# -*- 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.scheduling_db import MysqlHelper from common.publish import Publish from common.public import random_title, get_config_from_mysql, download_rule, get_title_score from common.userAgent import get_random_user_agent from common.mq import MQ class KuaiShouRecommendScheduling: platform = "快手" # 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, our_uid, rule_dict, env): mq = MQ(topic_name="topic_crawler_etl_" + env) for page in range(1, 101): try: Common.logger(log_type, crawler).info(f"正在抓取第{page}页") Common.logging(log_type, crawler, env, f"正在抓取第{page}页") 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_7cdc486ebd1aba220455a7781d6ae5b5{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"' } 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() if response.status_code != 200: Common.logger(log_type, crawler).warning(f"get_videoList_response:{response.status_code}, {response.text}\n") Common.logging(log_type, crawler, env, f"get_videoList_response:{response.status_code}, {response.text}\n") continue elif 'data' not in response.json(): Common.logger(log_type, crawler).warning(f"get_videoList_response:{response.json()}\n") Common.logging(log_type, crawler, env, 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") Common.logging(log_type, crawler, env, 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") Common.logging(log_type, crawler, env, 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") Common.logging(log_type, crawler, env, "没有更多视频啦 ~\n") continue else: feeds = response.json()['data']['visionNewRecoFeed']['feeds'] for i in range(len(feeds)): try: video_title = feeds[i].get("photo", random_title(log_type, crawler, env, text='title')).get("caption", random_title(log_type, crawler, env, text='title')) video_title = cls.video_title(log_type, crawler, env, video_title) try: video_id = feeds[i].get("photo", {}).get("videoResource").get("h264", {}).get("videoId", "") video_width = feeds[i].get("photo", {}).get("videoResource").get("h264", {}).get("adaptationSet", {})[0].get("representation", {})[0].get("width", 0) video_height = feeds[i].get("photo", {}).get("videoResource").get("h264", {}).get("adaptationSet", {})[0].get("representation", {})[0].get("height", 0) except KeyError: video_id = feeds[i].get("photo", {}).get("videoResource").get("hevc", {}).get("videoId", "") video_width = feeds[i].get("photo", {}).get("videoResource").get("hevc", {}).get("adaptationSet", {})[0].get("representation", {})[0].get("width", 0) video_height = feeds[i].get("photo", {}).get("videoResource").get("hevc", {}).get("adaptationSet", {})[0].get("representation", {})[0].get("height", 0) publish_time_stamp = int(int(feeds[i].get('photo', {}).get('timestamp', 0)) / 1000) publish_time_str = time.strftime("%Y-%m-%d %H:%M:%S", time.localtime(publish_time_stamp)) video_dict = {'video_title': video_title, 'video_id': video_id, 'play_cnt': int(feeds[i].get('photo', {}).get('viewCount', 0)), 'like_cnt': int(feeds[i].get('photo', {}).get('realLikeCount', 0)), 'comment_cnt': 0, 'share_cnt': 0, 'video_width': video_width, 'video_height': video_height, 'duration': int(int(feeds[i].get('photo', {}).get('duration', 0)) / 1000), 'publish_time_stamp': publish_time_stamp, 'publish_time_str': publish_time_str, 'user_name': feeds[i].get('author', {}).get('name', ""), 'user_id': feeds[i].get('author', {}).get('id', ""), 'avatar_url': feeds[i].get('author', {}).get('headerUrl', ""), 'cover_url': feeds[i].get('photo', {}).get('coverUrl', ""), 'video_url': feeds[i].get('photo', {}).get('photoUrl', ""), 'session': f"kuaishou-{int(time.time())}"} for k, v in video_dict.items(): Common.logger(log_type, crawler).info(f"{k}:{v}") Common.logging(log_type, crawler, env, f"{video_dict}") if video_dict["video_id"] == '' or video_dict["cover_url"] == '' or video_dict["video_url"] == '': Common.logger(log_type, crawler).info('无效视频\n') Common.logging(log_type, crawler, env, '无效视频\n') elif download_rule(log_type=log_type, crawler=crawler, video_dict=video_dict, rule_dict=rule_dict) is False: Common.logger(log_type, crawler).info("不满足抓取规则\n") Common.logging(log_type, crawler, env, "不满足抓取规则\n") elif any(str(word) if str(word) in video_dict["video_title"] else False for word in get_config_from_mysql(log_type=log_type, source=crawler, env=env, text="filter", action="")) is True: Common.logger(log_type, crawler).info('已中过滤词\n') Common.logging(log_type, crawler, env, '已中过滤词\n') elif cls.repeat_video(log_type, crawler, video_dict["video_id"], env) != 0: Common.logger(log_type, crawler).info('视频已下载\n') Common.logging(log_type, crawler, env, '视频已下载\n') else: video_dict["out_user_id"] = video_dict["user_id"] video_dict["platform"] = crawler video_dict["strategy"] = log_type video_dict["out_video_id"] = video_dict["video_id"] video_dict["width"] = video_dict["video_width"] video_dict["height"] = video_dict["video_height"] video_dict["crawler_rule"] = json.dumps(rule_dict) video_dict["user_id"] = our_uid video_dict["publish_time"] = video_dict["publish_time_str"] mq.send_msg(video_dict) # title_score = get_title_score(log_type, crawler, "16QspO", "0usaDk", video_title) # if title_score <= 0.3: # Common.logger(log_type, crawler).info(f"权重分:{title_score}<=0.3\n") # Common.logging(log_type, crawler, env, f"权重分:{title_score}<=0.3\n") # continue # cls.download_publish(log_type=log_type, # crawler=crawler, # our_uid=our_uid, # video_dict=video_dict, # rule_dict=rule_dict, # title_score=title_score, # env=env) except Exception as e: Common.logger(log_type, crawler).error(f"抓取单条视频异常:{e}\n") Common.logging(log_type, crawler, env, f"抓取单条视频异常:{e}\n") except Exception as e: Common.logger(log_type, crawler).error(f"抓取第{page}页时异常:{e}\n") Common.logging(log_type, crawler, env, f"抓取第{page}页时异常:{e}\n") @classmethod def repeat_video(cls, log_type, crawler, video_id, env): 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) return len(repeat_video) @classmethod def download_publish(cls, log_type, crawler, our_uid, video_dict, rule_dict, title_score, env): # 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() try: 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") Common.logging(log_type, crawler, env, "视频size=0，删除成功\n") return except FileNotFoundError: # 删除视频文件夹 shutil.rmtree(f"./{crawler}/videos/{md_title}") Common.logger(log_type, crawler).info("视频文件不存在，删除文件夹成功\n") Common.logging(log_type, crawler, env, "视频文件不存在，删除文件夹成功\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...") Common.logging(log_type, crawler, env, "开始上传视频...") if env == "dev": oss_endpoint = "out" our_video_id = Publish.upload_and_publish(log_type=log_type, crawler=crawler, strategy="推荐抓取策略", our_uid=our_uid, env=env, oss_endpoint=oss_endpoint) our_video_link = f"https://testadmin.piaoquantv.com/cms/post-detail/{our_video_id}/info" else: oss_endpoint = "inner" our_video_id = Publish.upload_and_publish(log_type=log_type, crawler=crawler, strategy="推荐抓取策略", our_uid=our_uid, env=env, oss_endpoint=oss_endpoint) our_video_link = f"https://admin.piaoquantv.com/cms/post-detail/{our_video_id}/info" if our_video_id is None: try: # 删除视频文件夹 shutil.rmtree(f"./{crawler}/videos/{md_title}") return except FileNotFoundError: return # 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}", "推荐抓取策略", "{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'])}, '{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}") Common.logging(log_type, crawler, env, f"insert_sql:{insert_sql}") MysqlHelper.update_values(log_type, crawler, insert_sql, env, action="") Common.logger(log_type, crawler).info('视频信息写入数据库成功') Common.logging(log_type, crawler, env, '视频信息写入数据库成功') # 视频写入飞书 Feishu.insert_columns(log_type, crawler, "Aps2BI", "ROWS", 1, 2) upload_time = int(time.time()) values = [[title_score, our_video_id, time.strftime("%Y-%m-%d %H:%M:%S", time.localtime(upload_time)), "推荐抓取策略", 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(0.5) Feishu.update_values(log_type, crawler, "Aps2BI", "D2:Z2", values) Common.logger(log_type, crawler).info(f"视频已保存至云文档\n") Common.logging(log_type, crawler, env, f"视频已保存至云文档\n") if __name__ == "__main__": print(get_config_from_mysql("recommend", "kuaishou", "prod", "filter")) pass