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