[sendia] fix info extraction

This commit is contained in:
Remita Amine 2016-07-05 13:37:46 +01:00
parent 77082c7b9e
commit 94a5cff91d
1 changed files with 25 additions and 69 deletions

View File

@ -27,7 +27,8 @@ class SandiaIE(InfoExtractor):
'ext': 'mp4', 'ext': 'mp4',
'title': 'Xyce Software Training - Section 1', 'title': 'Xyce Software Training - Section 1',
'description': 're:(?s)SAND Number: SAND 2013-7800.{200,}', 'description': 're:(?s)SAND Number: SAND 2013-7800.{200,}',
'upload_date': '20120904', 'upload_date': '20120409',
'timestamp': 1333983600,
'duration': 7794, 'duration': 7794,
} }
} }
@ -35,81 +36,36 @@ class SandiaIE(InfoExtractor):
def _real_extract(self, url): def _real_extract(self, url):
video_id = self._match_id(url) video_id = self._match_id(url)
req = sanitized_Request(url) presentation_data = self._download_json(
req.add_header('Cookie', 'MediasitePlayerCaps=ClientPlugins=4') 'http://digitalops.sandia.gov/Mediasite/PlayerService/PlayerService.svc/json/GetPlayerOptions',
webpage = self._download_webpage(req, video_id) video_id, data=json.dumps({
'getPlayerOptionsRequest': {
'ResourceId': video_id,
'QueryString': '',
}
}), headers={
'Content-Type': 'application/json; charset=utf-8',
})['d']['Presentation']
js_path = self._search_regex( title = presentation_data['Title']
r'<script type="text/javascript" src="(/Mediasite/FileServer/Presentation/[^"]+)"',
webpage, 'JS code URL')
js_url = compat_urlparse.urljoin(url, js_path)
js_code = self._download_webpage(
js_url, video_id, note='Downloading player')
def extract_str(key, **args):
return self._search_regex(
r'Mediasite\.PlaybackManifest\.%s\s*=\s*(.+);\s*?\n' % re.escape(key),
js_code, key, **args)
def extract_data(key, **args):
data_json = extract_str(key, **args)
if data_json is None:
return data_json
return self._parse_json(
data_json, video_id, transform_source=js_to_json)
formats = [] formats = []
for i in itertools.count(): for stream in presentation_data.get('Streams', []):
fd = extract_data('VideoUrls[%d]' % i, default=None) for fd in stream.get('VideoUrls', []):
if fd is None: formats.append({
break 'format_id': fd['MediaType'],
formats.append({ 'format_note': fd['MimeType'].partition('/')[2],
'format_id': '%s' % i, 'ext': mimetype2ext(fd['MimeType']),
'format_note': fd['MimeType'].partition('/')[2], 'url': fd['Location'],
'ext': mimetype2ext(fd['MimeType']), 'protocol': 'f4m' if fd['MimeType'] == 'video/x-mp4-fragmented' else None,
'url': fd['Location'], })
'protocol': 'f4m' if fd['MimeType'] == 'video/x-mp4-fragmented' else None,
})
self._sort_formats(formats) self._sort_formats(formats)
slide_baseurl = compat_urlparse.urljoin(
url, extract_data('SlideBaseUrl'))
slide_template = slide_baseurl + re.sub(
r'\{0:D?([0-9+])\}', r'%0\1d', extract_data('SlideImageFileNameTemplate'))
slides = []
last_slide_time = 0
for i in itertools.count(1):
sd = extract_str('Slides[%d]' % i, default=None)
if sd is None:
break
timestamp = int_or_none(self._search_regex(
r'^Mediasite\.PlaybackManifest\.CreateSlide\("[^"]*"\s*,\s*([0-9]+),',
sd, 'slide %s timestamp' % i, fatal=False))
slides.append({
'url': slide_template % i,
'duration': timestamp - last_slide_time,
})
last_slide_time = timestamp
formats.append({
'format_id': 'slides',
'protocol': 'slideshow',
'url': json.dumps(slides),
'preference': -10000, # Downloader not yet written
})
self._sort_formats(formats)
title = extract_data('Title')
description = extract_data('Description', fatal=False)
duration = int_or_none(extract_data(
'Duration', fatal=False), scale=1000)
upload_date = unified_strdate(extract_data('AirDate', fatal=False))
return { return {
'id': video_id, 'id': video_id,
'title': title, 'title': title,
'description': description, 'description': presentation_data.get('Description'),
'formats': formats, 'formats': formats,
'upload_date': upload_date, 'timestamp': int_or_none(presentation_data.get('UnixTime'), 1000),
'duration': duration, 'duration': int_or_none(presentation_data.get('Duration'), 1000),
} }