Extract YouTube MPD Information

A Python script to extract youtubes MPD file and print the information to the console

Requirements

• (Linux)
• Python 2.7 or better
• libxml2 (e.g., sudo apt-get install python-libxml2)
• urllib2

Usage

Example Call (will download the MPD file to aspen.mpd)

python extract.py https://www.youtube.com/watch?v=GTGUa4J8XKw aspen.mpd

Output (of the above Example, URLs shortened/anonymized):

Downloading HTML of https://www.youtube.com/watch?v=GTGUa4J8XKw 
Extracted MPDURL from HTML:  
https://manifest.googlevideo.com/api/manifest/dash/sparams/..... 
AdaptationSet,RepresentationID,Bitrate,Codec,ExtraInformation 
audio/mp4,140,127570,mp4a.40.2
video/mp4,133,247800,avc1.4d4015,426/240/24
video/mp4,134,601944,avc1.4d401e,640/360/24
video/mp4,135,1103336,avc1.4d401e,854/480/24
video/mp4,160,109967,avc1.42c00c,256/144/12
video/mp4,136,2206969,avc1.4d401f,1280/720/24
video/mp4,137,4144774,avc1.640028,1920/1080/24

Output Information

MPDs usually contain multiple AdaptationSet XML tags. Those contain a MimeType of either:

• audio/mp4
• audio/webm
• video/mp4
• video/webm

Within one AdaptationSet there are multiple Representation tags, which then contain information for each representation:

• Bitrate (in Kilobit per second)
• Codec
• Information for Video Only (in ExtraInformation column of generated output):
  • Video Width
  • Video Height
  • Video Frame Rate

Acknowledgements

This work was partly funded by the Austrian Science Fund (FWF) under the CHIST-ERA project CONCERT (A Context-Adaptive Content Ecosystem Under Uncertainty), project number I1402 (see [http://www.concert-project]() for more details).