mirror of
https://github.com/yt-dlp/yt-dlp.git
synced 2024-11-18 15:11:45 +00:00
45 lines
1.5 KiB
Python
45 lines
1.5 KiB
Python
|
import codecs
|
||
|
|
||
|
from .common import InfoExtractor
|
||
|
|
||
|
|
||
|
class WebcameraplIE(InfoExtractor):
|
||
|
_VALID_URL = r'https?://(?P<id>[\w-]+)\.webcamera\.pl'
|
||
|
_TESTS = [{
|
||
|
'url': 'https://warszawa-plac-zamkowy.webcamera.pl',
|
||
|
'info_dict': {
|
||
|
'id': 'warszawa-plac-zamkowy',
|
||
|
'ext': 'mp4',
|
||
|
'title': r're:WIDOK NA PLAC ZAMKOWY W WARSZAWIE \d{4}-\d{2}-\d{2} \d{2}:\d{2}$',
|
||
|
'live_status': 'is_live',
|
||
|
}
|
||
|
}, {
|
||
|
'url': 'https://gdansk-stare-miasto.webcamera.pl/',
|
||
|
'info_dict': {
|
||
|
'id': 'gdansk-stare-miasto',
|
||
|
'ext': 'mp4',
|
||
|
'title': r're:GDAŃSK - widok na Stare Miasto \d{4}-\d{2}-\d{2} \d{2}:\d{2}$',
|
||
|
'live_status': 'is_live',
|
||
|
}
|
||
|
}]
|
||
|
|
||
|
def _real_extract(self, url):
|
||
|
video_id = self._match_id(url)
|
||
|
webpage = self._download_webpage(url, video_id)
|
||
|
|
||
|
rot13_m3u8_url = self._search_regex(r'data-src\s*=\s*"(uggc[^"]+\.z3h8)"',
|
||
|
webpage, 'm3u8 url', default=None)
|
||
|
if not rot13_m3u8_url:
|
||
|
self.raise_no_formats('No video/audio found at the provided url', expected=True)
|
||
|
|
||
|
m3u8_url = codecs.decode(rot13_m3u8_url, 'rot-13')
|
||
|
formats, subtitles = self._extract_m3u8_formats_and_subtitles(m3u8_url, video_id, live=True)
|
||
|
|
||
|
return {
|
||
|
'id': video_id,
|
||
|
'title': self._html_search_regex(r'<h1\b[^>]*>([^>]+)</h1>', webpage, 'title'),
|
||
|
'formats': formats,
|
||
|
'subtitles': subtitles,
|
||
|
'is_live': True,
|
||
|
}
|