paradizelost/youtube-dl
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

[motherless] Fix recent videos upload date extraction (closes #27661)

Browse Source 
Less than a week old videos use a '20h ago' or '1d ago' format.

I kept the support for 'Ago' with uppercase start at is was already in the code.
This commit is contained in:
cladmi 2020-12-16 09:54:48 +01:00 committed by Sergey M․
parent c3399cac19
commit f318882955
No known key found for this signature in database
GPG Key ID: 2C393E0F18A9236D
1 changed files with 14 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

19
youtube_dl/extractor/motherless.py
View File

@ -85,18 +85,27 @@ class MotherlessIE(InfoExtractor):
or 'http://cdn4.videos.motherlessmedia.com/videos/%s.mp4?fs=opencloud' % video_id) or 'http://cdn4.videos.motherlessmedia.com/videos/%s.mp4?fs=opencloud' % video_id)
age_limit = self._rta_search(webpage) age_limit = self._rta_search(webpage)
view_count = str_to_int(self._html_search_regex( view_count = str_to_int(self._html_search_regex(
(r'>(\d+)\s+Views<', r'<strong>Views</strong>\s+([^<]+)<'), (r'>([\d,.]+)\s+Views<', # 1,234,567 Views
r'<strong>Views</strong>\s+([^<]+)<'),
webpage, 'view count', fatal=False)) webpage, 'view count', fatal=False))
like_count = str_to_int(self._html_search_regex( like_count = str_to_int(self._html_search_regex(
(r'>(\d+)\s+Favorites<', r'<strong>Favorited</strong>\s+([^<]+)<'), (r'>([\d,.]+)\s+Favorites<', # 1,234 Favorites
r'<strong>Favorited</strong>\s+([^<]+)<'),
webpage, 'like count', fatal=False)) webpage, 'like count', fatal=False))
upload_date = self._html_search_regex( upload_date = self._html_search_regex(
(r'class=["\']count[^>]+>(\d+\s+[a-zA-Z]{3}\s+\d{4})<', (r'class=["\']count[^>]+>(\d+\s+[a-zA-Z]{3}\s+\d{4})<',
r'class=["\']count[^>]+>(\d+[hd])\s+[aA]go<', # 20h/1d ago
r'<strong>Uploaded</strong>\s+([^<]+)<'), webpage, 'upload date') r'<strong>Uploaded</strong>\s+([^<]+)<'), webpage, 'upload date')
if 'Ago' in upload_date: relative = re.match(r'(\d+)([hd])$', upload_date)
days = int(re.search(r'([0-9]+)', upload_date).group(1)) if relative:
upload_date = (datetime.datetime.now() - datetime.timedelta(days=days)).strftime('%Y%m%d') delta = int(relative.group(1))
unit = relative.group(2)
if unit == 'h':
delta_t = datetime.timedelta(hours=delta)
else: # unit == 'd'
delta_t = datetime.timedelta(days=delta)
upload_date = (datetime.datetime.now() - delta_t).strftime('%Y%m%d')
else: else:
upload_date = unified_strdate(upload_date) upload_date = unified_strdate(upload_date)