Python爬取全球最大视频网站YouTube视频

前言

作为目前全世界最大的视频网站，它几乎全是用Python来写的
该网站当前行业内在线视频服务提供商，该网站的系统每天要处理上千万个视频片段，为全球成千上万的用户提供高水平的视频上传、分发、展示、浏览服务。2015年2月，央视首次把春晚推送到该网站。
今天，我们就要用Python来快速批量下载该网站的视频

开发环境

版 本： python 3.8
编辑器：pycharm 2021.2
第三方模块：requests + tqdm

所需模块

import requests
import re
import json
from tqdm import tqdm
import os

 

开始代码编写

请求数据

headers = {
    'cookie': 'VISITOR_INFO1_LIVE=9qZVrzB27uI; PREF=f4=4000000&tz=Asia.Shanghai; _ga=GA1.2.621834420.1648121145; _gcl_au=1.1.1853038046.1648121145; NID=511=Zc1APdmEbCD-iqVNVgI_vD_0S3LVI3XSfl-wUZEvvMU2MLePFKsQCaKUlUtchHSg-kWEVMGOhWUbxpQMwHeIuLjhxaslwniMh1OsjVfmOeTfhpwcRYpMgqpZtNQ7qQApY21xEObCvIez6DCMbjRhRQ5P7siOD3X87QX0CFyUxmY; OTZ=6430350_24_24__24_; GPS=1; YSC=0E115KqM_-I; GOOGLE_ABUSE_EXEMPTION=ID=d02004902c3d0f4d:TM=1648620854:C=r:IP=47.57.243.77-:S=YmZXPW7dxbu83bDuauEpXpE; CONSISTENCY=AGDxDeNysJ2boEmzRP4v6cwgg4NsdN4-FYQKHCGhA0AeW1QjFIU1Ejq1j8l6lwAc6c-pYTJiSaQItZ1M6QeI1pQ3wictnWXTOZ6_y8EKlt0Y_JdakwW6srR39-NLuPgSgXrXwtS0XTUGXpdnt4k3JjQ',
    'referer': 'https://www.youtube.com/results?search_query=jk%E7%BE%8E%E5%A5%B3',
    'user-agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/99.0.4844.84 Safari/537.36'
}
url = 'https://www.有土比.com/watch?v=ImoXcSpR_io'
response = requests.get(url=url, headers=headers)

 

解析数据

json_str = re.findall('var ytInitialPlayerResponse = (.*?);var', response.text)[0]
json_data = json.loads(json_str)
video_url = json_data['streamingData']['adaptiveFormats'][0]['url']
audio_url = json_data['streamingData']['adaptiveFormats'][-2]['url']
title = json_data['videoDetails']['title']
title = title.replace(' ', '')
title = re.sub(r'[\/:|?*"<>]', '', title)

 

视频数据

video_pbar = tqdm(total=file_size)
with open(f'{title}.mp4', mode='wb') as f:
    for video_chunk in video.iter_content(1024*1024*2):
        f.write(video_chunk)
        video_pbar.set_description(f'正在下载{title}视频中......')
        video_pbar.update(1024*1024*2)
    video_pbar.set_description('下载完成！')
    video_pbar.close()

 

音频数据

audio = requests.get(audio_url, stream=True)
file_size = int(audio.headers.get('Content-Length'))
audio_pbar = tqdm(total=file_size)
with open(f'{title}.mp3', mode='wb') as f:
    for audio_chunk in audio.iter_content(1024*1024*2):
        f.write(audio_chunk)
        audio_pbar.set_description(f'正在下载{title}音频中......')
        audio_pbar.update(1024*1024*2)
    audio_pbar.set_description('下载完成！')
    audio_pbar.close()

 

合并音频和视频

def merge(title):
    ffmpeg = r'D:\Download\ffmpeg\bin\ffmpeg.exe -i ' + title + '.mp4 -i ' + title + '.mp3 -acodec copy -vcodec copy ' + title + '-out.mp4'
    os.popen(ffmpeg)