWS", 1, 2)
                upload_time = int(time.time())
                values = [[our_video_id,
                           time.strftime("%Y-%m-%d %H:%M:%S", time.localtime(upload_time)),
                           strategy,
                           str(video_dict['video_id']),
                           video_dict['video_title'],
                           our_video_link,
                           video_dict['play_cnt'],
                           video_dict['comment_cnt'],
                           video_dict['like_cnt'],
                           video_dict['share_cnt'],
                           video_dict['duration'],
                           f"{video_dict['video_width']}*{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, 'kuaishou', "Aps2BI", "E2:Z2", values)
                Common.logger(log_type, crawler).info(f"视频已保存至云文档\n")
                download_finished = True
            return download_finished
        except Exception as e:
            Common.logger(log_type, crawler).error(f"download_publish:{e}\n")

    @classmethod
    def get_recommend_videos(cls, log_type, crawler, task, oss_endpoint, env):
        user_list = get_user_from_mysql(log_type, crawler, crawler, env)
        strategy = '推荐抓取策略'
        for user in user_list:
            spider_link = user["link"]
            out_uid = spider_link
            user_name = user["nick_name"]
            our_uid = user["uid"]
            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,
                              task=task,
                              our_uid=our_uid,
                              oss_endpoint=oss_endpoint,
                              env=env)


if __name__ == "__main__":
    task = ''
    KuaiShouRecommendScheduling.get_recommend_videos('recommend', 'kuaishou', task, 'outer', 'prod')