]> Raphaƫl G. Git Repositories - youtubedl/blob - youtube_dl/extractor/youtube.py
c7922c533343ca162c922659d34dc867d6839864
[youtubedl] / youtube_dl / extractor / youtube.py
1 # coding: utf-8
2
3 import json
4 import netrc
5 import re
6 import socket
7
8 from .common import InfoExtractor, SearchInfoExtractor
9 from ..utils import (
10 compat_http_client,
11 compat_parse_qs,
12 compat_urllib_error,
13 compat_urllib_parse,
14 compat_urllib_request,
15 compat_str,
16
17 clean_html,
18 get_element_by_id,
19 ExtractorError,
20 unescapeHTML,
21 unified_strdate,
22 )
23
24
25 class YoutubeIE(InfoExtractor):
26 """Information extractor for youtube.com."""
27
28 _VALID_URL = r"""^
29 (
30 (?:https?://)? # http(s):// (optional)
31 (?:youtu\.be/|(?:\w+\.)?youtube(?:-nocookie)?\.com/|
32 tube\.majestyc\.net/) # the various hostnames, with wildcard subdomains
33 (?:.*?\#/)? # handle anchor (#/) redirect urls
34 (?: # the various things that can precede the ID:
35 (?:(?:v|embed|e)/) # v/ or embed/ or e/
36 |(?: # or the v= param in all its forms
37 (?:watch(?:_popup)?(?:\.php)?)? # preceding watch(_popup|.php) or nothing (like /?v=xxxx)
38 (?:\?|\#!?) # the params delimiter ? or # or #!
39 (?:.*?&)? # any other preceding param (like /?s=tuff&v=xxxx)
40 v=
41 )
42 )? # optional -> youtube.com/xxxx is OK
43 )? # all until now is optional -> you can pass the naked ID
44 ([0-9A-Za-z_-]+) # here is it! the YouTube video ID
45 (?(1).+)? # if we found the ID, everything can follow
46 $"""
47 _LANG_URL = r'https://www.youtube.com/?hl=en&persist_hl=1&gl=US&persist_gl=1&opt_out_ackd=1'
48 _LOGIN_URL = 'https://accounts.google.com/ServiceLogin'
49 _AGE_URL = 'http://www.youtube.com/verify_age?next_url=/&gl=US&hl=en'
50 _NEXT_URL_RE = r'[\?&]next_url=([^&]+)'
51 _NETRC_MACHINE = 'youtube'
52 # Listed in order of quality
53 _available_formats = ['38', '37', '46', '22', '45', '35', '44', '34', '18', '43', '6', '5', '17', '13']
54 _available_formats_prefer_free = ['38', '46', '37', '45', '22', '44', '35', '43', '34', '18', '6', '5', '17', '13']
55 _video_extensions = {
56 '13': '3gp',
57 '17': 'mp4',
58 '18': 'mp4',
59 '22': 'mp4',
60 '37': 'mp4',
61 '38': 'mp4',
62 '43': 'webm',
63 '44': 'webm',
64 '45': 'webm',
65 '46': 'webm',
66 }
67 _video_dimensions = {
68 '5': '240x400',
69 '6': '???',
70 '13': '???',
71 '17': '144x176',
72 '18': '360x640',
73 '22': '720x1280',
74 '34': '360x640',
75 '35': '480x854',
76 '37': '1080x1920',
77 '38': '3072x4096',
78 '43': '360x640',
79 '44': '480x854',
80 '45': '720x1280',
81 '46': '1080x1920',
82 }
83 IE_NAME = u'youtube'
84
85 @classmethod
86 def suitable(cls, url):
87 """Receives a URL and returns True if suitable for this IE."""
88 if YoutubePlaylistIE.suitable(url): return False
89 return re.match(cls._VALID_URL, url, re.VERBOSE) is not None
90
91 def report_lang(self):
92 """Report attempt to set language."""
93 self.to_screen(u'Setting language')
94
95 def report_login(self):
96 """Report attempt to log in."""
97 self.to_screen(u'Logging in')
98
99 def report_video_webpage_download(self, video_id):
100 """Report attempt to download video webpage."""
101 self.to_screen(u'%s: Downloading video webpage' % video_id)
102
103 def report_video_info_webpage_download(self, video_id):
104 """Report attempt to download video info webpage."""
105 self.to_screen(u'%s: Downloading video info webpage' % video_id)
106
107 def report_video_subtitles_download(self, video_id):
108 """Report attempt to download video info webpage."""
109 self.to_screen(u'%s: Checking available subtitles' % video_id)
110
111 def report_video_subtitles_request(self, video_id, sub_lang, format):
112 """Report attempt to download video info webpage."""
113 self.to_screen(u'%s: Downloading video subtitles for %s.%s' % (video_id, sub_lang, format))
114
115 def report_video_subtitles_available(self, video_id, sub_lang_list):
116 """Report available subtitles."""
117 sub_lang = ",".join(list(sub_lang_list.keys()))
118 self.to_screen(u'%s: Available subtitles for video: %s' % (video_id, sub_lang))
119
120 def report_information_extraction(self, video_id):
121 """Report attempt to extract video information."""
122 self.to_screen(u'%s: Extracting video information' % video_id)
123
124 def report_unavailable_format(self, video_id, format):
125 """Report extracted video URL."""
126 self.to_screen(u'%s: Format %s not available' % (video_id, format))
127
128 def report_rtmp_download(self):
129 """Indicate the download will use the RTMP protocol."""
130 self.to_screen(u'RTMP download detected')
131
132 def _decrypt_signature(self, s):
133 """Decrypt the key the two subkeys must have a length of 43"""
134
135 if len(s) == 88:
136 return s[48] + s[81:67:-1] + s[82] + s[66:62:-1] + s[85] + s[61:48:-1] + s[67] + s[47:12:-1] + s[3] + s[11:3:-1] + s[2] + s[12]
137 elif len(s) == 87:
138 return s[62] + s[82:62:-1] + s[83] + s[61:52:-1] + s[0] + s[51:2:-1]
139 elif len(s) == 86:
140 return s[2:63] + s[82] + s[64:82] + s[63]
141 elif len(s) == 85:
142 return s[76] + s[82:76:-1] + s[83] + s[75:60:-1] + s[0] + s[59:50:-1] + s[1] + s[49:2:-1]
143 elif len(s) == 84:
144 return s[83:36:-1] + s[2] + s[35:26:-1] + s[3] + s[25:3:-1] + s[26]
145 elif len(s) == 83:
146 return s[52] + s[81:55:-1] + s[2] + s[54:52:-1] + s[82] + s[51:36:-1] + s[55] + s[35:2:-1] + s[36]
147 elif len(s) == 82:
148 return s[36] + s[79:67:-1] + s[81] + s[66:40:-1] + s[33] + s[39:36:-1] + s[40] + s[35] + s[0] + s[67] + s[32:0:-1] + s[34]
149
150 else:
151 raise ExtractorError(u'Unable to decrypt signature, subkeys length %d not supported; retrying might work' % (len(s)))
152
153 def _get_available_subtitles(self, video_id):
154 self.report_video_subtitles_download(video_id)
155 request = compat_urllib_request.Request('http://video.google.com/timedtext?hl=en&type=list&v=%s' % video_id)
156 try:
157 sub_list = compat_urllib_request.urlopen(request).read().decode('utf-8')
158 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
159 return (u'unable to download video subtitles: %s' % compat_str(err), None)
160 sub_lang_list = re.findall(r'name="([^"]*)"[^>]+lang_code="([\w\-]+)"', sub_list)
161 sub_lang_list = dict((l[1], l[0]) for l in sub_lang_list)
162 if not sub_lang_list:
163 return (u'video doesn\'t have subtitles', None)
164 return sub_lang_list
165
166 def _list_available_subtitles(self, video_id):
167 sub_lang_list = self._get_available_subtitles(video_id)
168 self.report_video_subtitles_available(video_id, sub_lang_list)
169
170 def _request_subtitle(self, sub_lang, sub_name, video_id, format):
171 """
172 Return tuple:
173 (error_message, sub_lang, sub)
174 """
175 self.report_video_subtitles_request(video_id, sub_lang, format)
176 params = compat_urllib_parse.urlencode({
177 'lang': sub_lang,
178 'name': sub_name,
179 'v': video_id,
180 'fmt': format,
181 })
182 url = 'http://www.youtube.com/api/timedtext?' + params
183 try:
184 sub = compat_urllib_request.urlopen(url).read().decode('utf-8')
185 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
186 return (u'unable to download video subtitles: %s' % compat_str(err), None, None)
187 if not sub:
188 return (u'Did not fetch video subtitles', None, None)
189 return (None, sub_lang, sub)
190
191 def _request_automatic_caption(self, video_id, webpage):
192 """We need the webpage for getting the captions url, pass it as an
193 argument to speed up the process."""
194 sub_lang = self._downloader.params.get('subtitleslang') or 'en'
195 sub_format = self._downloader.params.get('subtitlesformat')
196 self.to_screen(u'%s: Looking for automatic captions' % video_id)
197 mobj = re.search(r';ytplayer.config = ({.*?});', webpage)
198 err_msg = u'Couldn\'t find automatic captions for "%s"' % sub_lang
199 if mobj is None:
200 return [(err_msg, None, None)]
201 player_config = json.loads(mobj.group(1))
202 try:
203 args = player_config[u'args']
204 caption_url = args[u'ttsurl']
205 timestamp = args[u'timestamp']
206 params = compat_urllib_parse.urlencode({
207 'lang': 'en',
208 'tlang': sub_lang,
209 'fmt': sub_format,
210 'ts': timestamp,
211 'kind': 'asr',
212 })
213 subtitles_url = caption_url + '&' + params
214 sub = self._download_webpage(subtitles_url, video_id, u'Downloading automatic captions')
215 return [(None, sub_lang, sub)]
216 except KeyError:
217 return [(err_msg, None, None)]
218
219 def _extract_subtitle(self, video_id):
220 """
221 Return a list with a tuple:
222 [(error_message, sub_lang, sub)]
223 """
224 sub_lang_list = self._get_available_subtitles(video_id)
225 sub_format = self._downloader.params.get('subtitlesformat')
226 if isinstance(sub_lang_list,tuple): #There was some error, it didn't get the available subtitles
227 return [(sub_lang_list[0], None, None)]
228 if self._downloader.params.get('subtitleslang', False):
229 sub_lang = self._downloader.params.get('subtitleslang')
230 elif 'en' in sub_lang_list:
231 sub_lang = 'en'
232 else:
233 sub_lang = list(sub_lang_list.keys())[0]
234 if not sub_lang in sub_lang_list:
235 return [(u'no closed captions found in the specified language "%s"' % sub_lang, None, None)]
236
237 subtitle = self._request_subtitle(sub_lang, sub_lang_list[sub_lang].encode('utf-8'), video_id, sub_format)
238 return [subtitle]
239
240 def _extract_all_subtitles(self, video_id):
241 sub_lang_list = self._get_available_subtitles(video_id)
242 sub_format = self._downloader.params.get('subtitlesformat')
243 if isinstance(sub_lang_list,tuple): #There was some error, it didn't get the available subtitles
244 return [(sub_lang_list[0], None, None)]
245 subtitles = []
246 for sub_lang in sub_lang_list:
247 subtitle = self._request_subtitle(sub_lang, sub_lang_list[sub_lang].encode('utf-8'), video_id, sub_format)
248 subtitles.append(subtitle)
249 return subtitles
250
251 def _print_formats(self, formats):
252 print('Available formats:')
253 for x in formats:
254 print('%s\t:\t%s\t[%s]' %(x, self._video_extensions.get(x, 'flv'), self._video_dimensions.get(x, '???')))
255
256 def _real_initialize(self):
257 if self._downloader is None:
258 return
259
260 username = None
261 password = None
262 downloader_params = self._downloader.params
263
264 # Attempt to use provided username and password or .netrc data
265 if downloader_params.get('username', None) is not None:
266 username = downloader_params['username']
267 password = downloader_params['password']
268 elif downloader_params.get('usenetrc', False):
269 try:
270 info = netrc.netrc().authenticators(self._NETRC_MACHINE)
271 if info is not None:
272 username = info[0]
273 password = info[2]
274 else:
275 raise netrc.NetrcParseError('No authenticators for %s' % self._NETRC_MACHINE)
276 except (IOError, netrc.NetrcParseError) as err:
277 self._downloader.report_warning(u'parsing .netrc: %s' % compat_str(err))
278 return
279
280 # Set language
281 request = compat_urllib_request.Request(self._LANG_URL)
282 try:
283 self.report_lang()
284 compat_urllib_request.urlopen(request).read()
285 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
286 self._downloader.report_warning(u'unable to set language: %s' % compat_str(err))
287 return
288
289 # No authentication to be performed
290 if username is None:
291 return
292
293 request = compat_urllib_request.Request(self._LOGIN_URL)
294 try:
295 login_page = compat_urllib_request.urlopen(request).read().decode('utf-8')
296 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
297 self._downloader.report_warning(u'unable to fetch login page: %s' % compat_str(err))
298 return
299
300 galx = None
301 dsh = None
302 match = re.search(re.compile(r'<input.+?name="GALX".+?value="(.+?)"', re.DOTALL), login_page)
303 if match:
304 galx = match.group(1)
305
306 match = re.search(re.compile(r'<input.+?name="dsh".+?value="(.+?)"', re.DOTALL), login_page)
307 if match:
308 dsh = match.group(1)
309
310 # Log in
311 login_form_strs = {
312 u'continue': u'https://www.youtube.com/signin?action_handle_signin=true&feature=sign_in_button&hl=en_US&nomobiletemp=1',
313 u'Email': username,
314 u'GALX': galx,
315 u'Passwd': password,
316 u'PersistentCookie': u'yes',
317 u'_utf8': u'霱',
318 u'bgresponse': u'js_disabled',
319 u'checkConnection': u'',
320 u'checkedDomains': u'youtube',
321 u'dnConn': u'',
322 u'dsh': dsh,
323 u'pstMsg': u'0',
324 u'rmShown': u'1',
325 u'secTok': u'',
326 u'signIn': u'Sign in',
327 u'timeStmp': u'',
328 u'service': u'youtube',
329 u'uilel': u'3',
330 u'hl': u'en_US',
331 }
332 # Convert to UTF-8 *before* urlencode because Python 2.x's urlencode
333 # chokes on unicode
334 login_form = dict((k.encode('utf-8'), v.encode('utf-8')) for k,v in login_form_strs.items())
335 login_data = compat_urllib_parse.urlencode(login_form).encode('ascii')
336 request = compat_urllib_request.Request(self._LOGIN_URL, login_data)
337 try:
338 self.report_login()
339 login_results = compat_urllib_request.urlopen(request).read().decode('utf-8')
340 if re.search(r'(?i)<form[^>]* id="gaia_loginform"', login_results) is not None:
341 self._downloader.report_warning(u'unable to log in: bad username or password')
342 return
343 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
344 self._downloader.report_warning(u'unable to log in: %s' % compat_str(err))
345 return
346
347 # Confirm age
348 age_form = {
349 'next_url': '/',
350 'action_confirm': 'Confirm',
351 }
352 request = compat_urllib_request.Request(self._AGE_URL, compat_urllib_parse.urlencode(age_form))
353 try:
354 self.report_age_confirmation()
355 compat_urllib_request.urlopen(request).read().decode('utf-8')
356 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
357 raise ExtractorError(u'Unable to confirm age: %s' % compat_str(err))
358
359 def _extract_id(self, url):
360 mobj = re.match(self._VALID_URL, url, re.VERBOSE)
361 if mobj is None:
362 raise ExtractorError(u'Invalid URL: %s' % url)
363 video_id = mobj.group(2)
364 return video_id
365
366 def _real_extract(self, url):
367 # Extract original video URL from URL with redirection, like age verification, using next_url parameter
368 mobj = re.search(self._NEXT_URL_RE, url)
369 if mobj:
370 url = 'https://www.youtube.com/' + compat_urllib_parse.unquote(mobj.group(1)).lstrip('/')
371 video_id = self._extract_id(url)
372
373 # Get video webpage
374 self.report_video_webpage_download(video_id)
375 url = 'https://www.youtube.com/watch?v=%s&gl=US&hl=en&has_verified=1' % video_id
376 request = compat_urllib_request.Request(url)
377 try:
378 video_webpage_bytes = compat_urllib_request.urlopen(request).read()
379 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
380 raise ExtractorError(u'Unable to download video webpage: %s' % compat_str(err))
381
382 video_webpage = video_webpage_bytes.decode('utf-8', 'ignore')
383
384 # Attempt to extract SWF player URL
385 mobj = re.search(r'swfConfig.*?"(http:\\/\\/.*?watch.*?-.*?\.swf)"', video_webpage)
386 if mobj is not None:
387 player_url = re.sub(r'\\(.)', r'\1', mobj.group(1))
388 else:
389 player_url = None
390
391 # Get video info
392 self.report_video_info_webpage_download(video_id)
393 for el_type in ['&el=embedded', '&el=detailpage', '&el=vevo', '']:
394 video_info_url = ('https://www.youtube.com/get_video_info?&video_id=%s%s&ps=default&eurl=&gl=US&hl=en'
395 % (video_id, el_type))
396 video_info_webpage = self._download_webpage(video_info_url, video_id,
397 note=False,
398 errnote='unable to download video info webpage')
399 video_info = compat_parse_qs(video_info_webpage)
400 if 'token' in video_info:
401 break
402 if 'token' not in video_info:
403 if 'reason' in video_info:
404 raise ExtractorError(u'YouTube said: %s' % video_info['reason'][0])
405 else:
406 raise ExtractorError(u'"token" parameter not in video info for unknown reason')
407
408 # Check for "rental" videos
409 if 'ypc_video_rental_bar_text' in video_info and 'author' not in video_info:
410 raise ExtractorError(u'"rental" videos not supported')
411
412 # Start extracting information
413 self.report_information_extraction(video_id)
414
415 # uploader
416 if 'author' not in video_info:
417 raise ExtractorError(u'Unable to extract uploader name')
418 video_uploader = compat_urllib_parse.unquote_plus(video_info['author'][0])
419
420 # uploader_id
421 video_uploader_id = None
422 mobj = re.search(r'<link itemprop="url" href="http://www.youtube.com/(?:user|channel)/([^"]+)">', video_webpage)
423 if mobj is not None:
424 video_uploader_id = mobj.group(1)
425 else:
426 self._downloader.report_warning(u'unable to extract uploader nickname')
427
428 # title
429 if 'title' not in video_info:
430 raise ExtractorError(u'Unable to extract video title')
431 video_title = compat_urllib_parse.unquote_plus(video_info['title'][0])
432
433 # thumbnail image
434 if 'thumbnail_url' not in video_info:
435 self._downloader.report_warning(u'unable to extract video thumbnail')
436 video_thumbnail = ''
437 else: # don't panic if we can't find it
438 video_thumbnail = compat_urllib_parse.unquote_plus(video_info['thumbnail_url'][0])
439
440 # upload date
441 upload_date = None
442 mobj = re.search(r'id="eow-date.*?>(.*?)</span>', video_webpage, re.DOTALL)
443 if mobj is not None:
444 upload_date = ' '.join(re.sub(r'[/,-]', r' ', mobj.group(1)).split())
445 upload_date = unified_strdate(upload_date)
446
447 # description
448 video_description = get_element_by_id("eow-description", video_webpage)
449 if video_description:
450 video_description = clean_html(video_description)
451 else:
452 fd_mobj = re.search(r'<meta name="description" content="([^"]+)"', video_webpage)
453 if fd_mobj:
454 video_description = unescapeHTML(fd_mobj.group(1))
455 else:
456 video_description = u''
457
458 # subtitles
459 video_subtitles = None
460
461 if self._downloader.params.get('writesubtitles', False):
462 video_subtitles = self._extract_subtitle(video_id)
463 if video_subtitles:
464 (sub_error, sub_lang, sub) = video_subtitles[0]
465 if sub_error:
466 self._downloader.report_warning(sub_error)
467
468 if self._downloader.params.get('writeautomaticsub', False):
469 video_subtitles = self._request_automatic_caption(video_id, video_webpage)
470 (sub_error, sub_lang, sub) = video_subtitles[0]
471 if sub_error:
472 self._downloader.report_warning(sub_error)
473
474 if self._downloader.params.get('allsubtitles', False):
475 video_subtitles = self._extract_all_subtitles(video_id)
476 for video_subtitle in video_subtitles:
477 (sub_error, sub_lang, sub) = video_subtitle
478 if sub_error:
479 self._downloader.report_warning(sub_error)
480
481 if self._downloader.params.get('listsubtitles', False):
482 self._list_available_subtitles(video_id)
483 return
484
485 if 'length_seconds' not in video_info:
486 self._downloader.report_warning(u'unable to extract video duration')
487 video_duration = ''
488 else:
489 video_duration = compat_urllib_parse.unquote_plus(video_info['length_seconds'][0])
490
491 # Decide which formats to download
492 req_format = self._downloader.params.get('format', None)
493
494 try:
495 mobj = re.search(r';ytplayer.config = ({.*?});', video_webpage)
496 if not mobj:
497 raise ValueError('Could not find vevo ID')
498 info = json.loads(mobj.group(1))
499 args = info['args']
500 # Easy way to know if the 's' value is in url_encoded_fmt_stream_map
501 # this signatures are encrypted
502 m_s = re.search(r'[&,]s=', args['url_encoded_fmt_stream_map'])
503 if m_s is not None:
504 self.to_screen(u'%s: Encrypted signatures detected.' % video_id)
505 video_info['url_encoded_fmt_stream_map'] = [args['url_encoded_fmt_stream_map']]
506 except ValueError:
507 pass
508
509 if 'conn' in video_info and video_info['conn'][0].startswith('rtmp'):
510 self.report_rtmp_download()
511 video_url_list = [(None, video_info['conn'][0])]
512 elif 'url_encoded_fmt_stream_map' in video_info and len(video_info['url_encoded_fmt_stream_map']) >= 1:
513 url_map = {}
514 for url_data_str in video_info['url_encoded_fmt_stream_map'][0].split(','):
515 url_data = compat_parse_qs(url_data_str)
516 if 'itag' in url_data and 'url' in url_data:
517 url = url_data['url'][0]
518 if 'sig' in url_data:
519 url += '&signature=' + url_data['sig'][0]
520 elif 's' in url_data:
521 if self._downloader.params.get('verbose'):
522 s = url_data['s'][0]
523 player = self._search_regex(r'html5player-(.+?)\.js', video_webpage,
524 'html5 player', fatal=False)
525 self.to_screen('encrypted signature length %d (%d.%d), itag %s, html5 player %s' %
526 (len(s), len(s.split('.')[0]), len(s.split('.')[1]), url_data['itag'][0], player))
527 signature = self._decrypt_signature(url_data['s'][0])
528 url += '&signature=' + signature
529 if 'ratebypass' not in url:
530 url += '&ratebypass=yes'
531 url_map[url_data['itag'][0]] = url
532
533 format_limit = self._downloader.params.get('format_limit', None)
534 available_formats = self._available_formats_prefer_free if self._downloader.params.get('prefer_free_formats', False) else self._available_formats
535 if format_limit is not None and format_limit in available_formats:
536 format_list = available_formats[available_formats.index(format_limit):]
537 else:
538 format_list = available_formats
539 existing_formats = [x for x in format_list if x in url_map]
540 if len(existing_formats) == 0:
541 raise ExtractorError(u'no known formats available for video')
542 if self._downloader.params.get('listformats', None):
543 self._print_formats(existing_formats)
544 return
545 if req_format is None or req_format == 'best':
546 video_url_list = [(existing_formats[0], url_map[existing_formats[0]])] # Best quality
547 elif req_format == 'worst':
548 video_url_list = [(existing_formats[len(existing_formats)-1], url_map[existing_formats[len(existing_formats)-1]])] # worst quality
549 elif req_format in ('-1', 'all'):
550 video_url_list = [(f, url_map[f]) for f in existing_formats] # All formats
551 else:
552 # Specific formats. We pick the first in a slash-delimeted sequence.
553 # For example, if '1/2/3/4' is requested and '2' and '4' are available, we pick '2'.
554 req_formats = req_format.split('/')
555 video_url_list = None
556 for rf in req_formats:
557 if rf in url_map:
558 video_url_list = [(rf, url_map[rf])]
559 break
560 if video_url_list is None:
561 raise ExtractorError(u'requested format not available')
562 else:
563 raise ExtractorError(u'no conn or url_encoded_fmt_stream_map information found in video info')
564
565 results = []
566 for format_param, video_real_url in video_url_list:
567 # Extension
568 video_extension = self._video_extensions.get(format_param, 'flv')
569
570 video_format = '{0} - {1}'.format(format_param if format_param else video_extension,
571 self._video_dimensions.get(format_param, '???'))
572
573 results.append({
574 'id': video_id,
575 'url': video_real_url,
576 'uploader': video_uploader,
577 'uploader_id': video_uploader_id,
578 'upload_date': upload_date,
579 'title': video_title,
580 'ext': video_extension,
581 'format': video_format,
582 'thumbnail': video_thumbnail,
583 'description': video_description,
584 'player_url': player_url,
585 'subtitles': video_subtitles,
586 'duration': video_duration
587 })
588 return results
589
590 class YoutubePlaylistIE(InfoExtractor):
591 """Information Extractor for YouTube playlists."""
592
593 _VALID_URL = r"""(?:
594 (?:https?://)?
595 (?:\w+\.)?
596 youtube\.com/
597 (?:
598 (?:course|view_play_list|my_playlists|artist|playlist|watch)
599 \? (?:.*?&)*? (?:p|a|list)=
600 | p/
601 )
602 ((?:PL|EC|UU)?[0-9A-Za-z-_]{10,})
603 .*
604 |
605 ((?:PL|EC|UU)[0-9A-Za-z-_]{10,})
606 )"""
607 _TEMPLATE_URL = 'https://gdata.youtube.com/feeds/api/playlists/%s?max-results=%i&start-index=%i&v=2&alt=json&safeSearch=none'
608 _MAX_RESULTS = 50
609 IE_NAME = u'youtube:playlist'
610
611 @classmethod
612 def suitable(cls, url):
613 """Receives a URL and returns True if suitable for this IE."""
614 return re.match(cls._VALID_URL, url, re.VERBOSE) is not None
615
616 def _real_extract(self, url):
617 # Extract playlist id
618 mobj = re.match(self._VALID_URL, url, re.VERBOSE)
619 if mobj is None:
620 raise ExtractorError(u'Invalid URL: %s' % url)
621
622 # Download playlist videos from API
623 playlist_id = mobj.group(1) or mobj.group(2)
624 page_num = 1
625 videos = []
626
627 while True:
628 url = self._TEMPLATE_URL % (playlist_id, self._MAX_RESULTS, self._MAX_RESULTS * (page_num - 1) + 1)
629 page = self._download_webpage(url, playlist_id, u'Downloading page #%s' % page_num)
630
631 try:
632 response = json.loads(page)
633 except ValueError as err:
634 raise ExtractorError(u'Invalid JSON in API response: ' + compat_str(err))
635
636 if 'feed' not in response:
637 raise ExtractorError(u'Got a malformed response from YouTube API')
638 playlist_title = response['feed']['title']['$t']
639 if 'entry' not in response['feed']:
640 # Number of videos is a multiple of self._MAX_RESULTS
641 break
642
643 for entry in response['feed']['entry']:
644 index = entry['yt$position']['$t']
645 if 'media$group' in entry and 'media$player' in entry['media$group']:
646 videos.append((index, entry['media$group']['media$player']['url']))
647
648 if len(response['feed']['entry']) < self._MAX_RESULTS:
649 break
650 page_num += 1
651
652 videos = [v[1] for v in sorted(videos)]
653
654 url_results = [self.url_result(url, 'Youtube') for url in videos]
655 return [self.playlist_result(url_results, playlist_id, playlist_title)]
656
657
658 class YoutubeChannelIE(InfoExtractor):
659 """Information Extractor for YouTube channels."""
660
661 _VALID_URL = r"^(?:https?://)?(?:youtu\.be|(?:\w+\.)?youtube(?:-nocookie)?\.com)/channel/([0-9A-Za-z_-]+)"
662 _TEMPLATE_URL = 'http://www.youtube.com/channel/%s/videos?sort=da&flow=list&view=0&page=%s&gl=US&hl=en'
663 _MORE_PAGES_INDICATOR = 'yt-uix-load-more'
664 _MORE_PAGES_URL = 'http://www.youtube.com/channel_ajax?action_load_more_videos=1&flow=list&paging=%s&view=0&sort=da&channel_id=%s'
665 IE_NAME = u'youtube:channel'
666
667 def extract_videos_from_page(self, page):
668 ids_in_page = []
669 for mobj in re.finditer(r'href="/watch\?v=([0-9A-Za-z_-]+)&?', page):
670 if mobj.group(1) not in ids_in_page:
671 ids_in_page.append(mobj.group(1))
672 return ids_in_page
673
674 def _real_extract(self, url):
675 # Extract channel id
676 mobj = re.match(self._VALID_URL, url)
677 if mobj is None:
678 raise ExtractorError(u'Invalid URL: %s' % url)
679
680 # Download channel page
681 channel_id = mobj.group(1)
682 video_ids = []
683 pagenum = 1
684
685 url = self._TEMPLATE_URL % (channel_id, pagenum)
686 page = self._download_webpage(url, channel_id,
687 u'Downloading page #%s' % pagenum)
688
689 # Extract video identifiers
690 ids_in_page = self.extract_videos_from_page(page)
691 video_ids.extend(ids_in_page)
692
693 # Download any subsequent channel pages using the json-based channel_ajax query
694 if self._MORE_PAGES_INDICATOR in page:
695 while True:
696 pagenum = pagenum + 1
697
698 url = self._MORE_PAGES_URL % (pagenum, channel_id)
699 page = self._download_webpage(url, channel_id,
700 u'Downloading page #%s' % pagenum)
701
702 page = json.loads(page)
703
704 ids_in_page = self.extract_videos_from_page(page['content_html'])
705 video_ids.extend(ids_in_page)
706
707 if self._MORE_PAGES_INDICATOR not in page['load_more_widget_html']:
708 break
709
710 self._downloader.to_screen(u'[youtube] Channel %s: Found %i videos' % (channel_id, len(video_ids)))
711
712 urls = ['http://www.youtube.com/watch?v=%s' % id for id in video_ids]
713 url_entries = [self.url_result(url, 'Youtube') for url in urls]
714 return [self.playlist_result(url_entries, channel_id)]
715
716
717 class YoutubeUserIE(InfoExtractor):
718 """Information Extractor for YouTube users."""
719
720 _VALID_URL = r'(?:(?:(?:https?://)?(?:\w+\.)?youtube\.com/user/)|ytuser:)([A-Za-z0-9_-]+)'
721 _TEMPLATE_URL = 'http://gdata.youtube.com/feeds/api/users/%s'
722 _GDATA_PAGE_SIZE = 50
723 _GDATA_URL = 'http://gdata.youtube.com/feeds/api/users/%s/uploads?max-results=%d&start-index=%d'
724 _VIDEO_INDICATOR = r'/watch\?v=(.+?)[\<&]'
725 IE_NAME = u'youtube:user'
726
727 def _real_extract(self, url):
728 # Extract username
729 mobj = re.match(self._VALID_URL, url)
730 if mobj is None:
731 raise ExtractorError(u'Invalid URL: %s' % url)
732
733 username = mobj.group(1)
734
735 # Download video ids using YouTube Data API. Result size per
736 # query is limited (currently to 50 videos) so we need to query
737 # page by page until there are no video ids - it means we got
738 # all of them.
739
740 video_ids = []
741 pagenum = 0
742
743 while True:
744 start_index = pagenum * self._GDATA_PAGE_SIZE + 1
745
746 gdata_url = self._GDATA_URL % (username, self._GDATA_PAGE_SIZE, start_index)
747 page = self._download_webpage(gdata_url, username,
748 u'Downloading video ids from %d to %d' % (start_index, start_index + self._GDATA_PAGE_SIZE))
749
750 # Extract video identifiers
751 ids_in_page = []
752
753 for mobj in re.finditer(self._VIDEO_INDICATOR, page):
754 if mobj.group(1) not in ids_in_page:
755 ids_in_page.append(mobj.group(1))
756
757 video_ids.extend(ids_in_page)
758
759 # A little optimization - if current page is not
760 # "full", ie. does not contain PAGE_SIZE video ids then
761 # we can assume that this page is the last one - there
762 # are no more ids on further pages - no need to query
763 # again.
764
765 if len(ids_in_page) < self._GDATA_PAGE_SIZE:
766 break
767
768 pagenum += 1
769
770 urls = ['http://www.youtube.com/watch?v=%s' % video_id for video_id in video_ids]
771 url_results = [self.url_result(url, 'Youtube') for url in urls]
772 return [self.playlist_result(url_results, playlist_title = username)]
773
774 class YoutubeSearchIE(SearchInfoExtractor):
775 """Information Extractor for YouTube search queries."""
776 _API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc'
777 _MAX_RESULTS = 1000
778 IE_NAME = u'youtube:search'
779 _SEARCH_KEY = 'ytsearch'
780
781 def report_download_page(self, query, pagenum):
782 """Report attempt to download search page with given number."""
783 self._downloader.to_screen(u'[youtube] query "%s": Downloading page %s' % (query, pagenum))
784
785 def _get_n_results(self, query, n):
786 """Get a specified number of results for a query"""
787
788 video_ids = []
789 pagenum = 0
790 limit = n
791
792 while (50 * pagenum) < limit:
793 self.report_download_page(query, pagenum+1)
794 result_url = self._API_URL % (compat_urllib_parse.quote_plus(query), (50*pagenum)+1)
795 request = compat_urllib_request.Request(result_url)
796 try:
797 data = compat_urllib_request.urlopen(request).read().decode('utf-8')
798 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
799 raise ExtractorError(u'Unable to download API page: %s' % compat_str(err))
800 api_response = json.loads(data)['data']
801
802 if not 'items' in api_response:
803 raise ExtractorError(u'[youtube] No video results')
804
805 new_ids = list(video['id'] for video in api_response['items'])
806 video_ids += new_ids
807
808 limit = min(n, api_response['totalItems'])
809 pagenum += 1
810
811 if len(video_ids) > n:
812 video_ids = video_ids[:n]
813 videos = [self.url_result('http://www.youtube.com/watch?v=%s' % id, 'Youtube') for id in video_ids]
814 return self.playlist_result(videos, query)