[Theta] Add extractor (#1068)

Authored by: alerikaisattera
This commit is contained in:
Aleri Kaisattera 2021-09-24 06:53:51 +06:00 committed by GitHub
parent 99e9e001de
commit eb6d4ad1ca
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
2 changed files with 52 additions and 0 deletions

View File

@ -1428,6 +1428,7 @@ from .theplatform import (
from .thescene import TheSceneIE
from .thestar import TheStarIE
from .thesun import TheSunIE
from .theta import ThetaIE
from .theweatherchannel import TheWeatherChannelIE
from .thisamericanlife import ThisAmericanLifeIE
from .thisav import ThisAVIE

51
yt_dlp/extractor/theta.py Normal file
View File

@ -0,0 +1,51 @@
# coding: utf-8
from __future__ import unicode_literals
from .common import InfoExtractor
from ..utils import try_get
class ThetaIE(InfoExtractor):
_VALID_URL = r'https?://(?:www\.)?theta\.tv/(?P<id>[a-z0-9]+)'
_TESTS = [{
'url': 'https://www.theta.tv/davirus',
'skip': 'The live may have ended',
'info_dict': {
'id': 'DaVirus',
'ext': 'mp4',
'title': 'I choose you - My Community is King -👀 - YO HABLO ESPANOL - CODE DAVIRUS',
'thumbnail': r're:https://live-thumbnails-prod-theta-tv\.imgix\.net/thumbnail/.+\.jpg',
}
}, {
'url': 'https://www.theta.tv/mst3k',
'note': 'This channel is live 24/7',
'info_dict': {
'id': 'MST3K',
'ext': 'mp4',
'title': 'Mystery Science Theatre 3000 24/7 Powered by the THETA Network.',
'thumbnail': r're:https://user-prod-theta-tv\.imgix\.net/.+\.jpg',
}
}]
def _real_extract(self, url):
channel_id = self._match_id(url)
info = self._download_json(f'https://api.theta.tv/v1/channel?alias={channel_id}', channel_id)['body']
m3u8_playlist = next(
data['url'] for data in info['live_stream']['video_urls']
if data.get('type') != 'embed' and data.get('resolution') in ('master', 'source'))
formats = self._extract_m3u8_formats(m3u8_playlist, channel_id, 'mp4', m3u8_id='hls', live=True)
self._sort_formats(formats)
channel = try_get(info, lambda x: x['user']['username']) # using this field instead of channel_id due to capitalization
return {
'id': channel,
'title': try_get(info, lambda x: x['live_stream']['title']),
'channel': channel,
'view_count': try_get(info, lambda x: x['live_stream']['view_count']),
'is_live': True,
'formats': formats,
'thumbnail': try_get(info, lambda x: x['live_stream']['thumbnail_url']),
}