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     compat_urllib_parse,
   8     compat_urllib_request,
   9 )
  10
  11
  12 class GDCVaultIE(InfoExtractor):
  13     _VALID_URL = r'https?://(?:www\.)?gdcvault\.com/play/(?P<id>\d+)/(?P<name>(\w|-)+)'
  14     _TESTS = [
  15         {
  16             'url': 'http://www.gdcvault.com/play/1019721/Doki-Doki-Universe-Sweet-Simple',
  17             'md5': '7ce8388f544c88b7ac11c7ab1b593704',
  18             'info_dict': {
  19                 'id': '1019721',
  20                 'ext': 'mp4',
  21                 'title': 'Doki-Doki Universe: Sweet, Simple and Genuine (GDC Next 10)'
  22             }
  23         },
  24         {
  25             'url': 'http://www.gdcvault.com/play/1015683/Embracing-the-Dark-Art-of',
  26             'info_dict': {
  27                 'id': '1015683',
  28                 'ext': 'flv',
  29                 'title': 'Embracing the Dark Art of Mathematical Modeling in AI'
  30             },
  31             'params': {
  32                 'skip_download': True,  # Requires rtmpdump
  33             }
  34         },
  35         {
  36             'url': 'http://www.gdcvault.com/play/1015301/Thexder-Meets-Windows-95-or',
  37             'md5': 'a5eb77996ef82118afbbe8e48731b98e',
  38             'info_dict': {
  39                 'id': '1015301',
  40                 'ext': 'flv',
  41                 'title': 'Thexder Meets Windows 95, or Writing Great Games in the Windows 95 Environment',
  42             }
  43         }
  44     ]
  45
  46     def _parse_mp4(self, xml_description):
  47         video_formats = []
  48         mp4_video = xml_description.find('./metadata/mp4video')
  49         if mp4_video is None:
  50             return None
  51
  52         mobj = re.match(r'(?P<root>https?://.*?/).*', mp4_video.text)
  53         video_root = mobj.group('root')
  54         formats = xml_description.findall('./metadata/MBRVideos/MBRVideo')
  55         for format in formats:
  56             mobj = re.match(r'mp4\:(?P<path>.*)', format.find('streamName').text)
  57             url = video_root + mobj.group('path')
  58             vbr = format.find('bitrate').text
  59             video_formats.append({
  60                 'url': url,
  61                 'vbr': int(vbr),
  62             })
  63         return video_formats
  64
  65     def _parse_flv(self, xml_description):
  66         video_formats = []
  67         akami_url = xml_description.find('./metadata/akamaiHost').text
  68         slide_video_path = xml_description.find('./metadata/slideVideo').text
  69         video_formats.append({
  70             'url': 'rtmp://' + akami_url + '/' + slide_video_path,
  71             'format_note': 'slide deck video',
  72             'quality': -2,
  73             'preference': -2,
  74             'format_id': 'slides',
  75         })
  76         speaker_video_path = xml_description.find('./metadata/speakerVideo').text
  77         video_formats.append({
  78             'url': 'rtmp://' + akami_url + '/' + speaker_video_path,
  79             'format_note': 'speaker video',
  80             'quality': -1,
  81             'preference': -1,
  82             'format_id': 'speaker',
  83         })
  84         return video_formats
  85
  86     def _login(self, webpage_url, video_id):
  87         (username, password) = self._get_login_info()
  88         if username is None or password is None:
  89             self.report_warning('It looks like ' + webpage_url + ' requires a login. Try specifying a username and password and try again.')
  90             return None
  91
  92         mobj = re.match(r'(?P<root_url>https?://.*?/).*', webpage_url)
  93         login_url = mobj.group('root_url') + 'api/login.php'
  94         logout_url = mobj.group('root_url') + 'logout'
  95
  96         login_form = {
  97             'email': username,
  98             'password': password,
  99         }
 100
 101         request = compat_urllib_request.Request(login_url, compat_urllib_parse.urlencode(login_form))
 102         request.add_header('Content-Type', 'application/x-www-form-urlencoded')
 103         self._download_webpage(request, video_id, 'Logging in')
 104         start_page = self._download_webpage(webpage_url, video_id, 'Getting authenticated video page')
 105         self._download_webpage(logout_url, video_id, 'Logging out')
 106
 107         return start_page
 108
 109     def _real_extract(self, url):
 110         mobj = re.match(self._VALID_URL, url)
 111
 112         video_id = mobj.group('id')
 113         webpage_url = 'http://www.gdcvault.com/play/' + video_id
 114         start_page = self._download_webpage(webpage_url, video_id)
 115
 116         direct_url = self._search_regex(
 117             r's1\.addVariable\("file",\s*encodeURIComponent\("(/[^"]+)"\)\);',
 118             start_page, 'url', default=None)
 119         if direct_url:
 120             video_url = 'http://www.gdcvault.com/' + direct_url
 121             title = self._html_search_regex(
 122                 r'<td><strong>Session Name</strong></td>\s*<td>(.*?)</td>',
 123                 start_page, 'title')
 124
 125             return {
 126                 'id': video_id,
 127                 'url': video_url,
 128                 'ext': 'flv',
 129                 'title': title,
 130             }
 131
 132         xml_root = self._html_search_regex(
 133             r'<iframe src="(?P<xml_root>.*?)player.html.*?".*?</iframe>',
 134             start_page, 'xml root', default=None)
 135         if xml_root is None:
 136             # Probably need to authenticate
 137             login_res = self._login(webpage_url, video_id)
 138             if login_res is None:
 139                 self.report_warning('Could not login.')
 140             else:
 141                 start_page = login_res
 142                 # Grab the url from the authenticated page
 143                 xml_root = self._html_search_regex(
 144                     r'<iframe src="(.*?)player.html.*?".*?</iframe>',
 145                     start_page, 'xml root')
 146
 147         xml_name = self._html_search_regex(
 148             r'<iframe src=".*?\?xml=(.+?\.xml).*?".*?</iframe>',
 149             start_page, 'xml filename', default=None)
 150         if xml_name is None:
 151             # Fallback to the older format
 152             xml_name = self._html_search_regex(r'<iframe src=".*?\?xmlURL=xml/(?P<xml_file>.+?\.xml).*?".*?</iframe>', start_page, 'xml filename')
 153
 154         xml_decription_url = xml_root + 'xml/' + xml_name
 155         xml_description = self._download_xml(xml_decription_url, video_id)
 156
 157         video_title = xml_description.find('./metadata/title').text
 158         video_formats = self._parse_mp4(xml_description)
 159         if video_formats is None:
 160             video_formats = self._parse_flv(xml_description)
 161
 162         return {
 163             'id': video_id,
 164             'title': video_title,
 165             'formats': video_formats,
 166         }