Add support for authenticated videos

youtube_dl/extractor/gdcvault.py

@@ -1,25 +1,17 @@
 from __future__ import unicode_literals
 
 import re
-import json
 import xml.etree.ElementTree
 
 from .common import InfoExtractor
-from ..utils import unified_strdate
+from ..utils import (
-
+    compat_urllib_parse,
+    compat_urllib_request,
+)
 
 class GDCVaultIE(InfoExtractor):
     _VALID_URL = r'https?://(?:www\.)?gdcvault\.com/play/(?P<id>\d+)/(?P<name>(\w|-)+)'
     _TESTS = [
-        {
-            u'url': u'http://www.gdcvault.com/play/1015683/Embracing-the-Dark-Art-of',
-            u'md5': u'05763e5edd1a74776999a12b02ee1c4e',
-            u'info_dict': {
-                u"id": u"1015683",
-                u"ext": u"flv",
-                u"title": u"Embracing the Dark Art of Mathematical Modeling in AI"
-            }
-        },
         {
             u'url': u'http://www.gdcvault.com/play/1019721/Doki-Doki-Universe-Sweet-Simple',
             u'md5': u'7ce8388f544c88b7ac11c7ab1b593704',
@@ -29,38 +21,23 @@ class GDCVaultIE(InfoExtractor):
                 u"title": u"Doki-Doki Universe: Sweet, Simple and Genuine (GDC Next 10)"
             }
         },
+        {
+            u'url': u'http://www.gdcvault.com/play/1015683/Embracing-the-Dark-Art-of',
+            u'md5': u'fca91078a90f28aa5164ef6b23b78654',
+            u'info_dict': {
+                u"id": u"1015683",
+                u"ext": u"flv",
+                u"title": u"Embracing the Dark Art of Mathematical Modeling in AI"
+            }
+        },
     ]
 
-    def _real_extract(self, url):
+    def _parse_mp4(self, xml_description):
-        mobj = re.match(self._VALID_URL, url)
-
-        video_id = mobj.group('id')
-        webpage_url = 'http://www.gdcvault.com/play/' + video_id
-
-        start_page = self._download_webpage(webpage_url, video_id)
-
-        self.report_extraction(video_id)
-
-        xml_root = self._html_search_regex(r'<iframe src="(?P<xml_root>.*?)player.html.*?".*?</iframe>', start_page, 'xml root')
-        xml_name = self._html_search_regex(r'<iframe src=".*?\?xml=(?P<xml_file>.+?\.xml).*?".*?</iframe>', start_page, 'xml filename', None, False)
-        if xml_name is None:
-            # Fallback to the older format
-            xml_name = self._html_search_regex(r'<iframe src=".*?\?xmlURL=xml/(?P<xml_file>.+?\.xml).*?".*?</iframe>', start_page, 'xml filename')
-
-        xml_decription_url = xml_root + 'xml/' + xml_name
-
-        xml_description = self._download_xml(xml_decription_url, video_id)
-
-        video_title = xml_description.find('./metadata/title').text
-
-        video_details = {
-            'id': video_id,
-            'title': video_title,
-        }
         video_formats = []
-
         mp4_video = xml_description.find('./metadata/mp4video')
-        if mp4_video is not None:
+        if mp4_video is None:
+            return None
+
         mobj = re.match(r'(?P<root>https?://.*?/).*', mp4_video.text)
         video_root = mobj.group('root')
         formats = xml_description.findall('./metadata/MBRVideos/MBRVideo')
@@ -72,9 +49,10 @@ class GDCVaultIE(InfoExtractor):
                 'url': url,
                 'vbr': int(vbr),
             })
-            video_details['formats'] = video_formats
+        return video_formats
-        else:
+
-            # Fallback to flv
+    def _parse_flv(self, xml_description):
+        video_formats = []
         akami_url = xml_description.find('./metadata/akamaiHost').text
         slide_video_path = xml_description.find('./metadata/slideVideo').text
         video_formats.append({
@@ -92,6 +70,67 @@ class GDCVaultIE(InfoExtractor):
                 'preference': -1,
                 'format_id': 'speaker',
             })
+        return video_formats
+
+    def _login(self, webpage_url, video_id):
+        (username, password) = self._get_login_info()
+        if username is None or password is None:
+            self.report_warning(u'It looks like ' + webpage_url + u' requires a login. Try specifying a username and password and try again.')
+            return None
+
+        mobj = re.match(r'(?P<root_url>https?://.*?/).*', webpage_url)
+        login_url = mobj.group('root_url') + 'api/login.php'
+        logout_url = mobj.group('root_url') + 'logout'
+
+        login_form = {
+            'email': username,
+            'password': password,
+        }
+
+        request = compat_urllib_request.Request(login_url, compat_urllib_parse.urlencode(login_form))
+        request.add_header('Content-Type', 'application/x-www-form-urlencoded')
+        login_content = self._download_webpage(request, video_id, 'Logging in')
+        start_page = self._download_webpage(webpage_url, video_id, 'Getting authenticated video page')
+        logout_content = self._download_webpage(logout_url, video_id, 'Logging out')
+
+        return start_page
+
+    def _real_extract(self, url):
+        mobj = re.match(self._VALID_URL, url)
+
+        video_id = mobj.group('id')
+        webpage_url = 'http://www.gdcvault.com/play/' + video_id
+        start_page = self._download_webpage(webpage_url, video_id)
+
+        xml_root = self._html_search_regex(r'<iframe src="(?P<xml_root>.*?)player.html.*?".*?</iframe>', start_page, 'xml root', None, False)
+
+        if xml_root is None:
+            # Probably need to authenticate
+            start_page = self._login(webpage_url, video_id)
+            if start_page is None:
+                self.report_warning(u'Could not login.')
+            else:
+                # Grab the url from the authenticated page
+                xml_root = self._html_search_regex(r'<iframe src="(?P<xml_root>.*?)player.html.*?".*?</iframe>', start_page, 'xml root')
+
+        self.report_extraction(video_id)
+        xml_name = self._html_search_regex(r'<iframe src=".*?\?xml=(?P<xml_file>.+?\.xml).*?".*?</iframe>', start_page, 'xml filename', None, False)
+        if xml_name is None:
+            # Fallback to the older format
+            xml_name = self._html_search_regex(r'<iframe src=".*?\?xmlURL=xml/(?P<xml_file>.+?\.xml).*?".*?</iframe>', start_page, 'xml filename')
+
+        xml_decription_url = xml_root + 'xml/' + xml_name
+        xml_description = self._download_xml(xml_decription_url, video_id)
+
+        video_title = xml_description.find('./metadata/title').text
+        video_details = {
+            'id': video_id,
+            'title': video_title,
+        }
+
+        video_formats = self._parse_mp4(xml_description)
+        if video_formats is None:
+            video_formats = self._parse_flv(xml_description)
 
         return [{
             'id': video_id,