youtube_dl/extractor/vimeo.py

   1 # encoding: utf-8
   2 import json
   3 import re
   4 import itertools
   5
   6 from .common import InfoExtractor
   7 from ..utils import (
   8     compat_urllib_parse,
   9     compat_urllib_request,
  10
  11     clean_html,
  12     get_element_by_attribute,
  13     ExtractorError,
  14     RegexNotFoundError,
  15     std_headers,
  16     unsmuggle_url,
  17 )
  18
  19
  20 class VimeoIE(InfoExtractor):
  21     """Information extractor for vimeo.com."""
  22
  23     # _VALID_URL matches Vimeo URLs
  24     _VALID_URL = r'''(?x)
  25         (?P<proto>https?://)?
  26         (?:(?:www|(?P<player>player))\.)?
  27         vimeo(?P<pro>pro)?\.com/
  28         (?:.*?/)?
  29         (?:(?:play_redirect_hls|moogaloop\.swf)\?clip_id=)?
  30         (?:videos?/)?
  31         (?P<id>[0-9]+)
  32         /?(?:[?&].*)?(?:[#].*)?$'''
  33     _NETRC_MACHINE = 'vimeo'
  34     IE_NAME = u'vimeo'
  35     _TESTS = [
  36         {
  37             u'url': u'http://vimeo.com/56015672#at=0',
  38             u'file': u'56015672.mp4',
  39             u'md5': u'8879b6cc097e987f02484baf890129e5',
  40             u'info_dict': {
  41                 u"upload_date": u"20121220",
  42                 u"description": u"This is a test case for youtube-dl.\nFor more information, see github.com/rg3/youtube-dl\nTest chars: \u2605 \" ' \u5e78 / \\ \u00e4 \u21ad \U0001d550",
  43                 u"uploader_id": u"user7108434",
  44                 u"uploader": u"Filippo Valsorda",
  45                 u"title": u"youtube-dl test video - \u2605 \" ' \u5e78 / \\ \u00e4 \u21ad \U0001d550",
  46             },
  47         },
  48         {
  49             u'url': u'http://vimeopro.com/openstreetmapus/state-of-the-map-us-2013/video/68093876',
  50             u'file': u'68093876.mp4',
  51             u'md5': u'3b5ca6aa22b60dfeeadf50b72e44ed82',
  52             u'note': u'Vimeo Pro video (#1197)',
  53             u'info_dict': {
  54                 u'uploader_id': u'openstreetmapus',
  55                 u'uploader': u'OpenStreetMap US',
  56                 u'title': u'Andy Allan - Putting the Carto into OpenStreetMap Cartography',
  57             },
  58         },
  59         {
  60             u'url': u'http://player.vimeo.com/video/54469442',
  61             u'file': u'54469442.mp4',
  62             u'md5': u'619b811a4417aa4abe78dc653becf511',
  63             u'note': u'Videos that embed the url in the player page',
  64             u'info_dict': {
  65                 u'title': u'Kathy Sierra: Building the minimum Badass User, Business of Software',
  66                 u'uploader': u'The BLN & Business of Software',
  67             },
  68         },
  69         {
  70             u'url': u'http://vimeo.com/68375962',
  71             u'file': u'68375962.mp4',
  72             u'md5': u'aaf896bdb7ddd6476df50007a0ac0ae7',
  73             u'note': u'Video protected with password',
  74             u'info_dict': {
  75                 u'title': u'youtube-dl password protected test video',
  76                 u'upload_date': u'20130614',
  77                 u'uploader_id': u'user18948128',
  78                 u'uploader': u'Jaime Marquínez Ferrándiz',
  79             },
  80             u'params': {
  81                 u'videopassword': u'youtube-dl',
  82             },
  83         },
  84     ]
  85
  86     def _login(self):
  87         (username, password) = self._get_login_info()
  88         if username is None:
  89             return
  90         self.report_login()
  91         login_url = 'https://vimeo.com/log_in'
  92         webpage = self._download_webpage(login_url, None, False)
  93         token = re.search(r'xsrft: \'(.*?)\'', webpage).group(1)
  94         data = compat_urllib_parse.urlencode({'email': username,
  95                                               'password': password,
  96                                               'action': 'login',
  97                                               'service': 'vimeo',
  98                                               'token': token,
  99                                               })
 100         login_request = compat_urllib_request.Request(login_url, data)
 101         login_request.add_header('Content-Type', 'application/x-www-form-urlencoded')
 102         login_request.add_header('Cookie', 'xsrft=%s' % token)
 103         self._download_webpage(login_request, None, False, u'Wrong login info')
 104
 105     def _verify_video_password(self, url, video_id, webpage):
 106         password = self._downloader.params.get('videopassword', None)
 107         if password is None:
 108             raise ExtractorError(u'This video is protected by a password, use the --video-password option')
 109         token = re.search(r'xsrft: \'(.*?)\'', webpage).group(1)
 110         data = compat_urllib_parse.urlencode({'password': password,
 111                                               'token': token})
 112         # I didn't manage to use the password with https
 113         if url.startswith('https'):
 114             pass_url = url.replace('https','http')
 115         else:
 116             pass_url = url
 117         password_request = compat_urllib_request.Request(pass_url+'/password', data)
 118         password_request.add_header('Content-Type', 'application/x-www-form-urlencoded')
 119         password_request.add_header('Cookie', 'xsrft=%s' % token)
 120         self._download_webpage(password_request, video_id,
 121                                u'Verifying the password',
 122                                u'Wrong password')
 123
 124     def _real_initialize(self):
 125         self._login()
 126
 127     def _real_extract(self, url):
 128         url, data = unsmuggle_url(url)
 129         headers = std_headers
 130         if data is not None:
 131             headers = headers.copy()
 132             headers.update(data)
 133
 134         # Extract ID from URL
 135         mobj = re.match(self._VALID_URL, url)
 136         if mobj is None:
 137             raise ExtractorError(u'Invalid URL: %s' % url)
 138
 139         video_id = mobj.group('id')
 140         if mobj.group('pro') or mobj.group('player'):
 141             url = 'http://player.vimeo.com/video/' + video_id
 142         else:
 143             url = 'https://vimeo.com/' + video_id
 144
 145         # Retrieve video webpage to extract further information
 146         request = compat_urllib_request.Request(url, None, headers)
 147         webpage = self._download_webpage(request, video_id)
 148
 149         # Now we begin extracting as much information as we can from what we
 150         # retrieved. First we extract the information common to all extractors,
 151         # and latter we extract those that are Vimeo specific.
 152         self.report_extraction(video_id)
 153
 154         # Extract the config JSON
 155         try:
 156             try:
 157                 config_url = self._html_search_regex(
 158                     r' data-config-url="(.+?)"', webpage, u'config URL')
 159                 config_json = self._download_webpage(config_url, video_id)
 160                 config = json.loads(config_json)
 161             except RegexNotFoundError:
 162                 # For pro videos or player.vimeo.com urls
 163                 # We try to find out to which variable is assigned the config dic
 164                 m_variable_name = re.search('(\w)\.video\.id', webpage)
 165                 if m_variable_name is not None:
 166                     config_re = r'%s=({.+?});' % re.escape(m_variable_name.group(1))
 167                 else:
 168                     config_re = [r' = {config:({.+?}),assets:', r'(?:[abc])=({.+?});']
 169                 config = self._search_regex(config_re, webpage, u'info section',
 170                     flags=re.DOTALL)
 171                 config = json.loads(config)
 172         except Exception as e:
 173             if re.search('The creator of this video has not given you permission to embed it on this domain.', webpage):
 174                 raise ExtractorError(u'The author has restricted the access to this video, try with the "--referer" option')
 175
 176             if re.search('<form[^>]+?id="pw_form"', webpage) is not None:
 177                 self._verify_video_password(url, video_id, webpage)
 178                 return self._real_extract(url)
 179             else:
 180                 raise ExtractorError(u'Unable to extract info section',
 181                                      cause=e)
 182
 183         # Extract title
 184         video_title = config["video"]["title"]
 185
 186         # Extract uploader and uploader_id
 187         video_uploader = config["video"]["owner"]["name"]
 188         video_uploader_id = config["video"]["owner"]["url"].split('/')[-1] if config["video"]["owner"]["url"] else None
 189
 190         # Extract video thumbnail
 191         video_thumbnail = config["video"].get("thumbnail")
 192         if video_thumbnail is None:
 193             _, video_thumbnail = sorted((int(width), t_url) for (width, t_url) in config["video"]["thumbs"].items())[-1]
 194
 195         # Extract video description
 196         video_description = None
 197         try:
 198             video_description = get_element_by_attribute("itemprop", "description", webpage)
 199             if video_description: video_description = clean_html(video_description)
 200         except AssertionError as err:
 201             # On some pages like (http://player.vimeo.com/video/54469442) the
 202             # html tags are not closed, python 2.6 cannot handle it
 203             if err.args[0] == 'we should not get here!':
 204                 pass
 205             else:
 206                 raise
 207
 208         # Extract upload date
 209         video_upload_date = None
 210         mobj = re.search(r'<meta itemprop="dateCreated" content="(\d{4})-(\d{2})-(\d{2})T', webpage)
 211         if mobj is not None:
 212             video_upload_date = mobj.group(1) + mobj.group(2) + mobj.group(3)
 213
 214         try:
 215             view_count = int(self._search_regex(r'UserPlays:(\d+)', webpage, u'view count'))
 216             like_count = int(self._search_regex(r'UserLikes:(\d+)', webpage, u'like count'))
 217             comment_count = int(self._search_regex(r'UserComments:(\d+)', webpage, u'comment count'))
 218         except RegexNotFoundError:
 219             # This info is only available in vimeo.com/{id} urls
 220             view_count = None
 221             like_count = None
 222             comment_count = None
 223
 224         # Vimeo specific: extract request signature and timestamp
 225         sig = config['request']['signature']
 226         timestamp = config['request']['timestamp']
 227
 228         # Vimeo specific: extract video codec and quality information
 229         # First consider quality, then codecs, then take everything
 230         codecs = [('vp6', 'flv'), ('vp8', 'flv'), ('h264', 'mp4')]
 231         files = {'hd': [], 'sd': [], 'other': []}
 232         config_files = config["video"].get("files") or config["request"].get("files")
 233         for codec_name, codec_extension in codecs:
 234             for quality in config_files.get(codec_name, []):
 235                 format_id = '-'.join((codec_name, quality)).lower()
 236                 key = quality if quality in files else 'other'
 237                 video_url = None
 238                 if isinstance(config_files[codec_name], dict):
 239                     file_info = config_files[codec_name][quality]
 240                     video_url = file_info.get('url')
 241                 else:
 242                     file_info = {}
 243                 if video_url is None:
 244                     video_url = "http://player.vimeo.com/play_redirect?clip_id=%s&sig=%s&time=%s&quality=%s&codecs=%s&type=moogaloop_local&embed_location=" \
 245                         %(video_id, sig, timestamp, quality, codec_name.upper())
 246
 247                 files[key].append({
 248                     'ext': codec_extension,
 249                     'url': video_url,
 250                     'format_id': format_id,
 251                     'width': file_info.get('width'),
 252                     'height': file_info.get('height'),
 253                 })
 254         formats = []
 255         for key in ('other', 'sd', 'hd'):
 256             formats += files[key]
 257         if len(formats) == 0:
 258             raise ExtractorError(u'No known codec found')
 259
 260         return {
 261             'id':       video_id,
 262             'uploader': video_uploader,
 263             'uploader_id': video_uploader_id,
 264             'upload_date':  video_upload_date,
 265             'title':    video_title,
 266             'thumbnail':    video_thumbnail,
 267             'description':  video_description,
 268             'formats': formats,
 269             'webpage_url': url,
 270             'view_count': view_count,
 271             'like_count': like_count,
 272             'comment_count': comment_count,
 273         }
 274
 275
 276 class VimeoChannelIE(InfoExtractor):
 277     IE_NAME = u'vimeo:channel'
 278     _VALID_URL = r'(?:https?://)?vimeo.\com/channels/(?P<id>[^/]+)'
 279     _MORE_PAGES_INDICATOR = r'<a.+?rel="next"'
 280     _TITLE_RE = r'<link rel="alternate"[^>]+?title="(.*?)"'
 281
 282     def _page_url(self, base_url, pagenum):
 283         return '%s/videos/page:%d/' % (base_url, pagenum)
 284
 285     def _extract_list_title(self, webpage):
 286         return self._html_search_regex(self._TITLE_RE, webpage, u'list title')
 287
 288     def _extract_videos(self, list_id, base_url):
 289         video_ids = []
 290         for pagenum in itertools.count(1):
 291             webpage = self._download_webpage(
 292                 self._page_url(base_url, pagenum) ,list_id,
 293                 u'Downloading page %s' % pagenum)
 294             video_ids.extend(re.findall(r'id="clip_(\d+?)"', webpage))
 295             if re.search(self._MORE_PAGES_INDICATOR, webpage, re.DOTALL) is None:
 296                 break
 297
 298         entries = [self.url_result('http://vimeo.com/%s' % video_id, 'Vimeo')
 299                    for video_id in video_ids]
 300         return {'_type': 'playlist',
 301                 'id': list_id,
 302                 'title': self._extract_list_title(webpage),
 303                 'entries': entries,
 304                 }
 305
 306     def _real_extract(self, url):
 307         mobj = re.match(self._VALID_URL, url)
 308         channel_id =  mobj.group('id')
 309         return self._extract_videos(channel_id, 'http://vimeo.com/channels/%s' % channel_id)
 310
 311
 312 class VimeoUserIE(VimeoChannelIE):
 313     IE_NAME = u'vimeo:user'
 314     _VALID_URL = r'(?:https?://)?vimeo.\com/(?P<name>[^/]+)'
 315     _TITLE_RE = r'<a[^>]+?class="user">([^<>]+?)</a>'
 316
 317     @classmethod
 318     def suitable(cls, url):
 319         if VimeoChannelIE.suitable(url) or VimeoIE.suitable(url) or VimeoAlbumIE.suitable(url) or VimeoGroupsIE.suitable(url):
 320             return False
 321         return super(VimeoUserIE, cls).suitable(url)
 322
 323     def _real_extract(self, url):
 324         mobj = re.match(self._VALID_URL, url)
 325         name = mobj.group('name')
 326         return self._extract_videos(name, 'http://vimeo.com/%s' % name)
 327
 328
 329 class VimeoAlbumIE(VimeoChannelIE):
 330     IE_NAME = u'vimeo:album'
 331     _VALID_URL = r'(?:https?://)?vimeo.\com/album/(?P<id>\d+)'
 332     _TITLE_RE = r'<header id="page_header">\n\s*<h1>(.*?)</h1>'
 333
 334     def _page_url(self, base_url, pagenum):
 335         return '%s/page:%d/' % (base_url, pagenum)
 336
 337     def _real_extract(self, url):
 338         mobj = re.match(self._VALID_URL, url)
 339         album_id =  mobj.group('id')
 340         return self._extract_videos(album_id, 'http://vimeo.com/album/%s' % album_id)
 341
 342
 343 class VimeoGroupsIE(VimeoAlbumIE):
 344     IE_NAME = u'vimeo:group'
 345     _VALID_URL = r'(?:https?://)?vimeo.\com/groups/(?P<name>[^/]+)'
 346
 347     def _extract_list_title(self, webpage):
 348         return self._og_search_title(webpage)
 349
 350     def _real_extract(self, url):
 351         mobj = re.match(self._VALID_URL, url)
 352         name = mobj.group('name')
 353         return self._extract_videos(name, 'http://vimeo.com/groups/%s' % name)