[yahoo] improve extraction
[youtube-dl] / youtube_dl / extractor / yahoo.py
1 # coding: utf-8
2 from __future__ import unicode_literals
3
4 import hashlib
5 import itertools
6 import re
7
8 from .common import InfoExtractor, SearchInfoExtractor
9 from ..compat import (
10     compat_str,
11     compat_urllib_parse,
12 )
13 from ..utils import (
14     clean_html,
15     int_or_none,
16     mimetype2ext,
17     parse_iso8601,
18     smuggle_url,
19     try_get,
20     url_or_none,
21 )
22
23 from .brightcove import BrightcoveNewIE
24
25
26 class YahooIE(InfoExtractor):
27     IE_DESC = 'Yahoo screen and movies'
28     _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)'
29     _TESTS = [{
30         'url': 'http://screen.yahoo.com/julian-smith-travis-legg-watch-214727115.html',
31         'info_dict': {
32             'id': '2d25e626-2378-391f-ada0-ddaf1417e588',
33             'ext': 'mp4',
34             'title': 'Julian Smith & Travis Legg Watch Julian Smith',
35             'description': 'Julian and Travis watch Julian Smith',
36             'duration': 6863,
37             'timestamp': 1369812016,
38             'upload_date': '20130529',
39         },
40     }, {
41         'url': 'https://screen.yahoo.com/community/community-sizzle-reel-203225340.html?format=embed',
42         'md5': '7993e572fac98e044588d0b5260f4352',
43         'info_dict': {
44             'id': '4fe78544-8d48-39d8-97cd-13f205d9fcdb',
45             'ext': 'mp4',
46             'title': "Yahoo Saves 'Community'",
47             'description': 'md5:4d4145af2fd3de00cbb6c1d664105053',
48             'duration': 170,
49             'timestamp': 1406838636,
50             'upload_date': '20140731',
51         },
52     }, {
53         'url': 'https://uk.screen.yahoo.com/editor-picks/cute-raccoon-freed-drain-using-091756545.html',
54         'md5': '0b51660361f0e27c9789e7037ef76f4b',
55         'info_dict': {
56             'id': 'b3affa53-2e14-3590-852b-0e0db6cd1a58',
57             'ext': 'mp4',
58             'title': 'Cute Raccoon Freed From Drain\u00a0Using Angle Grinder',
59             'description': 'md5:f66c890e1490f4910a9953c941dee944',
60             'duration': 97,
61             'timestamp': 1414489862,
62             'upload_date': '20141028',
63         }
64     }, {
65         'url': 'http://news.yahoo.com/video/china-moses-crazy-blues-104538833.html',
66         'md5': '88e209b417f173d86186bef6e4d1f160',
67         'info_dict': {
68             'id': 'f885cf7f-43d4-3450-9fac-46ac30ece521',
69             'ext': 'mp4',
70             'title': 'China Moses Is Crazy About the Blues',
71             'description': 'md5:9900ab8cd5808175c7b3fe55b979bed0',
72             'duration': 128,
73             'timestamp': 1385722202,
74             'upload_date': '20131129',
75         }
76     }, {
77         'url': 'https://www.yahoo.com/movies/v/true-story-trailer-173000497.html',
78         'md5': '2a9752f74cb898af5d1083ea9f661b58',
79         'info_dict': {
80             'id': '071c4013-ce30-3a93-a5b2-e0413cd4a9d1',
81             'ext': 'mp4',
82             'title': '\'True Story\' Trailer',
83             'description': 'True Story',
84             'duration': 150,
85             'timestamp': 1418919206,
86             'upload_date': '20141218',
87         },
88     }, {
89         'url': 'https://gma.yahoo.com/pizza-delivery-man-surprised-huge-tip-college-kids-195200785.html',
90         'only_matching': True,
91     }, {
92         'note': 'NBC Sports embeds',
93         'url': 'http://sports.yahoo.com/blogs/ncaab-the-dagger/tyler-kalinoski-s-buzzer-beater-caps-davidson-s-comeback-win-185609842.html?guid=nbc_cbk_davidsonbuzzerbeater_150313',
94         'info_dict': {
95             'id': '9CsDKds0kvHI',
96             'ext': 'flv',
97             'description': 'md5:df390f70a9ba7c95ff1daace988f0d8d',
98             'title': 'Tyler Kalinoski hits buzzer-beater to lift Davidson',
99             'upload_date': '20150313',
100             'uploader': 'NBCU-SPORTS',
101             'timestamp': 1426270238,
102         },
103     }, {
104         'url': 'https://tw.news.yahoo.com/-100120367.html',
105         'only_matching': True,
106     }, {
107         # Query result is embedded in webpage, but explicit request to video API fails with geo restriction
108         'url': 'https://screen.yahoo.com/community/communitary-community-episode-1-ladders-154501237.html',
109         'md5': '4fbafb9c9b6f07aa8f870629f6671b35',
110         'info_dict': {
111             'id': '1f32853c-a271-3eef-8cb6-f6d6872cb504',
112             'ext': 'mp4',
113             'title': 'Communitary - Community Episode 1: Ladders',
114             'description': 'md5:8fc39608213295748e1e289807838c97',
115             'duration': 1646,
116             'timestamp': 1440436550,
117             'upload_date': '20150824',
118             'series': 'Communitary',
119             'season_number': 6,
120             'episode_number': 1,
121         },
122     }, {
123         # ytwnews://cavideo/
124         'url': 'https://tw.video.yahoo.com/movie-tw/單車天使-中文版預-092316541.html',
125         'info_dict': {
126             'id': 'ba133ff2-0793-3510-b636-59dfe9ff6cff',
127             'ext': 'mp4',
128             'title': '單車天使 - 中文版預',
129             'description': '中文版預',
130             'timestamp': 1476696196,
131             'upload_date': '20161017',
132         },
133         'params': {
134             'skip_download': True,
135         },
136     }, {
137         # Contains both a Yahoo hosted video and multiple Youtube embeds
138         'url': 'https://www.yahoo.com/entertainment/gwen-stefani-reveals-the-pop-hit-she-passed-on-assigns-it-to-her-voice-contestant-instead-033045672.html',
139         'info_dict': {
140             'id': '46c5d95a-528f-3d03-b732-732fcadd51de',
141             'title': 'Gwen Stefani reveals the pop hit she passed on, assigns it to her \'Voice\' contestant instead',
142             'description': 'Gwen decided not to record this hit herself, but she decided it was the perfect fit for Kyndall Inskeep.',
143         },
144         'playlist': [{
145             'info_dict': {
146                 'id': '966d4262-4fd1-3aaa-b45b-049ca6e38ba6',
147                 'ext': 'mp4',
148                 'title': 'Gwen Stefani reveals she turned down one of Sia\'s best songs',
149                 'description': 'On "The Voice" Tuesday, Gwen Stefani told Taylor Swift which Sia hit was almost hers.',
150                 'timestamp': 1572406500,
151                 'upload_date': '20191030',
152             },
153         }, {
154             'info_dict': {
155                 'id': '352CFDOQrKg',
156                 'ext': 'mp4',
157                 'title': 'Kyndal Inskeep "Performs the Hell Out of" Sia\'s "Elastic Heart" - The Voice Knockouts 2019',
158                 'description': 'md5:35b61e94c2ae214bc965ff4245f80d11',
159                 'uploader': 'The Voice',
160                 'uploader_id': 'NBCTheVoice',
161                 'upload_date': '20191029',
162             },
163         }],
164         'params': {
165             'playlistend': 2,
166         },
167     }, {
168         'url': 'https://malaysia.news.yahoo.com/video/bystanders-help-ontario-policeman-bust-190932818.html',
169         'only_matching': True,
170     }, {
171         'url': 'https://es-us.noticias.yahoo.com/es-la-puerta-irrompible-que-110539379.html',
172         'only_matching': True,
173     }]
174
175     def _real_extract(self, url):
176         url, country, display_id = re.match(self._VALID_URL, url).groups()
177         if not country:
178             country = 'us'
179         else:
180             country = country.split('-')[0]
181         api_base = 'https://%s.yahoo.com/_td/api/resource/' % country
182
183         for i, uuid in enumerate(['url=' + url, 'ymedia-alias=' + display_id]):
184             content = self._download_json(
185                 api_base + 'content;getDetailView=true;uuids=["%s"]' % uuid,
186                 display_id, 'Downloading content JSON metadata', fatal=i == 1)
187             if content:
188                 item = content['items'][0]
189                 break
190
191         if item.get('type') != 'video':
192             entries = []
193
194             cover = item.get('cover') or {}
195             if cover.get('type') == 'yvideo':
196                 cover_url = cover.get('url')
197                 if cover_url:
198                     entries.append(self.url_result(
199                         cover_url, 'Yahoo', cover.get('uuid')))
200
201             for e in item.get('body', []):
202                 if e.get('type') == 'videoIframe':
203                     iframe_url = e.get('url')
204                     if not iframe_url:
205                         continue
206                     entries.append(self.url_result(iframe_url))
207
208             return self.playlist_result(
209                 entries, item.get('uuid'),
210                 item.get('title'), item.get('summary'))
211
212         video_id = item['uuid']
213         video = self._download_json(
214             api_base + 'VideoService.videos;view=full;video_ids=["%s"]' % video_id,
215             video_id, 'Downloading video JSON metadata')[0]
216         title = video['title']
217
218         if country == 'malaysia':
219             country = 'my'
220
221         is_live = video.get('live_state') == 'live'
222         fmts = ('m3u8',) if is_live else ('web', 'mp4')
223
224         urls = []
225         formats = []
226         subtitles = {}
227         for fmt in fmts:
228             media_obj = self._download_json(
229                 'https://video-api.yql.yahoo.com/v1/video/sapi/streams/' + video_id,
230                 video_id, 'Downloading %s JSON metadata' % fmt,
231                 headers=self.geo_verification_headers(), query={
232                     'format': fmt,
233                     'region': country.upper(),
234                 })['query']['results']['mediaObj'][0]
235             msg = media_obj.get('status', {}).get('msg')
236
237             for s in media_obj.get('streams', []):
238                 host = s.get('host')
239                 path = s.get('path')
240                 if not host or not path:
241                     continue
242                 s_url = host + path
243                 if s.get('format') == 'm3u8':
244                     formats.extend(self._extract_m3u8_formats(
245                         s_url, video_id, 'mp4', m3u8_id='hls', fatal=False))
246                     continue
247                 tbr = int_or_none(s.get('bitrate'))
248                 formats.append({
249                     'url': s_url,
250                     'format_id': fmt + ('-%d' % tbr if tbr else ''),
251                     'width': int_or_none(s.get('width')),
252                     'height': int_or_none(s.get('height')),
253                     'tbr': tbr,
254                     'fps': int_or_none(s.get('framerate')),
255                 })
256
257             for cc in media_obj.get('closedcaptions', []):
258                 cc_url = cc.get('url')
259                 if not cc_url or cc_url in urls:
260                     continue
261                 urls.append(cc_url)
262                 subtitles.setdefault(cc.get('lang') or 'en-US', []).append({
263                     'url': cc_url,
264                     'ext': mimetype2ext(cc.get('content_type')),
265                 })
266
267         streaming_url = video.get('streaming_url')
268         if streaming_url and not is_live:
269             formats.extend(self._extract_m3u8_formats(
270                 streaming_url, video_id, 'mp4',
271                 'm3u8_native', m3u8_id='hls', fatal=False))
272
273         if not formats and msg == 'geo restricted':
274             self.raise_geo_restricted()
275
276         self._sort_formats(formats)
277
278         thumbnails = []
279         for thumb in video.get('thumbnails', []):
280             thumb_url = thumb.get('url')
281             if not thumb_url:
282                 continue
283             thumbnails.append({
284                 'id': thumb.get('tag'),
285                 'url': thumb.get('url'),
286                 'width': int_or_none(thumb.get('width')),
287                 'height': int_or_none(thumb.get('height')),
288             })
289
290         series_info = video.get('series_info') or {}
291
292         return {
293             'id': video_id,
294             'title': self._live_title(title) if is_live else title,
295             'formats': formats,
296             'display_id': display_id,
297             'thumbnails': thumbnails,
298             'description': clean_html(video.get('description')),
299             'timestamp': parse_iso8601(video.get('publish_time')),
300             'subtitles': subtitles,
301             'duration': int_or_none(video.get('duration')),
302             'view_count': int_or_none(video.get('view_count')),
303             'is_live': is_live,
304             'series': video.get('show_name'),
305             'season_number': int_or_none(series_info.get('season_number')),
306             'episode_number': int_or_none(series_info.get('episode_number')),
307         }
308
309
310 class YahooSearchIE(SearchInfoExtractor):
311     IE_DESC = 'Yahoo screen search'
312     _MAX_RESULTS = 1000
313     IE_NAME = 'screen.yahoo:search'
314     _SEARCH_KEY = 'yvsearch'
315
316     def _get_n_results(self, query, n):
317         """Get a specified number of results for a query"""
318         entries = []
319         for pagenum in itertools.count(0):
320             result_url = 'http://video.search.yahoo.com/search/?p=%s&fr=screen&o=js&gs=0&b=%d' % (compat_urllib_parse.quote_plus(query), pagenum * 30)
321             info = self._download_json(result_url, query,
322                                        note='Downloading results page ' + str(pagenum + 1))
323             m = info['m']
324             results = info['results']
325
326             for (i, r) in enumerate(results):
327                 if (pagenum * 30) + i >= n:
328                     break
329                 mobj = re.search(r'(?P<url>screen\.yahoo\.com/.*?-\d*?\.html)"', r)
330                 e = self.url_result('http://' + mobj.group('url'), 'Yahoo')
331                 entries.append(e)
332             if (pagenum * 30 + i >= n) or (m['last'] >= (m['total'] - 1)):
333                 break
334
335         return {
336             '_type': 'playlist',
337             'id': query,
338             'entries': entries,
339         }
340
341
342 class YahooGyaOPlayerIE(InfoExtractor):
343     IE_NAME = 'yahoo:gyao:player'
344     _VALID_URL = r'https?://(?:gyao\.yahoo\.co\.jp/(?:player|episode/[^/]+)|streaming\.yahoo\.co\.jp/c/y)/(?P<id>\d+/v\d+/v\d+|[\da-f]{8}-[\da-f]{4}-[\da-f]{4}-[\da-f]{4}-[\da-f]{12})'
345     _TESTS = [{
346         'url': 'https://gyao.yahoo.co.jp/player/00998/v00818/v0000000000000008564/',
347         'info_dict': {
348             'id': '5993125228001',
349             'ext': 'mp4',
350             'title': 'フューリー 【字幕版】',
351             'description': 'md5:21e691c798a15330eda4db17a8fe45a5',
352             'uploader_id': '4235717419001',
353             'upload_date': '20190124',
354             'timestamp': 1548294365,
355         },
356         'params': {
357             # m3u8 download
358             'skip_download': True,
359         },
360     }, {
361         'url': 'https://streaming.yahoo.co.jp/c/y/01034/v00133/v0000000000000000706/',
362         'only_matching': True,
363     }, {
364         'url': 'https://gyao.yahoo.co.jp/episode/%E3%81%8D%E3%81%AE%E3%81%86%E4%BD%95%E9%A3%9F%E3%81%B9%E3%81%9F%EF%BC%9F%20%E7%AC%AC2%E8%A9%B1%202019%2F4%2F12%E6%94%BE%E9%80%81%E5%88%86/5cb02352-b725-409e-9f8d-88f947a9f682',
365         'only_matching': True,
366     }]
367
368     def _real_extract(self, url):
369         video_id = self._match_id(url).replace('/', ':')
370         video = self._download_json(
371             'https://gyao.yahoo.co.jp/dam/v1/videos/' + video_id,
372             video_id, query={
373                 'fields': 'longDescription,title,videoId',
374             }, headers={
375                 'X-User-Agent': 'Unknown Pc GYAO!/2.0.0 Web',
376             })
377         return {
378             '_type': 'url_transparent',
379             'id': video_id,
380             'title': video['title'],
381             'url': smuggle_url(
382                 'http://players.brightcove.net/4235717419001/default_default/index.html?videoId=' + video['videoId'],
383                 {'geo_countries': ['JP']}),
384             'description': video.get('longDescription'),
385             'ie_key': BrightcoveNewIE.ie_key(),
386         }
387
388
389 class YahooGyaOIE(InfoExtractor):
390     IE_NAME = 'yahoo:gyao'
391     _VALID_URL = r'https?://(?:gyao\.yahoo\.co\.jp/(?:p|title/[^/]+)|streaming\.yahoo\.co\.jp/p/y)/(?P<id>\d+/v\d+|[\da-f]{8}-[\da-f]{4}-[\da-f]{4}-[\da-f]{4}-[\da-f]{12})'
392     _TESTS = [{
393         'url': 'https://gyao.yahoo.co.jp/p/00449/v03102/',
394         'info_dict': {
395             'id': '00449:v03102',
396         },
397         'playlist_count': 2,
398     }, {
399         'url': 'https://streaming.yahoo.co.jp/p/y/01034/v00133/',
400         'only_matching': True,
401     }, {
402         'url': 'https://gyao.yahoo.co.jp/title/%E3%81%97%E3%82%83%E3%81%B9%E3%81%8F%E3%82%8A007/5b025a49-b2e5-4dc7-945c-09c6634afacf',
403         'only_matching': True,
404     }]
405
406     def _real_extract(self, url):
407         program_id = self._match_id(url).replace('/', ':')
408         videos = self._download_json(
409             'https://gyao.yahoo.co.jp/api/programs/%s/videos' % program_id, program_id)['videos']
410         entries = []
411         for video in videos:
412             video_id = video.get('id')
413             if not video_id:
414                 continue
415             entries.append(self.url_result(
416                 'https://gyao.yahoo.co.jp/player/%s/' % video_id.replace(':', '/'),
417                 YahooGyaOPlayerIE.ie_key(), video_id))
418         return self.playlist_result(entries, program_id)
419
420
421 class YahooJapanNewsIE(InfoExtractor):
422     IE_NAME = 'yahoo:japannews'
423     IE_DESC = 'Yahoo! Japan News'
424     _VALID_URL = r'https?://(?P<host>(?:news|headlines)\.yahoo\.co\.jp)[^\d]*(?P<id>\d[\d-]*\d)?'
425     _GEO_COUNTRIES = ['JP']
426     _TESTS = [{
427         'url': 'https://headlines.yahoo.co.jp/videonews/ann?a=20190716-00000071-ann-int',
428         'info_dict': {
429             'id': '1736242',
430             'ext': 'mp4',
431             'title': 'ムン大統領が対日批判を強化“現金化”効果は?(テレビ朝日系(ANN)) - Yahoo!ニュース',
432             'description': '韓国の元徴用工らを巡る裁判の原告が弁護士が差し押さえた三菱重工業の資産を売却して - Yahoo!ニュース(テレビ朝日系(ANN))',
433             'thumbnail': r're:^https?://.*\.[a-zA-Z\d]{3,4}$',
434         },
435         'params': {
436             'skip_download': True,
437         },
438     }, {
439         # geo restricted
440         'url': 'https://headlines.yahoo.co.jp/hl?a=20190721-00000001-oxv-l04',
441         'only_matching': True,
442     }, {
443         'url': 'https://headlines.yahoo.co.jp/videonews/',
444         'only_matching': True,
445     }, {
446         'url': 'https://news.yahoo.co.jp',
447         'only_matching': True,
448     }, {
449         'url': 'https://news.yahoo.co.jp/byline/hashimotojunji/20190628-00131977/',
450         'only_matching': True,
451     }, {
452         'url': 'https://news.yahoo.co.jp/feature/1356',
453         'only_matching': True
454     }]
455
456     def _extract_formats(self, json_data, content_id):
457         formats = []
458
459         video_data = try_get(
460             json_data,
461             lambda x: x['ResultSet']['Result'][0]['VideoUrlSet']['VideoUrl'],
462             list)
463         for vid in video_data or []:
464             delivery = vid.get('delivery')
465             url = url_or_none(vid.get('Url'))
466             if not delivery or not url:
467                 continue
468             elif delivery == 'hls':
469                 formats.extend(
470                     self._extract_m3u8_formats(
471                         url, content_id, 'mp4', 'm3u8_native',
472                         m3u8_id='hls', fatal=False))
473             else:
474                 formats.append({
475                     'url': url,
476                     'format_id': 'http-%s' % compat_str(vid.get('bitrate', '')),
477                     'height': int_or_none(vid.get('height')),
478                     'width': int_or_none(vid.get('width')),
479                     'tbr': int_or_none(vid.get('bitrate')),
480                 })
481         self._remove_duplicate_formats(formats)
482         self._sort_formats(formats)
483
484         return formats
485
486     def _real_extract(self, url):
487         mobj = re.match(self._VALID_URL, url)
488         host = mobj.group('host')
489         display_id = mobj.group('id') or host
490
491         webpage = self._download_webpage(url, display_id)
492
493         title = self._html_search_meta(
494             ['og:title', 'twitter:title'], webpage, 'title', default=None
495         ) or self._html_search_regex('<title>([^<]+)</title>', webpage, 'title')
496
497         if display_id == host:
498             # Headline page (w/ multiple BC playlists) ('news.yahoo.co.jp', 'headlines.yahoo.co.jp/videonews/', ...)
499             stream_plists = re.findall(r'plist=(\d+)', webpage) or re.findall(r'plist["\']:\s*["\']([^"\']+)', webpage)
500             entries = [
501                 self.url_result(
502                     smuggle_url(
503                         'http://players.brightcove.net/5690807595001/HyZNerRl7_default/index.html?playlistId=%s' % plist_id,
504                         {'geo_countries': ['JP']}),
505                     ie='BrightcoveNew', video_id=plist_id)
506                 for plist_id in stream_plists]
507             return self.playlist_result(entries, playlist_title=title)
508
509         # Article page
510         description = self._html_search_meta(
511             ['og:description', 'description', 'twitter:description'],
512             webpage, 'description', default=None)
513         thumbnail = self._og_search_thumbnail(
514             webpage, default=None) or self._html_search_meta(
515             'twitter:image', webpage, 'thumbnail', default=None)
516         space_id = self._search_regex([
517             r'<script[^>]+class=["\']yvpub-player["\'][^>]+spaceid=([^&"\']+)',
518             r'YAHOO\.JP\.srch\.\w+link\.onLoad[^;]+spaceID["\' ]*:["\' ]+([^"\']+)',
519             r'<!--\s+SpaceID=(\d+)'
520         ], webpage, 'spaceid')
521
522         content_id = self._search_regex(
523             r'<script[^>]+class=["\']yvpub-player["\'][^>]+contentid=(?P<contentid>[^&"\']+)',
524             webpage, 'contentid', group='contentid')
525
526         json_data = self._download_json(
527             'https://feapi-yvpub.yahooapis.jp/v1/content/%s' % content_id,
528             content_id,
529             query={
530                 'appid': 'dj0zaiZpPVZMTVFJR0FwZWpiMyZzPWNvbnN1bWVyc2VjcmV0Jng9YjU-',
531                 'output': 'json',
532                 'space_id': space_id,
533                 'domain': host,
534                 'ak': hashlib.md5('_'.join((space_id, host)).encode()).hexdigest(),
535                 'device_type': '1100',
536             })
537         formats = self._extract_formats(json_data, content_id)
538
539         return {
540             'id': content_id,
541             'title': title,
542             'description': description,
543             'thumbnail': thumbnail,
544             'formats': formats,
545         }