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

   1 import json
   2 import re
   3 import itertools
   4
   5 from .common import InfoExtractor
   6 from ..utils import (
   7     compat_urllib_parse,
   8     compat_urllib_request,
   9
  10     clean_html,
  11     get_element_by_attribute,
  12     ExtractorError,
  13     RegexNotFoundError,
  14     std_headers,
  15     unsmuggle_url,
  16 )
  17
  18 class VimeoIE(InfoExtractor):
  19     """Information extractor for vimeo.com."""
  20
  21     # _VALID_URL matches Vimeo URLs
  22     _VALID_URL = r'(?P<proto>https?://)?(?:(?:www|player)\.)?vimeo(?P<pro>pro)?\.com/(?:(?:(?:groups|album)/[^/]+)|(?:.*?)/)?(?P<direct_link>play_redirect_hls\?clip_id=)?(?:videos?/)?(?P<id>[0-9]+)/?(?:[?].*)?$'
  23     _NETRC_MACHINE = 'vimeo'
  24     IE_NAME = u'vimeo'
  25     _TESTS = [
  26         {
  27             u'url': u'http://vimeo.com/56015672',
  28             u'file': u'56015672.mp4',
  29             u'md5': u'ae7a1d8b183758a0506b0622f37dfa14',
  30             u'info_dict': {
  31                 u"upload_date": u"20121220",
  32                 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",
  33                 u"uploader_id": u"user7108434",
  34                 u"uploader": u"Filippo Valsorda",
  35                 u"title": u"youtube-dl test video - \u2605 \" ' \u5e78 / \\ \u00e4 \u21ad \U0001d550",
  36             },
  37         },
  38         {
  39             u'url': u'http://vimeopro.com/openstreetmapus/state-of-the-map-us-2013/video/68093876',
  40             u'file': u'68093876.mp4',
  41             u'md5': u'3b5ca6aa22b60dfeeadf50b72e44ed82',
  42             u'note': u'Vimeo Pro video (#1197)',
  43             u'info_dict': {
  44                 u'uploader_id': u'openstreetmapus',
  45                 u'uploader': u'OpenStreetMap US',
  46                 u'title': u'Andy Allan - Putting the Carto into OpenStreetMap Cartography',
  47             },
  48         },
  49         {
  50             u'url': u'http://player.vimeo.com/video/54469442',
  51             u'file': u'54469442.mp4',
  52             u'md5': u'619b811a4417aa4abe78dc653becf511',
  53             u'note': u'Videos that embed the url in the player page',
  54             u'info_dict': {
  55                 u'title': u'Kathy Sierra: Building the minimum Badass User, Business of Software',
  56                 u'uploader': u'The BLN & Business of Software',
  57             },
  58         }
  59     ]
  60
  61     def _login(self):
  62         (username, password) = self._get_login_info()
  63         if username is None:
  64             return
  65         self.report_login()
  66         login_url = 'https://vimeo.com/log_in'
  67         webpage = self._download_webpage(login_url, None, False)
  68         token = re.search(r'xsrft: \'(.*?)\'', webpage).group(1)
  69         data = compat_urllib_parse.urlencode({'email': username,
  70                                               'password': password,
  71                                               'action': 'login',
  72                                               'service': 'vimeo',
  73                                               'token': token,
  74                                               })
  75         login_request = compat_urllib_request.Request(login_url, data)
  76         login_request.add_header('Content-Type', 'application/x-www-form-urlencoded')
  77         login_request.add_header('Cookie', 'xsrft=%s' % token)
  78         self._download_webpage(login_request, None, False, u'Wrong login info')
  79
  80     def _verify_video_password(self, url, video_id, webpage):
  81         password = self._downloader.params.get('videopassword', None)
  82         if password is None:
  83             raise ExtractorError(u'This video is protected by a password, use the --video-password option')
  84         token = re.search(r'xsrft: \'(.*?)\'', webpage).group(1)
  85         data = compat_urllib_parse.urlencode({'password': password,
  86                                               'token': token})
  87         # I didn't manage to use the password with https
  88         if url.startswith('https'):
  89             pass_url = url.replace('https','http')
  90         else:
  91             pass_url = url
  92         password_request = compat_urllib_request.Request(pass_url+'/password', data)
  93         password_request.add_header('Content-Type', 'application/x-www-form-urlencoded')
  94         password_request.add_header('Cookie', 'xsrft=%s' % token)
  95         self._download_webpage(password_request, video_id,
  96                                u'Verifying the password',
  97                                u'Wrong password')
  98
  99     def _real_initialize(self):
 100         self._login()
 101
 102     def _real_extract(self, url, new_video=True):
 103         url, data = unsmuggle_url(url)
 104         headers = std_headers
 105         if data is not None:
 106             headers = headers.copy()
 107             headers.update(data)
 108
 109         # Extract ID from URL
 110         mobj = re.match(self._VALID_URL, url)
 111         if mobj is None:
 112             raise ExtractorError(u'Invalid URL: %s' % url)
 113
 114         video_id = mobj.group('id')
 115         if not mobj.group('proto'):
 116             url = 'https://' + url
 117         elif mobj.group('pro'):
 118             url = 'http://player.vimeo.com/video/' + video_id
 119         elif mobj.group('direct_link'):
 120             url = 'https://vimeo.com/' + video_id
 121
 122         # Retrieve video webpage to extract further information
 123         request = compat_urllib_request.Request(url, None, headers)
 124         webpage = self._download_webpage(request, video_id)
 125
 126         # Now we begin extracting as much information as we can from what we
 127         # retrieved. First we extract the information common to all extractors,
 128         # and latter we extract those that are Vimeo specific.
 129         self.report_extraction(video_id)
 130
 131         # Extract the config JSON
 132         try:
 133             config_url = self._html_search_regex(
 134                 r' data-config-url="(.+?)"', webpage, u'config URL')
 135             config_json = self._download_webpage(config_url, video_id)
 136             config = json.loads(config_json)
 137         except RegexNotFoundError:
 138             # For pro videos or player.vimeo.com urls
 139             config = self._search_regex([r' = {config:({.+?}),assets:', r'c=({.+?);'],
 140                 webpage, u'info section', flags=re.DOTALL)
 141             config = json.loads(config)
 142         except Exception as e:
 143             if re.search('The creator of this video has not given you permission to embed it on this domain.', webpage):
 144                 raise ExtractorError(u'The author has restricted the access to this video, try with the "--referer" option')
 145
 146             if re.search('If so please provide the correct password.', webpage):
 147                 self._verify_video_password(url, video_id, webpage)
 148                 return self._real_extract(url)
 149             else:
 150                 raise ExtractorError(u'Unable to extract info section',
 151                                      cause=e)
 152
 153         # Extract title
 154         video_title = config["video"]["title"]
 155
 156         # Extract uploader and uploader_id
 157         video_uploader = config["video"]["owner"]["name"]
 158         video_uploader_id = config["video"]["owner"]["url"].split('/')[-1] if config["video"]["owner"]["url"] else None
 159
 160         # Extract video thumbnail
 161         video_thumbnail = config["video"].get("thumbnail")
 162         if video_thumbnail is None:
 163             _, video_thumbnail = sorted((int(width), t_url) for (width, t_url) in config["video"]["thumbs"].items())[-1]
 164
 165         # Extract video description
 166         video_description = None
 167         try:
 168             video_description = get_element_by_attribute("itemprop", "description", webpage)
 169             if video_description: video_description = clean_html(video_description)
 170         except AssertionError as err:
 171             # On some pages like (http://player.vimeo.com/video/54469442) the
 172             # html tags are not closed, python 2.6 cannot handle it
 173             if err.args[0] == 'we should not get here!':
 174                 pass
 175             else:
 176                 raise
 177
 178         # Extract upload date
 179         video_upload_date = None
 180         mobj = re.search(r'<meta itemprop="dateCreated" content="(\d{4})-(\d{2})-(\d{2})T', webpage)
 181         if mobj is not None:
 182             video_upload_date = mobj.group(1) + mobj.group(2) + mobj.group(3)
 183
 184         # Vimeo specific: extract request signature and timestamp
 185         sig = config['request']['signature']
 186         timestamp = config['request']['timestamp']
 187
 188         # Vimeo specific: extract video codec and quality information
 189         # First consider quality, then codecs, then take everything
 190         codecs = [('vp6', 'flv'), ('vp8', 'flv'), ('h264', 'mp4')]
 191         files = { 'hd': [], 'sd': [], 'other': []}
 192         config_files = config["video"].get("files") or config["request"].get("files")
 193         for codec_name, codec_extension in codecs:
 194             for quality in config_files.get(codec_name, []):
 195                 format_id = '-'.join((codec_name, quality)).lower()
 196                 key = quality if quality in files else 'other'
 197                 video_url = None
 198                 if isinstance(config_files[codec_name], dict):
 199                     file_info = config_files[codec_name][quality]
 200                     video_url = file_info.get('url')
 201                 else:
 202                     file_info = {}
 203                 if video_url is None:
 204                     video_url = "http://player.vimeo.com/play_redirect?clip_id=%s&sig=%s&time=%s&quality=%s&codecs=%s&type=moogaloop_local&embed_location=" \
 205                         %(video_id, sig, timestamp, quality, codec_name.upper())
 206
 207                 files[key].append({
 208                     'ext': codec_extension,
 209                     'url': video_url,
 210                     'format_id': format_id,
 211                     'width': file_info.get('width'),
 212                     'height': file_info.get('height'),
 213                 })
 214         formats = []
 215         for key in ('other', 'sd', 'hd'):
 216             formats += files[key]
 217         if len(formats) == 0:
 218             raise ExtractorError(u'No known codec found')
 219
 220         return [{
 221             'id':       video_id,
 222             'uploader': video_uploader,
 223             'uploader_id': video_uploader_id,
 224             'upload_date':  video_upload_date,
 225             'title':    video_title,
 226             'thumbnail':    video_thumbnail,
 227             'description':  video_description,
 228             'formats': formats,
 229         }]
 230
 231
 232 class VimeoChannelIE(InfoExtractor):
 233     IE_NAME = u'vimeo:channel'
 234     _VALID_URL = r'(?:https?://)?vimeo.\com/channels/(?P<id>[^/]+)'
 235     _MORE_PAGES_INDICATOR = r'<a.+?rel="next"'
 236
 237     def _real_extract(self, url):
 238         mobj = re.match(self._VALID_URL, url)
 239         channel_id =  mobj.group('id')
 240         video_ids = []
 241
 242         for pagenum in itertools.count(1):
 243             webpage = self._download_webpage('http://vimeo.com/channels/%s/videos/page:%d' % (channel_id, pagenum),
 244                                              channel_id, u'Downloading page %s' % pagenum)
 245             video_ids.extend(re.findall(r'id="clip_(\d+?)"', webpage))
 246             if re.search(self._MORE_PAGES_INDICATOR, webpage, re.DOTALL) is None:
 247                 break
 248
 249         entries = [self.url_result('http://vimeo.com/%s' % video_id, 'Vimeo')
 250                    for video_id in video_ids]
 251         channel_title = self._html_search_regex(r'<a href="/channels/%s">(.*?)</a>' % channel_id,
 252                                                 webpage, u'channel title')
 253         return {'_type': 'playlist',
 254                 'id': channel_id,
 255                 'title': channel_title,
 256                 'entries': entries,
 257                 }