]> Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/yahoo.py
Prepare to release.
[youtubedl] / youtube_dl / extractor / yahoo.py
1 # coding: utf-8
2 from __future__ import unicode_literals
3
4 import itertools
5 import json
6 import re
7
8 from .common import InfoExtractor, SearchInfoExtractor
9 from ..compat import (
10 compat_urllib_parse,
11 compat_urlparse,
12 )
13 from ..utils import (
14 clean_html,
15 unescapeHTML,
16 ExtractorError,
17 int_or_none,
18 mimetype2ext,
19 determine_ext,
20 )
21
22 from .brightcove import (
23 BrightcoveLegacyIE,
24 BrightcoveNewIE,
25 )
26 from .nbc import NBCSportsVPlayerIE
27
28
29 class YahooIE(InfoExtractor):
30 IE_DESC = 'Yahoo screen and movies'
31 _VALID_URL = r'(?P<url>(?P<host>https?://(?:[a-zA-Z]{2}\.)?[\da-zA-Z_-]+\.yahoo\.com)/(?:[^/]+/)*(?P<display_id>.+)?-(?P<id>[0-9]+)(?:-[a-z]+)?(?:\.html)?)'
32 _TESTS = [
33 {
34 'url': 'http://screen.yahoo.com/julian-smith-travis-legg-watch-214727115.html',
35 'info_dict': {
36 'id': '2d25e626-2378-391f-ada0-ddaf1417e588',
37 'ext': 'mp4',
38 'title': 'Julian Smith & Travis Legg Watch Julian Smith',
39 'description': 'Julian and Travis watch Julian Smith',
40 'duration': 6863,
41 },
42 },
43 {
44 'url': 'http://screen.yahoo.com/wired/codefellas-s1-ep12-cougar-lies-103000935.html',
45 'md5': '251af144a19ebc4a033e8ba91ac726bb',
46 'info_dict': {
47 'id': 'd1dedf8c-d58c-38c3-8963-e899929ae0a9',
48 'ext': 'mp4',
49 'title': 'Codefellas - The Cougar Lies with Spanish Moss',
50 'description': 'md5:66b627ab0a282b26352136ca96ce73c1',
51 'duration': 151,
52 },
53 },
54 {
55 'url': 'https://screen.yahoo.com/community/community-sizzle-reel-203225340.html?format=embed',
56 'md5': '7993e572fac98e044588d0b5260f4352',
57 'info_dict': {
58 'id': '4fe78544-8d48-39d8-97cd-13f205d9fcdb',
59 'ext': 'mp4',
60 'title': "Yahoo Saves 'Community'",
61 'description': 'md5:4d4145af2fd3de00cbb6c1d664105053',
62 'duration': 170,
63 }
64 },
65 {
66 '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',
67 'md5': '45c024bad51e63e9b6f6fad7a43a8c23',
68 'info_dict': {
69 'id': 'cac903b3-fcf4-3c14-b632-643ab541712f',
70 'ext': 'mp4',
71 'title': '敢問市長/黃秀霜批賴清德「非常高傲」',
72 'description': '直言台南沒捷運 交通居五都之末',
73 'duration': 396,
74 },
75 },
76 {
77 'url': 'https://uk.screen.yahoo.com/editor-picks/cute-raccoon-freed-drain-using-091756545.html',
78 'md5': '71298482f7c64cbb7fa064e4553ff1c1',
79 'info_dict': {
80 'id': 'b3affa53-2e14-3590-852b-0e0db6cd1a58',
81 'ext': 'webm',
82 'title': 'Cute Raccoon Freed From Drain\u00a0Using Angle Grinder',
83 'description': 'md5:f66c890e1490f4910a9953c941dee944',
84 'duration': 97,
85 }
86 },
87 {
88 'url': 'https://ca.sports.yahoo.com/video/program-makes-hockey-more-affordable-013127711.html',
89 'md5': '57e06440778b1828a6079d2f744212c4',
90 'info_dict': {
91 'id': 'c9fa2a36-0d4d-3937-b8f6-cc0fb1881e73',
92 'ext': 'mp4',
93 'title': 'Program that makes hockey more affordable not offered in Manitoba',
94 'description': 'md5:c54a609f4c078d92b74ffb9bf1f496f4',
95 'duration': 121,
96 },
97 'skip': 'Video gone',
98 }, {
99 'url': 'https://ca.finance.yahoo.com/news/hackers-sony-more-trouble-well-154609075.html',
100 'info_dict': {
101 'id': '154609075',
102 },
103 'playlist': [{
104 'md5': '000887d0dc609bc3a47c974151a40fb8',
105 'info_dict': {
106 'id': 'e624c4bc-3389-34de-9dfc-025f74943409',
107 'ext': 'mp4',
108 'title': '\'The Interview\' TV Spot: War',
109 'description': 'The Interview',
110 'duration': 30,
111 },
112 }, {
113 'md5': '81bc74faf10750fe36e4542f9a184c66',
114 'info_dict': {
115 'id': '1fc8ada0-718e-3abe-a450-bf31f246d1a9',
116 'ext': 'mp4',
117 'title': '\'The Interview\' TV Spot: Guys',
118 'description': 'The Interview',
119 'duration': 30,
120 },
121 }],
122 }, {
123 'url': 'http://news.yahoo.com/video/china-moses-crazy-blues-104538833.html',
124 'md5': '88e209b417f173d86186bef6e4d1f160',
125 'info_dict': {
126 'id': 'f885cf7f-43d4-3450-9fac-46ac30ece521',
127 'ext': 'mp4',
128 'title': 'China Moses Is Crazy About the Blues',
129 'description': 'md5:9900ab8cd5808175c7b3fe55b979bed0',
130 'duration': 128,
131 }
132 }, {
133 'url': 'https://in.lifestyle.yahoo.com/video/connect-dots-dark-side-virgo-090247395.html',
134 'md5': 'd9a083ccf1379127bf25699d67e4791b',
135 'info_dict': {
136 'id': '52aeeaa3-b3d1-30d8-9ef8-5d0cf05efb7c',
137 'ext': 'mp4',
138 'title': 'Connect the Dots: Dark Side of Virgo',
139 'description': 'md5:1428185051cfd1949807ad4ff6d3686a',
140 'duration': 201,
141 },
142 'skip': 'Domain name in.lifestyle.yahoo.com gone',
143 }, {
144 'url': 'https://www.yahoo.com/movies/v/true-story-trailer-173000497.html',
145 'md5': '2a9752f74cb898af5d1083ea9f661b58',
146 'info_dict': {
147 'id': '071c4013-ce30-3a93-a5b2-e0413cd4a9d1',
148 'ext': 'mp4',
149 'title': '\'True Story\' Trailer',
150 'description': 'True Story',
151 'duration': 150,
152 },
153 }, {
154 'url': 'https://gma.yahoo.com/pizza-delivery-man-surprised-huge-tip-college-kids-195200785.html',
155 'only_matching': True,
156 }, {
157 'note': 'NBC Sports embeds',
158 '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',
159 'info_dict': {
160 'id': '9CsDKds0kvHI',
161 'ext': 'flv',
162 'description': 'md5:df390f70a9ba7c95ff1daace988f0d8d',
163 'title': 'Tyler Kalinoski hits buzzer-beater to lift Davidson',
164 'upload_date': '20150313',
165 'uploader': 'NBCU-SPORTS',
166 'timestamp': 1426270238,
167 }
168 }, {
169 'url': 'https://tw.news.yahoo.com/-100120367.html',
170 'only_matching': True,
171 }, {
172 # Query result is embedded in webpage, but explicit request to video API fails with geo restriction
173 'url': 'https://screen.yahoo.com/community/communitary-community-episode-1-ladders-154501237.html',
174 'md5': '4fbafb9c9b6f07aa8f870629f6671b35',
175 'info_dict': {
176 'id': '1f32853c-a271-3eef-8cb6-f6d6872cb504',
177 'ext': 'mp4',
178 'title': 'Communitary - Community Episode 1: Ladders',
179 'description': 'md5:8fc39608213295748e1e289807838c97',
180 'duration': 1646,
181 },
182 }, {
183 # it uses an alias to get the video_id
184 'url': 'https://www.yahoo.com/movies/the-stars-of-daddys-home-have-very-different-212843197.html',
185 'info_dict': {
186 'id': '40eda9c8-8e5f-3552-8745-830f67d0c737',
187 'ext': 'mp4',
188 'title': 'Will Ferrell & Mark Wahlberg Are Pro-Spanking',
189 'description': 'While they play feuding fathers in \'Daddy\'s Home,\' star Will Ferrell & Mark Wahlberg share their true feelings on parenthood.',
190 },
191 },
192 {
193 # config['models']['applet_model']['data']['sapi'] has no query
194 'url': 'https://www.yahoo.com/music/livenation/event/galactic-2016',
195 'md5': 'dac0c72d502bc5facda80c9e6d5c98db',
196 'info_dict': {
197 'id': 'a6015640-e9e5-3efb-bb60-05589a183919',
198 'ext': 'mp4',
199 'description': 'Galactic',
200 'title': 'Dolla Diva (feat. Maggie Koerner)',
201 },
202 'skip': 'redirect to https://www.yahoo.com/music',
203 },
204 {
205 # yahoo://article/
206 'url': 'https://www.yahoo.com/movies/video/true-story-trailer-173000497.html',
207 'info_dict': {
208 'id': '071c4013-ce30-3a93-a5b2-e0413cd4a9d1',
209 'ext': 'mp4',
210 'title': "'True Story' Trailer",
211 'description': 'True Story',
212 },
213 'params': {
214 'skip_download': True,
215 },
216 },
217 {
218 # ytwnews://cavideo/
219 'url': 'https://tw.video.yahoo.com/movie-tw/單車天使-中文版預-092316541.html',
220 'info_dict': {
221 'id': 'ba133ff2-0793-3510-b636-59dfe9ff6cff',
222 'ext': 'mp4',
223 'title': '單車天使 - 中文版預',
224 'description': '中文版預',
225 },
226 'params': {
227 'skip_download': True,
228 },
229 },
230 ]
231
232 def _real_extract(self, url):
233 mobj = re.match(self._VALID_URL, url)
234 display_id = mobj.group('display_id') or self._match_id(url)
235 page_id = mobj.group('id')
236 url = mobj.group('url')
237 host = mobj.group('host')
238 webpage, urlh = self._download_webpage_handle(url, display_id)
239 if 'err=404' in urlh.geturl():
240 raise ExtractorError('Video gone', expected=True)
241
242 # Look for iframed media first
243 entries = []
244 iframe_urls = re.findall(r'<iframe[^>]+src="(/video/.+?-\d+\.html\?format=embed.*?)"', webpage)
245 for idx, iframe_url in enumerate(iframe_urls):
246 entries.append(self.url_result(host + iframe_url, 'Yahoo'))
247 if entries:
248 return self.playlist_result(entries, page_id)
249
250 # Look for NBCSports iframes
251 nbc_sports_url = NBCSportsVPlayerIE._extract_url(webpage)
252 if nbc_sports_url:
253 return self.url_result(nbc_sports_url, NBCSportsVPlayerIE.ie_key())
254
255 # Look for Brightcove Legacy Studio embeds
256 bc_url = BrightcoveLegacyIE._extract_brightcove_url(webpage)
257 if bc_url:
258 return self.url_result(bc_url, BrightcoveLegacyIE.ie_key())
259
260 # Look for Brightcove New Studio embeds
261 bc_url = BrightcoveNewIE._extract_url(webpage)
262 if bc_url:
263 return self.url_result(bc_url, BrightcoveNewIE.ie_key())
264
265 # Query result is often embedded in webpage as JSON. Sometimes explicit requests
266 # to video API results in a failure with geo restriction reason therefore using
267 # embedded query result when present sounds reasonable.
268 config_json = self._search_regex(
269 r'window\.Af\.bootstrap\[[^\]]+\]\s*=\s*({.*?"applet_type"\s*:\s*"td-applet-videoplayer".*?});(?:</script>|$)',
270 webpage, 'videoplayer applet', default=None)
271 if config_json:
272 config = self._parse_json(config_json, display_id, fatal=False)
273 if config:
274 sapi = config.get('models', {}).get('applet_model', {}).get('data', {}).get('sapi')
275 if sapi and 'query' in sapi:
276 info = self._extract_info(display_id, sapi, webpage)
277 self._sort_formats(info['formats'])
278 return info
279
280 items_json = self._search_regex(
281 r'mediaItems: ({.*?})$', webpage, 'items', flags=re.MULTILINE,
282 default=None)
283 if items_json is None:
284 alias = self._search_regex(
285 r'"aliases":{"video":"(.*?)"', webpage, 'alias', default=None)
286 if alias is not None:
287 alias_info = self._download_json(
288 'https://www.yahoo.com/_td/api/resource/VideoService.videos;video_aliases=["%s"]' % alias,
289 display_id, 'Downloading alias info')
290 video_id = alias_info[0]['id']
291 else:
292 CONTENT_ID_REGEXES = [
293 r'YUI\.namespace\("Media"\)\.CONTENT_ID\s*=\s*"([^"]+)"',
294 r'root\.App\.Cache\.context\.videoCache\.curVideo = \{"([^"]+)"',
295 r'"first_videoid"\s*:\s*"([^"]+)"',
296 r'%s[^}]*"ccm_id"\s*:\s*"([^"]+)"' % re.escape(page_id),
297 r'<article[^>]data-uuid=["\']([^"\']+)',
298 r'<meta[^<>]+yahoo://article/view\?.*\buuid=([^&"\']+)',
299 r'<meta[^<>]+["\']ytwnews://cavideo/(?:[^/]+/)+([\da-fA-F-]+)[&"\']',
300 ]
301 video_id = self._search_regex(
302 CONTENT_ID_REGEXES, webpage, 'content ID')
303 else:
304 items = json.loads(items_json)
305 info = items['mediaItems']['query']['results']['mediaObj'][0]
306 # The 'meta' field is not always in the video webpage, we request it
307 # from another page
308 video_id = info['id']
309 return self._get_info(video_id, display_id, webpage)
310
311 def _extract_info(self, display_id, query, webpage):
312 info = query['query']['results']['mediaObj'][0]
313 meta = info.get('meta')
314 video_id = info.get('id')
315
316 if not meta:
317 msg = info['status'].get('msg')
318 if msg:
319 raise ExtractorError(
320 '%s returned error: %s' % (self.IE_NAME, msg), expected=True)
321 raise ExtractorError('Unable to extract media object meta')
322
323 formats = []
324 for s in info['streams']:
325 tbr = int_or_none(s.get('bitrate'))
326 format_info = {
327 'width': int_or_none(s.get('width')),
328 'height': int_or_none(s.get('height')),
329 'tbr': tbr,
330 }
331
332 host = s['host']
333 path = s['path']
334 if host.startswith('rtmp'):
335 fmt = 'rtmp'
336 format_info.update({
337 'url': host,
338 'play_path': path,
339 'ext': 'flv',
340 })
341 else:
342 if s.get('format') == 'm3u8_playlist':
343 fmt = 'hls'
344 format_info.update({
345 'protocol': 'm3u8_native',
346 'ext': 'mp4',
347 })
348 else:
349 fmt = format_info['ext'] = determine_ext(path)
350 format_url = compat_urlparse.urljoin(host, path)
351 format_info['url'] = format_url
352 format_info['format_id'] = fmt + ('-%d' % tbr if tbr else '')
353 formats.append(format_info)
354
355 closed_captions = self._html_search_regex(
356 r'"closedcaptions":(\[[^\]]+\])', webpage, 'closed captions',
357 default='[]')
358
359 cc_json = self._parse_json(closed_captions, video_id, fatal=False)
360 subtitles = {}
361 if cc_json:
362 for closed_caption in cc_json:
363 lang = closed_caption['lang']
364 if lang not in subtitles:
365 subtitles[lang] = []
366 subtitles[lang].append({
367 'url': closed_caption['url'],
368 'ext': mimetype2ext(closed_caption['content_type']),
369 })
370
371 return {
372 'id': video_id,
373 'display_id': display_id,
374 'title': unescapeHTML(meta['title']),
375 'formats': formats,
376 'description': clean_html(meta['description']),
377 'thumbnail': meta['thumbnail'] if meta.get('thumbnail') else self._og_search_thumbnail(webpage),
378 'duration': int_or_none(meta.get('duration')),
379 'subtitles': subtitles,
380 }
381
382 def _get_info(self, video_id, display_id, webpage):
383 region = self._search_regex(
384 r'\\?"region\\?"\s*:\s*\\?"([^"]+?)\\?"',
385 webpage, 'region', fatal=False, default='US').upper()
386 formats = []
387 info = {}
388 for fmt in ('webm', 'mp4'):
389 query_result = self._download_json(
390 'https://video.media.yql.yahoo.com/v1/video/sapi/streams/' + video_id,
391 display_id, 'Downloading %s video info' % fmt, query={
392 'protocol': 'http',
393 'region': region,
394 'format': fmt,
395 })
396 info = self._extract_info(display_id, query_result, webpage)
397 formats.extend(info['formats'])
398 formats.extend(self._extract_m3u8_formats(
399 'http://video.media.yql.yahoo.com/v1/hls/%s?region=%s' % (video_id, region),
400 video_id, 'mp4', 'm3u8_native', m3u8_id='hls', fatal=False))
401 self._sort_formats(formats)
402 info['formats'] = formats
403 return info
404
405
406 class YahooSearchIE(SearchInfoExtractor):
407 IE_DESC = 'Yahoo screen search'
408 _MAX_RESULTS = 1000
409 IE_NAME = 'screen.yahoo:search'
410 _SEARCH_KEY = 'yvsearch'
411
412 def _get_n_results(self, query, n):
413 """Get a specified number of results for a query"""
414 entries = []
415 for pagenum in itertools.count(0):
416 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)
417 info = self._download_json(result_url, query,
418 note='Downloading results page ' + str(pagenum + 1))
419 m = info['m']
420 results = info['results']
421
422 for (i, r) in enumerate(results):
423 if (pagenum * 30) + i >= n:
424 break
425 mobj = re.search(r'(?P<url>screen\.yahoo\.com/.*?-\d*?\.html)"', r)
426 e = self.url_result('http://' + mobj.group('url'), 'Yahoo')
427 entries.append(e)
428 if (pagenum * 30 + i >= n) or (m['last'] >= (m['total'] - 1)):
429 break
430
431 return {
432 '_type': 'playlist',
433 'id': query,
434 'entries': entries,
435 }