crawler_xigua/main/demo.py

# -*- coding: utf-8 -*-
# @Author: wangkun
# @Time: 2022/8/23
import ast
import base64
import json
import os
import subprocess
import requests
import urllib3
from main.feishu import Feishu

proxies = {'http:': None, 'https': None}


class Demo:
    @classmethod
    def base64_decode(cls):
        s = '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'
        s1 = 'aHR0cDovL3Y5LXhnLXdlYi1wYy5peGlndWEuY29tLzRhNDA2MjBiNjY0OTA4ZmYzOGNmNGM0MWQ0NTM5YzlkLzYzN2M5NDUxL3ZpZGVvL3Rvcy9jbi90b3MtY24tdi0wMDI2LzgyZjY4N2E3NjRhYjQyNmM5NzI0YTEwMzc4MjM5NTA5L21lZGlhLWF1ZGlvLXVuZC1tcDRhLz9hPTE3NjgmY2g9MCZjcj0wJmRyPTAmZXI9MCZjZD0wJTdDMCU3QzAlN0MwJmN2PTEmYnI9MCZidD0wJmNzPTAmbWltZV90eXBlPXZpZGVvX21wNCZxcz0wJnJjPVFHa3phalYwTnpwbWRtaG5Nek0wWkRNelFHRWpjMkF2YjNJMGJ6RnhZQzB0WkM0dmMzTSUzRCZsPTIwMjIxMTIyMTYwOTM0MDEwMjA0MDU1MTQ1MDU0NDQ1MEEmYnRhZz0zODAwMA=='
        if len(s) % 3 == 1:
            s += '=='
        elif len(s) % 3 == 2:
            s += '='
        elif len(s1) % 3 == 1:
            s1 += '=='
        elif len(s1) % 3 == 2:
            s1 += '='
        base64_url = base64.b64decode(s).decode('utf8')
        base64_url1 = base64.b64decode(s1).decode('utf8')
        print(base64_url)
        print(base64_url1)

    @classmethod
    def download_video(cls):
        video_url = 'http://v9-xg-web-pc.ixigua.com/69637e5454682779d9ba8529eeb3de3d/637dbed1/video/tos/cn/tos-cn-v-0026/e62a566034754f70a31b277040e934a1/media-video-avc1/?a=1768&ch=0&cr=0&dr=0&er=0&cd=0%7C0%7C0%7C0&cv=1&br=2974&bt=2974&cs=0&ds=4&mime_type=video_mp4&qs=0&rc=Z2lnNzw4NjU0N2c7ZWRpZUBpM2szOTo6ZjxlZzMzNGQzM0AtMDIwYV4zNTAxM140NmJfYSNyM3FocjRnaHJgLS1kLi9zcw%3D%3D&l=20221123132351010159037053022B9FD3&btag=38000'.replace('http://', 'https://')
        video_name = 'video.mp4'
        # video_url = 'http://v9-xg-web-pc.ixigua.com/08ef1be3a3b73c16f414bddcd7810362/637cc021/video/tos/cn/tos-cn-v-0026/82f687a764ab426c9724a10378239509/media-audio-und-mp4a/?a=1768&ch=0&cr=0&dr=0&er=0&cd=0%7C0%7C0%7C0&cv=1&br=0&bt=0&cs=0&mime_type=video_mp4&qs=0&rc=QGkzajV0NzpmdmhnMzM0ZDMzQGEjc2Avb3I0bzFxYC0tZC4vc3M%3D&l=20221122191629010210039035028580DD&btag=38000'
        # video_name = 'audio.mp4'

        urllib3.disable_warnings()
        # 设置请求头信息
        headers = {
            'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko)'
                          ' Chrome/96.0.4664.110 Safari/537.36 Edg/96.0.1054.62'
        }
        video_content = requests.get(video_url, headers=headers, stream=True, proxies=proxies, verify=False)
        try:
            with open('./' + video_name, "wb") as f:
                for chunk in video_content.iter_content(chunk_size=1024*1024):
                    if chunk:
                        f.write(chunk)
                        # 记录已下载视频长度，实时输出下载进度
                        downSize = 0
                        contentLength = int(video_content.headers['content-length'])
                        downSize += len(chunk)
                        print('进度：{:.2%}'.format(downSize / contentLength), end='\r')
            print('下载完成')
        except Exception as e:
            print(f"视频下载失败：{e}\n")

    @classmethod
    def ffmpeg(cls):
        video_path = '../videos/陕西第二支预告片，武则天乾陵，气势磅礴，夕阳下拍到神奇的一幕/video1.mp4'
        audio_path = '../videos/陕西第二支预告片，武则天乾陵，气势磅礴，夕阳下拍到神奇的一幕/audio1.mp4'
        out_path = '../videos/陕西第二支预告片，武则天乾陵，气势磅礴，夕阳下拍到神奇的一幕/video.mp4'
        cmd = 'ffmpeg -i ' + video_path + ' -i ' + audio_path + ' -c:v copy -c:a aac -strict experimental -map 0:v:0 -map 1:a:0 ' + out_path
        print(cmd)
        subprocess.call(cmd, shell=True)

        print(os.listdir('../videos/陕西第二支预告片，武则天乾陵，气势磅礴，夕阳下拍到神奇的一幕/'))
        for file in os.listdir('../videos/陕西第二支预告片，武则天乾陵，气势磅礴，夕阳下拍到神奇的一幕/'):
            # print(file.title())
            if file.split('.mp4')[0] == 'video1' or file.split('.mp4')[0] == 'audio1':
                os.remove('../videos/陕西第二支预告片，武则天乾陵，气势磅礴，夕阳下拍到神奇的一幕/' + file)

    @classmethod
    def sheet(cls, log_type):
        sheet = Feishu.get_values_batch(log_type, 'xigua', 'e075e9')
        print(sheet)

    @classmethod
    def get_video_info(cls, gid):
        url = 'https://www.ixigua.com/api/mixVideo/information?'
        headers = {
            "accept-encoding": "gzip, deflate, br",
            "accept-language": "zh-CN,zh-Hans;q=0.9",
            "user-agent": "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) "
                          "AppleWebKit/605.1.15 (KHTML, like Gecko) Version/15.5 Safari/605.1.15",
            "referer": "https://www.ixigua.com/7102614741050196520?logTag=0531c88ac04f38ab2c62",
        }
        params = {
            'mixId': gid,
            'msToken': 'IlG0wd0Pylyw9ghcYiB2YseUmTwrsrqqhXrbIcsSaTcLTJyVlbYJzk20zw3UO-CfrfC'
                       'NVVIOBNjIl7vfBoxnVUwO9ZyzAI3umSKsT5-pef_RRfQCJwmA',
            'X-Bogus': 'DFSzswVupYTANCJOSBk0P53WxM-r',
            '_signature': '_02B4Z6wo0000119LvEwAAIDCuktNZ0y5wkdfS7jAALThuOR8D9yWNZ.EmWHKV0WSn6Px'
                          'fPsH9-BldyxVje0f49ryXgmn7Tzk-swEHNb15TiGqa6YF.cX0jW8Eds1TtJOIZyfc9s5emH7gdWN94',
        }
        cookies = {
            'ixigua-a-s': '1',
            'msToken': 'IlG0wd0Pylyw9ghcYiB2YseUmTwrsrqqhXrbIcsSaTcLTJyVlbYJzk20zw3UO-CfrfCNVVIOB'
                       'NjIl7vfBoxnVUwO9ZyzAI3umSKsT5-pef_RRfQCJwmA',
            'ttwid': '1%7C_yXQeHWwLZgCsgHClOwTCdYSOt_MjdOkgnPIkpi-Sr8%7C1661241238%7Cf57d0c5ef3f1d7'
                     '6e049fccdca1ac54887c34d1f8731c8e51a49780ff0ceab9f8',
            'tt_scid': 'QZ4l8KXDG0YAEaMCSbADdcybdKbUfG4BC6S4OBv9lpRS5VyqYLX2bIR8CTeZeGHR9ee3',
            'MONITOR_WEB_ID': '0a49204a-7af5-4e96-95f0-f4bafb7450ad',
            '__ac_nonce': '06304878000964fdad287',
            '__ac_signature': '_02B4Z6wo00f017Rcr3AAAIDCUVxeW1tOKEu0fKvAAI4cvoYzV-wBhq7B6D8k0no7lb'
                              'FlvYoinmtK6UXjRIYPXnahUlFTvmWVtb77jsMkKAXzAEsLE56m36RlvL7ky.M3Xn52r9t1IEb7IR3ke8',
            'ttcid': 'e56fabf6e85d4adf9e4d91902496a0e882',
            '_tea_utm_cache_1300': 'undefined',
            'support_avif': 'false',
            'support_webp': 'false',
            'xiguavideopcwebid': '7134967546256016900',
            'xiguavideopcwebid.sig': 'xxRww5R1VEMJN_dQepHorEu_eAc',
        }
        urllib3.disable_warnings()
        response = requests.get(url=url, headers=headers, params=params, cookies=cookies, proxies=proxies, verify=False)
        if 'data' not in response.json() or response.json()['data'] == '':
            print(f'response: {response}')
        else:
            video_info = response.json()['data']['gidInformation']['packerData']['video']

            video_url = ''
            audio_url = ''
            video_width = ''
            video_height = ''

            # video_url
            if 'videoResource' not in video_info:
                video_url = 0
                audio_url = 0
                video_width = 0
                video_height = 0
            elif 'dash' in video_info['videoResource']:
                video_url = video_info['videoResource']['dash']['dynamic_video']['dynamic_video_list'][-1]['main_url']
                audio_url = video_info['videoResource']['dash']['dynamic_video']['dynamic_audio_list'][-1]['main_url']
                video_url = base64.b64decode(video_url).decode('utf8')
                audio_url = base64.b64decode(audio_url).decode('utf8')
                video_width = video_info['videoResource']['dash']['dynamic_video']['dynamic_video_list'][-1]['vwidth']
                video_height = video_info['videoResource']['dash']['dynamic_video']['dynamic_video_list'][-1]['vheight']

            elif 'normal' in video_info['videoResource']:
                video_list = video_info['videoResource']['normal']['video_list']
                if 'video_4' in video_list.keys():
                    video_url = video_list['video_4']['main_url']
                    audio_url = video_list['video_4']['main_url']
                    video_url = base64.b64decode(video_url).decode('utf8')
                    audio_url = base64.b64decode(audio_url).decode('utf8')
                    video_width = video_list['video_4']['vwidth']
                    video_height = video_list['video_4']['vheight']
                elif 'video_3' in video_list.keys():
                    video_url = video_list['video_3']['main_url']
                    audio_url = video_list['video_3']['main_url']
                    video_url = base64.b64decode(video_url).decode('utf8')
                    audio_url = base64.b64decode(audio_url).decode('utf8')
                    video_width = video_list['video_3']['vwidth']
                    video_height = video_list['video_3']['vheight']
                elif 'video_2' in video_list.keys():
                    video_url = video_list['video_2']['main_url']
                    audio_url = video_list['video_2']['main_url']
                    video_url = base64.b64decode(video_url).decode('utf8')
                    audio_url = base64.b64decode(audio_url).decode('utf8')
                    video_width = video_list['video_2']['vwidth']
                    video_height = video_list['video_2']['vheight']
                elif 'video_1' in video_list.keys():
                    video_url = video_list['video_1']['main_url']
                    audio_url = video_list['video_1']['main_url']
                    video_url = base64.b64decode(video_url).decode('utf8')
                    audio_url = base64.b64decode(audio_url).decode('utf8')
                    video_width = video_list['video_1']['vwidth']
                    video_height = video_list['video_1']['vheight']
                else:
                    video_url = 0
                    audio_url = 0
                    video_width = 0
                    video_height = 0

            print(f'video_url:{video_url}')
            print(f'audio_url:{audio_url}')
            print(f'video_width:{video_width}')
            print(f'video_height:{video_height}')


if __name__ == '__main__':
    # Demo.base64_decode()
    # Demo.ffmpeg()
    Demo.sheet('xigua')
    # Demo.get_video_info('7133153239867851295')
    # Demo.get_signature()
    # Demo.get_logs()
    # Demo.download_video()

    pass