R-yan/l1ving_youtube-dl
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

Facebook timestamp fix (#320)

Browse Source 
* missing metadate fix

* timestamp fix

* timestamp conditions fix

* timestamp conditions fix

* timestamp conditions fix

Co-authored-by: bhodaya <bhodaya@videocites.com>
This commit is contained in:
hodayabu 2020-05-25 16:17:52 +03:00 committed by GitHub
parent 5fada70cbf
commit 846b092130
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 24 additions and 10 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

34
youtube_dl/extractor/facebook.py
View File

@ -454,16 +454,30 @@ class FacebookIE(InfoExtractor):
self._search_regex(r'ownerName"\s*:\s*"([^"]+)"', webpage, 'uploader', default=None) or \
self._og_search_title(webpage, default=None)
if webpage.find('Paid Partnership'):
timestamp = self._search_regex(
r'datePublished":"(.+?)"', webpage,
'timestamp', default=None)
timestamp = parse_iso8601(timestamp)
else:
timestamp = int_or_none(
self._search_regex(r'data-utime=\\\"(\d+)\\\"', tahoe_data.secondary,'timestamp', default=None)
or self._search_regex(r'<abbr[^>]+data-utime=["\'](\d+)', webpage, 'timestamp', default=None)
) or int_or_none(self._search_regex(r'publish_time&quot;:([\d]+)', webpage, 'timestamp', default=None))
timestamp = self._search_regex(
r'datePublished":"(.+?)"', webpage,'timestamp', default=None)\
or self._search_regex(r'datePublished":"(.+?)"', tahoe_data.secondary, 'timestamp', default=None)\
or self._search_regex(r'datePublished":"(.+?)"', tahoe_data.primary, 'timestamp', default=None)
timestamp = parse_iso8601(timestamp)
if timestamp == None and webpage.find('Paid Partnership') == -1 or\
(timestamp == None and webpage.find('Paid Partnership') > -1 and
'cookiefile' in self._downloader.params):
regex_search_result_date_time = self._search_regex(r'data-utime=\\\"(\d+)\\\"', tahoe_data.secondary, 'timestamp', default=None)\
or self._search_regex(r'data-utime=\\\"(\d+)\\\"', tahoe_data.primary, 'timestamp', default=None)\
or self._search_regex(r'data-utime=\\\"(\d+)\\\"', webpage,'timestamp', default=None)\
or self._search_regex(r'<abbr[^>]+data-utime=["\'](\d+)', webpage, 'timestamp', default=None)\
or self._search_regex(r'<abbr[^>]+data-utime=["\'](\d+)', tahoe_data.secondary, 'timestamp', default=None)\
or self._search_regex(r'<abbr[^>]+data-utime=["\'](\d+)', tahoe_data.primary, 'timestamp', default=None)
regex_search_result_publish_time = self._search_regex(r'publish_time&quot;:([\d]+)', webpage, 'timestamp', default=None)\
or self._search_regex(r'publish_time&quot;:([\d]+)', tahoe_data.primary, 'timestamp', default=None)\
or self._search_regex(r'publish_time&quot;:([\d]+)', tahoe_data.secondary, 'timestamp', default=None)
timestamp = int_or_none(regex_search_result_date_time) or int_or_none(regex_search_result_publish_time)
uploader_id = self._search_regex(