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

   1 from __future__ import unicode_literals
   2
   3 import re
   4
   5 from .common import InfoExtractor
   6 from ..utils import (
   7     remove_end,
   8     HEADRequest,
   9     sanitized_Request,
  10     urlencode_postdata,
  11 )
  12
  13
  14 class GDCVaultIE(InfoExtractor):
  15     _VALID_URL = r'https?://(?:www\.)?gdcvault\.com/play/(?P<id>\d+)/(?P<name>(\w|-)+)?'
  16     _NETRC_MACHINE = 'gdcvault'
  17     _TESTS = [
  18         {
  19             'url': 'http://www.gdcvault.com/play/1019721/Doki-Doki-Universe-Sweet-Simple',
  20             'md5': '7ce8388f544c88b7ac11c7ab1b593704',
  21             'info_dict': {
  22                 'id': '1019721',
  23                 'display_id': 'Doki-Doki-Universe-Sweet-Simple',
  24                 'ext': 'mp4',
  25                 'title': 'Doki-Doki Universe: Sweet, Simple and Genuine (GDC Next 10)'
  26             }
  27         },
  28         {
  29             'url': 'http://www.gdcvault.com/play/1015683/Embracing-the-Dark-Art-of',
  30             'info_dict': {
  31                 'id': '1015683',
  32                 'display_id': 'Embracing-the-Dark-Art-of',
  33                 'ext': 'flv',
  34                 'title': 'Embracing the Dark Art of Mathematical Modeling in AI'
  35             },
  36             'params': {
  37                 'skip_download': True,  # Requires rtmpdump
  38             }
  39         },
  40         {
  41             'url': 'http://www.gdcvault.com/play/1015301/Thexder-Meets-Windows-95-or',
  42             'md5': 'a5eb77996ef82118afbbe8e48731b98e',
  43             'info_dict': {
  44                 'id': '1015301',
  45                 'display_id': 'Thexder-Meets-Windows-95-or',
  46                 'ext': 'flv',
  47                 'title': 'Thexder Meets Windows 95, or Writing Great Games in the Windows 95 Environment',
  48             },
  49             'skip': 'Requires login',
  50         },
  51         {
  52             'url': 'http://gdcvault.com/play/1020791/',
  53             'only_matching': True,
  54         },
  55         {
  56             'url': 'http://gdcvault.com/play/1023460/Tenacious-Design-and-The-Interface',
  57             'md5': 'a8efb6c31ed06ca8739294960b2dbabd',
  58             'info_dict': {
  59                 'id': '1023460',
  60                 'ext': 'mp4',
  61                 'display_id': 'Tenacious-Design-and-The-Interface',
  62                 'title': 'Tenacious Design and The Interface of \'Destiny\'',
  63             },
  64         },
  65     ]
  66
  67     def _parse_mp4(self, xml_description):
  68         video_formats = []
  69         video_root = None
  70
  71         mp4_video = xml_description.find('./metadata/mp4video')
  72         if mp4_video is not None:
  73             mobj = re.match(r'(?P<root>https?://.*?/).*', mp4_video.text)
  74             video_root = mobj.group('root')
  75         if video_root is None:
  76             # Hard-coded in http://evt.dispeak.com/ubm/gdc/sf16/custom/player2.js
  77             video_root = 'http://s3-2u.digitallyspeaking.com/'
  78
  79         formats = xml_description.findall('./metadata/MBRVideos/MBRVideo')
  80         if not formats:
  81             return None
  82         for format in formats:
  83             mobj = re.match(r'mp4\:(?P<path>.*)', format.find('streamName').text)
  84             url = video_root + mobj.group('path')
  85             vbr = format.find('bitrate').text
  86             video_formats.append({
  87                 'url': url,
  88                 'vbr': int(vbr),
  89             })
  90         return video_formats
  91
  92     def _parse_flv(self, xml_description):
  93         formats = []
  94         akamai_url = xml_description.find('./metadata/akamaiHost').text
  95         audios = xml_description.find('./metadata/audios')
  96         if audios is not None:
  97             for audio in audios:
  98                 formats.append({
  99                     'url': 'rtmp://%s/ondemand?ovpfv=1.1' % akamai_url,
 100                     'play_path': remove_end(audio.get('url'), '.flv'),
 101                     'ext': 'flv',
 102                     'vcodec': 'none',
 103                     'format_id': audio.get('code'),
 104                 })
 105         slide_video_path = xml_description.find('./metadata/slideVideo').text
 106         formats.append({
 107             'url': 'rtmp://%s/ondemand?ovpfv=1.1' % akamai_url,
 108             'play_path': remove_end(slide_video_path, '.flv'),
 109             'ext': 'flv',
 110             'format_note': 'slide deck video',
 111             'quality': -2,
 112             'preference': -2,
 113             'format_id': 'slides',
 114         })
 115         speaker_video_path = xml_description.find('./metadata/speakerVideo').text
 116         formats.append({
 117             'url': 'rtmp://%s/ondemand?ovpfv=1.1' % akamai_url,
 118             'play_path': remove_end(speaker_video_path, '.flv'),
 119             'ext': 'flv',
 120             'format_note': 'speaker video',
 121             'quality': -1,
 122             'preference': -1,
 123             'format_id': 'speaker',
 124         })
 125         return formats
 126
 127     def _login(self, webpage_url, display_id):
 128         (username, password) = self._get_login_info()
 129         if username is None or password is None:
 130             self.report_warning('It looks like ' + webpage_url + ' requires a login. Try specifying a username and password and try again.')
 131             return None
 132
 133         mobj = re.match(r'(?P<root_url>https?://.*?/).*', webpage_url)
 134         login_url = mobj.group('root_url') + 'api/login.php'
 135         logout_url = mobj.group('root_url') + 'logout'
 136
 137         login_form = {
 138             'email': username,
 139             'password': password,
 140         }
 141
 142         request = sanitized_Request(login_url, urlencode_postdata(login_form))
 143         request.add_header('Content-Type', 'application/x-www-form-urlencoded')
 144         self._download_webpage(request, display_id, 'Logging in')
 145         start_page = self._download_webpage(webpage_url, display_id, 'Getting authenticated video page')
 146         self._download_webpage(logout_url, display_id, 'Logging out')
 147
 148         return start_page
 149
 150     def _real_extract(self, url):
 151         mobj = re.match(self._VALID_URL, url)
 152
 153         video_id = mobj.group('id')
 154         display_id = mobj.group('name') or video_id
 155
 156         webpage_url = 'http://www.gdcvault.com/play/' + video_id
 157         start_page = self._download_webpage(webpage_url, display_id)
 158
 159         direct_url = self._search_regex(
 160             r's1\.addVariable\("file",\s*encodeURIComponent\("(/[^"]+)"\)\);',
 161             start_page, 'url', default=None)
 162         if direct_url:
 163             title = self._html_search_regex(
 164                 r'<td><strong>Session Name</strong></td>\s*<td>(.*?)</td>',
 165                 start_page, 'title')
 166             video_url = 'http://www.gdcvault.com' + direct_url
 167             # resolve the url so that we can detect the correct extension
 168             head = self._request_webpage(HEADRequest(video_url), video_id)
 169             video_url = head.geturl()
 170
 171             return {
 172                 'id': video_id,
 173                 'display_id': display_id,
 174                 'url': video_url,
 175                 'title': title,
 176             }
 177
 178         PLAYER_REGEX = r'<iframe src="(?P<xml_root>.+?)/player.*?\.html.*?".*?</iframe>'
 179
 180         xml_root = self._html_search_regex(
 181             PLAYER_REGEX, start_page, 'xml root', default=None)
 182         if xml_root is None:
 183             # Probably need to authenticate
 184             login_res = self._login(webpage_url, display_id)
 185             if login_res is None:
 186                 self.report_warning('Could not login.')
 187             else:
 188                 start_page = login_res
 189                 # Grab the url from the authenticated page
 190                 xml_root = self._html_search_regex(
 191                     PLAYER_REGEX, start_page, 'xml root')
 192
 193         xml_name = self._html_search_regex(
 194             r'<iframe src=".*?\?xml=(.+?\.xml).*?".*?</iframe>',
 195             start_page, 'xml filename', default=None)
 196         if xml_name is None:
 197             # Fallback to the older format
 198             xml_name = self._html_search_regex(
 199                 r'<iframe src=".*?\?xmlURL=xml/(?P<xml_file>.+?\.xml).*?".*?</iframe>',
 200                 start_page, 'xml filename')
 201
 202         xml_description = self._download_xml(
 203             '%s/xml/%s' % (xml_root, xml_name), display_id)
 204
 205         video_title = xml_description.find('./metadata/title').text
 206         video_formats = self._parse_mp4(xml_description)
 207         if video_formats is None:
 208             video_formats = self._parse_flv(xml_description)
 209
 210         return {
 211             'id': video_id,
 212             'display_id': display_id,
 213             'title': video_title,
 214             'formats': video_formats,
 215         }