mirror of
https://github.com/yt-dlp/yt-dlp.git
synced 2024-11-25 00:00:48 +01:00
86e5f3ed2e
Using https://github.com/asottile/pyupgrade 1. `__future__` imports and `coding: utf-8` were removed 2. Files were rewritten with `pyupgrade --py36-plus --keep-percent-format` 3. f-strings were cherry-picked from `pyupgrade --py36-plus` Extractors are left untouched (except removing header) to avoid unnecessary merge conflicts
40 lines
1.4 KiB
Python
40 lines
1.4 KiB
Python
from .common import InfoExtractor
|
|
from ..compat import compat_str
|
|
from ..utils import int_or_none
|
|
|
|
|
|
class LiveJournalIE(InfoExtractor):
|
|
_VALID_URL = r'https?://(?:[^.]+\.)?livejournal\.com/video/album/\d+.+?\bid=(?P<id>\d+)'
|
|
_TEST = {
|
|
'url': 'https://andrei-bt.livejournal.com/video/album/407/?mode=view&id=51272',
|
|
'md5': 'adaf018388572ced8a6f301ace49d4b2',
|
|
'info_dict': {
|
|
'id': '1263729',
|
|
'ext': 'mp4',
|
|
'title': 'Истребители против БПЛА',
|
|
'upload_date': '20190624',
|
|
'timestamp': 1561406715,
|
|
}
|
|
}
|
|
|
|
def _real_extract(self, url):
|
|
video_id = self._match_id(url)
|
|
webpage = self._download_webpage(url, video_id)
|
|
record = self._parse_json(self._search_regex(
|
|
r'Site\.page\s*=\s*({.+?});', webpage,
|
|
'page data'), video_id)['video']['record']
|
|
storage_id = compat_str(record['storageid'])
|
|
title = record.get('name')
|
|
if title:
|
|
# remove filename extension(.mp4, .mov, etc...)
|
|
title = title.rsplit('.', 1)[0]
|
|
return {
|
|
'_type': 'url_transparent',
|
|
'id': video_id,
|
|
'title': title,
|
|
'thumbnail': record.get('thumbnail'),
|
|
'timestamp': int_or_none(record.get('timecreate')),
|
|
'url': 'eagleplatform:vc.videos.livejournal.com:' + storage_id,
|
|
'ie_key': 'EaglePlatform',
|
|
}
|