archive
/
yt-dlp
mirror of https://github.com/yt-dlp/yt-dlp
1
0
Fork 0
Code Issues Projects Releases Wiki Activity

[yahoo] improve extraction

Browse Source 
- add support for live streams(closes #3597)(closes #3779)(closes #22178)
- bypass cookie consent page for european domains(closes #16948)(closes #22576)
- add generic support for embeds(closes #20332)
pull/8/head
Remita Amine 5 years ago
parent 9a621ddc3a
commit 45f4a43389
1 changed files with 254 additions and 398 deletions
Show Stats Download Patch File Download Diff File

486
youtube_dl/extractor/yahoo.py
Unescape Escape View File

@ -3,40 +3,30 @@ from __future__ import unicode_literals
import hashlib
import itertools
import json
import re
from .common import InfoExtractor, SearchInfoExtractor
from ..compat import (
compat_str,
compat_urllib_parse,
compat_urlparse,
)
from ..utils import (
clean_html,
determine_ext,
ExtractorError,
extract_attributes,
int_or_none,
mimetype2ext,
parse_iso8601,
smuggle_url,
try_get,
unescapeHTML,
url_or_none,
)
from .brightcove import (
BrightcoveLegacyIE,
BrightcoveNewIE,
)
from .nbc import NBCSportsVPlayerIE
from .brightcove import BrightcoveNewIE
class YahooIE(InfoExtractor):
IE_DESC = 'Yahoo screen and movies'
_VALID_URL = r'(?P<host>https?://(?:(?P<country>[a-zA-Z]{2})\.)?[\da-zA-Z_-]+\.yahoo\.com)/(?:[^/]+/)*(?:(?P<display_id>.+)?-)?(?P<id>[0-9]+)(?:-[a-z]+)?(?:\.html)?'
_TESTS = [
{
_VALID_URL = r'(?P<url>https?://(?:(?P<country>[a-zA-Z]{2}(?:-[a-zA-Z]{2})?|malaysia)\.)?(?:[\da-zA-Z_-]+\.)?yahoo\.com/(?:[^/]+/)*(?P<id>[^?&#]*-[0-9]+)\.html)'
_TESTS = [{
'url': 'http://screen.yahoo.com/julian-smith-travis-legg-watch-214727115.html',
'info_dict': {
'id': '2d25e626-2378-391f-ada0-ddaf1417e588',
@ -44,21 +34,10 @@ class YahooIE(InfoExtractor):
'title': 'Julian Smith & Travis Legg Watch Julian Smith',
'description': 'Julian and Travis watch Julian Smith',
'duration': 6863,
'timestamp': 1369812016,
'upload_date': '20130529',
},
},
{
'url': 'http://screen.yahoo.com/wired/codefellas-s1-ep12-cougar-lies-103000935.html',
'md5': '251af144a19ebc4a033e8ba91ac726bb',
'info_dict': {
'id': 'd1dedf8c-d58c-38c3-8963-e899929ae0a9',
'ext': 'mp4',
'title': 'Codefellas - The Cougar Lies with Spanish Moss',
'description': 'md5:66b627ab0a282b26352136ca96ce73c1',
'duration': 151,
},
'skip': 'HTTP Error 404',
},
{
}, {
'url': 'https://screen.yahoo.com/community/community-sizzle-reel-203225340.html?format=embed',
'md5': '7993e572fac98e044588d0b5260f4352',
'info_dict': {
@ -67,65 +46,21 @@ class YahooIE(InfoExtractor):
'title': "Yahoo Saves 'Community'",
'description': 'md5:4d4145af2fd3de00cbb6c1d664105053',
'duration': 170,
}
},
{
'url': 'https://tw.news.yahoo.com/%E6%95%A2%E5%95%8F%E5%B8%82%E9%95%B7%20%E9%BB%83%E7%A7%80%E9%9C%9C%E6%89%B9%E8%B3%B4%E6%B8%85%E5%BE%B7%20%E9%9D%9E%E5%B8%B8%E9%AB%98%E5%82%B2-034024051.html',
'md5': '45c024bad51e63e9b6f6fad7a43a8c23',
'info_dict': {
'id': 'cac903b3-fcf4-3c14-b632-643ab541712f',
'ext': 'mp4',
'title': '敢問市長／黃秀霜批賴清德「非常高傲」',
'description': '直言台南沒捷運 交通居五都之末',
'duration': 396,
'timestamp': 1406838636,
'upload_date': '20140731',
},
},
{
}, {
'url': 'https://uk.screen.yahoo.com/editor-picks/cute-raccoon-freed-drain-using-091756545.html',
'md5': '71298482f7c64cbb7fa064e4553ff1c1',
'md5': '0b51660361f0e27c9789e7037ef76f4b',
'info_dict': {
'id': 'b3affa53-2e14-3590-852b-0e0db6cd1a58',
'ext': 'webm',
'ext': 'mp4',
'title': 'Cute Raccoon Freed From Drain\u00a0Using Angle Grinder',
'description': 'md5:f66c890e1490f4910a9953c941dee944',
'duration': 97,
'timestamp': 1414489862,
'upload_date': '20141028',
}
},
{
'url': 'https://ca.sports.yahoo.com/video/program-makes-hockey-more-affordable-013127711.html',
'md5': '57e06440778b1828a6079d2f744212c4',
'info_dict': {
'id': 'c9fa2a36-0d4d-3937-b8f6-cc0fb1881e73',
'ext': 'mp4',
'title': 'Program that makes hockey more affordable not offered in Manitoba',
'description': 'md5:c54a609f4c078d92b74ffb9bf1f496f4',
'duration': 121,
},
'skip': 'Video gone',
}, {
'url': 'https://ca.finance.yahoo.com/news/hackers-sony-more-trouble-well-154609075.html',
'info_dict': {
'id': '154609075',
},
'playlist': [{
'md5': '000887d0dc609bc3a47c974151a40fb8',
'info_dict': {
'id': 'e624c4bc-3389-34de-9dfc-025f74943409',
'ext': 'mp4',
'title': '\'The Interview\' TV Spot: War',
'description': 'The Interview',
'duration': 30,
},
}, {
'md5': '81bc74faf10750fe36e4542f9a184c66',
'info_dict': {
'id': '1fc8ada0-718e-3abe-a450-bf31f246d1a9',
'ext': 'mp4',
'title': '\'The Interview\' TV Spot: Guys',
'description': 'The Interview',
'duration': 30,
},
}],
}, {
'url': 'http://news.yahoo.com/video/china-moses-crazy-blues-104538833.html',
'md5': '88e209b417f173d86186bef6e4d1f160',
@ -135,27 +70,20 @@ class YahooIE(InfoExtractor):
'title': 'China Moses Is Crazy About the Blues',
'description': 'md5:9900ab8cd5808175c7b3fe55b979bed0',
'duration': 128,
'timestamp': 1385722202,
'upload_date': '20131129',
}
}, {
'url': 'https://in.lifestyle.yahoo.com/video/connect-dots-dark-side-virgo-090247395.html',
'md5': 'd9a083ccf1379127bf25699d67e4791b',
'info_dict': {
'id': '52aeeaa3-b3d1-30d8-9ef8-5d0cf05efb7c',
'ext': 'mp4',
'title': 'Connect the Dots: Dark Side of Virgo',
'description': 'md5:1428185051cfd1949807ad4ff6d3686a',
'duration': 201,
},
'skip': 'Domain name in.lifestyle.yahoo.com gone',
}, {
'url': 'https://www.yahoo.com/movies/v/true-story-trailer-173000497.html',
'md5': '989396ae73d20c6f057746fb226aa215',
'md5': '2a9752f74cb898af5d1083ea9f661b58',
'info_dict': {
'id': '071c4013-ce30-3a93-a5b2-e0413cd4a9d1',
'ext': 'mp4',
'title': '\'True Story\' Trailer',
'description': 'True Story',
'duration': 150,
'timestamp': 1418919206,
'upload_date': '20141218',
},
}, {
'url': 'https://gma.yahoo.com/pizza-delivery-man-surprised-huge-tip-college-kids-195200785.html',
@ -171,7 +99,7 @@ class YahooIE(InfoExtractor):
'upload_date': '20150313',
'uploader': 'NBCU-SPORTS',
'timestamp': 1426270238,
}
},
}, {
'url': 'https://tw.news.yahoo.com/-100120367.html',
'only_matching': True,
@ -185,43 +113,13 @@ class YahooIE(InfoExtractor):
'title': 'Communitary - Community Episode 1: Ladders',
'description': 'md5:8fc39608213295748e1e289807838c97',
'duration': 1646,
'timestamp': 1440436550,
'upload_date': '20150824',
'series': 'Communitary',
'season_number': 6,
'episode_number': 1,
},
}, {
# it uses an alias to get the video_id
'url': 'https://www.yahoo.com/movies/the-stars-of-daddys-home-have-very-different-212843197.html',
'info_dict': {
'id': '40eda9c8-8e5f-3552-8745-830f67d0c737',
'ext': 'mp4',
'title': 'Will Ferrell & Mark Wahlberg Are Pro-Spanking',
'description': 'While they play feuding fathers in \'Daddy\'s Home,\' star Will Ferrell & Mark Wahlberg share their true feelings on parenthood.',
},
},
{
# config['models']['applet_model']['data']['sapi'] has no query
'url': 'https://www.yahoo.com/music/livenation/event/galactic-2016',
'md5': 'dac0c72d502bc5facda80c9e6d5c98db',
'info_dict': {
'id': 'a6015640-e9e5-3efb-bb60-05589a183919',
'ext': 'mp4',
'description': 'Galactic',
'title': 'Dolla Diva (feat. Maggie Koerner)',
},
'skip': 'redirect to https://www.yahoo.com/music',
},
{
# yahoo://article/
'url': 'https://www.yahoo.com/movies/video/true-story-trailer-173000497.html',
'info_dict': {
'id': '071c4013-ce30-3a93-a5b2-e0413cd4a9d1',
'ext': 'mp4',
'title': "'True Story' Trailer",
'description': 'True Story',
},
'params': {
'skip_download': True,
},
},
{
# ytwnews://cavideo/
'url': 'https://tw.video.yahoo.com/movie-tw/單車天使-中文版預-092316541.html',
'info_dict': {
@ -229,227 +127,185 @@ class YahooIE(InfoExtractor):
'ext': 'mp4',
'title': '單車天使 - 中文版預',
'description': '中文版預',
'timestamp': 1476696196,
'upload_date': '20161017',
},
'params': {
'skip_download': True,
},
}, {
# Contains both a Yahoo hosted video and multiple Youtube embeds
'url': 'https://www.yahoo.com/entertainment/gwen-stefani-reveals-the-pop-hit-she-passed-on-assigns-it-to-her-voice-contestant-instead-033045672.html',
'info_dict': {
'id': '46c5d95a-528f-3d03-b732-732fcadd51de',
'title': 'Gwen Stefani reveals the pop hit she passed on, assigns it to her \'Voice\' contestant instead',
'description': 'Gwen decided not to record this hit herself, but she decided it was the perfect fit for Kyndall Inskeep.',
},
{
# custom brightcove
'url': 'https://au.tv.yahoo.com/plus7/sunrise/-/watch/37083565/clown-entertainers-say-it-is-hurting-their-business/',
'playlist': [{
'info_dict': {
'id': '5575377707001',
'id': '966d4262-4fd1-3aaa-b45b-049ca6e38ba6',
'ext': 'mp4',
'title': "Clown entertainers say 'It' is hurting their business",
'description': 'Stephen King s horror film has much to answer for. Jelby and Mr Loopy the Clowns join us.',
'timestamp': 1505341164,
'upload_date': '20170913',
'uploader_id': '2376984109001',
'title': 'Gwen Stefani reveals she turned down one of Sia\'s best songs',
'description': 'On "The Voice" Tuesday, Gwen Stefani told Taylor Swift which Sia hit was almost hers.',
'timestamp': 1572406500,
'upload_date': '20191030',
},
'params': {
'skip_download': True,
}, {
'info_dict': {
'id': '352CFDOQrKg',
'ext': 'mp4',
'title': 'Kyndal Inskeep "Performs the Hell Out of" Sia\'s "Elastic Heart" - The Voice Knockouts 2019',
'description': 'md5:35b61e94c2ae214bc965ff4245f80d11',
'uploader': 'The Voice',
'uploader_id': 'NBCTheVoice',
'upload_date': '20191029',
},
}],
'params': {
'playlistend': 2,
},
{
# custom brightcove, geo-restricted to Australia, bypassable
'url': 'https://au.tv.yahoo.com/plus7/sunrise/-/watch/37263964/sunrise-episode-wed-27-sep/',
}, {
'url': 'https://malaysia.news.yahoo.com/video/bystanders-help-ontario-policeman-bust-190932818.html',
'only_matching': True,
}
]
}, {
'url': 'https://es-us.noticias.yahoo.com/es-la-puerta-irrompible-que-110539379.html',
'only_matching': True,
}]
def _real_extract(self, url):
mobj = re.match(self._VALID_URL, url)
page_id = mobj.group('id')
display_id = mobj.group('display_id') or page_id
host = mobj.group('host')
webpage, urlh = self._download_webpage_handle(url, display_id)
if 'err=404' in urlh.geturl():
raise ExtractorError('Video gone', expected=True)
url, country, display_id = re.match(self._VALID_URL, url).groups()
if not country:
country = 'us'
else:
country = country.split('-')[0]
api_base = 'https://%s.yahoo.com/_td/api/resource/' % country
for i, uuid in enumerate(['url=' + url, 'ymedia-alias=' + display_id]):
content = self._download_json(
api_base + 'content;getDetailView=true;uuids=["%s"]' % uuid,
display_id, 'Downloading content JSON metadata', fatal=i == 1)
if content:
item = content['items'][0]
break
# Look for iframed media first
if item.get('type') != 'video':
entries = []
iframe_urls = re.findall(r'<iframe[^>]+src="(/video/.+?-\d+\.html\?format=embed.*?)"', webpage)
for idx, iframe_url in enumerate(iframe_urls):
entries.append(self.url_result(host + iframe_url, 'Yahoo'))
if entries:
return self.playlist_result(entries, page_id)
# Look for NBCSports iframes
nbc_sports_url = NBCSportsVPlayerIE._extract_url(webpage)
if nbc_sports_url:
return self.url_result(nbc_sports_url, NBCSportsVPlayerIE.ie_key())
# Look for Brightcove Legacy Studio embeds
bc_url = BrightcoveLegacyIE._extract_brightcove_url(webpage)
if bc_url:
return self.url_result(bc_url, BrightcoveLegacyIE.ie_key())
def brightcove_url_result(bc_url):
return self.url_result(
smuggle_url(bc_url, {'geo_countries': [mobj.group('country')]}),
BrightcoveNewIE.ie_key())
# Look for Brightcove New Studio embeds
bc_url = BrightcoveNewIE._extract_url(self, webpage)
if bc_url:
return brightcove_url_result(bc_url)
brightcove_iframe = self._search_regex(
r'(<iframe[^>]+data-video-id=["\']\d+[^>]+>)', webpage,
'brightcove iframe', default=None)
if brightcove_iframe:
attr = extract_attributes(brightcove_iframe)
src = attr.get('src')
if src:
parsed_src = compat_urlparse.urlparse(src)
qs = compat_urlparse.parse_qs(parsed_src.query)
account_id = qs.get('accountId', ['2376984109001'])[0]
brightcove_id = attr.get('data-video-id') or qs.get('videoId', [None])[0]
if account_id and brightcove_id:
return brightcove_url_result(
'http://players.brightcove.net/%s/default_default/index.html?videoId=%s'
% (account_id, brightcove_id))
# Query result is often embedded in webpage as JSON. Sometimes explicit requests
# to video API results in a failure with geo restriction reason therefore using
# embedded query result when present sounds reasonable.
config_json = self._search_regex(
r'window\.Af\.bootstrap\[[^\]]+\]\s*=\s*({.*?"applet_type"\s*:\s*"td-applet-videoplayer".*?});(?:</script>|$)',
webpage, 'videoplayer applet', default=None)
if config_json:
config = self._parse_json(config_json, display_id, fatal=False)
if config:
sapi = config.get('models', {}).get('applet_model', {}).get('data', {}).get('sapi')
if sapi and 'query' in sapi:
info = self._extract_info(display_id, sapi, webpage)
self._sort_formats(info['formats'])
return info
items_json = self._search_regex(
r'mediaItems: ({.*?})$', webpage, 'items', flags=re.MULTILINE,
default=None)
if items_json is None:
alias = self._search_regex(
r'"aliases":{"video":"(.*?)"', webpage, 'alias', default=None)
if alias is not None:
alias_info = self._download_json(
'https://www.yahoo.com/_td/api/resource/VideoService.videos;video_aliases=["%s"]' % alias,
display_id, 'Downloading alias info')
video_id = alias_info[0]['id']
else:
CONTENT_ID_REGEXES = [
r'YUI\.namespace\("Media"\)\.CONTENT_ID\s*=\s*"([^"]+)"',
r'root\.App\.Cache\.context\.videoCache\.curVideo = \{"([^"]+)"',
r'"first_videoid"\s*:\s*"([^"]+)"',
r'%s[^}]*"ccm_id"\s*:\s*"([^"]+)"' % re.escape(page_id),
r'<article[^>]data-uuid=["\']([^"\']+)',
r'<meta[^<>]+yahoo://article/view\?.*\buuid=([^&"\']+)',
r'<meta[^<>]+["\']ytwnews://cavideo/(?:[^/]+/)+([\da-fA-F-]+)[&"\']',
]
video_id = self._search_regex(
CONTENT_ID_REGEXES, webpage, 'content ID')
else:
items = json.loads(items_json)
info = items['mediaItems']['query']['results']['mediaObj'][0]
# The 'meta' field is not always in the video webpage, we request it
# from another page
video_id = info['id']
return self._get_info(video_id, display_id, webpage)
def _extract_info(self, display_id, query, webpage):
info = query['query']['results']['mediaObj'][0]
meta = info.get('meta')
video_id = info.get('id')
if not meta:
msg = info['status'].get('msg')
if msg:
raise ExtractorError(
'%s returned error: %s' % (self.IE_NAME, msg), expected=True)
raise ExtractorError('Unable to extract media object meta')
cover = item.get('cover') or {}
if cover.get('type') == 'yvideo':
cover_url = cover.get('url')
if cover_url:
entries.append(self.url_result(
cover_url, 'Yahoo', cover.get('uuid')))
for e in item.get('body', []):
if e.get('type') == 'videoIframe':
iframe_url = e.get('url')
if not iframe_url:
continue
entries.append(self.url_result(iframe_url))
return self.playlist_result(
entries, item.get('uuid'),
item.get('title'), item.get('summary'))
video_id = item['uuid']
video = self._download_json(
api_base + 'VideoService.videos;view=full;video_ids=["%s"]' % video_id,
video_id, 'Downloading video JSON metadata')[0]
title = video['title']
if country == 'malaysia':
country = 'my'
is_live = video.get('live_state') == 'live'
fmts = ('m3u8',) if is_live else ('web', 'mp4')
urls = []
formats = []
for s in info['streams']:
subtitles = {}
for fmt in fmts:
media_obj = self._download_json(
'https://video-api.yql.yahoo.com/v1/video/sapi/streams/' + video_id,
video_id, 'Downloading %s JSON metadata' % fmt,
headers=self.geo_verification_headers(), query={
'format': fmt,
'region': country.upper(),
})['query']['results']['mediaObj'][0]
msg = media_obj.get('status', {}).get('msg')
for s in media_obj.get('streams', []):
host = s.get('host')
path = s.get('path')
if not host or not path:
continue
s_url = host + path
if s.get('format') == 'm3u8':
formats.extend(self._extract_m3u8_formats(
s_url, video_id, 'mp4', m3u8_id='hls', fatal=False))
continue
tbr = int_or_none(s.get('bitrate'))
format_info = {
formats.append({
'url': s_url,
'format_id': fmt + ('-%d' % tbr if tbr else ''),
'width': int_or_none(s.get('width')),
'height': int_or_none(s.get('height')),
'tbr': tbr,
}
host = s['host']
path = s['path']
if host.startswith('rtmp'):
fmt = 'rtmp'
format_info.update({
'url': host,
'play_path': path,
'ext': 'flv',
'fps': int_or_none(s.get('framerate')),
})
else:
if s.get('format') == 'm3u8_playlist':
fmt = 'hls'
format_info.update({
'protocol': 'm3u8_native',
'ext': 'mp4',
for cc in media_obj.get('closedcaptions', []):
cc_url = cc.get('url')
if not cc_url or cc_url in urls:
continue
urls.append(cc_url)
subtitles.setdefault(cc.get('lang') or 'en-US', []).append({
'url': cc_url,
'ext': mimetype2ext(cc.get('content_type')),
})
else:
fmt = format_info['ext'] = determine_ext(path)
format_url = compat_urlparse.urljoin(host, path)
format_info['url'] = format_url
format_info['format_id'] = fmt + ('-%d' % tbr if tbr else '')
formats.append(format_info)
closed_captions = self._html_search_regex(
r'"closedcaptions":(\[[^\]]+\])', webpage, 'closed captions',
default='[]')
streaming_url = video.get('streaming_url')
if streaming_url and not is_live:
formats.extend(self._extract_m3u8_formats(
streaming_url, video_id, 'mp4',
'm3u8_native', m3u8_id='hls', fatal=False))
if not formats and msg == 'geo restricted':
self.raise_geo_restricted()
cc_json = self._parse_json(closed_captions, video_id, fatal=False)
subtitles = {}
if cc_json:
for closed_caption in cc_json:
lang = closed_caption['lang']
if lang not in subtitles:
subtitles[lang] = []
subtitles[lang].append({
'url': closed_caption['url'],
'ext': mimetype2ext(closed_caption['content_type']),
self._sort_formats(formats)
thumbnails = []
for thumb in video.get('thumbnails', []):
thumb_url = thumb.get('url')
if not thumb_url:
continue
thumbnails.append({
'id': thumb.get('tag'),
'url': thumb.get('url'),
'width': int_or_none(thumb.get('width')),
'height': int_or_none(thumb.get('height')),
})
series_info = video.get('series_info') or {}
return {
'id': video_id,
'display_id': display_id,
'title': unescapeHTML(meta['title']),
'title': self._live_title(title) if is_live else title,
'formats': formats,
'description': clean_html(meta['description']),
'thumbnail': meta['thumbnail'] if meta.get('thumbnail') else self._og_search_thumbnail(webpage),
'duration': int_or_none(meta.get('duration')),
'display_id': display_id,
'thumbnails': thumbnails,
'description': clean_html(video.get('description')),
'timestamp': parse_iso8601(video.get('publish_time')),
'subtitles': subtitles,
'duration': int_or_none(video.get('duration')),
'view_count': int_or_none(video.get('view_count')),
'is_live': is_live,
'series': video.get('show_name'),
'season_number': int_or_none(series_info.get('season_number')),
'episode_number': int_or_none(series_info.get('episode_number')),
}
def _get_info(self, video_id, display_id, webpage):
region = self._search_regex(
r'\\?"region\\?"\s*:\s*\\?"([^"]+?)\\?"',
webpage, 'region', fatal=False, default='US').upper()
formats = []
info = {}
for fmt in ('webm', 'mp4'):
query_result = self._download_json(
'https://video.media.yql.yahoo.com/v1/video/sapi/streams/' + video_id,
display_id, 'Downloading %s video info' % fmt, query={
'protocol': 'http',
'region': region,
'format': fmt,
})
info = self._extract_info(display_id, query_result, webpage)
formats.extend(info['formats'])
formats.extend(self._extract_m3u8_formats(
'http://video.media.yql.yahoo.com/v1/hls/%s?region=%s' % (video_id, region),
video_id, 'mp4', 'm3u8_native', m3u8_id='hls', fatal=False))
self._sort_formats(formats)
info['formats'] = formats
return info
class YahooSearchIE(SearchInfoExtractor):
IE_DESC = 'Yahoo screen search'

Write Preview
Loading…
Cancel
Save