video_rewriting/data_channel/qpl_ks.py

import random
import time
from datetime import datetime

import requests
import json

from common import Feishu
from common.sql_help import sqlCollect


class QplKs:
    @classmethod
    def ks_data_list(cls):
        content_type = [{'美食': '673'}, {'短剧': '674'}, {'时尚': '675'}, {'母婴亲子': '676'}, {'生活': '677'},
                        {'汽车': '678'}, {'旅游': '679'}, {'体育运动': '680'}, {'艺术文化': '682'}, {'房产家居': '683'},
                        {'健康医疗': '684'}, {'教育培训': '685'}, {'高新数码': '686'}, {'音乐': '687'}, {'舞蹈': '688'}, {'读书': '689'},
                        {'摄影': '704'},
                        {'财经投资': '690'}, {'萌宠': '691'}, {'时政资讯': '692'}, {'科学与法律': '693'}, {'三农': '694'}, {'搞笑': '695'},
                        {'情感': '696'},
                        {'人文': '697'}, {'星座命理': '699'}, {'奇人异象': '700'}, {'军事': '701'}, {'影视娱乐': '702'},
                        {'民生资讯': '703'}, {'健身达人': '706'}, {'纪实类': '705'}]
        # fans_count = [{0: 100000}, {100000: 1000000}, {1000000: 3000000}, {3000000: 5000000}, {5000000: 10000000},
        #               {10000000: 0}]


        fans_count = [{10000000: 0}, {5000000: 10000000}]
        url = "https://k.kuaishou.com/rest/web/star/list"
        headers = {
            'Accept': 'application/json',
            'Accept-Language': 'zh-CN,zh;q=0.9',
            'Content-Type': 'application/json',
            'Cookie': 'did=web_9c6a04a4004fdb7c95a658a56ed275b6; didv=1711519980000; apdid=328ac94f-4040-41fe-a038-b60140291aca99fb22e9862c74736f53d57b666ee53b:1719580529:1; language=zh-CN; ud=2205012540; app_id=ks686235996059836592; expire_time=1800; userId=2574854626; access_token=ChFvYXV0aC5hY2Nlc3NUb2tlbhIwIRvxysHMF86NQiMYMtw1_s0zM4gAEA9_4VrJixHfjYqD8RBNius0YXow_kWI0B8sGhINVY3cQVhC4YDnv5YNyfZ0IKwiIObPEx_s6MCNTBHJclEO0mnOPHspy1vnwt-9536AWyqYKAUwAQ; nc_user_id=CiVhZC5ub3RpZnkuY2VudGVyLm9hdXRoLnVzZXIuaWQuc2VjcmV0EiAtVd5hI6zlT9/UG3HxzRPGDpAxZul8rkwLeQzpGx4y5xoS0fOAqkTADrggSZ6ZpciI2NO9IiBQFbr8Uk6ApRwM5LcFRtS+AAj/QbpiELCRBW7L200+nCgFMAE=; kuaishou.ad.social_st=ChVrdWFpc2hvdS5hZC5zb2NpYWwuc3QSoAFByWIrb3A2KGuX_eJCqlJa6wQTlc2l7mp_MQN_5RLJ4INIE2O3-v18nVEBafp8VK9sY5-ExW_XMpfak7gT8EoI0ft5snPqXur9Ki4gVWIbK-Z-VkyQFkAaD6qB-VtBsqiUoHEmAI9s1H-1kLMKYTCwBUbPH8TsBbnPQw0Lg8NvSQ2Gl4_DP0uA4ouuPFjKJtcoEs9SEz9VbPFfC7X0iDQFGhIk8-y_Rw9CUpd3ZdyZev2LvtUiIKLFKmL283X9GwjqcIMqjlF3erFhqI1QIA9aK-z1gQh3KAUwAQ; kuaishou.ad.social_ph=8764c8e6b60f8cff08b469a56da3fdbe2482; did=web_9c6a04a4004fdb7c95a658a56ed275b6; didv=1711522726000',
            'Origin': 'https://k.kuaishou.com',
            'Pragma': 'no-cache',
            'Referer': 'https://k.kuaishou.com/',
            'User-Agent': 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/126.0.0.0 Safari/537.36',
        }
        # for count in fans_count:
        #     (fans_min_num, fans_max_num), = count.items()
        for content in content_type:
            (content_name, content_id), = content.items()
            current_page = 1
            while True:
                payload_data = {
                    "currentPage": current_page,
                    "pageSize": 20,
                    "starOrderTag": 3,
                    "taskType": 1,
                    "marketingGoal": 1,
                    "viewerAgeList": [  # 观众画像 50+
                        {
                            "from": "0.50",
                            "to": "1.00",
                            "queryName": "50+"
                        }
                    ],
                    "userName": "",
                    "vitalityTags": [  # 活跃度 近期短视频活跃
                        "668"
                    ],
                    "contentTagIdList": [  # 内容类型
                        {
                            "id": str(content_id),
                            "selectFirstLevelId": True
                        }
                    ],
                    "starTagIds": [  # 高调性
                        13
                    ]
                    # "fansMinNum": 1000000,  # 粉丝数量
                    # "fansMaxNum": 3000000,  # 粉丝数量
                }
                # if fans_min_num > 0:
                #     payload_data["fansMinNum"] = fans_min_num
                # if fans_max_num > 0:
                #     payload_data["fansMaxNum"] = fans_max_num
                payload = json.dumps(payload_data)

                time.sleep(random.randint(10, 20))
                response = requests.request("POST", url, headers=headers, data=payload)
                response = response.json()
                print(f"开始扫描{content}")

                result = response["result"]
                if result == 1:
                    total = response["total"]  # 总条数
                    if total == 0:
                        print(f"没有扫描到数据{content}")
                        break
                    star_list = response["starList"]
                    if len(star_list) == 0 or star_list == []:
                        break
                    for star in star_list:
                        print(f"扫描到一条数据{content}")
                        user_id = star["userId"]
                        star_id = star["starId"]
                        name = star["name"]  # 用户名
                        kwai_id = star["kwaiId"]  # 用户名id
                        gender = star["gender"]  # 性别
                        fans_number = star["fansNumber"]  # 粉丝数
                        profile_id = star["profileId"]  # 主页id
                        star_tag_str = star["starTagStr"]  # 内容类型1
                        industry_tag_str = star["industryTagStr"]  # 内容类型2
                        photo_expect_play = star["photoExpectPlay"]  # 预期播放量
                        photo_expect_cpm = star["photoExpectCpm"]  # 预期CPM
                        photo_interaction_rate = star["photoInteractionRate"]  # 互动率
                        photo_complete_play_rate = star["photoCompletePlayRate"]  # 完播率
                        fans_increase_num = star["fansIncreaseNum"]  # 粉丝增长量
                        fans_increase_rate = star["fansIncreaseRate"]  # 粉丝增长率
                        # res = sqlCollect.insert_ks_qpl_data(user_id, star_id, name, kwai_id, gender, fans_number, profile_id, star_tag_str, industry_tag_str, photo_expect_play, photo_expect_cpm, photo_interaction_rate, photo_complete_play_rate, fans_increase_num, fans_increase_rate)
                        # if res == 1:
                        current_time = datetime.now()
                        formatted_time = current_time.strftime("%Y-%m-%d %H:%M:%S")
                            # if fans_min_num == 0:
                            #     fans = f"{fans_max_num}及以下"
                            # if fans_max_num == 0:
                            #     fans = f"{fans_min_num}及以上"
                            # else:
                            #     fans = f"{fans_min_num}~{fans_max_num}"
                        values = [
                            [
                                name,
                                gender,
                                fans_number,
                                profile_id,
                                "https://www.kuaishou.com/profile/"+profile_id,
                                star_tag_str,
                                industry_tag_str,
                                photo_expect_play,
                                photo_expect_cpm,
                                photo_interaction_rate,
                                photo_complete_play_rate,
                                fans_increase_num,
                                fans_increase_rate,
                                # fans,
                                formatted_time
                            ]
                        ]
                        Feishu.insert_columns("GjGZsmW2ahaCe4tmzDTc58tVnbe", "emf56g", "ROWS", 1, 2)
                        time.sleep(0.5)
                        Feishu.update_values("GjGZsmW2ahaCe4tmzDTc58tVnbe", "emf56g", "A2:Z2", values)
                        print(f"入库到一条数据{content}")
                    current_page += 1
                    if total < 20:
                        break
                    else:
                        page = int(total)/20
                        if current_page > int(page)+1:
                            break











if __name__ == '__main__':
    QplKs.ks_data_list()