如何在Python中从视频链接下载音频

Question

我有一个正在迭代的链接列表，例如下面的

https://www.loc.gov/item/2015669100/
https://www.loc.gov/item/2015669101/
https://www.loc.gov/item/2015669102/
https://www.loc.gov/item/2015669103/
https://www.loc.gov/item/2015669104/
https://www.loc.gov/item/2015669105/
https://www.loc.gov/item/2015669106/
https://www.loc.gov/item/2015669107/
https://www.loc.gov/item/2015669108/
https://www.loc.gov/item/2015669109/

如果您查看这些链接，您可以看到它有一个视频和一个可下载的XML文件。我的任务是从视频中下载音频，并从一个页面一起下载XML文件。

我的问题是，如何从这样的音频文件中获取音频？

下面是我目前的代码。

from bs4 import BeautifulSoup
from urllib.request import Request, urlopen
base_html = "https://www.loc.gov/collections/civil-rights-history-project/?sp={}"
for i in range(1,8):
html = base_html.format(i)
req = Request(html, headers={'User-Agent': 'Mozilla/5.0'})
soup = BeautifulSoup(urlopen(req).read(), 'html.parser')

pages = soup.findAll('div', attrs={'class' : 'item-description'})
for div in pages:
crawl_p = div.find('a')['href']
#some logic here

Answer 1

查看该网站，它看起来像是通过传输段(.ts文件(流式传输的视频和音频文件，形成m3u8url。

<source>包含m3u8流的url。标记是使用其在标记中的属性来识别的。(这里是类型属性(。

import subprocess
from bs4 import BeautifulSoup
import requests as r
sess = r.session()
site_url = "https://www.loc.gov/item/2015669100/"
request = sess.get(site_url)
#print(request.content)
soup = BeautifulSoup(request.content, 'html5lib')
m3u8_url = soup.find('source',attrs={'type' : 'application/x-mpegURL'})['src']
print(str(m3u8_url))

这将废弃m3u8 url、https://tile.loc.gov/streaming-services/iiif/service:afc:afc2010039:afc2010039_crhp0001:afc2010039_crhp0001_mv04/full/full/0/full/default.m3u8

ffmpeg可以从m3u8文件下载流(视频或音频(。它也可以从python中运行。

subprocess.call(['ffmpeg','-i',m3u8_url,'-vn','-map','a','output.ts' ])

不要忘记用[]包装命令。引号内的每个单词表示一个空格分隔的命令。完整的ffmpeg命令是ffmpeg -i m3u8_url -vn -map a output.ts

这是完整的代码。只需确保包含ffmpeg路径变量，否则子进程将抛出错误。文件大小较大，因此下载音频文件可能需要一段时间。

import subprocess
from bs4 import BeautifulSoup
import requests as r
sess = r.session()
site_url = "https://www.loc.gov/item/2015669100/"
request = sess.get(site_url)
#print(request.content)
soup = BeautifulSoup(request.content, 'html5lib')
m3u8_url = soup.find('source',attrs={'type' : 'application/x-mpegURL'})['src']
print(str(m3u8_url))
subprocess.call(['ffmpeg','-i',m3u8_url,'-vn','-map','a','output.ts' ])