From ec6504b39c9b3d9766c551691a96e481278fa659 Mon Sep 17 00:00:00 2001
From: remitamine <remitamine@gmail.com>
Date: Sat, 19 Dec 2015 23:28:54 +0100
Subject: [PATCH] [gputechconf] Add new extractor(closes #5775)

---
 youtube_dl/extractor/__init__.py    |  1 +
 youtube_dl/extractor/gputechconf.py | 57 +++++++++++++++++++++++++++++
 2 files changed, 58 insertions(+)
 create mode 100644 youtube_dl/extractor/gputechconf.py

diff --git a/youtube_dl/extractor/__init__.py b/youtube_dl/extractor/__init__.py
index fef2cff62b..65c5e1c920 100644
--- a/youtube_dl/extractor/__init__.py
+++ b/youtube_dl/extractor/__init__.py
@@ -235,6 +235,7 @@ from .golem import GolemIE
 from .googleplus import GooglePlusIE
 from .googlesearch import GoogleSearchIE
 from .goshgay import GoshgayIE
+from .gputechconf import GPUTechConfIE
 from .groupon import GrouponIE
 from .hark import HarkIE
 from .hearthisat import HearThisAtIE
diff --git a/youtube_dl/extractor/gputechconf.py b/youtube_dl/extractor/gputechconf.py
new file mode 100644
index 0000000000..becc61f1e3
--- /dev/null
+++ b/youtube_dl/extractor/gputechconf.py
@@ -0,0 +1,57 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+from .common import InfoExtractor
+from ..utils import (
+    xpath_element,
+    xpath_text,
+    int_or_none,
+    parse_duration,
+)
+
+
+class GPUTechConfIE(InfoExtractor):
+    _VALID_URL = r'https?://on-demand\.gputechconf\.com/gtc/2015/video/S(?P<id>\d+)\.html'
+    _TEST = {
+        'url': 'http://on-demand.gputechconf.com/gtc/2015/video/S5156.html',
+        'md5': 'a8862a00a0fd65b8b43acc5b8e33f798',
+        'info_dict': {
+            'id': '5156',
+            'ext': 'mp4',
+            'title': 'Coordinating More Than 3 Million CUDA Threads for Social Network Analysis',
+            'duration': 1219,
+        }
+    }
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        webpage = self._download_webpage(url, video_id)
+
+        root_path = self._search_regex(r'var\s+rootPath\s*=\s*"([^"]+)', webpage, 'root path', 'http://evt.dispeak.com/nvidia/events/gtc15/')
+        xml_file_id = self._search_regex(r'var\s+xmlFileId\s*=\s*"([^"]+)', webpage, 'xml file id')
+
+        doc = self._download_xml('%sxml/%s.xml' % (root_path, xml_file_id), video_id)
+
+        metadata = xpath_element(doc, 'metadata')
+        http_host = xpath_text(metadata, 'httpHost')
+        mbr_videos = xpath_element(metadata, 'MBRVideos')
+
+        formats = []
+        for mbr_video in mbr_videos.findall('MBRVideo'):
+            stream_name = xpath_text(mbr_video, 'streamName')
+            if stream_name:
+                bitrate = int_or_none(xpath_text(mbr_video, 'bitrate'))
+                formats.append({
+                    'url': 'http://%s/%s' % (http_host, stream_name.replace('mp4:', '')),
+                    'tbr': bitrate,
+                    'format_id': 'http-%d' % bitrate,
+                })
+        self._sort_formats(formats)
+
+        return {
+            'id': video_id,
+            'title': xpath_text(metadata, 'title'),
+            'duration': parse_duration(xpath_text(metadata, 'endTime')),
+            'creator': xpath_text(metadata, 'speaker'),
+            'formats': formats,
+        }