_ Git - youtube-dl/blob - youtube_dl/extractor/vimeo.py

   1 # encoding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import json
   5 import re
   6 import itertools
   7
   8 from .common import InfoExtractor
   9 from .subtitles import SubtitlesInfoExtractor
  10 from ..utils import (
  11     compat_HTTPError,
  12     compat_urllib_parse,
  13     compat_urllib_request,
  14     clean_html,
  15     get_element_by_attribute,
  16     ExtractorError,
  17     RegexNotFoundError,
  18     std_headers,
  19     unsmuggle_url,
  20     urlencode_postdata,
  21     int_or_none,
  22 )
  23
  24
  25 class VimeoBaseInfoExtractor(InfoExtractor):
  26     _NETRC_MACHINE = 'vimeo'
  27     _LOGIN_REQUIRED = False
  28
  29     def _login(self):
  30         (username, password) = self._get_login_info()
  31         if username is None:
  32             if self._LOGIN_REQUIRED:
  33                 raise ExtractorError('No login info available, needed for using %s.' % self.IE_NAME, expected=True)
  34             return
  35         self.report_login()
  36         login_url = 'https://vimeo.com/log_in'
  37         webpage = self._download_webpage(login_url, None, False)
  38         token = self._search_regex(r'xsrft: \'(.*?)\'', webpage, 'login token')
  39         data = urlencode_postdata({
  40             'email': username,
  41             'password': password,
  42             'action': 'login',
  43             'service': 'vimeo',
  44             'token': token,
  45         })
  46         login_request = compat_urllib_request.Request(login_url, data)
  47         login_request.add_header('Content-Type', 'application/x-www-form-urlencoded')
  48         login_request.add_header('Cookie', 'xsrft=%s' % token)
  49         self._download_webpage(login_request, None, False, 'Wrong login info')
  50
  51
  52 class VimeoIE(VimeoBaseInfoExtractor, SubtitlesInfoExtractor):
  53     """Information extractor for vimeo.com."""
  54
  55     # _VALID_URL matches Vimeo URLs
  56     _VALID_URL = r'''(?x)
  57         (?P<proto>(?:https?:)?//)?
  58         (?:(?:www|(?P<player>player))\.)?
  59         vimeo(?P<pro>pro)?\.com/
  60         (?:.*?/)?
  61         (?:(?:play_redirect_hls|moogaloop\.swf)\?clip_id=)?
  62         (?:videos?/)?
  63         (?P<id>[0-9]+)
  64         /?(?:[?&].*)?(?:[#].*)?$'''
  65     IE_NAME = 'vimeo'
  66     _TESTS = [
  67         {
  68             'url': 'http://vimeo.com/56015672#at=0',
  69             'md5': '8879b6cc097e987f02484baf890129e5',
  70             'info_dict': {
  71                 'id': '56015672',
  72                 'ext': 'mp4',
  73                 "upload_date": "20121220",
  74                 "description": "This is a test case for youtube-dl.\nFor more information, see github.com/rg3/youtube-dl\nTest chars: \u2605 \" ' \u5e78 / \\ \u00e4 \u21ad \U0001d550",
  75                 "uploader_id": "user7108434",
  76                 "uploader": "Filippo Valsorda",
  77                 "title": "youtube-dl test video - \u2605 \" ' \u5e78 / \\ \u00e4 \u21ad \U0001d550",
  78                 "duration": 10,
  79             },
  80         },
  81         {
  82             'url': 'http://vimeopro.com/openstreetmapus/state-of-the-map-us-2013/video/68093876',
  83             'md5': '3b5ca6aa22b60dfeeadf50b72e44ed82',
  84             'note': 'Vimeo Pro video (#1197)',
  85             'info_dict': {
  86                 'id': '68093876',
  87                 'ext': 'mp4',
  88                 'uploader_id': 'openstreetmapus',
  89                 'uploader': 'OpenStreetMap US',
  90                 'title': 'Andy Allan - Putting the Carto into OpenStreetMap Cartography',
  91                 'duration': 1595,
  92             },
  93         },
  94         {
  95             'url': 'http://player.vimeo.com/video/54469442',
  96             'md5': '619b811a4417aa4abe78dc653becf511',
  97             'note': 'Videos that embed the url in the player page',
  98             'info_dict': {
  99                 'id': '54469442',
 100                 'ext': 'mp4',
 101                 'title': 'Kathy Sierra: Building the minimum Badass User, Business of Software 2012',
 102                 'uploader': 'The BLN & Business of Software',
 103                 'uploader_id': 'theblnbusinessofsoftware',
 104                 'duration': 3610,
 105             },
 106         },
 107         {
 108             'url': 'http://vimeo.com/68375962',
 109             'md5': 'aaf896bdb7ddd6476df50007a0ac0ae7',
 110             'note': 'Video protected with password',
 111             'info_dict': {
 112                 'id': '68375962',
 113                 'ext': 'mp4',
 114                 'title': 'youtube-dl password protected test video',
 115                 'upload_date': '20130614',
 116                 'uploader_id': 'user18948128',
 117                 'uploader': 'Jaime Marquínez Ferrándiz',
 118                 'duration': 10,
 119             },
 120             'params': {
 121                 'videopassword': 'youtube-dl',
 122             },
 123         },
 124         {
 125             'url': 'http://vimeo.com/channels/keypeele/75629013',
 126             'md5': '2f86a05afe9d7abc0b9126d229bbe15d',
 127             'note': 'Video is freely available via original URL '
 128                     'and protected with password when accessed via http://vimeo.com/75629013',
 129             'info_dict': {
 130                 'id': '75629013',
 131                 'ext': 'mp4',
 132                 'title': 'Key & Peele: Terrorist Interrogation',
 133                 'description': 'md5:8678b246399b070816b12313e8b4eb5c',
 134                 'uploader_id': 'atencio',
 135                 'uploader': 'Peter Atencio',
 136                 'duration': 187,
 137             },
 138         },
 139         {
 140             'url': 'http://vimeo.com/76979871',
 141             'md5': '3363dd6ffebe3784d56f4132317fd446',
 142             'note': 'Video with subtitles',
 143             'info_dict': {
 144                 'id': '76979871',
 145                 'ext': 'mp4',
 146                 'title': 'The New Vimeo Player (You Know, For Videos)',
 147                 'description': 'md5:2ec900bf97c3f389378a96aee11260ea',
 148                 'upload_date': '20131015',
 149                 'uploader_id': 'staff',
 150                 'uploader': 'Vimeo Staff',
 151                 'duration': 62,
 152             }
 153         },
 154     ]
 155
 156     @classmethod
 157     def suitable(cls, url):
 158         if VimeoChannelIE.suitable(url):
 159             # Otherwise channel urls like http://vimeo.com/channels/31259 would
 160             # match
 161             return False
 162         else:
 163             return super(VimeoIE, cls).suitable(url)
 164
 165     def _verify_video_password(self, url, video_id, webpage):
 166         password = self._downloader.params.get('videopassword', None)
 167         if password is None:
 168             raise ExtractorError('This video is protected by a password, use the --video-password option')
 169         token = self._search_regex(r'xsrft: \'(.*?)\'', webpage, 'login token')
 170         data = compat_urllib_parse.urlencode({
 171             'password': password,
 172             'token': token,
 173         })
 174         # I didn't manage to use the password with https
 175         if url.startswith('https'):
 176             pass_url = url.replace('https', 'http')
 177         else:
 178             pass_url = url
 179         password_request = compat_urllib_request.Request(pass_url + '/password', data)
 180         password_request.add_header('Content-Type', 'application/x-www-form-urlencoded')
 181         password_request.add_header('Cookie', 'xsrft=%s' % token)
 182         self._download_webpage(password_request, video_id,
 183                                'Verifying the password',
 184                                'Wrong password')
 185
 186     def _verify_player_video_password(self, url, video_id):
 187         password = self._downloader.params.get('videopassword', None)
 188         if password is None:
 189             raise ExtractorError('This video is protected by a password, use the --video-password option')
 190         data = compat_urllib_parse.urlencode({'password': password})
 191         pass_url = url + '/check-password'
 192         password_request = compat_urllib_request.Request(pass_url, data)
 193         password_request.add_header('Content-Type', 'application/x-www-form-urlencoded')
 194         return self._download_json(
 195             password_request, video_id,
 196             'Verifying the password',
 197             'Wrong password')
 198
 199     def _real_initialize(self):
 200         self._login()
 201
 202     def _real_extract(self, url):
 203         url, data = unsmuggle_url(url)
 204         headers = std_headers
 205         if data is not None:
 206             headers = headers.copy()
 207             headers.update(data)
 208         if 'Referer' not in headers:
 209             headers['Referer'] = url
 210
 211         # Extract ID from URL
 212         mobj = re.match(self._VALID_URL, url)
 213         video_id = mobj.group('id')
 214         if mobj.group('pro') or mobj.group('player'):
 215             url = 'http://player.vimeo.com/video/' + video_id
 216
 217         # Retrieve video webpage to extract further information
 218         request = compat_urllib_request.Request(url, None, headers)
 219         try:
 220             webpage = self._download_webpage(request, video_id)
 221         except ExtractorError as ee:
 222             if isinstance(ee.cause, compat_HTTPError) and ee.cause.code == 403:
 223                 errmsg = ee.cause.read()
 224                 if b'Because of its privacy settings, this video cannot be played here' in errmsg:
 225                     raise ExtractorError(
 226                         'Cannot download embed-only video without embedding '
 227                         'URL. Please call youtube-dl with the URL of the page '
 228                         'that embeds this video.',
 229                         expected=True)
 230             raise
 231
 232         # Now we begin extracting as much information as we can from what we
 233         # retrieved. First we extract the information common to all extractors,
 234         # and latter we extract those that are Vimeo specific.
 235         self.report_extraction(video_id)
 236
 237         # Extract the config JSON
 238         try:
 239             try:
 240                 config_url = self._html_search_regex(
 241                     r' data-config-url="(.+?)"', webpage, 'config URL')
 242                 config_json = self._download_webpage(config_url, video_id)
 243                 config = json.loads(config_json)
 244             except RegexNotFoundError:
 245                 # For pro videos or player.vimeo.com urls
 246                 # We try to find out to which variable is assigned the config dic
 247                 m_variable_name = re.search('(\w)\.video\.id', webpage)
 248                 if m_variable_name is not None:
 249                     config_re = r'%s=({.+?});' % re.escape(m_variable_name.group(1))
 250                 else:
 251                     config_re = [r' = {config:({.+?}),assets:', r'(?:[abc])=({.+?});']
 252                 config = self._search_regex(config_re, webpage, 'info section',
 253                     flags=re.DOTALL)
 254                 config = json.loads(config)
 255         except Exception as e:
 256             if re.search('The creator of this video has not given you permission to embed it on this domain.', webpage):
 257                 raise ExtractorError('The author has restricted the access to this video, try with the "--referer" option')
 258
 259             if re.search('<form[^>]+?id="pw_form"', webpage) is not None:
 260                 self._verify_video_password(url, video_id, webpage)
 261                 return self._real_extract(url)
 262             else:
 263                 raise ExtractorError('Unable to extract info section',
 264                                      cause=e)
 265         else:
 266             if config.get('view') == 4:
 267                 config = self._verify_player_video_password(url, video_id)
 268
 269         # Extract title
 270         video_title = config["video"]["title"]
 271
 272         # Extract uploader and uploader_id
 273         video_uploader = config["video"]["owner"]["name"]
 274         video_uploader_id = config["video"]["owner"]["url"].split('/')[-1] if config["video"]["owner"]["url"] else None
 275
 276         # Extract video thumbnail
 277         video_thumbnail = config["video"].get("thumbnail")
 278         if video_thumbnail is None:
 279             video_thumbs = config["video"].get("thumbs")
 280             if video_thumbs and isinstance(video_thumbs, dict):
 281                 _, video_thumbnail = sorted((int(width if width.isdigit() else 0), t_url) for (width, t_url) in video_thumbs.items())[-1]
 282
 283         # Extract video description
 284         video_description = None
 285         try:
 286             video_description = get_element_by_attribute("class", "description_wrapper", webpage)
 287             if video_description:
 288                 video_description = clean_html(video_description)
 289         except AssertionError as err:
 290             # On some pages like (http://player.vimeo.com/video/54469442) the
 291             # html tags are not closed, python 2.6 cannot handle it
 292             if err.args[0] == 'we should not get here!':
 293                 pass
 294             else:
 295                 raise
 296
 297         # Extract video duration
 298         video_duration = int_or_none(config["video"].get("duration"))
 299
 300         # Extract upload date
 301         video_upload_date = None
 302         mobj = re.search(r'<meta itemprop="dateCreated" content="(\d{4})-(\d{2})-(\d{2})T', webpage)
 303         if mobj is not None:
 304             video_upload_date = mobj.group(1) + mobj.group(2) + mobj.group(3)
 305
 306         try:
 307             view_count = int(self._search_regex(r'UserPlays:(\d+)', webpage, 'view count'))
 308             like_count = int(self._search_regex(r'UserLikes:(\d+)', webpage, 'like count'))
 309             comment_count = int(self._search_regex(r'UserComments:(\d+)', webpage, 'comment count'))
 310         except RegexNotFoundError:
 311             # This info is only available in vimeo.com/{id} urls
 312             view_count = None
 313             like_count = None
 314             comment_count = None
 315
 316         # Vimeo specific: extract request signature and timestamp
 317         sig = config['request']['signature']
 318         timestamp = config['request']['timestamp']
 319
 320         # Vimeo specific: extract video codec and quality information
 321         # First consider quality, then codecs, then take everything
 322         codecs = [('vp6', 'flv'), ('vp8', 'flv'), ('h264', 'mp4')]
 323         files = {'hd': [], 'sd': [], 'other': []}
 324         config_files = config["video"].get("files") or config["request"].get("files")
 325         for codec_name, codec_extension in codecs:
 326             for quality in config_files.get(codec_name, []):
 327                 format_id = '-'.join((codec_name, quality)).lower()
 328                 key = quality if quality in files else 'other'
 329                 video_url = None
 330                 if isinstance(config_files[codec_name], dict):
 331                     file_info = config_files[codec_name][quality]
 332                     video_url = file_info.get('url')
 333                 else:
 334                     file_info = {}
 335                 if video_url is None:
 336                     video_url = "http://player.vimeo.com/play_redirect?clip_id=%s&sig=%s&time=%s&quality=%s&codecs=%s&type=moogaloop_local&embed_location=" \
 337                         % (video_id, sig, timestamp, quality, codec_name.upper())
 338
 339                 files[key].append({
 340                     'ext': codec_extension,
 341                     'url': video_url,
 342                     'format_id': format_id,
 343                     'width': file_info.get('width'),
 344                     'height': file_info.get('height'),
 345                 })
 346         formats = []
 347         for key in ('other', 'sd', 'hd'):
 348             formats += files[key]
 349         if len(formats) == 0:
 350             raise ExtractorError('No known codec found')
 351
 352         subtitles = {}
 353         text_tracks = config['request'].get('text_tracks')
 354         if text_tracks:
 355             for tt in text_tracks:
 356                 subtitles[tt['lang']] = 'http://vimeo.com' + tt['url']
 357
 358         video_subtitles = self.extract_subtitles(video_id, subtitles)
 359         if self._downloader.params.get('listsubtitles', False):
 360             self._list_available_subtitles(video_id, subtitles)
 361             return
 362
 363         return {
 364             'id': video_id,
 365             'uploader': video_uploader,
 366             'uploader_id': video_uploader_id,
 367             'upload_date': video_upload_date,
 368             'title': video_title,
 369             'thumbnail': video_thumbnail,
 370             'description': video_description,
 371             'duration': video_duration,
 372             'formats': formats,
 373             'webpage_url': url,
 374             'view_count': view_count,
 375             'like_count': like_count,
 376             'comment_count': comment_count,
 377             'subtitles': video_subtitles,
 378         }
 379
 380
 381 class VimeoChannelIE(InfoExtractor):
 382     IE_NAME = 'vimeo:channel'
 383     _VALID_URL = r'(?:https?://)?vimeo\.com/channels/(?P<id>[^/]+)/?(\?.*)?$'
 384     _MORE_PAGES_INDICATOR = r'<a.+?rel="next"'
 385     _TITLE_RE = r'<link rel="alternate"[^>]+?title="(.*?)"'
 386
 387     def _page_url(self, base_url, pagenum):
 388         return '%s/videos/page:%d/' % (base_url, pagenum)
 389
 390     def _extract_list_title(self, webpage):
 391         return self._html_search_regex(self._TITLE_RE, webpage, 'list title')
 392
 393     def _extract_videos(self, list_id, base_url):
 394         video_ids = []
 395         for pagenum in itertools.count(1):
 396             webpage = self._download_webpage(
 397                 self._page_url(base_url, pagenum), list_id,
 398                 'Downloading page %s' % pagenum)
 399             video_ids.extend(re.findall(r'id="clip_(\d+?)"', webpage))
 400             if re.search(self._MORE_PAGES_INDICATOR, webpage, re.DOTALL) is None:
 401                 break
 402
 403         entries = [self.url_result('http://vimeo.com/%s' % video_id, 'Vimeo')
 404                    for video_id in video_ids]
 405         return {'_type': 'playlist',
 406                 'id': list_id,
 407                 'title': self._extract_list_title(webpage),
 408                 'entries': entries,
 409                 }
 410
 411     def _real_extract(self, url):
 412         mobj = re.match(self._VALID_URL, url)
 413         channel_id = mobj.group('id')
 414         return self._extract_videos(channel_id, 'http://vimeo.com/channels/%s' % channel_id)
 415
 416
 417 class VimeoUserIE(VimeoChannelIE):
 418     IE_NAME = 'vimeo:user'
 419     _VALID_URL = r'(?:https?://)?vimeo\.com/(?P<name>[^/]+)(?:/videos|[#?]|$)'
 420     _TITLE_RE = r'<a[^>]+?class="user">([^<>]+?)</a>'
 421
 422     @classmethod
 423     def suitable(cls, url):
 424         if VimeoChannelIE.suitable(url) or VimeoIE.suitable(url) or VimeoAlbumIE.suitable(url) or VimeoGroupsIE.suitable(url):
 425             return False
 426         return super(VimeoUserIE, cls).suitable(url)
 427
 428     def _real_extract(self, url):
 429         mobj = re.match(self._VALID_URL, url)
 430         name = mobj.group('name')
 431         return self._extract_videos(name, 'http://vimeo.com/%s' % name)
 432
 433
 434 class VimeoAlbumIE(VimeoChannelIE):
 435     IE_NAME = 'vimeo:album'
 436     _VALID_URL = r'(?:https?://)?vimeo\.com/album/(?P<id>\d+)'
 437     _TITLE_RE = r'<header id="page_header">\n\s*<h1>(.*?)</h1>'
 438
 439     def _page_url(self, base_url, pagenum):
 440         return '%s/page:%d/' % (base_url, pagenum)
 441
 442     def _real_extract(self, url):
 443         mobj = re.match(self._VALID_URL, url)
 444         album_id = mobj.group('id')
 445         return self._extract_videos(album_id, 'http://vimeo.com/album/%s' % album_id)
 446
 447
 448 class VimeoGroupsIE(VimeoAlbumIE):
 449     IE_NAME = 'vimeo:group'
 450     _VALID_URL = r'(?:https?://)?vimeo\.com/groups/(?P<name>[^/]+)'
 451
 452     def _extract_list_title(self, webpage):
 453         return self._og_search_title(webpage)
 454
 455     def _real_extract(self, url):
 456         mobj = re.match(self._VALID_URL, url)
 457         name = mobj.group('name')
 458         return self._extract_videos(name, 'http://vimeo.com/groups/%s' % name)
 459
 460
 461 class VimeoReviewIE(InfoExtractor):
 462     IE_NAME = 'vimeo:review'
 463     IE_DESC = 'Review pages on vimeo'
 464     _VALID_URL = r'https?://vimeo\.com/[^/]+/review/(?P<id>[^/]+)'
 465     _TESTS = [{
 466         'url': 'https://vimeo.com/user21297594/review/75524534/3c257a1b5d',
 467         'file': '75524534.mp4',
 468         'md5': 'c507a72f780cacc12b2248bb4006d253',
 469         'info_dict': {
 470             'title': "DICK HARDWICK 'Comedian'",
 471             'uploader': 'Richard Hardwick',
 472         }
 473     }, {
 474         'note': 'video player needs Referer',
 475         'url': 'http://vimeo.com/user22258446/review/91613211/13f927e053',
 476         'md5': '6295fdab8f4bf6a002d058b2c6dce276',
 477         'info_dict': {
 478             'id': '91613211',
 479             'ext': 'mp4',
 480             'title': 'Death by dogma versus assembling agile - Sander Hoogendoorn',
 481             'uploader': 'DevWeek Events',
 482             'duration': 2773,
 483             'thumbnail': 're:^https?://.*\.jpg$',
 484         }
 485     }]
 486
 487     def _real_extract(self, url):
 488         mobj = re.match(self._VALID_URL, url)
 489         video_id = mobj.group('id')
 490         player_url = 'https://player.vimeo.com/player/' + video_id
 491         return self.url_result(player_url, 'Vimeo', video_id)
 492
 493
 494 class VimeoWatchLaterIE(VimeoBaseInfoExtractor, VimeoChannelIE):
 495     IE_NAME = 'vimeo:watchlater'
 496     IE_DESC = 'Vimeo watch later list, "vimeowatchlater" keyword (requires authentication)'
 497     _VALID_URL = r'https?://vimeo\.com/home/watchlater|:vimeowatchlater'
 498     _LOGIN_REQUIRED = True
 499     _TITLE_RE = r'href="/home/watchlater".*?>(.*?)<'
 500
 501     def _real_initialize(self):
 502         self._login()
 503
 504     def _page_url(self, base_url, pagenum):
 505         url = '%s/page:%d/' % (base_url, pagenum)
 506         request = compat_urllib_request.Request(url)
 507         # Set the header to get a partial html page with the ids,
 508         # the normal page doesn't contain them.
 509         request.add_header('X-Requested-With', 'XMLHttpRequest')
 510         return request
 511
 512     def _real_extract(self, url):
 513         return self._extract_videos('watchlater', 'https://vimeo.com/home/watchlater')