Add support for FLV videos with speaker decks

This commit is contained in:
David Wagner 2014-02-27 00:20:34 +00:00
parent 9e68f9fdf1
commit 83cebb8b7a
1 changed files with 63 additions and 24 deletions

View File

@ -10,8 +10,17 @@ from ..utils import unified_strdate
class GDCVaultIE(InfoExtractor): 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|-)+)'
_TESTS = [
_TEST = { {
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'url': u'http://www.gdcvault.com/play/1019721/Doki-Doki-Universe-Sweet-Simple',
u'md5': u'7ce8388f544c88b7ac11c7ab1b593704', u'md5': u'7ce8388f544c88b7ac11c7ab1b593704',
u'info_dict': { u'info_dict': {
@ -19,7 +28,8 @@ class GDCVaultIE(InfoExtractor):
u"ext": u"mp4", u"ext": u"mp4",
u"title": u"Doki-Doki Universe: Sweet, Simple and Genuine (GDC Next 10)" u"title": u"Doki-Doki Universe: Sweet, Simple and Genuine (GDC Next 10)"
} }
} },
]
def _real_extract(self, url): def _real_extract(self, url):
mobj = re.match(self._VALID_URL, url) mobj = re.match(self._VALID_URL, url)
@ -32,19 +42,28 @@ class GDCVaultIE(InfoExtractor):
self.report_extraction(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_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') 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_decription_url = xml_root + 'xml/' + xml_name
xml_description = self._download_xml(xml_decription_url, video_id) xml_description = self._download_xml(xml_decription_url, video_id)
video_title = xml_description.find('./metadata/title').text video_title = xml_description.find('./metadata/title').text
mp4_video = xml_description.find('./metadata/mp4video').text video_details = {
mobj = re.match(r'(?P<root>https?://.*?/).*', mp4_video) 'id': video_id,
video_root = mobj.group('root') 'title': video_title,
}
formats = xml_description.findall('./metadata/MBRVideos/MBRVideo')
video_formats = [] video_formats = []
mp4_video = xml_description.find('./metadata/mp4video')
if mp4_video is not 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: for format in formats:
mobj = re.match(r'mp4\:(?P<path>.*)', format.find('streamName').text) mobj = re.match(r'mp4\:(?P<path>.*)', format.find('streamName').text)
url = video_root + mobj.group('path') url = video_root + mobj.group('path')
@ -53,9 +72,29 @@ class GDCVaultIE(InfoExtractor):
'url': url, 'url': url,
'vbr': int(vbr), 'vbr': int(vbr),
}) })
video_details['formats'] = video_formats
else:
# Fallback to flv
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 [{ return [{
'id': video_id, 'id': video_id,
'formats': video_formats,
'title': video_title, 'title': video_title,
'formats': video_formats,
}] }]