heckyel/yt-local
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Handle case where embedded player response missing

Browse Source 
Change so it extracts other stuff from regular playerResponse
Extract formats from embedded player response, but fallback to
regular one if that doesn't work.
Sometimes there is no 'player' at top_level and the urls are in
the regular playerResponse
This commit is contained in:
James Taylor 2020-06-28 13:18:54 -07:00
parent 0b5d6fe1ed
commit 6e14a8547d
1 changed files with 10 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
youtube/yt_data_extract/watch_extraction.py
View File

@ -415,8 +415,14 @@ def extract_watch_info(polymer_json):
if error:
info['playability_error'] = error
player_response = top_level.get('playerResponse', {})
# usually, only the embedded one has the urls
player_args = deep_get(top_level, 'player', 'args', default={})
player_response = json.loads(player_args['player_response']) if 'player_response' in player_args else {}
if 'player_response' in player_args:
embedded_player_response = json.loads(player_args['player_response'])
else:
embedded_player_response = {}
# captions
info['automatic_caption_languages'] = []
@ -446,7 +452,9 @@ def extract_watch_info(polymer_json):
print('WARNING: Found non-translatable caption language')
# formats
_extract_formats(info, player_response)
_extract_formats(info, embedded_player_response)
if not info['formats']:
_extract_formats(info, player_response)
# playability errors
_extract_playability_error(info, player_response)