0
0
Fork 0
mirror of https://github.com/yt-dlp/yt-dlp.git synced 2024-11-30 03:33:02 +00:00

[extractor/epoch] Support videos without data-trailer (#5387)

Closes #5359
Authored by: gibson042, pukkandan
This commit is contained in:
Richard Gibson 2022-11-06 12:23:16 -05:00 committed by GitHub
parent 049565df2e
commit 7053aa3a48
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
2 changed files with 11 additions and 1 deletions

View file

@ -1,4 +1,5 @@
from .common import InfoExtractor from .common import InfoExtractor
from ..utils import extract_attributes, get_element_html_by_id
class EpochIE(InfoExtractor): class EpochIE(InfoExtractor):
@ -28,13 +29,21 @@ class EpochIE(InfoExtractor):
'title': 'Kash Patel: A 6-Year-Saga of Government Corruption, From Russiagate to Mar-a-Lago', 'title': 'Kash Patel: A 6-Year-Saga of Government Corruption, From Russiagate to Mar-a-Lago',
} }
}, },
{
'url': 'https://www.theepochtimes.com/dick-morris-discusses-his-book-the-return-trumps-big-2024-comeback_4819205.html',
'info_dict': {
'id': '9489f994-2a20-4812-b233-ac0e5c345632',
'ext': 'mp4',
'title': 'Dick Morris Discusses His Book The Return: Trumps Big 2024 Comeback',
}
},
] ]
def _real_extract(self, url): def _real_extract(self, url):
video_id = self._match_id(url) video_id = self._match_id(url)
webpage = self._download_webpage(url, video_id) webpage = self._download_webpage(url, video_id)
youmaker_video_id = self._search_regex(r'data-trailer="[\w-]+" data-id="([\w-]+)"', webpage, 'url') youmaker_video_id = extract_attributes(get_element_html_by_id('videobox', webpage))['data-id']
formats, subtitles = self._extract_m3u8_formats_and_subtitles( formats, subtitles = self._extract_m3u8_formats_and_subtitles(
f'http://vs1.youmaker.com/assets/{youmaker_video_id}/playlist.m3u8', video_id, 'mp4', m3u8_id='hls') f'http://vs1.youmaker.com/assets/{youmaker_video_id}/playlist.m3u8', video_id, 'mp4', m3u8_id='hls')

View file

@ -524,6 +524,7 @@ def __init__(self):
def handle_starttag(self, tag, attrs): def handle_starttag(self, tag, attrs):
self.attrs = dict(attrs) self.attrs = dict(attrs)
raise compat_HTMLParseError('done')
class HTMLListAttrsParser(html.parser.HTMLParser): class HTMLListAttrsParser(html.parser.HTMLParser):