mirror of
https://github.com/yt-dlp/yt-dlp.git
synced 2024-11-24 02:35:12 +00:00
19dbaeece3
Now _sort_formats should be called explicitly. _sort_formats has been added to all the necessary places in code. Closes #8051
56 lines
1.8 KiB
Python
56 lines
1.8 KiB
Python
# coding: utf-8
|
||
from __future__ import unicode_literals
|
||
|
||
import random
|
||
|
||
from .common import InfoExtractor
|
||
from ..compat import compat_urllib_parse_urlencode
|
||
from ..utils import (
|
||
sanitized_Request,
|
||
xpath_text,
|
||
)
|
||
|
||
|
||
class MatchTVIE(InfoExtractor):
|
||
_VALID_URL = r'https?://matchtv\.ru/?#live-player'
|
||
_TEST = {
|
||
'url': 'http://matchtv.ru/#live-player',
|
||
'info_dict': {
|
||
'id': 'matchtv-live',
|
||
'ext': 'flv',
|
||
'title': 're:^Матч ТВ - Прямой эфир \d{4}-\d{2}-\d{2} \d{2}:\d{2}$',
|
||
'is_live': True,
|
||
},
|
||
'params': {
|
||
'skip_download': True,
|
||
},
|
||
}
|
||
|
||
def _real_extract(self, url):
|
||
video_id = 'matchtv-live'
|
||
request = sanitized_Request(
|
||
'http://player.matchtv.ntvplus.tv/player/smil?%s' % compat_urllib_parse_urlencode({
|
||
'ts': '',
|
||
'quality': 'SD',
|
||
'contentId': '561d2c0df7159b37178b4567',
|
||
'sign': '',
|
||
'includeHighlights': '0',
|
||
'userId': '',
|
||
'sessionId': random.randint(1, 1000000000),
|
||
'contentType': 'channel',
|
||
'timeShift': '0',
|
||
'platform': 'portal',
|
||
}),
|
||
headers={
|
||
'Referer': 'http://player.matchtv.ntvplus.tv/embed-player/NTVEmbedPlayer.swf',
|
||
})
|
||
video_url = self._download_json(request, video_id)['data']['videoUrl']
|
||
f4m_url = xpath_text(self._download_xml(video_url, video_id), './to')
|
||
formats = self._extract_f4m_formats(f4m_url, video_id)
|
||
self._sort_formats(formats)
|
||
return {
|
||
'id': video_id,
|
||
'title': self._live_title('Матч ТВ - Прямой эфир'),
|
||
'is_live': True,
|
||
'formats': formats,
|
||
}
|