[ooyala] extract more formats and metadata

This commit is contained in:
remitamine 2015-10-15 14:28:56 +01:00
parent fafc7950e2
commit 90bddb6cdd
1 changed files with 53 additions and 98 deletions

View File

@ -1,108 +1,64 @@
from __future__ import unicode_literals from __future__ import unicode_literals
import re import re
import json
import base64 import base64
from .common import InfoExtractor from .common import InfoExtractor
from ..utils import ( from ..utils import (
unescapeHTML,
ExtractorError,
determine_ext,
int_or_none, int_or_none,
float_or_none,
) )
class OoyalaBaseIE(InfoExtractor): class OoyalaBaseIE(InfoExtractor):
def _extract_result(self, info, more_info): def _extract(self, player_url, video_id):
embedCode = info['embedCode'] print(player_url)
video_url = info.get('ipad_url') or info['url'] content_tree = self._download_json(player_url, video_id)['content_tree']
metadata = content_tree[list(content_tree)[0]]
if determine_ext(video_url) == 'm3u8': embed_code = metadata['embed_code']
formats = self._extract_m3u8_formats(video_url, embedCode, ext='mp4') pcode = metadata.get('asset_pcode') or embed_code
else: video_info = {
formats = [{ 'id': embed_code,
'url': video_url, 'title': metadata['title'],
'ext': 'mp4', 'description': metadata.get('description'),
}] 'thumbnail': metadata.get('thumbnail_image') or metadata.get('promo_image'),
'duration': int_or_none(metadata.get('duration')),
return {
'id': embedCode,
'title': unescapeHTML(info['title']),
'formats': formats,
'description': unescapeHTML(more_info['description']),
'thumbnail': more_info['promo'],
} }
def _extract(self, player_url, video_id): formats = []
player = self._download_webpage(player_url, video_id) for supported_format in ('mp4', 'm3u8', 'hds', 'rtmp'):
mobile_url = self._search_regex(r'mobile_player_url="(.+?)&device="',
player, 'mobile player url')
# Looks like some videos are only available for particular devices
# (e.g. http://player.ooyala.com/player.js?embedCode=x1b3lqZDq9y_7kMyC2Op5qo-p077tXD0
# is only available for ipad)
# Working around with fetching URLs for all the devices found starting with 'unknown'
# until we succeed or eventually fail for each device.
devices = re.findall(r'device\s*=\s*"([^"]+)";', player)
devices.remove('unknown')
devices.insert(0, 'unknown')
for device in devices:
mobile_player = self._download_webpage(
'%s&device=%s' % (mobile_url, device), video_id,
'Downloading mobile player JS for %s device' % device)
videos_info = self._search_regex(
r'var streams=window.oo_testEnv\?\[\]:eval\("\((\[{.*?}\])\)"\);',
mobile_player, 'info', fatal=False, default=None)
if videos_info:
break
if not videos_info:
formats = []
auth_data = self._download_json( auth_data = self._download_json(
'http://player.ooyala.com/sas/player_api/v1/authorization/embed_code/%s/%s?domain=www.example.org&supportedFormats=mp4,webm' % (video_id, video_id), 'http://player.ooyala.com/sas/player_api/v1/authorization/embed_code/%s/%s?domain=www.example.org&supportedFormats=%s' % (pcode, embed_code, supported_format),
video_id) video_id, 'Downloading %s JSON' % supported_format)
cur_auth_data = auth_data['authorization_data'][video_id] cur_auth_data = auth_data['authorization_data'][embed_code]
for stream in cur_auth_data['streams']: for stream in cur_auth_data['streams']:
formats.append({ url = base64.b64decode(stream['url']['data'].encode('ascii')).decode('utf-8')
'url': base64.b64decode(stream['url']['data'].encode('ascii')).decode('utf-8'), delivery_type = stream['delivery_type']
'ext': stream.get('delivery_type'), if delivery_type == 'remote_asset':
'format': stream.get('video_codec'), video_info['url'] = url
'format_id': stream.get('profile'), return video_info
'width': int_or_none(stream.get('width')), if delivery_type == 'hls':
'height': int_or_none(stream.get('height')), formats.extend(self._extract_m3u8_formats(url, embed_code, 'mp4', 'm3u8_native', 0, m3u8_id='hls', fatal=False))
'abr': int_or_none(stream.get('audio_bitrate')), elif delivery_type == 'hds':
'vbr': int_or_none(stream.get('video_bitrate')), formats.extend(self._extract_f4m_formats(url, embed_code, f4m_id='hds', fatal=False))
}) else:
if formats: formats.append({
return { 'url': url,
'id': video_id, 'ext': stream.get('delivery_type'),
'formats': formats, 'vcodec': stream.get('video_codec'),
'title': 'Ooyala video', 'format_id': stream.get('profile'),
} 'width': int_or_none(stream.get('width')),
'height': int_or_none(stream.get('height')),
'abr': int_or_none(stream.get('audio_bitrate')),
'vbr': int_or_none(stream.get('video_bitrate')),
'fps': float_or_none(stream.get('framerate')),
})
self._sort_formats(formats)
if not cur_auth_data['authorized']: video_info['formats'] = formats
raise ExtractorError(cur_auth_data['message'], expected=True) return video_info
if not videos_info:
raise ExtractorError('Unable to extract info')
videos_info = videos_info.replace('\\"', '"')
videos_more_info = self._search_regex(
r'eval\("\(({.*?\\"promo\\".*?})\)"', mobile_player, 'more info').replace('\\"', '"')
videos_info = json.loads(videos_info)
videos_more_info = json.loads(videos_more_info)
if videos_more_info.get('lineup'):
videos = [self._extract_result(info, more_info) for (info, more_info) in zip(videos_info, videos_more_info['lineup'])]
return {
'_type': 'playlist',
'id': video_id,
'title': unescapeHTML(videos_more_info['title']),
'entries': videos,
}
else:
return self._extract_result(videos_info[0], videos_more_info)
class OoyalaIE(OoyalaBaseIE): class OoyalaIE(OoyalaBaseIE):
@ -117,6 +73,7 @@ class OoyalaIE(OoyalaBaseIE):
'ext': 'mp4', 'ext': 'mp4',
'title': 'Explaining Data Recovery from Hard Drives and SSDs', 'title': 'Explaining Data Recovery from Hard Drives and SSDs',
'description': 'How badly damaged does a drive have to be to defeat Russell and his crew? Apparently, smashed to bits.', 'description': 'How badly damaged does a drive have to be to defeat Russell and his crew? Apparently, smashed to bits.',
'duration': 853386,
}, },
}, { }, {
# Only available for ipad # Only available for ipad
@ -125,7 +82,7 @@ class OoyalaIE(OoyalaBaseIE):
'id': 'x1b3lqZDq9y_7kMyC2Op5qo-p077tXD0', 'id': 'x1b3lqZDq9y_7kMyC2Op5qo-p077tXD0',
'ext': 'mp4', 'ext': 'mp4',
'title': 'Simulation Overview - Levels of Simulation', 'title': 'Simulation Overview - Levels of Simulation',
'description': '', 'duration': 194948,
}, },
}, },
{ {
@ -136,7 +93,8 @@ class OoyalaIE(OoyalaBaseIE):
'info_dict': { 'info_dict': {
'id': 'FiOG81ZTrvckcchQxmalf4aQj590qTEx', 'id': 'FiOG81ZTrvckcchQxmalf4aQj590qTEx',
'ext': 'mp4', 'ext': 'mp4',
'title': 'Ooyala video', 'title': 'Divide Tool Path.mp4',
'duration': 204405,
} }
} }
] ]
@ -152,8 +110,8 @@ class OoyalaIE(OoyalaBaseIE):
def _real_extract(self, url): def _real_extract(self, url):
embed_code = self._match_id(url) embed_code = self._match_id(url)
player_url = 'http://player.ooyala.com/player.js?embedCode=%s' % embed_code content_tree_url = 'http://player.ooyala.com/player_api/v1/content_tree/embed_code/%s/%s' % (embed_code, embed_code)
return self._extract(player_url, embed_code) return self._extract(content_tree_url, embed_code)
class OoyalaExternalIE(OoyalaBaseIE): class OoyalaExternalIE(OoyalaBaseIE):
@ -170,7 +128,7 @@ class OoyalaExternalIE(OoyalaBaseIE):
.*?&pcode= .*?&pcode=
) )
(?P<pcode>.+?) (?P<pcode>.+?)
(&|$) (?:&|$)
''' '''
_TEST = { _TEST = {
@ -179,7 +137,7 @@ class OoyalaExternalIE(OoyalaBaseIE):
'id': 'FkYWtmazr6Ed8xmvILvKLWjd4QvYZpzG', 'id': 'FkYWtmazr6Ed8xmvILvKLWjd4QvYZpzG',
'ext': 'mp4', 'ext': 'mp4',
'title': 'dm_140128_30for30Shorts___JudgingJewellv2', 'title': 'dm_140128_30for30Shorts___JudgingJewellv2',
'description': '', 'duration': 1302000,
}, },
'params': { 'params': {
# m3u8 download # m3u8 download
@ -188,9 +146,6 @@ class OoyalaExternalIE(OoyalaBaseIE):
} }
def _real_extract(self, url): def _real_extract(self, url):
mobj = re.match(self._VALID_URL, url) partner_id, video_id, pcode = re.match(self._VALID_URL, url).groups()
partner_id = mobj.group('partner_id') content_tree_url = 'http://player.ooyala.com/player_api/v1/content_tree/external_id/%s/%s:%s' % (pcode, partner_id, video_id)
video_id = mobj.group('id') return self._extract(content_tree_url, video_id)
pcode = mobj.group('pcode')
player_url = 'http://player.ooyala.com/player.js?externalId=%s:%s&pcode=%s' % (partner_id, video_id, pcode)
return self._extract(player_url, video_id)