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