_ 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     unsmuggle_url,
  15 )
  16
  17 class VimeoIE(InfoExtractor):
  18     """Information extractor for vimeo.com."""
  19
  20     # _VALID_URL matches Vimeo URLs
  21     _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]+)/?(?:[?].*)?$'
  22     _NETRC_MACHINE = 'vimeo'
  23     IE_NAME = u'vimeo'
  24     _TESTS = [
  25         {
  26             u'url': u'http://vimeo.com/56015672',
  27             u'file': u'56015672.mp4',
  28             u'md5': u'8879b6cc097e987f02484baf890129e5',
  29             u'info_dict': {
  30                 u"upload_date": u"20121220",
  31                 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",
  32                 u"uploader_id": u"user7108434",
  33                 u"uploader": u"Filippo Valsorda",
  34                 u"title": u"youtube-dl test video - \u2605 \" ' \u5e78 / \\ \u00e4 \u21ad \U0001d550",
  35             },
  36         },
  37         {
  38             u'url': u'http://vimeopro.com/openstreetmapus/state-of-the-map-us-2013/video/68093876',
  39             u'file': u'68093876.mp4',
  40             u'md5': u'3b5ca6aa22b60dfeeadf50b72e44ed82',
  41             u'note': u'Vimeo Pro video (#1197)',
  42             u'info_dict': {
  43                 u'uploader_id': u'openstreetmapus',
  44                 u'uploader': u'OpenStreetMap US',
  45                 u'title': u'Andy Allan - Putting the Carto into OpenStreetMap Cartography',
  46             },
  47         },
  48         {
  49             u'url': u'http://player.vimeo.com/video/54469442',
  50             u'file': u'54469442.mp4',
  51             u'md5': u'619b811a4417aa4abe78dc653becf511',
  52             u'note': u'Videos that embed the url in the player page',
  53             u'info_dict': {
  54                 u'title': u'Kathy Sierra: Building the minimum Badass User, Business of Software',
  55                 u'uploader': u'The BLN & Business of Software',
  56             },
  57         }
  58     ]
  59
  60     def _login(self):
  61         (username, password) = self._get_login_info()
  62         if username is None:
  63             return
  64         self.report_login()
  65         login_url = 'https://vimeo.com/log_in'
  66         webpage = self._download_webpage(login_url, None, False)
  67         token = re.search(r'xsrft: \'(.*?)\'', webpage).group(1)
  68         data = compat_urllib_parse.urlencode({'email': username,
  69                                               'password': password,
  70                                               'action': 'login',
  71                                               'service': 'vimeo',
  72                                               'token': token,
  73                                               })
  74         login_request = compat_urllib_request.Request(login_url, data)
  75         login_request.add_header('Content-Type', 'application/x-www-form-urlencoded')
  76         login_request.add_header('Cookie', 'xsrft=%s' % token)
  77         self._download_webpage(login_request, None, False, u'Wrong login info')
  78
  79     def _verify_video_password(self, url, video_id, webpage):
  80         password = self._downloader.params.get('videopassword', None)
  81         if password is None:
  82             raise ExtractorError(u'This video is protected by a password, use the --video-password option')
  83         token = re.search(r'xsrft: \'(.*?)\'', webpage).group(1)
  84         data = compat_urllib_parse.urlencode({'password': password,
  85                                               'token': token})
  86         # I didn't manage to use the password with https
  87         if url.startswith('https'):
  88             pass_url = url.replace('https','http')
  89         else:
  90             pass_url = url
  91         password_request = compat_urllib_request.Request(pass_url+'/password', data)
  92         password_request.add_header('Content-Type', 'application/x-www-form-urlencoded')
  93         password_request.add_header('Cookie', 'xsrft=%s' % token)
  94         self._download_webpage(password_request, video_id,
  95                                u'Verifying the password',
  96                                u'Wrong password')
  97
  98     def _real_initialize(self):
  99         self._login()
 100
 101     def _real_extract(self, url, new_video=True):
 102         url, data = unsmuggle_url(url)
 103         headers = std_headers
 104         if data is not None:
 105             headers = headers.copy()
 106             headers.update(data)
 107
 108         # Extract ID from URL
 109         mobj = re.match(self._VALID_URL, url)
 110         if mobj is None:
 111             raise ExtractorError(u'Invalid URL: %s' % url)
 112
 113         video_id = mobj.group('id')
 114         if not mobj.group('proto'):
 115             url = 'https://' + url
 116         elif mobj.group('pro'):
 117             url = 'http://player.vimeo.com/video/' + video_id
 118         elif mobj.group('direct_link'):
 119             url = 'https://vimeo.com/' + video_id
 120
 121         # Retrieve video webpage to extract further information
 122         request = compat_urllib_request.Request(url, None, headers)
 123         webpage = self._download_webpage(request, video_id)
 124
 125         # Now we begin extracting as much information as we can from what we
 126         # retrieved. First we extract the information common to all extractors,
 127         # and latter we extract those that are Vimeo specific.
 128         self.report_extraction(video_id)
 129
 130         # Extract the config JSON
 131         try:
 132             config = self._search_regex([r' = {config:({.+?}),assets:', r'c=({.+?);'],
 133                 webpage, u'info section', flags=re.DOTALL)
 134             config = json.loads(config)
 135         except:
 136             if re.search('The creator of this video has not given you permission to embed it on this domain.', webpage):
 137                 raise ExtractorError(u'The author has restricted the access to this video, try with the "--referer" option')
 138
 139             if re.search('If so please provide the correct password.', webpage):
 140                 self._verify_video_password(url, video_id, webpage)
 141                 return self._real_extract(url)
 142             else:
 143                 raise ExtractorError(u'Unable to extract info section')
 144
 145         # Extract title
 146         video_title = config["video"]["title"]
 147
 148         # Extract uploader and uploader_id
 149         video_uploader = config["video"]["owner"]["name"]
 150         video_uploader_id = config["video"]["owner"]["url"].split('/')[-1] if config["video"]["owner"]["url"] else None
 151
 152         # Extract video thumbnail
 153         video_thumbnail = config["video"].get("thumbnail")
 154         if video_thumbnail is None:
 155             _, video_thumbnail = sorted((int(width), t_url) for (width, t_url) in config["video"]["thumbs"].items())[-1]
 156
 157         # Extract video description
 158         video_description = None
 159         try:
 160             video_description = get_element_by_attribute("itemprop", "description", webpage)
 161             if video_description: video_description = clean_html(video_description)
 162         except AssertionError as err:
 163             # On some pages like (http://player.vimeo.com/video/54469442) the
 164             # html tags are not closed, python 2.6 cannot handle it
 165             if err.args[0] == 'we should not get here!':
 166                 pass
 167             else:
 168                 raise
 169
 170         # Extract upload date
 171         video_upload_date = None
 172         mobj = re.search(r'<meta itemprop="dateCreated" content="(\d{4})-(\d{2})-(\d{2})T', webpage)
 173         if mobj is not None:
 174             video_upload_date = mobj.group(1) + mobj.group(2) + mobj.group(3)
 175
 176         # Vimeo specific: extract request signature and timestamp
 177         sig = config['request']['signature']
 178         timestamp = config['request']['timestamp']
 179
 180         # Vimeo specific: extract video codec and quality information
 181         # First consider quality, then codecs, then take everything
 182         # TODO bind to format param
 183         codecs = [('h264', 'mp4'), ('vp8', 'flv'), ('vp6', 'flv')]
 184         files = { 'hd': [], 'sd': [], 'other': []}
 185         config_files = config["video"].get("files") or config["request"].get("files")
 186         for codec_name, codec_extension in codecs:
 187             if codec_name in config_files:
 188                 if 'hd' in config_files[codec_name]:
 189                     files['hd'].append((codec_name, codec_extension, 'hd'))
 190                 elif 'sd' in config_files[codec_name]:
 191                     files['sd'].append((codec_name, codec_extension, 'sd'))
 192                 else:
 193                     files['other'].append((codec_name, codec_extension, config_files[codec_name][0]))
 194
 195         for quality in ('hd', 'sd', 'other'):
 196             if len(files[quality]) > 0:
 197                 video_quality = files[quality][0][2]
 198                 video_codec = files[quality][0][0]
 199                 video_extension = files[quality][0][1]
 200                 self.to_screen(u'%s: Downloading %s file at %s quality' % (video_id, video_codec.upper(), video_quality))
 201                 break
 202         else:
 203             raise ExtractorError(u'No known codec found')
 204
 205         video_url = None
 206         if isinstance(config_files[video_codec], dict):
 207             video_url = config_files[video_codec][video_quality].get("url")
 208         if video_url is None:
 209             video_url = "http://player.vimeo.com/play_redirect?clip_id=%s&sig=%s&time=%s&quality=%s&codecs=%s&type=moogaloop_local&embed_location=" \
 210                         %(video_id, sig, timestamp, video_quality, video_codec.upper())
 211
 212         return [{
 213             'id':       video_id,
 214             'url':      video_url,
 215             'uploader': video_uploader,
 216             'uploader_id': video_uploader_id,
 217             'upload_date':  video_upload_date,
 218             'title':    video_title,
 219             'ext':      video_extension,
 220             'thumbnail':    video_thumbnail,
 221             'description':  video_description,
 222         }]
 223
 224
 225 class VimeoChannelIE(InfoExtractor):
 226     IE_NAME = u'vimeo:channel'
 227     _VALID_URL = r'(?:https?://)?vimeo.\com/channels/(?P<id>[^/]+)'
 228     _MORE_PAGES_INDICATOR = r'<a.+?rel="next"'
 229
 230     def _real_extract(self, url):
 231         mobj = re.match(self._VALID_URL, url)
 232         channel_id =  mobj.group('id')
 233         video_ids = []
 234
 235         for pagenum in itertools.count(1):
 236             webpage = self._download_webpage('http://vimeo.com/channels/%s/videos/page:%d' % (channel_id, pagenum),
 237                                              channel_id, u'Downloading page %s' % pagenum)
 238             video_ids.extend(re.findall(r'id="clip_(\d+?)"', webpage))
 239             if re.search(self._MORE_PAGES_INDICATOR, webpage, re.DOTALL) is None:
 240                 break
 241
 242         entries = [self.url_result('http://vimeo.com/%s' % video_id, 'Vimeo')
 243                    for video_id in video_ids]
 244         channel_title = self._html_search_regex(r'<a href="/channels/%s">(.*?)</a>' % channel_id,
 245                                                 webpage, u'channel title')
 246         return {'_type': 'playlist',
 247                 'id': channel_id,
 248                 'title': channel_title,
 249                 'entries': entries,
 250                 }