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