mirror of
https://github.com/ytdl-org/youtube-dl
synced 2024-12-18 07:26:41 +00:00
[playwire] Add extractor
This commit is contained in:
parent
f28fe66970
commit
12a129ec6d
@ -364,6 +364,7 @@ from .planetaplay import PlanetaPlayIE
|
||||
from .played import PlayedIE
|
||||
from .playfm import PlayFMIE
|
||||
from .playvid import PlayvidIE
|
||||
from .playwire import PlaywireIE
|
||||
from .podomatic import PodomaticIE
|
||||
from .pornhd import PornHdIE
|
||||
from .pornhub import (
|
||||
|
78
youtube_dl/extractor/playwire.py
Normal file
78
youtube_dl/extractor/playwire.py
Normal file
@ -0,0 +1,78 @@
|
||||
from __future__ import unicode_literals
|
||||
|
||||
import re
|
||||
|
||||
from .common import InfoExtractor
|
||||
from ..utils import (
|
||||
xpath_text,
|
||||
float_or_none,
|
||||
int_or_none,
|
||||
)
|
||||
|
||||
|
||||
class PlaywireIE(InfoExtractor):
|
||||
_VALID_URL = r'https?://(?:config|cdn)\.playwire\.com(?:/v2)?/(?P<publisher_id>\d+)/(?:videos/v2|embed|config)/(?P<id>\d+)'
|
||||
_TESTS = [{
|
||||
'url': 'http://config.playwire.com/14907/videos/v2/3353705/player.json',
|
||||
'md5': 'e6398701e3595888125729eaa2329ed9',
|
||||
'info_dict': {
|
||||
'id': '3353705',
|
||||
'ext': 'mp4',
|
||||
'title': 'S04_RM_UCL_Rus',
|
||||
'thumbnail': 're:^http://.*\.png$',
|
||||
'duration': 145.94,
|
||||
},
|
||||
}, {
|
||||
'url': 'http://cdn.playwire.com/11625/embed/85228.html',
|
||||
'only_matching': True,
|
||||
}, {
|
||||
'url': 'http://config.playwire.com/12421/videos/v2/3389892/zeus.json',
|
||||
'only_matching': True,
|
||||
}, {
|
||||
'url': 'http://cdn.playwire.com/v2/12342/config/1532636.json',
|
||||
'only_matching': True,
|
||||
}]
|
||||
|
||||
def _real_extract(self, url):
|
||||
mobj = re.match(self._VALID_URL, url)
|
||||
publisher_id, video_id = mobj.group('publisher_id'), mobj.group('id')
|
||||
|
||||
player = self._download_json(
|
||||
'http://config.playwire.com/%s/videos/v2/%s/zeus.json' % (publisher_id, video_id),
|
||||
video_id)
|
||||
|
||||
title = player['settings']['title']
|
||||
duration = float_or_none(player.get('duration'), 1000)
|
||||
|
||||
content = player['content']
|
||||
thumbnail = content.get('poster')
|
||||
src = content['media']['f4m']
|
||||
|
||||
f4m = self._download_xml(src, video_id)
|
||||
base_url = xpath_text(f4m, './{http://ns.adobe.com/f4m/1.0}baseURL', 'base url', fatal=True)
|
||||
formats = []
|
||||
for media in f4m.findall('./{http://ns.adobe.com/f4m/1.0}media'):
|
||||
media_url = media.get('url')
|
||||
if not media_url:
|
||||
continue
|
||||
tbr = int_or_none(media.get('bitrate'))
|
||||
width = int_or_none(media.get('width'))
|
||||
height = int_or_none(media.get('height'))
|
||||
f = {
|
||||
'url': '%s/%s' % (base_url, media.attrib['url']),
|
||||
'tbr': tbr,
|
||||
'width': width,
|
||||
'height': height,
|
||||
}
|
||||
if not (tbr or width or height):
|
||||
f['quality'] = 1 if '-hd.' in media_url else 0
|
||||
formats.append(f)
|
||||
self._sort_formats(formats)
|
||||
|
||||
return {
|
||||
'id': video_id,
|
||||
'title': title,
|
||||
'thumbnail': thumbnail,
|
||||
'duration': duration,
|
||||
'formats': formats,
|
||||
}
|
Loading…
Reference in New Issue
Block a user