github/yt-dlp
1
0
Fork 0
mirror of https://github.com/yt-dlp/yt-dlp.git synced 2024-11-14 20:38:11 -05:00
Code Issues Projects Releases Packages Wiki Activity

[ie/archive.org] Fix comments extraction (#11527)

Browse source 
Closes #11526
Authored by: jshumphrey
This commit is contained in:
Jackson Humphrey 2024-11-12 17:26:18 -06:00 committed by GitHub
parent bacc31b05a
commit f2a4983df7
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
1 changed files with 21 additions and 1 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

22
yt_dlp/extractor/archiveorg.py
View file

@ -205,6 +205,26 @@ class ArchiveOrgIE(InfoExtractor):
}, },
}, },
], ],
}, {
# The reviewbody is None for one of the reviews; just need to extract data without crashing
'url': 'https://archive.org/details/gd95-04-02.sbd.11622.sbeok.shnf/gd95-04-02d1t04.shn',
'info_dict': {
'id': 'gd95-04-02.sbd.11622.sbeok.shnf/gd95-04-02d1t04.shn',
'ext': 'mp3',
'title': 'Stuck Inside of Mobile with the Memphis Blues Again',
'creators': ['Grateful Dead'],
'duration': 338.31,
'track': 'Stuck Inside of Mobile with the Memphis Blues Again',
'description': 'md5:764348a470b986f1217ffd38d6ac7b72',
'display_id': 'gd95-04-02d1t04.shn',
'location': 'Pyramid Arena',
'uploader': 'jon@archive.org',
'album': '1995-04-02 - Pyramid Arena',
'upload_date': '20040519',
'track_number': 4,
'release_date': '19950402',
'timestamp': 1084927901,
},
}] }]
@staticmethod @staticmethod
@ -335,7 +355,7 @@ def _real_extract(self, url):
info['comments'].append({ info['comments'].append({
'id': review.get('review_id'), 'id': review.get('review_id'),
'author': review.get('reviewer'), 'author': review.get('reviewer'),
'text': str_or_none(review.get('reviewtitle'), '') + '\n\n' + review.get('reviewbody'), 'text': join_nonempty('reviewtitle', 'reviewbody', from_dict=review, delim='\n\n'),
'timestamp': unified_timestamp(review.get('createdate')), 'timestamp': unified_timestamp(review.get('createdate')),
'parent': 'root'}) 'parent': 'root'})