]> Raphaƫl G. Git Repositories - youtubedl/blob - youtube_dl/extractor/ted.py
d/rules: Run offline tests.
[youtubedl] / youtube_dl / extractor / ted.py
1 from __future__ import unicode_literals
2
3 import json
4 import re
5
6 from .common import InfoExtractor
7
8 from ..compat import (
9 compat_str,
10 compat_urlparse
11 )
12 from ..utils import (
13 extract_attributes,
14 float_or_none,
15 int_or_none,
16 try_get,
17 url_or_none,
18 )
19
20
21 class TEDIE(InfoExtractor):
22 IE_NAME = 'ted'
23 _VALID_URL = r'''(?x)
24 (?P<proto>https?://)
25 (?P<type>www|embed(?:-ssl)?)(?P<urlmain>\.ted\.com/
26 (
27 (?P<type_playlist>playlists(?:/(?P<playlist_id>\d+))?) # We have a playlist
28 |
29 ((?P<type_talk>talks)) # We have a simple talk
30 |
31 (?P<type_watch>watch)/[^/]+/[^/]+
32 )
33 (/lang/(.*?))? # The url may contain the language
34 /(?P<name>[\w-]+) # Here goes the name and then ".html"
35 .*)$
36 '''
37 _TESTS = [{
38 'url': 'http://www.ted.com/talks/dan_dennett_on_our_consciousness.html',
39 'md5': 'b0ce2b05ca215042124fbc9e3886493a',
40 'info_dict': {
41 'id': '102',
42 'ext': 'mp4',
43 'title': 'The illusion of consciousness',
44 'description': ('Philosopher Dan Dennett makes a compelling '
45 'argument that not only don\'t we understand our own '
46 'consciousness, but that half the time our brains are '
47 'actively fooling us.'),
48 'uploader': 'Dan Dennett',
49 'width': 853,
50 'duration': 1308,
51 'view_count': int,
52 'comment_count': int,
53 'tags': list,
54 },
55 'params': {
56 'skip_download': True,
57 },
58 }, {
59 # missing HTTP bitrates
60 'url': 'https://www.ted.com/talks/vishal_sikka_the_beauty_and_power_of_algorithms',
61 'info_dict': {
62 'id': '6069',
63 'ext': 'mp4',
64 'title': 'The beauty and power of algorithms',
65 'thumbnail': r're:^https?://.+\.jpg',
66 'description': 'md5:734e352710fb00d840ab87ae31aaf688',
67 'uploader': 'Vishal Sikka',
68 },
69 'params': {
70 'skip_download': True,
71 },
72 }, {
73 'url': 'http://www.ted.com/talks/gabby_giffords_and_mark_kelly_be_passionate_be_courageous_be_your_best',
74 'md5': 'e6b9617c01a7970ceac8bb2c92c346c0',
75 'info_dict': {
76 'id': '1972',
77 'ext': 'mp4',
78 'title': 'Be passionate. Be courageous. Be your best.',
79 'uploader': 'Gabby Giffords and Mark Kelly',
80 'description': 'md5:5174aed4d0f16021b704120360f72b92',
81 'duration': 1128,
82 },
83 'params': {
84 'skip_download': True,
85 },
86 }, {
87 'url': 'http://www.ted.com/playlists/who_are_the_hackers',
88 'info_dict': {
89 'id': '10',
90 'title': 'Who are the hackers?',
91 'description': 'md5:49a0dbe8fb76d81a0e64b4a80af7f15a'
92 },
93 'playlist_mincount': 6,
94 }, {
95 # contains a youtube video
96 'url': 'https://www.ted.com/talks/douglas_adams_parrots_the_universe_and_everything',
97 'add_ie': ['Youtube'],
98 'info_dict': {
99 'id': '_ZG8HBuDjgc',
100 'ext': 'webm',
101 'title': 'Douglas Adams: Parrots the Universe and Everything',
102 'description': 'md5:01ad1e199c49ac640cb1196c0e9016af',
103 'uploader': 'University of California Television (UCTV)',
104 'uploader_id': 'UCtelevision',
105 'upload_date': '20080522',
106 },
107 'params': {
108 'skip_download': True,
109 },
110 }, {
111 # no nativeDownloads
112 'url': 'https://www.ted.com/talks/tom_thum_the_orchestra_in_my_mouth',
113 'info_dict': {
114 'id': '1792',
115 'ext': 'mp4',
116 'title': 'The orchestra in my mouth',
117 'description': 'md5:5d1d78650e2f8dfcbb8ebee2951ac29a',
118 'uploader': 'Tom Thum',
119 'view_count': int,
120 'comment_count': int,
121 'tags': list,
122 },
123 'params': {
124 'skip_download': True,
125 },
126 }]
127
128 _NATIVE_FORMATS = {
129 'low': {'width': 320, 'height': 180},
130 'medium': {'width': 512, 'height': 288},
131 'high': {'width': 854, 'height': 480},
132 }
133
134 def _extract_info(self, webpage):
135 info_json = self._search_regex(
136 r'(?s)q\(\s*"\w+.init"\s*,\s*({.+?})\)\s*</script>',
137 webpage, 'info json')
138 return json.loads(info_json)
139
140 def _real_extract(self, url):
141 m = re.match(self._VALID_URL, url, re.VERBOSE)
142 if m.group('type').startswith('embed'):
143 desktop_url = m.group('proto') + 'www' + m.group('urlmain')
144 return self.url_result(desktop_url, 'TED')
145 name = m.group('name')
146 if m.group('type_talk'):
147 return self._talk_info(url, name)
148 elif m.group('type_watch'):
149 return self._watch_info(url, name)
150 else:
151 return self._playlist_videos_info(url, name)
152
153 def _playlist_videos_info(self, url, name):
154 '''Returns the videos of the playlist'''
155
156 webpage = self._download_webpage(url, name,
157 'Downloading playlist webpage')
158
159 playlist_entries = []
160 for entry in re.findall(r'(?s)<[^>]+data-ga-context=["\']playlist["\'][^>]*>', webpage):
161 attrs = extract_attributes(entry)
162 entry_url = compat_urlparse.urljoin(url, attrs['href'])
163 playlist_entries.append(self.url_result(entry_url, self.ie_key()))
164
165 final_url = self._og_search_url(webpage, fatal=False)
166 playlist_id = (
167 re.match(self._VALID_URL, final_url).group('playlist_id')
168 if final_url else None)
169
170 return self.playlist_result(
171 playlist_entries, playlist_id=playlist_id,
172 playlist_title=self._og_search_title(webpage, fatal=False),
173 playlist_description=self._og_search_description(webpage))
174
175 def _talk_info(self, url, video_name):
176 webpage = self._download_webpage(url, video_name)
177
178 info = self._extract_info(webpage)
179
180 data = try_get(info, lambda x: x['__INITIAL_DATA__'], dict) or info
181 talk_info = data['talks'][0]
182
183 title = talk_info['title'].strip()
184
185 downloads = talk_info.get('downloads') or {}
186 native_downloads = downloads.get('nativeDownloads') or talk_info.get('nativeDownloads') or {}
187
188 formats = [{
189 'url': format_url,
190 'format_id': format_id,
191 } for (format_id, format_url) in native_downloads.items() if format_url is not None]
192
193 subtitled_downloads = downloads.get('subtitledDownloads') or {}
194 for lang, subtitled_download in subtitled_downloads.items():
195 for q in self._NATIVE_FORMATS:
196 q_url = subtitled_download.get(q)
197 if not q_url:
198 continue
199 formats.append({
200 'url': q_url,
201 'format_id': '%s-%s' % (q, lang),
202 'language': lang,
203 })
204
205 if formats:
206 for f in formats:
207 finfo = self._NATIVE_FORMATS.get(f['format_id'].split('-')[0])
208 if finfo:
209 f.update(finfo)
210
211 player_talk = talk_info['player_talks'][0]
212
213 external = player_talk.get('external')
214 if isinstance(external, dict):
215 service = external.get('service')
216 if isinstance(service, compat_str):
217 ext_url = None
218 if service.lower() == 'youtube':
219 ext_url = external.get('code')
220
221 return self.url_result(ext_url or external['uri'])
222
223 resources_ = player_talk.get('resources') or talk_info.get('resources')
224
225 http_url = None
226 for format_id, resources in resources_.items():
227 if format_id == 'hls':
228 if not isinstance(resources, dict):
229 continue
230 stream_url = url_or_none(resources.get('stream'))
231 if not stream_url:
232 continue
233 formats.extend(self._extract_m3u8_formats(
234 stream_url, video_name, 'mp4', m3u8_id=format_id,
235 fatal=False))
236 else:
237 if not isinstance(resources, list):
238 continue
239 if format_id == 'h264':
240 for resource in resources:
241 h264_url = resource.get('file')
242 if not h264_url:
243 continue
244 bitrate = int_or_none(resource.get('bitrate'))
245 formats.append({
246 'url': h264_url,
247 'format_id': '%s-%sk' % (format_id, bitrate),
248 'tbr': bitrate,
249 })
250 if re.search(r'\d+k', h264_url):
251 http_url = h264_url
252 elif format_id == 'rtmp':
253 streamer = talk_info.get('streamer')
254 if not streamer:
255 continue
256 for resource in resources:
257 formats.append({
258 'format_id': '%s-%s' % (format_id, resource.get('name')),
259 'url': streamer,
260 'play_path': resource['file'],
261 'ext': 'flv',
262 'width': int_or_none(resource.get('width')),
263 'height': int_or_none(resource.get('height')),
264 'tbr': int_or_none(resource.get('bitrate')),
265 })
266
267 m3u8_formats = list(filter(
268 lambda f: f.get('protocol') == 'm3u8' and f.get('vcodec') != 'none',
269 formats))
270 if http_url:
271 for m3u8_format in m3u8_formats:
272 bitrate = self._search_regex(r'(\d+k)', m3u8_format['url'], 'bitrate', default=None)
273 if not bitrate:
274 continue
275 bitrate_url = re.sub(r'\d+k', bitrate, http_url)
276 if not self._is_valid_url(
277 bitrate_url, video_name, '%s bitrate' % bitrate):
278 continue
279 f = m3u8_format.copy()
280 f.update({
281 'url': bitrate_url,
282 'format_id': m3u8_format['format_id'].replace('hls', 'http'),
283 'protocol': 'http',
284 })
285 if f.get('acodec') == 'none':
286 del f['acodec']
287 formats.append(f)
288
289 audio_download = talk_info.get('audioDownload')
290 if audio_download:
291 formats.append({
292 'url': audio_download,
293 'format_id': 'audio',
294 'vcodec': 'none',
295 })
296
297 self._sort_formats(formats)
298
299 video_id = compat_str(talk_info['id'])
300
301 return {
302 'id': video_id,
303 'title': title,
304 'uploader': player_talk.get('speaker') or talk_info.get('speaker'),
305 'thumbnail': player_talk.get('thumb') or talk_info.get('thumb'),
306 'description': self._og_search_description(webpage),
307 'subtitles': self._get_subtitles(video_id, talk_info),
308 'formats': formats,
309 'duration': float_or_none(talk_info.get('duration')),
310 'view_count': int_or_none(data.get('viewed_count')),
311 'comment_count': int_or_none(
312 try_get(data, lambda x: x['comments']['count'])),
313 'tags': try_get(talk_info, lambda x: x['tags'], list),
314 }
315
316 def _get_subtitles(self, video_id, talk_info):
317 sub_lang_list = {}
318 for language in try_get(
319 talk_info,
320 (lambda x: x['downloads']['languages'],
321 lambda x: x['languages']), list):
322 lang_code = language.get('languageCode') or language.get('ianaCode')
323 if not lang_code:
324 continue
325 sub_lang_list[lang_code] = [
326 {
327 'url': 'http://www.ted.com/talks/subtitles/id/%s/lang/%s/format/%s' % (video_id, lang_code, ext),
328 'ext': ext,
329 }
330 for ext in ['ted', 'srt']
331 ]
332 return sub_lang_list
333
334 def _watch_info(self, url, name):
335 webpage = self._download_webpage(url, name)
336
337 config_json = self._html_search_regex(
338 r'"pages\.jwplayer"\s*,\s*({.+?})\s*\)\s*</script>',
339 webpage, 'config', default=None)
340 if not config_json:
341 embed_url = self._search_regex(
342 r"<iframe[^>]+class='pages-video-embed__video__object'[^>]+src='([^']+)'", webpage, 'embed url')
343 return self.url_result(self._proto_relative_url(embed_url))
344 config = json.loads(config_json)['config']
345 video_url = config['video']['url']
346 thumbnail = config.get('image', {}).get('url')
347
348 title = self._html_search_regex(
349 r"(?s)<h1(?:\s+class='[^']+')?>(.+?)</h1>", webpage, 'title')
350 description = self._html_search_regex(
351 [
352 r'(?s)<h4 class="[^"]+" id="h3--about-this-talk">.*?</h4>(.*?)</div>',
353 r'(?s)<p><strong>About this talk:</strong>\s+(.*?)</p>',
354 ],
355 webpage, 'description', fatal=False)
356
357 return {
358 'id': name,
359 'url': video_url,
360 'title': title,
361 'thumbnail': thumbnail,
362 'description': description,
363 }