]> Raphaƫl G. Git Repositories - youtubedl/blobdiff - youtube_dl/extractor/gdcvault.py
Merge tag 'upstream/2016.06.25'
[youtubedl] / youtube_dl / extractor / gdcvault.py
index fed968f5179ebf6159212da5ab75b024b3bc0a03..3136427db39a2f1739fa0a791bd2cc85f1eedd02 100644 (file)
@@ -3,20 +3,23 @@ from __future__ import unicode_literals
 import re
 
 from .common import InfoExtractor
-from ..compat import (
-    compat_urllib_parse,
-    compat_urllib_request,
+from ..utils import (
+    HEADRequest,
+    sanitized_Request,
+    urlencode_postdata,
 )
 
 
 class GDCVaultIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?gdcvault\.com/play/(?P<id>\d+)/(?P<name>(\w|-)+)'
+    _VALID_URL = r'https?://(?:www\.)?gdcvault\.com/play/(?P<id>\d+)/(?P<name>(\w|-)+)?'
+    _NETRC_MACHINE = 'gdcvault'
     _TESTS = [
         {
             'url': 'http://www.gdcvault.com/play/1019721/Doki-Doki-Universe-Sweet-Simple',
             'md5': '7ce8388f544c88b7ac11c7ab1b593704',
             'info_dict': {
                 'id': '1019721',
+                'display_id': 'Doki-Doki-Universe-Sweet-Simple',
                 'ext': 'mp4',
                 'title': 'Doki-Doki Universe: Sweet, Simple and Genuine (GDC Next 10)'
             }
@@ -25,6 +28,7 @@ class GDCVaultIE(InfoExtractor):
             'url': 'http://www.gdcvault.com/play/1015683/Embracing-the-Dark-Art-of',
             'info_dict': {
                 'id': '1015683',
+                'display_id': 'Embracing-the-Dark-Art-of',
                 'ext': 'flv',
                 'title': 'Embracing the Dark Art of Mathematical Modeling in AI'
             },
@@ -37,54 +41,43 @@ class GDCVaultIE(InfoExtractor):
             'md5': 'a5eb77996ef82118afbbe8e48731b98e',
             'info_dict': {
                 'id': '1015301',
+                'display_id': 'Thexder-Meets-Windows-95-or',
                 'ext': 'flv',
                 'title': 'Thexder Meets Windows 95, or Writing Great Games in the Windows 95 Environment',
             },
             'skip': 'Requires login',
-        }
+        },
+        {
+            'url': 'http://gdcvault.com/play/1020791/',
+            'only_matching': True,
+        },
+        {
+            # Hard-coded hostname
+            'url': 'http://gdcvault.com/play/1023460/Tenacious-Design-and-The-Interface',
+            'md5': 'a8efb6c31ed06ca8739294960b2dbabd',
+            'info_dict': {
+                'id': '1023460',
+                'ext': 'mp4',
+                'display_id': 'Tenacious-Design-and-The-Interface',
+                'title': 'Tenacious Design and The Interface of \'Destiny\'',
+            },
+        },
+        {
+            # Multiple audios
+            'url': 'http://www.gdcvault.com/play/1014631/Classic-Game-Postmortem-PAC',
+            'info_dict': {
+                'id': '1014631',
+                'ext': 'flv',
+                'title': 'How to Create a Good Game - From My Experience of Designing Pac-Man',
+            },
+            'params': {
+                'skip_download': True,  # Requires rtmpdump
+                'format': 'jp',  # The japanese audio
+            }
+        },
     ]
 
-    def _parse_mp4(self, xml_description):
-        video_formats = []
-        mp4_video = xml_description.find('./metadata/mp4video')
-        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')
-        for format in formats:
-            mobj = re.match(r'mp4\:(?P<path>.*)', format.find('streamName').text)
-            url = video_root + mobj.group('path')
-            vbr = format.find('bitrate').text
-            video_formats.append({
-                'url': url,
-                'vbr': int(vbr),
-            })
-        return video_formats
-
-    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({
-            'url': 'rtmp://' + akami_url + '/' + slide_video_path,
-            'format_note': 'slide deck video',
-            'quality': -2,
-            'preference': -2,
-            'format_id': 'slides',
-        })
-        speaker_video_path = xml_description.find('./metadata/speakerVideo').text
-        video_formats.append({
-            'url': 'rtmp://' + akami_url + '/' + speaker_video_path,
-            'format_note': 'speaker video',
-            'quality': -1,
-            'preference': -1,
-            'format_id': 'speaker',
-        })
-        return video_formats
-
-    def _login(self, webpage_url, video_id):
+    def _login(self, webpage_url, display_id):
         (username, password) = self._get_login_info()
         if username is None or password is None:
             self.report_warning('It looks like ' + webpage_url + ' requires a login. Try specifying a username and password and try again.')
@@ -99,11 +92,11 @@ class GDCVaultIE(InfoExtractor):
             'password': password,
         }
 
-        request = compat_urllib_request.Request(login_url, compat_urllib_parse.urlencode(login_form))
+        request = sanitized_Request(login_url, urlencode_postdata(login_form))
         request.add_header('Content-Type', 'application/x-www-form-urlencoded')
-        self._download_webpage(request, video_id, 'Logging in')
-        start_page = self._download_webpage(webpage_url, video_id, 'Getting authenticated video page')
-        self._download_webpage(logout_url, video_id, 'Logging out')
+        self._download_webpage(request, display_id, 'Logging in')
+        start_page = self._download_webpage(webpage_url, display_id, 'Getting authenticated video page')
+        self._download_webpage(logout_url, display_id, 'Logging out')
 
         return start_page
 
@@ -111,57 +104,58 @@ class GDCVaultIE(InfoExtractor):
         mobj = re.match(self._VALID_URL, url)
 
         video_id = mobj.group('id')
+        display_id = mobj.group('name') or video_id
+
         webpage_url = 'http://www.gdcvault.com/play/' + video_id
-        start_page = self._download_webpage(webpage_url, video_id)
+        start_page = self._download_webpage(webpage_url, display_id)
 
         direct_url = self._search_regex(
             r's1\.addVariable\("file",\s*encodeURIComponent\("(/[^"]+)"\)\);',
             start_page, 'url', default=None)
         if direct_url:
-            video_url = 'http://www.gdcvault.com/' + direct_url
             title = self._html_search_regex(
                 r'<td><strong>Session Name</strong></td>\s*<td>(.*?)</td>',
                 start_page, 'title')
+            video_url = 'http://www.gdcvault.com' + direct_url
+            # resolve the url so that we can detect the correct extension
+            head = self._request_webpage(HEADRequest(video_url), video_id)
+            video_url = head.geturl()
 
             return {
                 'id': video_id,
+                'display_id': display_id,
                 'url': video_url,
-                'ext': 'flv',
                 'title': title,
             }
 
+        PLAYER_REGEX = r'<iframe src="(?P<xml_root>.+?)/player.*?\.html.*?".*?</iframe>'
+
         xml_root = self._html_search_regex(
-            r'<iframe src="(?P<xml_root>.*?)player.html.*?".*?</iframe>',
-            start_page, 'xml root', default=None)
+            PLAYER_REGEX, start_page, 'xml root', default=None)
         if xml_root is None:
             # Probably need to authenticate
-            login_res = self._login(webpage_url, video_id)
+            login_res = self._login(webpage_url, display_id)
             if login_res is None:
                 self.report_warning('Could not login.')
             else:
                 start_page = login_res
                 # Grab the url from the authenticated page
                 xml_root = self._html_search_regex(
-                    r'<iframe src="(.*?)player.html.*?".*?</iframe>',
-                    start_page, 'xml root')
+                    PLAYER_REGEX, start_page, 'xml root')
 
         xml_name = self._html_search_regex(
             r'<iframe src=".*?\?xml=(.+?\.xml).*?".*?</iframe>',
             start_page, 'xml filename', default=None)
         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_formats = self._parse_mp4(xml_description)
-        if video_formats is None:
-            video_formats = self._parse_flv(xml_description)
+            xml_name = self._html_search_regex(
+                r'<iframe src=".*?\?xmlURL=xml/(?P<xml_file>.+?\.xml).*?".*?</iframe>',
+                start_page, 'xml filename')
 
         return {
+            '_type': 'url_transparent',
             'id': video_id,
-            'title': video_title,
-            'formats': video_formats,
+            'display_id': display_id,
+            'url': '%s/xml/%s' % (xml_root, xml_name),
+            'ie_key': 'DigitallySpeaking',
         }