春江暮客

春江暮客的个人学习分享网站

Downloading m3u8 Streaming Media Using Python and Embedding into AMP Webpages

2023-08-01 Technology
Downloading m3u8 Streaming Media Using Python and Embedding into AMP Webpages

In short video and live-streaming workflows, you often do not get a direct MP4 file. Instead, you get an m3u8 playlist plus a set of .ts segments. If you want to archive the content locally, process it with scripts, or embed the result into your own page, you need to handle that pipeline yourself.

This article focuses on two practical steps:

  1. Download the TS segments behind an m3u8 playlist with Python
  2. Merge them into MP4 and embed the output into an AMP page

If what you want is a newer browser-based M3U8 to MP4 workflow, I also link that updated guide near the end.

0. Setup

Install the Python packages used in the example:

pip install m3u8 requests

If you also want to merge TS files into MP4, make sure ffmpeg is installed.

1. What is m3u8?

The M3U8 format is based on HTTP Live Streaming (HLS) technology, which splits the entire video into a series of small files, then creates an index file (.m3u8). Through this index file, streaming media can be loaded. This means that when playing a video, the client only needs to load the index file and then download the small video files one by one according to the index file for playback, enabling streaming playback while downloading.

2. Downloading all video files from an m3u8 file

First, we need to obtain the m3u8 file and parse the URLs of the video segments (.TS files) inside it. To handle network requests, we use Python’s requests library. Here is a snippet of code:

import m3u8
import requests
import os

def check_path(path):
    if not os.path.exists(path):
        os.makedirs(path)

headers = {
    "User-Agent": "Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101 Firefox/102.0"
}

domain = 'www.bobobk.com'

def download_m3u8(m3u8url):
    print(m3u8url)
    playlist_filename = os.path.join(*m3u8url.split('/')[3:])
    if os.path.exists(playlist_filename):
        pass
    check_path(os.path.dirname(playlist_filename))

    host = m3u8url.split("://")[1].split('/')[0]
    baseurl, m3u8url = m3u8url.split("://")
    if not os.path.exists(playlist_filename):
        text = requests.get(m3u8url, headers=headers).text
        open(playlist_filename, 'w').write(text.replace(host, domain))

    m3u8_obj = m3u8.load(playlist_filename)
    print(f"playlist:{playlist_filename}")
    baseurl += "://" + m3u8url.split("/")[0]
    print("start check playlist")
    if len(m3u8_obj.playlists) != 0:
        m3u8url = m3u8_obj.playlists[0].uri
        if '://' not in m3u8url:
            if m3u8url[0] == '/':
                m3u8url = baseurl + m3u8_obj.playlists[0].uri
            else:
                m3u8url = baseurl + '/' + os.path.dirname(playlist_filename)+ '/' + m3u8_obj.playlists[0].uri
        print(m3u8url)
        text = requests.get(m3u8url, headers=headers).text
        playlist_filename = os.path.join(*m3u8url.split('/')[3:])
        check_path(os.path.dirname(playlist_filename))
        open(playlist_filename, 'w').write(text.replace(host, domain))
        m3u8_obj = m3u8.load(playlist_filename)
        baseurl, m3u8url = m3u8url.split("://")
        host = m3u8url.split('/')[0]
        print(f"playlist:{playlist_filename}")

    output = os.path.join(*m3u8_obj.segments[0].absolute_uri.split('/')[3:])
    check_path(os.path.dirname(output))
    print("check subdir")
    for segment in m3u8_obj.segments:
        segment_url = segment.absolute_uri
        segment_file_name = os.path.join(*segment_url.split('/')[3:])
        r = requests.get(segment_url, headers=headers, stream=True)
        print(f"{segment_url}t{segment_file_name}")
        if os.path.exists(segment_file_name):
            continue

        with open(segment_file_name, 'wb') as f:
            for chunk in r.iter_content(chunk_size=1024):
                if chunk:
                    f.write(chunk)


m3u8url = "https://pptv.sd-play.com/202308/01/PfdfHWiJa43/video/index.m3u8"
download_m3u8(m3u8url)

One useful part of this script is that it handles master playlists by continuing down to the actual media playlist before downloading the .ts segments.

3. Embedding m3u8 video into an AMP webpage

AMP (Accelerated Mobile Pages) is an open-source framework optimized for mobile devices. Using this framework, you can create fast and visually appealing web applications. It standardizes but also flexibly defines HTML usage, making responsive web development easier.

Directly embedding m3u8 videos in AMP has a limitation: AMP currently does not support direct playback of m3u8 files. Therefore, we need to convert the video into MP4 format before embedding it into the AMP webpage.

We can use the ffmpeg tool to merge .ts files and convert them into .mp4 files. Here is the code:

for tsfile in "202308/01/PfdfHWiJa43/video/900k_0X480_64k_25/hls/"*.ts;do
echo "file '${tsfile}'" >> tsfile.txt
done

ffmpeg -f concat -safe 0 -i tsfile.txt -c copy 202308/01/PfdfHWiJa43/video/900k_0X480_64k_25/hls/202308.mp4

Then, we can use the amp-video component in AMP webpages to play the mp4 video:

https://www.bobobk.com/m3u8.html

<!doctype html> 
<html amp> 
<head> 
<meta charset="utf-8"> 
<script async src="https://cdn.ampproject.org/v0.js"></script> 
<title>Downloading m3u8 Streaming Media Using Python and Embedding into AMP Webpages (amp video page)</title> 
<link rel="canonical" href="self.html"> 
<meta name="viewport" content="width=device-width,minimum-scale=1,initial-scale=1"> 
<style amp-boilerplate>body{-webkit-animation:-amp-start 8s steps(1,end) 0s 1 normal both;-moz-animation:-amp-start 8s steps(1,end) 0s 1 normal both;-ms-animation:-amp-start 8s steps(1,end) 0s 1 normal both;animation:-amp-start 8s steps(1,end) 0s 1 normal both}@-webkit-keyframes -amp-start{from{visibility:hidden}to{visibility:visible}}@-moz-keyframes -amp-start{from{visibility:hidden}to{visibility:visible}}@-ms-keyframes -amp-start{from{visibility:hidden}to{visibility:visible}}@-o-keyframes -amp-start{from{visibility:hidden}to{visibility:visible}}@keyframes -amp-start{from{visibility:hidden}to{visibility:visible}}</style>
<noscript><style amp-boilerplate>body{-webkit-animation:none;-moz-animation:none;-ms-animation:none;animation:none}</style></noscript> 
<script async custom-element="amp-video" src="https://cdn.ampproject.org/v0/amp-video-0.1.js"></script> 
</head> 
<body> 
<amp-video controls width="640" height="360" layout="responsive" > <source src="/202308/01/PfdfHWiJa43/video/900k_0X480_64k_25/hls/202308.mp4" type="video/mp4"> </amp-video> 
</body> 
</html>

4. How to validate the download and embed flow

Check at least these three things:

  1. The target directory now contains the expected .ts segment files
  2. The merged MP4 plays normally
  3. The AMP page loads the converted MP4 correctly on mobile

The fastest local checks are:

find 202308/01/PfdfHWiJa43/video/ -name '*.ts' | wc -l
ffmpeg -i 202308/01/PfdfHWiJa43/video/900k_0X480_64k_25/hls/202308.mp4

5. Common problems and fixes

1. Download stops halfway

Most common cause: source-site rate limits, expired URLs, or inaccessible segments.

Fix:

  1. Use a more stable request header set
  2. Refresh the source m3u8 URL
  3. Resume by skipping already-downloaded segments

2. ffmpeg merge fails

Most common cause: incomplete tsfile.txt input or missing segments.

Fix:

  1. Confirm all .ts files are present
  2. Rebuild tsfile.txt
  3. Run the concat command again

3. AMP cannot play m3u8 directly

This is an AMP limitation, not a Python issue.

Fix:

  • Convert the stream to MP4 first
  • Then embed the MP4 with amp-video

7. Conclusion

This article provided a detailed walkthrough of downloading m3u8 streaming media using Python and embedding it into an AMP webpage. I hope it helps you. In practical use, please ensure you fully comply with the streaming service’s terms and respect relevant copyright laws.

相关文章

m3u8 amp python

最新文章

分类

标签

1024 adsense agents.md ai ai-agent ai-seo algorithm amp automation background bioinformatics bootstrapping boxes c-index cca cdn cli cloudflare codex copy cpu monitoring cuda datascience datavisualization deployment devtools disown docker dovecot download economics fable faceswap fastmcp ffmpeg flask folium frontend game generator git github-actions google grep hls hugo indexnow javascript json k-means kaggle leecode linux list litellm llm llms.txt logs lollipop m3u8 manacher matplotlib mcp mirror mp3 mp4 multiomics mutation mysql networkx nginx normalize password pep-723 phaser pillow pip postfix preprocessing print probability proxy pyecharts pyqt python python3 r raincloud requests rg ripgrep roundcube s-tui sampling scale scrapy screen seaborn security seo sklearn solana somaticsignatures spl standardize static-site tensorflow trade tron tronpy turtle usdt uv vite wallet webp wordcloud wordpress workflow yaml 迅雷解析

友情链接

其它