_ 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     std_headers,
  14 )
  15
  16 class VimeoIE(InfoExtractor):
  17     """Information extractor for vimeo.com."""
  18
  19     # _VALID_URL matches Vimeo URLs
  20     _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]+)(?:[?].*)?$'
  21     _NETRC_MACHINE = 'vimeo'
  22     IE_NAME = u'vimeo'
  23     _TESTS = [
  24         {
  25             u'url': u'http://vimeo.com/56015672',
  26             u'file': u'56015672.mp4',
  27             u'md5': u'8879b6cc097e987f02484baf890129e5',
  28             u'info_dict': {
  29                 u"upload_date": u"20121220",
  30                 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",
  31                 u"uploader_id": u"user7108434",
  32                 u"uploader": u"Filippo Valsorda",
  33                 u"title": u"youtube-dl test video - \u2605 \" ' \u5e78 / \\ \u00e4 \u21ad \U0001d550",
  34             },
  35         },
  36         {
  37             u'url': u'http://vimeopro.com/openstreetmapus/state-of-the-map-us-2013/video/68093876',
  38             u'file': u'68093876.mp4',
  39             u'md5': u'3b5ca6aa22b60dfeeadf50b72e44ed82',
  40             u'note': u'Vimeo Pro video (#1197)',
  41             u'info_dict': {
  42                 u'uploader_id': u'openstreetmapus',
  43                 u'uploader': u'OpenStreetMap US',
  44                 u'title': u'Andy Allan - Putting the Carto into OpenStreetMap Cartography',
  45             },
  46         },
  47         {
  48             u'url': u'http://player.vimeo.com/video/54469442',
  49             u'file': u'54469442.mp4',
  50             u'md5': u'619b811a4417aa4abe78dc653becf511',
  51             u'note': u'Videos that embed the url in the player page',
  52             u'info_dict': {
  53                 u'title': u'Kathy Sierra: Building the minimum Badass User, Business of Software',
  54                 u'uploader': u'The BLN & Business of Software',
  55             },
  56         },
  57     ]
  58
  59     def _login(self):
  60         (username, password) = self._get_login_info()
  61         if username is None:
  62             return
  63         self.report_login()
  64         login_url = 'https://vimeo.com/log_in'
  65         webpage = self._download_webpage(login_url, None, False)
  66         token = re.search(r'xsrft: \'(.*?)\'', webpage).group(1)
  67         data = compat_urllib_parse.urlencode({'email': username,
  68                                               'password': password,
  69                                               'action': 'login',
  70                                               'service': 'vimeo',
  71                                               'token': token,
  72                                               })
  73         login_request = compat_urllib_request.Request(login_url, data)
  74         login_request.add_header('Content-Type', 'application/x-www-form-urlencoded')
  75         login_request.add_header('Cookie', 'xsrft=%s' % token)
  76         self._download_webpage(login_request, None, False, u'Wrong login info')
  77
  78     def _verify_video_password(self, url, video_id, webpage):
  79         password = self._downloader.params.get('videopassword', None)
  80         if password is None:
  81             raise ExtractorError(u'This video is protected by a password, use the --video-password option')
  82         token = re.search(r'xsrft: \'(.*?)\'', webpage).group(1)
  83         data = compat_urllib_parse.urlencode({'password': password,
  84                                               'token': token})
  85         # I didn't manage to use the password with https
  86         if url.startswith('https'):
  87             pass_url = url.replace('https','http')
  88         else:
  89             pass_url = url
  90         password_request = compat_urllib_request.Request(pass_url+'/password', data)
  91         password_request.add_header('Content-Type', 'application/x-www-form-urlencoded')
  92         password_request.add_header('Cookie', 'xsrft=%s' % token)
  93         self._download_webpage(password_request, video_id,
  94                                u'Verifying the password',
  95                                u'Wrong password')
  96
  97     def _real_initialize(self):
  98         self._login()
  99
 100     def _real_extract(self, url, new_video=True):
 101         # Extract ID from URL
 102         mobj = re.match(self._VALID_URL, url)
 103         if mobj is None:
 104             raise ExtractorError(u'Invalid URL: %s' % url)
 105
 106         video_id = mobj.group('id')
 107         if not mobj.group('proto'):
 108             url = 'https://' + url
 109         elif mobj.group('pro'):
 110             url = 'http://player.vimeo.com/video/' + video_id
 111         elif mobj.group('direct_link'):
 112             url = 'https://vimeo.com/' + video_id
 113
 114         # Retrieve video webpage to extract further information
 115         request = compat_urllib_request.Request(url, None, std_headers)
 116         webpage = self._download_webpage(request, video_id)
 117
 118         # Now we begin extracting as much information as we can from what we
 119         # retrieved. First we extract the information common to all extractors,
 120         # and latter we extract those that are Vimeo specific.
 121         self.report_extraction(video_id)
 122
 123         # Extract the config JSON
 124         try:
 125             config = self._search_regex([r' = {config:({.+?}),assets:', r'c=({.+?);'],
 126                 webpage, u'info section', flags=re.DOTALL)
 127             config = json.loads(config)
 128         except:
 129             if re.search('The creator of this video has not given you permission to embed it on this domain.', webpage):
 130                 raise ExtractorError(u'The author has restricted the access to this video, try with the "--referer" option')
 131
 132             if re.search('If so please provide the correct password.', webpage):
 133                 self._verify_video_password(url, video_id, webpage)
 134                 return self._real_extract(url)
 135             else:
 136                 raise ExtractorError(u'Unable to extract info section')
 137
 138         # Extract title
 139         video_title = config["video"]["title"]
 140
 141         # Extract uploader and uploader_id
 142         video_uploader = config["video"]["owner"]["name"]
 143         video_uploader_id = config["video"]["owner"]["url"].split('/')[-1] if config["video"]["owner"]["url"] else None
 144
 145         # Extract video thumbnail
 146         video_thumbnail = config["video"].get("thumbnail")
 147         if video_thumbnail is None:
 148             _, video_thumbnail = sorted((int(width), t_url) for (width, t_url) in config["video"]["thumbs"].items())[-1]
 149
 150         # Extract video description
 151         video_description = None
 152         try:
 153             video_description = get_element_by_attribute("itemprop", "description", webpage)
 154             if video_description: video_description = clean_html(video_description)
 155         except AssertionError as err:
 156             # On some pages like (http://player.vimeo.com/video/54469442) the
 157             # html tags are not closed, python 2.6 cannot handle it
 158             if err.args[0] == 'we should not get here!':
 159                 pass
 160             else:
 161                 raise
 162
 163         # Extract upload date
 164         video_upload_date = None
 165         mobj = re.search(r'<meta itemprop="dateCreated" content="(\d{4})-(\d{2})-(\d{2})T', webpage)
 166         if mobj is not None:
 167             video_upload_date = mobj.group(1) + mobj.group(2) + mobj.group(3)
 168
 169         # Vimeo specific: extract request signature and timestamp
 170         sig = config['request']['signature']
 171         timestamp = config['request']['timestamp']
 172
 173         # Vimeo specific: extract video codec and quality information
 174         # First consider quality, then codecs, then take everything
 175         # TODO bind to format param
 176         codecs = [('h264', 'mp4'), ('vp8', 'flv'), ('vp6', 'flv')]
 177         files = { 'hd': [], 'sd': [], 'other': []}
 178         config_files = config["video"].get("files") or config["request"].get("files")
 179         for codec_name, codec_extension in codecs:
 180             if codec_name in config_files:
 181                 if 'hd' in config_files[codec_name]:
 182                     files['hd'].append((codec_name, codec_extension, 'hd'))
 183                 elif 'sd' in config_files[codec_name]:
 184                     files['sd'].append((codec_name, codec_extension, 'sd'))
 185                 else:
 186                     files['other'].append((codec_name, codec_extension, config_files[codec_name][0]))
 187
 188         for quality in ('hd', 'sd', 'other'):
 189             if len(files[quality]) > 0:
 190                 video_quality = files[quality][0][2]
 191                 video_codec = files[quality][0][0]
 192                 video_extension = files[quality][0][1]
 193                 self.to_screen(u'%s: Downloading %s file at %s quality' % (video_id, video_codec.upper(), video_quality))
 194                 break
 195         else:
 196             raise ExtractorError(u'No known codec found')
 197
 198         video_url = None
 199         if isinstance(config_files[video_codec], dict):
 200             video_url = config_files[video_codec][video_quality].get("url")
 201         if video_url is None:
 202             video_url = "http://player.vimeo.com/play_redirect?clip_id=%s&sig=%s&time=%s&quality=%s&codecs=%s&type=moogaloop_local&embed_location=" \
 203                         %(video_id, sig, timestamp, video_quality, video_codec.upper())
 204
 205         return [{
 206             'id':       video_id,
 207             'url':      video_url,
 208             'uploader': video_uploader,
 209             'uploader_id': video_uploader_id,
 210             'upload_date':  video_upload_date,
 211             'title':    video_title,
 212             'ext':      video_extension,
 213             'thumbnail':    video_thumbnail,
 214             'description':  video_description,
 215         }]
 216
 217
 218 class VimeoChannelIE(InfoExtractor):
 219     IE_NAME = u'vimeo:channel'
 220     _VALID_URL = r'(?:https?://)?vimeo.\com/channels/(?P<id>[^/]+)'
 221     _MORE_PAGES_INDICATOR = r'<a.+?rel="next"'
 222
 223     def _real_extract(self, url):
 224         mobj = re.match(self._VALID_URL, url)
 225         channel_id =  mobj.group('id')
 226         video_ids = []
 227
 228         for pagenum in itertools.count(1):
 229             webpage = self._download_webpage('http://vimeo.com/channels/%s/videos/page:%d' % (channel_id, pagenum),
 230                                              channel_id, u'Downloading page %s' % pagenum)
 231             video_ids.extend(re.findall(r'id="clip_(\d+?)"', webpage))
 232             if re.search(self._MORE_PAGES_INDICATOR, webpage, re.DOTALL) is None:
 233                 break
 234
 235         entries = [self.url_result('http://vimeo.com/%s' % video_id, 'Vimeo')
 236                    for video_id in video_ids]
 237         channel_title = self._html_search_regex(r'<a href="/channels/%s">(.*?)</a>' % channel_id,
 238                                                 webpage, u'channel title')
 239         return {'_type': 'playlist',
 240                 'id': channel_id,
 241                 'title': channel_title,
 242                 'entries': entries,
 243                 }