mirror of
https://github.com/ytdl-org/youtube-dl
synced 2024-11-23 18:06:50 +01:00
[picarto] Add extractor
This commit is contained in:
parent
8e41c9ad01
commit
d6166a7602
@ -815,6 +815,10 @@ from .periscope import (
|
||||
from .philharmoniedeparis import PhilharmonieDeParisIE
|
||||
from .phoenix import PhoenixIE
|
||||
from .photobucket import PhotobucketIE
|
||||
from .picarto import (
|
||||
PicartoVodIE,
|
||||
PicartoIE,
|
||||
)
|
||||
from .piksel import PikselIE
|
||||
from .pinkbike import PinkbikeIE
|
||||
from .pladform import PladformIE
|
||||
|
87
youtube_dl/extractor/picarto.py
Executable file
87
youtube_dl/extractor/picarto.py
Executable file
@ -0,0 +1,87 @@
|
||||
# coding: utf-8
|
||||
from __future__ import unicode_literals
|
||||
|
||||
from .common import InfoExtractor
|
||||
from ..utils import ExtractorError, js_to_json, urlencode_postdata
|
||||
|
||||
|
||||
class PicartoIE(InfoExtractor):
|
||||
_VALID_URL = r'https?://(?:www.)?picarto\.tv/(?P<id>[a-zA-Z0-9]+)[^/]*$'
|
||||
_TEST = {
|
||||
'url': 'https://picarto.tv/Setz',
|
||||
'info_dict': {
|
||||
'id': 'Setz',
|
||||
'ext': 'mp4',
|
||||
'title': 're:^Setz [0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2}$',
|
||||
'timestamp': int,
|
||||
'is_live': True
|
||||
},
|
||||
'params': {
|
||||
'skip_download': True
|
||||
}
|
||||
}
|
||||
|
||||
def _real_extract(self, url):
|
||||
channel_id = self._match_id(url)
|
||||
stream_page = self._download_webpage(url, channel_id)
|
||||
|
||||
if 'This channel does not exist.' in stream_page:
|
||||
raise ExtractorError('Channel does not exist', expected=True)
|
||||
|
||||
player_settings_js = self._html_search_regex(
|
||||
r'(?s)playerSettings\[1\]\s*=\s*(\{.+?\}\n)', stream_page, 'player-settings')
|
||||
player_settings = self._parse_json(player_settings_js, channel_id,
|
||||
transform_source=js_to_json)
|
||||
if not player_settings.get('online'):
|
||||
raise ExtractorError('Stream is offline', expected=True)
|
||||
|
||||
cdn_data = self._download_json('https://picarto.tv/process/channel', channel_id,
|
||||
data=urlencode_postdata({'loadbalancinginfo': channel_id}),
|
||||
note='Fetching load balancer info')
|
||||
edge = [edge['ep'] for edge in cdn_data['edges'] if edge['id'] == cdn_data['preferedEdge']][0]
|
||||
|
||||
formats = self._extract_m3u8_formats('https://%s/hls/%s/index.m3u8' % (edge, channel_id),
|
||||
channel_id, 'mp4')
|
||||
formats.append({'url': 'https://%s/mp4/%s.mp4' % (edge, channel_id)})
|
||||
self._sort_formats(formats)
|
||||
|
||||
return {
|
||||
'id': channel_id,
|
||||
'formats': formats,
|
||||
'ext': 'mp4',
|
||||
'title': self._live_title(channel_id),
|
||||
'is_live': True,
|
||||
'thumbnail': player_settings.get('vodThumb'),
|
||||
'age_limit': 18 if player_settings.get('mature') else None,
|
||||
}
|
||||
|
||||
|
||||
class PicartoVodIE(InfoExtractor):
|
||||
_VALID_URL = r'https?://(?:www.)?picarto\.tv/videopopout/(?P<id>[a-zA-Z0-9_\-\.]+).flv'
|
||||
_TEST = {
|
||||
'url': 'https://picarto.tv/videopopout/Carrot_2018.01.11.07.55.12.flv',
|
||||
'md5': '80765b67813053ff31d4df2bd5e900ce',
|
||||
'info_dict': {
|
||||
'id': 'Carrot_2018.01.11.07.55.12',
|
||||
'ext': 'mp4',
|
||||
'title': 'Carrot_2018.01.11.07.55.12',
|
||||
'thumbnail': r're:^https?://.*\.jpg$'
|
||||
}
|
||||
}
|
||||
|
||||
def _real_extract(self, url):
|
||||
video_id = self._match_id(url)
|
||||
webpage = self._download_webpage(url, video_id)
|
||||
|
||||
vod_info_js = self._html_search_regex(r'(?s)"#vod-player",\s*(\{.+?\})\)',
|
||||
webpage, video_id)
|
||||
vod_info = self._parse_json(vod_info_js, video_id, transform_source=js_to_json)
|
||||
|
||||
return {
|
||||
'id': video_id,
|
||||
'title': video_id,
|
||||
'ext': 'mp4',
|
||||
'protocol': 'm3u8',
|
||||
'url': vod_info['vod'],
|
||||
'thumbnail': vod_info.get('vodThumb'),
|
||||
}
|
Loading…
Reference in New Issue
Block a user