]> Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/yahoo.py
b0679dfb70868f39b1190b8b04696787c02d75b1
[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_urllib_parse_urlencode,
12 compat_urlparse,
13 )
14 from ..utils import (
15 clean_html,
16 unescapeHTML,
17 ExtractorError,
18 int_or_none,
19 mimetype2ext,
20 )
21
22 from .brightcove import BrightcoveNewIE
23 from .nbc import NBCSportsVPlayerIE
24
25
26 class YahooIE(InfoExtractor):
27 IE_DESC = 'Yahoo screen and movies'
28 _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)?)'
29 _TESTS = [
30 {
31 'url': 'http://screen.yahoo.com/julian-smith-travis-legg-watch-214727115.html',
32 'info_dict': {
33 'id': '2d25e626-2378-391f-ada0-ddaf1417e588',
34 'ext': 'mp4',
35 'title': 'Julian Smith & Travis Legg Watch Julian Smith',
36 'description': 'Julian and Travis watch Julian Smith',
37 'duration': 6863,
38 },
39 },
40 {
41 'url': 'http://screen.yahoo.com/wired/codefellas-s1-ep12-cougar-lies-103000935.html',
42 'md5': 'c3466d2b6d5dd6b9f41ba9ed04c24b23',
43 'info_dict': {
44 'id': 'd1dedf8c-d58c-38c3-8963-e899929ae0a9',
45 'ext': 'mp4',
46 'title': 'Codefellas - The Cougar Lies with Spanish Moss',
47 'description': 'md5:66b627ab0a282b26352136ca96ce73c1',
48 'duration': 151,
49 },
50 },
51 {
52 'url': 'https://screen.yahoo.com/community/community-sizzle-reel-203225340.html?format=embed',
53 'md5': '75ffabdb87c16d4ffe8c036dc4d1c136',
54 'info_dict': {
55 'id': '4fe78544-8d48-39d8-97cd-13f205d9fcdb',
56 'ext': 'mp4',
57 'title': "Yahoo Saves 'Community'",
58 'description': 'md5:4d4145af2fd3de00cbb6c1d664105053',
59 'duration': 170,
60 }
61 },
62 {
63 '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',
64 'md5': '9035d38f88b1782682a3e89f985be5bb',
65 'info_dict': {
66 'id': 'cac903b3-fcf4-3c14-b632-643ab541712f',
67 'ext': 'mp4',
68 'title': '敢問市長/黃秀霜批賴清德「非常高傲」',
69 'description': '直言台南沒捷運 交通居五都之末',
70 'duration': 396,
71 },
72 },
73 {
74 'url': 'https://uk.screen.yahoo.com/editor-picks/cute-raccoon-freed-drain-using-091756545.html',
75 'md5': '0b51660361f0e27c9789e7037ef76f4b',
76 'info_dict': {
77 'id': 'b3affa53-2e14-3590-852b-0e0db6cd1a58',
78 'ext': 'mp4',
79 'title': 'Cute Raccoon Freed From Drain\u00a0Using Angle Grinder',
80 'description': 'md5:f66c890e1490f4910a9953c941dee944',
81 'duration': 97,
82 }
83 },
84 {
85 'url': 'https://ca.sports.yahoo.com/video/program-makes-hockey-more-affordable-013127711.html',
86 'md5': '57e06440778b1828a6079d2f744212c4',
87 'info_dict': {
88 'id': 'c9fa2a36-0d4d-3937-b8f6-cc0fb1881e73',
89 'ext': 'mp4',
90 'title': 'Program that makes hockey more affordable not offered in Manitoba',
91 'description': 'md5:c54a609f4c078d92b74ffb9bf1f496f4',
92 'duration': 121,
93 },
94 'skip': 'Video gone',
95 }, {
96 'url': 'https://ca.finance.yahoo.com/news/hackers-sony-more-trouble-well-154609075.html',
97 'info_dict': {
98 'id': '154609075',
99 },
100 'playlist': [{
101 'md5': 'f8e336c6b66f503282e5f719641d6565',
102 'info_dict': {
103 'id': 'e624c4bc-3389-34de-9dfc-025f74943409',
104 'ext': 'mp4',
105 'title': '\'The Interview\' TV Spot: War',
106 'description': 'The Interview',
107 'duration': 30,
108 },
109 }, {
110 'md5': '958bcb90b4d6df71c56312137ee1cd5a',
111 'info_dict': {
112 'id': '1fc8ada0-718e-3abe-a450-bf31f246d1a9',
113 'ext': 'mp4',
114 'title': '\'The Interview\' TV Spot: Guys',
115 'description': 'The Interview',
116 'duration': 30,
117 },
118 }],
119 }, {
120 'url': 'http://news.yahoo.com/video/china-moses-crazy-blues-104538833.html',
121 'md5': '88e209b417f173d86186bef6e4d1f160',
122 'info_dict': {
123 'id': 'f885cf7f-43d4-3450-9fac-46ac30ece521',
124 'ext': 'mp4',
125 'title': 'China Moses Is Crazy About the Blues',
126 'description': 'md5:9900ab8cd5808175c7b3fe55b979bed0',
127 'duration': 128,
128 }
129 }, {
130 'url': 'https://in.lifestyle.yahoo.com/video/connect-dots-dark-side-virgo-090247395.html',
131 'md5': 'd9a083ccf1379127bf25699d67e4791b',
132 'info_dict': {
133 'id': '52aeeaa3-b3d1-30d8-9ef8-5d0cf05efb7c',
134 'ext': 'mp4',
135 'title': 'Connect the Dots: Dark Side of Virgo',
136 'description': 'md5:1428185051cfd1949807ad4ff6d3686a',
137 'duration': 201,
138 },
139 'skip': 'Domain name in.lifestyle.yahoo.com gone',
140 }, {
141 'url': 'https://www.yahoo.com/movies/v/true-story-trailer-173000497.html',
142 'md5': 'b17ac378b1134fa44370fb27db09a744',
143 'info_dict': {
144 'id': '071c4013-ce30-3a93-a5b2-e0413cd4a9d1',
145 'ext': 'mp4',
146 'title': '\'True Story\' Trailer',
147 'description': 'True Story',
148 'duration': 150,
149 },
150 }, {
151 'url': 'https://gma.yahoo.com/pizza-delivery-man-surprised-huge-tip-college-kids-195200785.html',
152 'only_matching': True,
153 }, {
154 'note': 'NBC Sports embeds',
155 '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',
156 'info_dict': {
157 'id': '9CsDKds0kvHI',
158 'ext': 'flv',
159 'description': 'md5:df390f70a9ba7c95ff1daace988f0d8d',
160 'title': 'Tyler Kalinoski hits buzzer-beater to lift Davidson',
161 'upload_date': '20150313',
162 'uploader': 'NBCU-SPORTS',
163 'timestamp': 1426270238,
164 }
165 }, {
166 'url': 'https://tw.news.yahoo.com/-100120367.html',
167 'only_matching': True,
168 }, {
169 # Query result is embedded in webpage, but explicit request to video API fails with geo restriction
170 'url': 'https://screen.yahoo.com/community/communitary-community-episode-1-ladders-154501237.html',
171 'md5': '1ddbf7c850777548438e5c4f147c7b8c',
172 'info_dict': {
173 'id': '1f32853c-a271-3eef-8cb6-f6d6872cb504',
174 'ext': 'mp4',
175 'title': 'Communitary - Community Episode 1: Ladders',
176 'description': 'md5:8fc39608213295748e1e289807838c97',
177 'duration': 1646,
178 },
179 }, {
180 # it uses an alias to get the video_id
181 'url': 'https://www.yahoo.com/movies/the-stars-of-daddys-home-have-very-different-212843197.html',
182 'info_dict': {
183 'id': '40eda9c8-8e5f-3552-8745-830f67d0c737',
184 'ext': 'mp4',
185 'title': 'Will Ferrell & Mark Wahlberg Are Pro-Spanking',
186 'description': 'While they play feuding fathers in \'Daddy\'s Home,\' star Will Ferrell & Mark Wahlberg share their true feelings on parenthood.',
187 },
188 },
189 {
190 # config['models']['applet_model']['data']['sapi'] has no query
191 'url': 'https://www.yahoo.com/music/livenation/event/galactic-2016',
192 'md5': 'dac0c72d502bc5facda80c9e6d5c98db',
193 'info_dict': {
194 'id': 'a6015640-e9e5-3efb-bb60-05589a183919',
195 'ext': 'mp4',
196 'description': 'Galactic',
197 'title': 'Dolla Diva (feat. Maggie Koerner)',
198 },
199 },
200 ]
201
202 def _real_extract(self, url):
203 mobj = re.match(self._VALID_URL, url)
204 display_id = mobj.group('display_id') or self._match_id(url)
205 page_id = mobj.group('id')
206 url = mobj.group('url')
207 host = mobj.group('host')
208 webpage, urlh = self._download_webpage_handle(url, display_id)
209 if 'err=404' in urlh.geturl():
210 raise ExtractorError('Video gone', expected=True)
211
212 # Look for iframed media first
213 entries = []
214 iframe_urls = re.findall(r'<iframe[^>]+src="(/video/.+?-\d+\.html\?format=embed.*?)"', webpage)
215 for idx, iframe_url in enumerate(iframe_urls):
216 iframepage = self._download_webpage(
217 host + iframe_url, display_id,
218 note='Downloading iframe webpage for video #%d' % idx)
219 items_json = self._search_regex(
220 r'mediaItems: (\[.+?\])$', iframepage, 'items', flags=re.MULTILINE, default=None)
221 if items_json:
222 items = json.loads(items_json)
223 video_id = items[0]['id']
224 entries.append(self._get_info(video_id, display_id, webpage))
225 if entries:
226 return self.playlist_result(entries, page_id)
227
228 # Look for NBCSports iframes
229 nbc_sports_url = NBCSportsVPlayerIE._extract_url(webpage)
230 if nbc_sports_url:
231 return self.url_result(nbc_sports_url, NBCSportsVPlayerIE.ie_key())
232
233 # Look for Brightcove New Studio embeds
234 bc_url = BrightcoveNewIE._extract_url(webpage)
235 if bc_url:
236 return self.url_result(bc_url, BrightcoveNewIE.ie_key())
237
238 # Query result is often embedded in webpage as JSON. Sometimes explicit requests
239 # to video API results in a failure with geo restriction reason therefore using
240 # embedded query result when present sounds reasonable.
241 config_json = self._search_regex(
242 r'window\.Af\.bootstrap\[[^\]]+\]\s*=\s*({.*?"applet_type"\s*:\s*"td-applet-videoplayer".*?});(?:</script>|$)',
243 webpage, 'videoplayer applet', default=None)
244 if config_json:
245 config = self._parse_json(config_json, display_id, fatal=False)
246 if config:
247 sapi = config.get('models', {}).get('applet_model', {}).get('data', {}).get('sapi')
248 if sapi and 'query' in sapi:
249 return self._extract_info(display_id, sapi, webpage)
250
251 items_json = self._search_regex(
252 r'mediaItems: ({.*?})$', webpage, 'items', flags=re.MULTILINE,
253 default=None)
254 if items_json is None:
255 alias = self._search_regex(
256 r'"aliases":{"video":"(.*?)"', webpage, 'alias', default=None)
257 if alias is not None:
258 alias_info = self._download_json(
259 'https://www.yahoo.com/_td/api/resource/VideoService.videos;video_aliases=["%s"]' % alias,
260 display_id, 'Downloading alias info')
261 video_id = alias_info[0]['id']
262 else:
263 CONTENT_ID_REGEXES = [
264 r'YUI\.namespace\("Media"\)\.CONTENT_ID\s*=\s*"([^"]+)"',
265 r'root\.App\.Cache\.context\.videoCache\.curVideo = \{"([^"]+)"',
266 r'"first_videoid"\s*:\s*"([^"]+)"',
267 r'%s[^}]*"ccm_id"\s*:\s*"([^"]+)"' % re.escape(page_id),
268 r'<article[^>]data-uuid=["\']([^"\']+)',
269 r'yahoo://article/view\?.*\buuid=([^&"\']+)',
270 ]
271 video_id = self._search_regex(
272 CONTENT_ID_REGEXES, webpage, 'content ID')
273 else:
274 items = json.loads(items_json)
275 info = items['mediaItems']['query']['results']['mediaObj'][0]
276 # The 'meta' field is not always in the video webpage, we request it
277 # from another page
278 video_id = info['id']
279 return self._get_info(video_id, display_id, webpage)
280
281 def _extract_info(self, display_id, query, webpage):
282 info = query['query']['results']['mediaObj'][0]
283 meta = info.get('meta')
284 video_id = info.get('id')
285
286 if not meta:
287 msg = info['status'].get('msg')
288 if msg:
289 raise ExtractorError(
290 '%s returned error: %s' % (self.IE_NAME, msg), expected=True)
291 raise ExtractorError('Unable to extract media object meta')
292
293 formats = []
294 for s in info['streams']:
295 format_info = {
296 'width': int_or_none(s.get('width')),
297 'height': int_or_none(s.get('height')),
298 'tbr': int_or_none(s.get('bitrate')),
299 }
300
301 host = s['host']
302 path = s['path']
303 if host.startswith('rtmp'):
304 format_info.update({
305 'url': host,
306 'play_path': path,
307 'ext': 'flv',
308 })
309 else:
310 if s.get('format') == 'm3u8_playlist':
311 format_info['protocol'] = 'm3u8_native'
312 format_info['ext'] = 'mp4'
313 format_url = compat_urlparse.urljoin(host, path)
314 format_info['url'] = format_url
315 formats.append(format_info)
316
317 self._sort_formats(formats)
318
319 closed_captions = self._html_search_regex(
320 r'"closedcaptions":(\[[^\]]+\])', webpage, 'closed captions',
321 default='[]')
322
323 cc_json = self._parse_json(closed_captions, video_id, fatal=False)
324 subtitles = {}
325 if cc_json:
326 for closed_caption in cc_json:
327 lang = closed_caption['lang']
328 if lang not in subtitles:
329 subtitles[lang] = []
330 subtitles[lang].append({
331 'url': closed_caption['url'],
332 'ext': mimetype2ext(closed_caption['content_type']),
333 })
334
335 return {
336 'id': video_id,
337 'display_id': display_id,
338 'title': unescapeHTML(meta['title']),
339 'formats': formats,
340 'description': clean_html(meta['description']),
341 'thumbnail': meta['thumbnail'] if meta.get('thumbnail') else self._og_search_thumbnail(webpage),
342 'duration': int_or_none(meta.get('duration')),
343 'subtitles': subtitles,
344 }
345
346 def _get_info(self, video_id, display_id, webpage):
347 region = self._search_regex(
348 r'\\?"region\\?"\s*:\s*\\?"([^"]+?)\\?"',
349 webpage, 'region', fatal=False, default='US')
350 data = compat_urllib_parse_urlencode({
351 'protocol': 'http',
352 'region': region.upper(),
353 })
354 query_url = (
355 'https://video.media.yql.yahoo.com/v1/video/sapi/streams/'
356 '{id}?{data}'.format(id=video_id, data=data))
357 query_result = self._download_json(
358 query_url, display_id, 'Downloading video info')
359 return self._extract_info(display_id, query_result, webpage)
360
361
362 class YahooSearchIE(SearchInfoExtractor):
363 IE_DESC = 'Yahoo screen search'
364 _MAX_RESULTS = 1000
365 IE_NAME = 'screen.yahoo:search'
366 _SEARCH_KEY = 'yvsearch'
367
368 def _get_n_results(self, query, n):
369 """Get a specified number of results for a query"""
370 entries = []
371 for pagenum in itertools.count(0):
372 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)
373 info = self._download_json(result_url, query,
374 note='Downloading results page ' + str(pagenum + 1))
375 m = info['m']
376 results = info['results']
377
378 for (i, r) in enumerate(results):
379 if (pagenum * 30) + i >= n:
380 break
381 mobj = re.search(r'(?P<url>screen\.yahoo\.com/.*?-\d*?\.html)"', r)
382 e = self.url_result('http://' + mobj.group('url'), 'Yahoo')
383 entries.append(e)
384 if (pagenum * 30 + i >= n) or (m['last'] >= (m['total'] - 1)):
385 break
386
387 return {
388 '_type': 'playlist',
389 'id': query,
390 'entries': entries,
391 }