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

   1 from __future__ import unicode_literals
   2
   3 import re
   4 import xml.etree.ElementTree
   5
   6 from .common import InfoExtractor
   7 from ..utils import (
   8     compat_urllib_parse,
   9     compat_urllib_request,
  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             u'url': u'http://www.gdcvault.com/play/1019721/Doki-Doki-Universe-Sweet-Simple',
  17             u'md5': u'7ce8388f544c88b7ac11c7ab1b593704',
  18             u'info_dict': {
  19                 u"id": u"1019721",
  20                 u"ext": u"mp4",
  21                 u"title": u"Doki-Doki Universe: Sweet, Simple and Genuine (GDC Next 10)"
  22             }
  23         },
  24         {
  25             u'url': u'http://www.gdcvault.com/play/1015683/Embracing-the-Dark-Art-of',
  26             u'md5': u'fca91078a90f28aa5164ef6b23b78654',
  27             u'info_dict': {
  28                 u"id": u"1015683",
  29                 u"ext": u"flv",
  30                 u"title": u"Embracing the Dark Art of Mathematical Modeling in AI"
  31             }
  32         },
  33     ]
  34
  35     def _parse_mp4(self, xml_description):
  36         video_formats = []
  37         mp4_video = xml_description.find('./metadata/mp4video')
  38         if mp4_video is None:
  39             return None
  40
  41         mobj = re.match(r'(?P<root>https?://.*?/).*', mp4_video.text)
  42         video_root = mobj.group('root')
  43         formats = xml_description.findall('./metadata/MBRVideos/MBRVideo')
  44         for format in formats:
  45             mobj = re.match(r'mp4\:(?P<path>.*)', format.find('streamName').text)
  46             url = video_root + mobj.group('path')
  47             vbr = format.find('bitrate').text
  48             video_formats.append({
  49                 'url': url,
  50                 'vbr': int(vbr),
  51             })
  52         return video_formats
  53
  54     def _parse_flv(self, xml_description):
  55         video_formats = []
  56         akami_url = xml_description.find('./metadata/akamaiHost').text
  57         slide_video_path = xml_description.find('./metadata/slideVideo').text
  58         video_formats.append({
  59                 'url': 'rtmp://' + akami_url + '/' + slide_video_path,
  60                 'format_note': 'slide deck video',
  61                 'quality': -2,
  62                 'preference': -2,
  63                 'format_id': 'slides',
  64             })
  65         speaker_video_path = xml_description.find('./metadata/speakerVideo').text
  66         video_formats.append({
  67                 'url': 'rtmp://' + akami_url + '/' + speaker_video_path,
  68                 'format_note': 'speaker video',
  69                 'quality': -1,
  70                 'preference': -1,
  71                 'format_id': 'speaker',
  72             })
  73         return video_formats
  74
  75     def _login(self, webpage_url, video_id):
  76         (username, password) = self._get_login_info()
  77         if username is None or password is None:
  78             self.report_warning(u'It looks like ' + webpage_url + u' requires a login. Try specifying a username and password and try again.')
  79             return None
  80
  81         mobj = re.match(r'(?P<root_url>https?://.*?/).*', webpage_url)
  82         login_url = mobj.group('root_url') + 'api/login.php'
  83         logout_url = mobj.group('root_url') + 'logout'
  84
  85         login_form = {
  86             'email': username,
  87             'password': password,
  88         }
  89
  90         request = compat_urllib_request.Request(login_url, compat_urllib_parse.urlencode(login_form))
  91         request.add_header('Content-Type', 'application/x-www-form-urlencoded')
  92         login_content = self._download_webpage(request, video_id, 'Logging in')
  93         start_page = self._download_webpage(webpage_url, video_id, 'Getting authenticated video page')
  94         logout_content = self._download_webpage(logout_url, video_id, 'Logging out')
  95
  96         return start_page
  97
  98     def _real_extract(self, url):
  99         mobj = re.match(self._VALID_URL, url)
 100
 101         video_id = mobj.group('id')
 102         webpage_url = 'http://www.gdcvault.com/play/' + video_id
 103         start_page = self._download_webpage(webpage_url, video_id)
 104
 105         xml_root = self._html_search_regex(r'<iframe src="(?P<xml_root>.*?)player.html.*?".*?</iframe>', start_page, 'xml root', None, False)
 106
 107         if xml_root is None:
 108             # Probably need to authenticate
 109             start_page = self._login(webpage_url, video_id)
 110             if start_page is None:
 111                 self.report_warning(u'Could not login.')
 112             else:
 113                 # Grab the url from the authenticated page
 114                 xml_root = self._html_search_regex(r'<iframe src="(?P<xml_root>.*?)player.html.*?".*?</iframe>', start_page, 'xml root')
 115
 116         self.report_extraction(video_id)
 117         xml_name = self._html_search_regex(r'<iframe src=".*?\?xml=(?P<xml_file>.+?\.xml).*?".*?</iframe>', start_page, 'xml filename', None, False)
 118         if xml_name is None:
 119             # Fallback to the older format
 120             xml_name = self._html_search_regex(r'<iframe src=".*?\?xmlURL=xml/(?P<xml_file>.+?\.xml).*?".*?</iframe>', start_page, 'xml filename')
 121
 122         xml_decription_url = xml_root + 'xml/' + xml_name
 123         xml_description = self._download_xml(xml_decription_url, video_id)
 124
 125         video_title = xml_description.find('./metadata/title').text
 126         video_details = {
 127             'id': video_id,
 128             'title': video_title,
 129         }
 130
 131         video_formats = self._parse_mp4(xml_description)
 132         if video_formats is None:
 133             video_formats = self._parse_flv(xml_description)
 134
 135         return [{
 136             'id': video_id,
 137             'title': video_title,
 138             'formats': video_formats,
 139         }]