Compare commits

..

6 Commits

Author SHA1 Message Date
Remita Amine
90a271e914 [nba] rewrite extractor 2020-12-19 20:14:44 +01:00
Remita Amine
172d270607 [turner] improve info extraction 2020-12-19 20:14:44 +01:00
Remita Amine
22feed08a1 [common] remove unwanted query params from unsigned akamai manifest URLs 2020-12-19 20:14:44 +01:00
Sergey M․
942b8ca3be
[youtube] Improve xsrf token extraction (closes #27442) 2020-12-20 00:48:44 +07:00
Sergey M․
3729c52f9d
[generic] Improve RSS age limit extraction 2020-12-19 23:24:52 +07:00
renalid
71679eaee8
[generic] Fix RSS itunes thumbnail extraction (#27405) 2020-12-19 23:18:51 +07:00
8 changed files with 492 additions and 136 deletions

View File

@ -96,7 +96,10 @@ class CNNIE(TurnerBaseIE):
config['data_src'] % path, page_title, {
'default': {
'media_src': config['media_src'],
}
},
'f4m': {
'host': 'cnn-vh.akamaihd.net',
},
})

View File

@ -2605,6 +2605,13 @@ class InfoExtractor(object):
return entries
def _extract_akamai_formats(self, manifest_url, video_id, hosts={}):
signed = 'hdnea=' in manifest_url
if not signed:
# https://learn.akamai.com/en-us/webhelp/media-services-on-demand/stream-packaging-user-guide/GUID-BE6C0F73-1E06-483B-B0EA-57984B91B7F9.html
manifest_url = re.sub(
r'(?:b=[\d,-]+|(?:__a__|attributes)=off|__b__=\d+)&?',
'', manifest_url).strip('?')
formats = []
hdcore_sign = 'hdcore=3.7.0'
@ -2630,7 +2637,7 @@ class InfoExtractor(object):
formats.extend(m3u8_formats)
http_host = hosts.get('http')
if http_host and m3u8_formats and 'hdnea=' not in m3u8_url:
if http_host and m3u8_formats and not signed:
REPL_REGEX = r'https?://[^/]+/i/([^,]+),([^/]+),([^/]+)\.csmil/.+'
qualities = re.match(REPL_REGEX, m3u8_url).group(2).split(',')
qualities_length = len(qualities)

View File

@ -679,7 +679,14 @@ from .nationalgeographic import (
NationalGeographicTVIE,
)
from .naver import NaverIE
from .nba import NBAIE
from .nba import (
NBAWatchEmbedIE,
NBAWatchIE,
NBAWatchCollectionIE,
NBAEmbedIE,
NBAIE,
NBAChannelIE,
)
from .nbc import (
CSNNEIE,
NBCIE,

View File

@ -35,6 +35,7 @@ from ..utils import (
unsmuggle_url,
UnsupportedError,
url_or_none,
xpath_attr,
xpath_text,
xpath_with_ns,
)
@ -217,6 +218,33 @@ class GenericIE(InfoExtractor):
},
}],
},
# RSS feed with item with description and thumbnails
{
'url': 'https://anchor.fm/s/dd00e14/podcast/rss',
'info_dict': {
'id': 'https://anchor.fm/s/dd00e14/podcast/rss',
'title': 're:.*100% Hydrogen.*',
'description': 're:.*In this episode.*',
},
'playlist': [{
'info_dict': {
'ext': 'm4a',
'id': 'c1c879525ce2cb640b344507e682c36d',
'title': 're:Hydrogen!',
'description': 're:.*In this episode we are going.*',
'timestamp': 1567977776,
'upload_date': '20190908',
'duration': 459,
'thumbnail': r're:^https?://.*\.jpg$',
'episode_number': 1,
'season_number': 1,
'age_limit': 0,
},
}],
'params': {
'skip_download': True,
},
},
# RSS feed with enclosures and unsupported link URLs
{
'url': 'http://www.hellointernet.fm/podcast?format=rss',
@ -2218,10 +2246,10 @@ class GenericIE(InfoExtractor):
default=None)
duration = itunes('duration')
explicit = itunes('explicit')
if explicit == 'true':
explicit = (itunes('explicit') or '').lower()
if explicit in ('true', 'yes'):
age_limit = 18
elif explicit == 'false':
elif explicit in ('false', 'no'):
age_limit = 0
else:
age_limit = None
@ -2234,7 +2262,7 @@ class GenericIE(InfoExtractor):
'timestamp': unified_timestamp(
xpath_text(it, 'pubDate', default=None)),
'duration': int_or_none(duration) or parse_duration(duration),
'thumbnail': url_or_none(itunes('image')),
'thumbnail': url_or_none(xpath_attr(it, xpath_with_ns('./itunes:image', NS_MAP), 'href')),
'episode': itunes('title'),
'episode_number': int_or_none(itunes('episode')),
'season_number': int_or_none(itunes('season')),

View File

@ -5,33 +5,137 @@ import re
from .turner import TurnerBaseIE
from ..compat import (
compat_urllib_parse_urlencode,
compat_urlparse,
compat_parse_qs,
compat_str,
compat_urllib_parse_unquote,
compat_urllib_parse_urlparse,
)
from ..utils import (
int_or_none,
merge_dicts,
OnDemandPagedList,
remove_start,
parse_duration,
parse_iso8601,
try_get,
update_url_query,
urljoin,
)
class NBAIE(TurnerBaseIE):
_VALID_URL = r'https?://(?:watch\.|www\.)?nba\.com/(?P<path>(?:[^/]+/)+(?P<id>[^?]*?))/?(?:/index\.html)?(?:\?.*)?$'
class NBACVPBaseIE(TurnerBaseIE):
def _extract_nba_cvp_info(self, path, video_id, fatal=False):
return self._extract_cvp_info(
'http://secure.nba.com/%s' % path, video_id, {
'default': {
'media_src': 'http://nba.cdn.turner.com/nba/big',
},
'm3u8': {
'media_src': 'http://nbavod-f.akamaihd.net',
},
}, fatal=fatal)
class NBAWatchBaseIE(NBACVPBaseIE):
_VALID_URL_BASE = r'https?://(?:(?:www\.)?nba\.com(?:/watch)?|watch\.nba\.com)/'
def _extract_video(self, filter_key, filter_value):
video = self._download_json(
'https://neulionscnbav2-a.akamaihd.net/solr/nbad_program/usersearch',
filter_value, query={
'fl': 'description,image,name,pid,releaseDate,runtime,tags,seoName',
'q': filter_key + ':' + filter_value,
'wt': 'json',
})['response']['docs'][0]
video_id = str(video['pid'])
title = video['name']
formats = []
m3u8_url = (self._download_json(
'https://watch.nba.com/service/publishpoint', video_id, query={
'type': 'video',
'format': 'json',
'id': video_id,
}, headers={
'User-Agent': 'Mozilla/5.0 (iPhone; CPU iPhone OS 11_0_1 like Mac OS X) AppleWebKit/604.1.38 (KHTML, like Gecko) Version/11.0 Mobile/15A402 Safari/604.1',
}, fatal=False) or {}).get('path')
if m3u8_url:
m3u8_formats = self._extract_m3u8_formats(
re.sub(r'_(?:pc|iphone)\.', '.', m3u8_url), video_id, 'mp4',
'm3u8_native', m3u8_id='hls', fatal=False)
formats.extend(m3u8_formats)
for f in m3u8_formats:
http_f = f.copy()
http_f.update({
'format_id': http_f['format_id'].replace('hls-', 'http-'),
'protocol': 'http',
'url': http_f['url'].replace('.m3u8', ''),
})
formats.append(http_f)
info = {
'id': video_id,
'title': title,
'thumbnail': urljoin('https://nbadsdmt.akamaized.net/media/nba/nba/thumbs/', video.get('image')),
'description': video.get('description'),
'duration': int_or_none(video.get('runtime')),
'timestamp': parse_iso8601(video.get('releaseDate')),
'tags': video.get('tags'),
}
seo_name = video.get('seoName')
if seo_name and re.search(r'\d{4}/\d{2}/\d{2}/', seo_name):
base_path = ''
if seo_name.startswith('teams/'):
base_path += seo_name.split('/')[1] + '/'
base_path += 'video/'
cvp_info = self._extract_nba_cvp_info(
base_path + seo_name + '.xml', video_id, False)
if cvp_info:
formats.extend(cvp_info['formats'])
info = merge_dicts(info, cvp_info)
self._sort_formats(formats)
info['formats'] = formats
return info
class NBAWatchEmbedIE(NBAWatchBaseIE):
IENAME = 'nba:watch:embed'
_VALID_URL = NBAWatchBaseIE._VALID_URL_BASE + r'embed\?.*?\bid=(?P<id>\d+)'
_TESTS = [{
'url': 'http://watch.nba.com/embed?id=659395',
'md5': 'b7e3f9946595f4ca0a13903ce5edd120',
'info_dict': {
'id': '659395',
'ext': 'mp4',
'title': 'Mix clip: More than 7 points of Joe Ingles, Luc Mbah a Moute, Blake Griffin and 6 more in Utah Jazz vs. the Clippers, 4/15/2017',
'description': 'Mix clip: More than 7 points of Joe Ingles, Luc Mbah a Moute, Blake Griffin and 6 more in Utah Jazz vs. the Clippers, 4/15/2017',
'timestamp': 1492228800,
'upload_date': '20170415',
},
}]
def _real_extract(self, url):
video_id = self._match_id(url)
return self._extract_video('pid', video_id)
class NBAWatchIE(NBAWatchBaseIE):
IE_NAME = 'nba:watch'
_VALID_URL = NBAWatchBaseIE._VALID_URL_BASE + r'(?:nba/)?video/(?P<id>.+?(?=/index\.html)|(?:[^/]+/)*[^/?#&]+)'
_TESTS = [{
'url': 'http://www.nba.com/video/games/nets/2012/12/04/0021200253-okc-bkn-recap.nba/index.html',
'md5': '9e7729d3010a9c71506fd1248f74e4f4',
'md5': '9d902940d2a127af3f7f9d2f3dc79c96',
'info_dict': {
'id': '0021200253-okc-bkn-recap',
'id': '70946',
'ext': 'mp4',
'title': 'Thunder vs. Nets',
'description': 'Kevin Durant scores 32 points and dishes out six assists as the Thunder beat the Nets in Brooklyn.',
'duration': 181,
'timestamp': 1354638466,
'timestamp': 1354597200,
'upload_date': '20121204',
},
'params': {
# m3u8 download
'skip_download': True,
},
}, {
'url': 'http://www.nba.com/video/games/hornets/2014/12/05/0021400276-nyk-cha-play5.nba/',
'only_matching': True,
@ -39,116 +143,286 @@ class NBAIE(TurnerBaseIE):
'url': 'http://watch.nba.com/video/channels/playoffs/2015/05/20/0041400301-cle-atl-recap.nba',
'md5': 'b2b39b81cf28615ae0c3360a3f9668c4',
'info_dict': {
'id': 'channels/playoffs/2015/05/20/0041400301-cle-atl-recap.nba',
'id': '330865',
'ext': 'mp4',
'title': 'Hawks vs. Cavaliers Game 1',
'description': 'md5:8094c3498d35a9bd6b1a8c396a071b4d',
'duration': 228,
'timestamp': 1432134543,
'upload_date': '20150520',
},
'expected_warnings': ['Unable to download f4m manifest'],
}, {
'url': 'http://www.nba.com/clippers/news/doc-rivers-were-not-trading-blake',
'info_dict': {
'id': 'teams/clippers/2016/02/17/1455672027478-Doc_Feb16_720.mov-297324',
'ext': 'mp4',
'title': 'Practice: Doc Rivers - 2/16/16',
'description': 'Head Coach Doc Rivers addresses the media following practice.',
'upload_date': '20160216',
'timestamp': 1455672000,
},
'params': {
# m3u8 download
'skip_download': True,
},
'expected_warnings': ['Unable to download f4m manifest'],
}, {
'url': 'http://www.nba.com/timberwolves/wiggins-shootaround#',
'info_dict': {
'id': 'timberwolves',
'title': 'Shootaround Access - Dec. 12 | Andrew Wiggins',
},
'playlist_count': 30,
'params': {
# Download the whole playlist takes too long time
'playlist_items': '1-30',
'timestamp': 1432094400,
'upload_date': '20150521',
},
}, {
'url': 'http://www.nba.com/timberwolves/wiggins-shootaround#',
'info_dict': {
'id': 'teams/timberwolves/2014/12/12/Wigginsmp4-3462601',
'ext': 'mp4',
'title': 'Shootaround Access - Dec. 12 | Andrew Wiggins',
'description': 'Wolves rookie Andrew Wiggins addresses the media after Friday\'s shootaround.',
'upload_date': '20141212',
'timestamp': 1418418600,
},
'params': {
'noplaylist': True,
# m3u8 download
'skip_download': True,
},
'expected_warnings': ['Unable to download f4m manifest'],
'url': 'http://watch.nba.com/nba/video/channels/nba_tv/2015/06/11/YT_go_big_go_home_Game4_061115',
'only_matching': True,
}, {
# only CVP mp4 format available
'url': 'https://watch.nba.com/video/teams/cavaliers/2012/10/15/sloan121015mov-2249106',
'only_matching': True,
}, {
'url': 'https://watch.nba.com/video/top-100-dunks-from-the-2019-20-season?plsrc=nba&collection=2019-20-season-highlights',
'only_matching': True,
}]
_PAGE_SIZE = 30
def _real_extract(self, url):
display_id = self._match_id(url)
collection_id = compat_parse_qs(compat_urllib_parse_urlparse(url).query).get('collection', [None])[0]
if collection_id:
if self._downloader.params.get('noplaylist'):
self.to_screen('Downloading just video %s because of --no-playlist' % display_id)
else:
self.to_screen('Downloading playlist %s - add --no-playlist to just download video' % collection_id)
return self.url_result(
'https://www.nba.com/watch/list/collection/' + collection_id,
NBAWatchCollectionIE.ie_key(), collection_id)
return self._extract_video('seoName', display_id)
def _fetch_page(self, team, video_id, page):
search_url = 'http://searchapp2.nba.com/nba-search/query.jsp?' + compat_urllib_parse_urlencode({
'type': 'teamvideo',
'start': page * self._PAGE_SIZE + 1,
'npp': (page + 1) * self._PAGE_SIZE + 1,
'sort': 'recent',
'output': 'json',
'site': team,
})
results = self._download_json(
search_url, video_id, note='Download page %d of playlist data' % page)['results'][0]
for item in results:
yield self.url_result(compat_urlparse.urljoin('http://www.nba.com/', item['url']))
def _extract_playlist(self, orig_path, video_id, webpage):
team = orig_path.split('/')[0]
class NBAWatchCollectionIE(NBAWatchBaseIE):
IE_NAME = 'nba:watch:collection'
_VALID_URL = NBAWatchBaseIE._VALID_URL_BASE + r'list/collection/(?P<id>[^/?#&]+)'
_TESTS = [{
'url': 'https://watch.nba.com/list/collection/season-preview-2020',
'info_dict': {
'id': 'season-preview-2020',
},
'playlist_mincount': 43,
}]
_PAGE_SIZE = 100
if self._downloader.params.get('noplaylist'):
self.to_screen('Downloading just video because of --no-playlist')
video_path = self._search_regex(
r'nbaVideoCore\.firstVideo\s*=\s*\'([^\']+)\';', webpage, 'video path')
video_url = 'http://www.nba.com/%s/video/%s' % (team, video_path)
return self.url_result(video_url)
self.to_screen('Downloading playlist - add --no-playlist to just download video')
playlist_title = self._og_search_title(webpage, fatal=False)
entries = OnDemandPagedList(
functools.partial(self._fetch_page, team, video_id),
self._PAGE_SIZE)
return self.playlist_result(entries, team, playlist_title)
def _fetch_page(self, collection_id, page):
page += 1
videos = self._download_json(
'https://content-api-prod.nba.com/public/1/endeavor/video-list/collection/' + collection_id,
collection_id, 'Downloading page %d JSON metadata' % page, query={
'count': self._PAGE_SIZE,
'page': page,
})['results']['videos']
for video in videos:
program = video.get('program') or {}
seo_name = program.get('seoName') or program.get('slug')
if not seo_name:
continue
yield {
'_type': 'url',
'id': program.get('id'),
'title': program.get('title') or video.get('title'),
'url': 'https://www.nba.com/watch/video/' + seo_name,
'thumbnail': video.get('image'),
'description': program.get('description') or video.get('description'),
'duration': parse_duration(program.get('runtimeHours')),
'timestamp': parse_iso8601(video.get('releaseDate')),
}
def _real_extract(self, url):
path, video_id = re.match(self._VALID_URL, url).groups()
orig_path = path
if path.startswith('nba/'):
path = path[3:]
collection_id = self._match_id(url)
entries = OnDemandPagedList(
functools.partial(self._fetch_page, collection_id),
self._PAGE_SIZE)
return self.playlist_result(entries, collection_id)
if 'video/' not in path:
webpage = self._download_webpage(url, video_id)
path = remove_start(self._search_regex(r'data-videoid="([^"]+)"', webpage, 'video id'), '/')
if path == '{{id}}':
return self._extract_playlist(orig_path, video_id, webpage)
class NBABaseIE(NBACVPBaseIE):
_VALID_URL_BASE = r'''(?x)
https?://(?:www\.)?nba\.com/
(?P<team>
blazers|
bucks|
bulls|
cavaliers|
celtics|
clippers|
grizzlies|
hawks|
heat|
hornets|
jazz|
kings|
knicks|
lakers|
magic|
mavericks|
nets|
nuggets|
pacers|
pelicans|
pistons|
raptors|
rockets|
sixers|
spurs|
suns|
thunder|
timberwolves|
warriors|
wizards
)
(?:/play\#)?/'''
_CHANNEL_PATH_REGEX = r'video/channel|series'
# See prepareContentId() of pkgCvp.js
if path.startswith('video/teams'):
path = 'video/channels/proxy/' + path[6:]
def _embed_url_result(self, team, content_id):
return self.url_result(update_url_query(
'https://secure.nba.com/assets/amp/include/video/iframe.html', {
'contentId': content_id,
'team': team,
}), NBAEmbedIE.ie_key())
return self._extract_cvp_info(
'http://www.nba.com/%s.xml' % path, video_id, {
'default': {
'media_src': 'http://nba.cdn.turner.com/nba/big',
},
'm3u8': {
'media_src': 'http://nbavod-f.akamaihd.net',
},
def _call_api(self, team, content_id, query, resource):
return self._download_json(
'https://api.nba.net/2/%s/video,imported_video,wsc/' % team,
content_id, 'Download %s JSON metadata' % resource,
query=query, headers={
'accessToken': 'internal|bb88df6b4c2244e78822812cecf1ee1b',
})['response']['result']
def _extract_video(self, video, team, extract_all=True):
video_id = compat_str(video['nid'])
team = video['brand']
info = {
'id': video_id,
'title': video.get('title') or video.get('headline') or video['shortHeadline'],
'description': video.get('description'),
'timestamp': parse_iso8601(video.get('published')),
}
subtitles = {}
captions = try_get(video, lambda x: x['videoCaptions']['sidecars'], dict) or {}
for caption_url in captions.values():
subtitles.setdefault('en', []).append({'url': caption_url})
formats = []
mp4_url = video.get('mp4')
if mp4_url:
formats.append({
'url': mp4_url,
})
if extract_all:
source_url = video.get('videoSource')
if source_url and not source_url.startswith('s3://') and self._is_valid_url(source_url, video_id, 'source'):
formats.append({
'format_id': 'source',
'url': source_url,
'preference': 1,
})
m3u8_url = video.get('m3u8')
if m3u8_url:
if '.akamaihd.net/i/' in m3u8_url:
formats.extend(self._extract_akamai_formats(
m3u8_url, video_id, {'http': 'pmd.cdn.turner.com'}))
else:
formats.extend(self._extract_m3u8_formats(
m3u8_url, video_id, 'mp4',
'm3u8_native', m3u8_id='hls', fatal=False))
content_xml = video.get('contentXml')
if team and content_xml:
cvp_info = self._extract_nba_cvp_info(
team + content_xml, video_id, fatal=False)
if cvp_info:
formats.extend(cvp_info['formats'])
subtitles = self._merge_subtitles(subtitles, cvp_info['subtitles'])
info = merge_dicts(info, cvp_info)
self._sort_formats(formats)
else:
info.update(self._embed_url_result(team, video['videoId']))
info.update({
'formats': formats,
'subtitles': subtitles,
})
return info
def _real_extract(self, url):
team, display_id = re.match(self._VALID_URL, url).groups()
if '/play#/' in url:
display_id = compat_urllib_parse_unquote(display_id)
else:
webpage = self._download_webpage(url, display_id)
display_id = self._search_regex(
self._CONTENT_ID_REGEX + r'\s*:\s*"([^"]+)"', webpage, 'video id')
return self._extract_url_results(team, display_id)
class NBAEmbedIE(NBABaseIE):
IENAME = 'nba:embed'
_VALID_URL = r'https?://secure\.nba\.com/assets/amp/include/video/(?:topI|i)frame\.html\?.*?\bcontentId=(?P<id>[^?#&]+)'
_TESTS = [{
'url': 'https://secure.nba.com/assets/amp/include/video/topIframe.html?contentId=teams/bulls/2020/12/04/3478774/1607105587854-20201204_SCHEDULE_RELEASE_FINAL_DRUPAL-3478774&team=bulls&adFree=false&profile=71&videoPlayerName=TAMPCVP&baseUrl=&videoAdsection=nba.com_mobile_web_teamsites_chicagobulls&ampEnv=',
'only_matching': True,
}, {
'url': 'https://secure.nba.com/assets/amp/include/video/iframe.html?contentId=2016/10/29/0021600027boschaplay7&adFree=false&profile=71&team=&videoPlayerName=LAMPCVP',
'only_matching': True,
}]
def _real_extract(self, url):
qs = compat_parse_qs(compat_urllib_parse_urlparse(url).query)
content_id = qs['contentId'][0]
team = qs.get('team', [None])[0]
if not team:
return self.url_result(
'https://watch.nba.com/video/' + content_id, NBAWatchIE.ie_key())
video = self._call_api(team, content_id, {'videoid': content_id}, 'video')[0]
return self._extract_video(video, team)
class NBAIE(NBABaseIE):
IENAME = 'nba'
_VALID_URL = NBABaseIE._VALID_URL_BASE + '(?!%s)video/(?P<id>(?:[^/]+/)*[^/?#&]+)' % NBABaseIE._CHANNEL_PATH_REGEX
_TESTS = [{
'url': 'https://www.nba.com/bulls/video/teams/bulls/2020/12/04/3478774/1607105587854-20201204schedulereleasefinaldrupal-3478774',
'info_dict': {
'id': '45039',
'ext': 'mp4',
'title': 'AND WE BACK.',
'description': 'Part 1 of our 2020-21 schedule is here! Watch our games on NBC Sports Chicago.',
'duration': 94,
'timestamp': 1607112000,
'upload_date': '20201218',
},
}, {
'url': 'https://www.nba.com/bucks/play#/video/teams%2Fbucks%2F2020%2F12%2F17%2F64860%2F1608252863446-Op_Dream_16x9-64860',
'only_matching': True,
}, {
'url': 'https://www.nba.com/bucks/play#/video/wsc%2Fteams%2F2787C911AA1ACD154B5377F7577CCC7134B2A4B0',
'only_matching': True,
}]
_CONTENT_ID_REGEX = r'videoID'
def _extract_url_results(self, team, content_id):
return self._embed_url_result(team, content_id)
class NBAChannelIE(NBABaseIE):
IENAME = 'nba:channel'
_VALID_URL = NBABaseIE._VALID_URL_BASE + '(?:%s)/(?P<id>[^/?#&]+)' % NBABaseIE._CHANNEL_PATH_REGEX
_TESTS = [{
'url': 'https://www.nba.com/blazers/video/channel/summer_league',
'info_dict': {
'title': 'Summer League',
},
'playlist_mincount': 138,
}, {
'url': 'https://www.nba.com/bucks/play#/series/On%20This%20Date',
'only_matching': True,
}]
_CONTENT_ID_REGEX = r'videoSubCategory'
_PAGE_SIZE = 100
def _fetch_page(self, team, channel, page):
results = self._call_api(team, channel, {
'channels': channel,
'count': self._PAGE_SIZE,
'offset': page * self._PAGE_SIZE,
}, 'page %d' % (page + 1))
for video in results:
yield self._extract_video(video, team, False)
def _extract_url_results(self, team, content_id):
entries = OnDemandPagedList(
functools.partial(self._fetch_page, team, content_id),
self._PAGE_SIZE)
return self.playlist_result(entries, playlist_title=content_id)

View File

@ -33,8 +33,7 @@ class NRKBaseIE(InfoExtractor):
def _extract_nrk_formats(self, asset_url, video_id):
if re.match(r'https?://[^/]+\.akamaihd\.net/i/', asset_url):
return self._extract_akamai_formats(
re.sub(r'(?:b=\d+-\d+|__a__=off)&?', '', asset_url), video_id)
return self._extract_akamai_formats(asset_url, video_id)
asset_url = re.sub(r'(?:bw_(?:low|high)=\d+|no_audio_only)&?', '', asset_url)
formats = self._extract_m3u8_formats(
asset_url, video_id, 'mp4', 'm3u8_native', fatal=False)

View File

@ -6,6 +6,7 @@ import re
from .adobepass import AdobePassIE
from ..compat import compat_str
from ..utils import (
fix_xml_ampersands,
xpath_text,
int_or_none,
determine_ext,
@ -49,8 +50,13 @@ class TurnerBaseIE(AdobePassIE):
self._AKAMAI_SPE_TOKEN_CACHE[secure_path] = token
return video_url + '?hdnea=' + token
def _extract_cvp_info(self, data_src, video_id, path_data={}, ap_data={}):
video_data = self._download_xml(data_src, video_id)
def _extract_cvp_info(self, data_src, video_id, path_data={}, ap_data={}, fatal=False):
video_data = self._download_xml(
data_src, video_id,
transform_source=lambda s: fix_xml_ampersands(s).strip(),
fatal=fatal)
if not video_data:
return {}
video_id = video_data.attrib['id']
title = xpath_text(video_data, 'headline', fatal=True)
content_id = xpath_text(video_data, 'contentId') or video_id
@ -63,12 +69,14 @@ class TurnerBaseIE(AdobePassIE):
urls = []
formats = []
thumbnails = []
subtitles = {}
rex = re.compile(
r'(?P<width>[0-9]+)x(?P<height>[0-9]+)(?:_(?P<bitrate>[0-9]+))?')
# Possible formats locations: files/file, files/groupFiles/files
# and maybe others
for video_file in video_data.findall('.//file'):
video_url = video_file.text.strip()
video_url = url_or_none(video_file.text.strip())
if not video_url:
continue
ext = determine_ext(video_url)
@ -108,9 +116,28 @@ class TurnerBaseIE(AdobePassIE):
continue
urls.append(video_url)
format_id = video_file.get('bitrate')
if ext == 'smil':
if ext in ('scc', 'srt', 'vtt'):
subtitles.setdefault('en', []).append({
'ext': ext,
'url': video_url,
})
elif ext == 'png':
thumbnails.append({
'id': format_id,
'url': video_url,
})
elif ext == 'smil':
formats.extend(self._extract_smil_formats(
video_url, video_id, fatal=False))
elif re.match(r'https?://[^/]+\.akamaihd\.net/[iz]/', video_url):
formats.extend(self._extract_akamai_formats(
video_url, video_id, {
'hds': path_data.get('f4m', {}).get('host'),
# nba.cdn.turner.com, ht.cdn.turner.com, ht2.cdn.turner.com
# ht3.cdn.turner.com, i.cdn.turner.com, s.cdn.turner.com
# ssl.cdn.turner.com
'http': 'pmd.cdn.turner.com',
}))
elif ext == 'm3u8':
m3u8_formats = self._extract_m3u8_formats(
video_url, video_id, 'mp4',
@ -129,7 +156,7 @@ class TurnerBaseIE(AdobePassIE):
'url': video_url,
'ext': ext,
}
mobj = rex.search(format_id + video_url)
mobj = rex.search(video_url)
if mobj:
f.update({
'width': int(mobj.group('width')),
@ -152,7 +179,6 @@ class TurnerBaseIE(AdobePassIE):
formats.append(f)
self._sort_formats(formats)
subtitles = {}
for source in video_data.findall('closedCaptions/source'):
for track in source.findall('track'):
track_url = url_or_none(track.get('url'))
@ -168,12 +194,12 @@ class TurnerBaseIE(AdobePassIE):
}.get(source.get('format'))
})
thumbnails = [{
'id': image.get('cut'),
thumbnails.extend({
'id': image.get('cut') or image.get('name'),
'url': image.text,
'width': int_or_none(image.get('width')),
'height': int_or_none(image.get('height')),
} for image in video_data.findall('images/image')]
} for image in video_data.findall('images/image'))
is_live = xpath_text(video_data, 'isLive') == 'true'

View File

@ -300,6 +300,12 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
self._YT_INITIAL_DATA_RE), webpage, 'yt initial data'),
video_id)
def _extract_ytcfg(self, video_id, webpage):
return self._parse_json(
self._search_regex(
r'ytcfg\.set\s*\(\s*({.+?})\s*\)\s*;', webpage, 'ytcfg',
default='{}'), video_id, fatal=False)
class YoutubeIE(YoutubeBaseInfoExtractor):
IE_DESC = 'YouTube.com'
@ -2283,16 +2289,25 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
# annotations
video_annotations = None
if self._downloader.params.get('writeannotations', False):
xsrf_token = self._search_regex(
r'([\'"])XSRF_TOKEN\1\s*:\s*([\'"])(?P<xsrf_token>[A-Za-z0-9+/=]+)\2',
video_webpage, 'xsrf token', group='xsrf_token', fatal=False)
xsrf_token = None
ytcfg = self._extract_ytcfg(video_id, video_webpage)
if ytcfg:
xsrf_token = try_get(ytcfg, lambda x: x['XSRF_TOKEN'], compat_str)
if not xsrf_token:
xsrf_token = self._search_regex(
r'([\'"])XSRF_TOKEN\1\s*:\s*([\'"])(?P<xsrf_token>(?:(?!\2).)+)\2',
video_webpage, 'xsrf token', group='xsrf_token', fatal=False)
invideo_url = try_get(
player_response, lambda x: x['annotations'][0]['playerAnnotationsUrlsRenderer']['invideoUrl'], compat_str)
if xsrf_token and invideo_url:
xsrf_field_name = self._search_regex(
r'([\'"])XSRF_FIELD_NAME\1\s*:\s*([\'"])(?P<xsrf_field_name>\w+)\2',
video_webpage, 'xsrf field name',
group='xsrf_field_name', default='session_token')
xsrf_field_name = None
if ytcfg:
xsrf_field_name = try_get(ytcfg, lambda x: x['XSRF_FIELD_NAME'], compat_str)
if not xsrf_field_name:
xsrf_field_name = self._search_regex(
r'([\'"])XSRF_FIELD_NAME\1\s*:\s*([\'"])(?P<xsrf_field_name>\w+)\2',
video_webpage, 'xsrf field name',
group='xsrf_field_name', default='session_token')
video_annotations = self._download_webpage(
self._proto_relative_url(invideo_url),
video_id, note='Downloading annotations',
@ -3130,10 +3145,7 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor):
playlist_title=title)
def _extract_identity_token(self, webpage, item_id):
ytcfg = self._parse_json(
self._search_regex(
r'ytcfg\.set\s*\(\s*({.+?})\s*\)\s*;', webpage, 'ytcfg',
default='{}'), item_id, fatal=False)
ytcfg = self._extract_ytcfg(item_id, webpage)
if ytcfg:
token = try_get(ytcfg, lambda x: x['ID_TOKEN'], compat_str)
if token: