_ 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     _TEST = {
  24         u'url': u'http://vimeo.com/56015672',
  25         u'file': u'56015672.mp4',
  26         u'md5': u'8879b6cc097e987f02484baf890129e5',
  27         u'info_dict': {
  28             u"upload_date": u"20121220",
  29             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",
  30             u"uploader_id": u"user7108434",
  31             u"uploader": u"Filippo Valsorda",
  32             u"title": u"youtube-dl test video - \u2605 \" ' \u5e78 / \\ \u00e4 \u21ad \U0001d550"
  33         }
  34     }
  35
  36     def _login(self):
  37         (username, password) = self._get_login_info()
  38         if username is None:
  39             return
  40         self.report_login()
  41         login_url = 'https://vimeo.com/log_in'
  42         webpage = self._download_webpage(login_url, None, False)
  43         token = re.search(r'xsrft: \'(.*?)\'', webpage).group(1)
  44         data = compat_urllib_parse.urlencode({'email': username,
  45                                               'password': password,
  46                                               'action': 'login',
  47                                               'service': 'vimeo',
  48                                               'token': token,
  49                                               })
  50         login_request = compat_urllib_request.Request(login_url, data)
  51         login_request.add_header('Content-Type', 'application/x-www-form-urlencoded')
  52         login_request.add_header('Cookie', 'xsrft=%s' % token)
  53         self._download_webpage(login_request, None, False, u'Wrong login info')
  54
  55     def _verify_video_password(self, url, video_id, webpage):
  56         password = self._downloader.params.get('videopassword', None)
  57         if password is None:
  58             raise ExtractorError(u'This video is protected by a password, use the --video-password option')
  59         token = re.search(r'xsrft: \'(.*?)\'', webpage).group(1)
  60         data = compat_urllib_parse.urlencode({'password': password,
  61                                               'token': token})
  62         # I didn't manage to use the password with https
  63         if url.startswith('https'):
  64             pass_url = url.replace('https','http')
  65         else:
  66             pass_url = url
  67         password_request = compat_urllib_request.Request(pass_url+'/password', data)
  68         password_request.add_header('Content-Type', 'application/x-www-form-urlencoded')
  69         password_request.add_header('Cookie', 'xsrft=%s' % token)
  70         self._download_webpage(password_request, video_id,
  71                                u'Verifying the password',
  72                                u'Wrong password')
  73
  74     def _real_initialize(self):
  75         self._login()
  76
  77     def _real_extract(self, url, new_video=True):
  78         # Extract ID from URL
  79         mobj = re.match(self._VALID_URL, url)
  80         if mobj is None:
  81             raise ExtractorError(u'Invalid URL: %s' % url)
  82
  83         video_id = mobj.group('id')
  84         if not mobj.group('proto'):
  85             url = 'https://' + url
  86         if mobj.group('direct_link') or mobj.group('pro'):
  87             url = 'https://vimeo.com/' + video_id
  88
  89         # Retrieve video webpage to extract further information
  90         request = compat_urllib_request.Request(url, None, std_headers)
  91         webpage = self._download_webpage(request, video_id)
  92
  93         # Now we begin extracting as much information as we can from what we
  94         # retrieved. First we extract the information common to all extractors,
  95         # and latter we extract those that are Vimeo specific.
  96         self.report_extraction(video_id)
  97
  98         # Extract the config JSON
  99         try:
 100             config = webpage.split(' = {config:')[1].split(',assets:')[0]
 101             config = json.loads(config)
 102         except:
 103             if re.search('The creator of this video has not given you permission to embed it on this domain.', webpage):
 104                 raise ExtractorError(u'The author has restricted the access to this video, try with the "--referer" option')
 105
 106             if re.search('If so please provide the correct password.', webpage):
 107                 self._verify_video_password(url, video_id, webpage)
 108                 return self._real_extract(url)
 109             else:
 110                 raise ExtractorError(u'Unable to extract info section')
 111
 112         # Extract title
 113         video_title = config["video"]["title"]
 114
 115         # Extract uploader and uploader_id
 116         video_uploader = config["video"]["owner"]["name"]
 117         video_uploader_id = config["video"]["owner"]["url"].split('/')[-1] if config["video"]["owner"]["url"] else None
 118
 119         # Extract video thumbnail
 120         video_thumbnail = config["video"]["thumbnail"]
 121
 122         # Extract video description
 123         video_description = get_element_by_attribute("itemprop", "description", webpage)
 124         if video_description: video_description = clean_html(video_description)
 125         else: video_description = u''
 126
 127         # Extract upload date
 128         video_upload_date = None
 129         mobj = re.search(r'<meta itemprop="dateCreated" content="(\d{4})-(\d{2})-(\d{2})T', webpage)
 130         if mobj is not None:
 131             video_upload_date = mobj.group(1) + mobj.group(2) + mobj.group(3)
 132
 133         # Vimeo specific: extract request signature and timestamp
 134         sig = config['request']['signature']
 135         timestamp = config['request']['timestamp']
 136
 137         # Vimeo specific: extract video codec and quality information
 138         # First consider quality, then codecs, then take everything
 139         # TODO bind to format param
 140         codecs = [('h264', 'mp4'), ('vp8', 'flv'), ('vp6', 'flv')]
 141         files = { 'hd': [], 'sd': [], 'other': []}
 142         for codec_name, codec_extension in codecs:
 143             if codec_name in config["video"]["files"]:
 144                 if 'hd' in config["video"]["files"][codec_name]:
 145                     files['hd'].append((codec_name, codec_extension, 'hd'))
 146                 elif 'sd' in config["video"]["files"][codec_name]:
 147                     files['sd'].append((codec_name, codec_extension, 'sd'))
 148                 else:
 149                     files['other'].append((codec_name, codec_extension, config["video"]["files"][codec_name][0]))
 150
 151         for quality in ('hd', 'sd', 'other'):
 152             if len(files[quality]) > 0:
 153                 video_quality = files[quality][0][2]
 154                 video_codec = files[quality][0][0]
 155                 video_extension = files[quality][0][1]
 156                 self.to_screen(u'%s: Downloading %s file at %s quality' % (video_id, video_codec.upper(), video_quality))
 157                 break
 158         else:
 159             raise ExtractorError(u'No known codec found')
 160
 161         video_url = "http://player.vimeo.com/play_redirect?clip_id=%s&sig=%s&time=%s&quality=%s&codecs=%s&type=moogaloop_local&embed_location=" \
 162                     %(video_id, sig, timestamp, video_quality, video_codec.upper())
 163
 164         return [{
 165             'id':       video_id,
 166             'url':      video_url,
 167             'uploader': video_uploader,
 168             'uploader_id': video_uploader_id,
 169             'upload_date':  video_upload_date,
 170             'title':    video_title,
 171             'ext':      video_extension,
 172             'thumbnail':    video_thumbnail,
 173             'description':  video_description,
 174         }]
 175
 176
 177 class VimeoChannelIE(InfoExtractor):
 178     IE_NAME = u'vimeo:channel'
 179     _VALID_URL = r'(?:https?://)?vimeo.\com/channels/(?P<id>[^/]+)'
 180     _MORE_PAGES_INDICATOR = r'<a.+?rel="next"'
 181
 182     def _real_extract(self, url):
 183         mobj = re.match(self._VALID_URL, url)
 184         channel_id =  mobj.group('id')
 185         video_ids = []
 186
 187         for pagenum in itertools.count(1):
 188             webpage = self._download_webpage('http://vimeo.com/channels/%s/videos/page:%d' % (channel_id, pagenum),
 189                                              channel_id, u'Downloading page %s' % pagenum)
 190             video_ids.extend(re.findall(r'id="clip_(\d+?)"', webpage))
 191             if re.search(self._MORE_PAGES_INDICATOR, webpage, re.DOTALL) is None:
 192                 break
 193
 194         entries = [self.url_result('http://vimeo.com/%s' % video_id, 'Vimeo')
 195                    for video_id in video_ids]
 196         channel_title = self._html_search_regex(r'<a href="/channels/%s">(.*?)</a>' % channel_id,
 197                                                 webpage, u'channel title')
 198         return {'_type': 'playlist',
 199                 'id': channel_id,
 200                 'title': channel_title,
 201                 'entries': entries,
 202                 }