Browse Source

[nexx] Add support for free cdn (closes #16538)

master
Sergey M․ 6 years ago
parent
commit
06ea7bdd99
No known key found for this signature in database GPG Key ID: 2C393E0F18A9236D
1 changed files with 144 additions and 65 deletions
  1. 209
      youtube_dl/extractor/nexx.py

209
youtube_dl/extractor/nexx.py

@ -77,6 +77,21 @@ class NexxIE(InfoExtractor):
'timestamp': 1518614955,
'upload_date': '20180214',
},
}, {
# free cdn from http://www.spiegel.de/video/eifel-zoo-aufregung-um-ausgebrochene-raubtiere-video-99018031.html
'url': 'nexx:747:1533779',
'md5': '6bf6883912b82b7069fb86c2297e9893',
'info_dict': {
'id': '1533779',
'ext': 'mp4',
'title': 'Aufregung um ausgebrochene Raubtiere',
'alt_title': 'Eifel-Zoo',
'description': 'md5:f21375c91c74ad741dcb164c427999d2',
'thumbnail': r're:^https?://.*\.jpg$',
'duration': 111,
'timestamp': 1527874460,
'upload_date': '20180601',
},
}, {
'url': 'https://api.nexxcdn.com/v3/748/videos/byid/128907',
'only_matching': True,
@ -141,6 +156,127 @@ class NexxIE(InfoExtractor):
self._handle_error(result)
return result['result']
def _extract_free_formats(self, video, video_id):
stream_data = video['streamdata']
cdn = stream_data['cdnType']
assert cdn == 'free'
hash = video['general']['hash']
ps = compat_str(stream_data['originalDomain'])
if stream_data['applyFolderHierarchy'] == 1:
s = ('%04d' % int(video_id))[::-1]
ps += '/%s/%s' % (s[0:2], s[2:4])
ps += '/%s/%s_' % (video_id, hash)
formats = [{
'url': 'http://%s%s2500_var.mp4' % (stream_data['cdnPathHTTP'], ps),
'format_id': '%s-http' % cdn,
}]
def make_url(root, protocol):
t = 'http://' + root + ps
fd = stream_data['azureFileDistribution'].split(',')
cdn_provider = stream_data['cdnProvider']
def p0(p):
return '_%s' % int(p[0]) if stream_data['applyAzureStructure'] == 1 else ''
if cdn_provider == 'ak':
t += ','
for i in fd:
p = i.split(':')
t += p[1] + p0(p) + ','
t += '.mp4.csmil/master.m3u8'
elif cdn_provider == 'ce':
k = t.split('/')
h = k.pop()
t = '/'.join(k)
t += '/asset.ism/manifest.' + ('m3u8' if protocol == 'hls' else 'mpd') + '?dcp_ver=aos4&videostream='
for i in fd:
p = i.split(':')
a = '%s%s%s.mp4:%s' % (h, p[1], p0(p), int(p[0]) * 1000)
t += a + ','
t = t[:-1] + '&audiostream=' + a.split(':')[0]
return t
formats.extend(self._extract_mpd_formats(
make_url(stream_data['cdnPathDASH'], 'dash'), video_id,
mpd_id='%s-dash' % cdn, fatal=False))
formats.extend(self._extract_m3u8_formats(
make_url(stream_data['cdnPathHLS'], 'hls'), video_id, 'mp4',
entry_protocol='m3u8_native', m3u8_id='%s-hls' % cdn, fatal=False))
return formats
def _extract_azure_formats(self, video, video_id):
stream_data = video['streamdata']
cdn = stream_data['cdnType']
assert cdn == 'azure'
azure_locator = stream_data['azureLocator']
def get_cdn_shield_base(shield_type='', static=False):
for secure in ('', 's'):
cdn_shield = stream_data.get('cdnShield%sHTTP%s' % (shield_type, secure.upper()))
if cdn_shield:
return 'http%s://%s' % (secure, cdn_shield)
else:
if 'fb' in stream_data['azureAccount']:
prefix = 'df' if static else 'f'
else:
prefix = 'd' if static else 'p'
account = int(stream_data['azureAccount'].replace('nexxplayplus', '').replace('nexxplayfb', ''))
return 'http://nx-%s%02d.akamaized.net/' % (prefix, account)
language = video['general'].get('language_raw') or ''
azure_stream_base = get_cdn_shield_base()
is_ml = ',' in language
azure_manifest_url = '%s%s/%s_src%s.ism/Manifest' % (
azure_stream_base, azure_locator, video_id, ('_manifest' if is_ml else '')) + '%s'
protection_token = try_get(
video, lambda x: x['protectiondata']['token'], compat_str)
if protection_token:
azure_manifest_url += '?hdnts=%s' % protection_token
formats = self._extract_m3u8_formats(
azure_manifest_url % '(format=m3u8-aapl)',
video_id, 'mp4', 'm3u8_native',
m3u8_id='%s-hls' % cdn, fatal=False)
formats.extend(self._extract_mpd_formats(
azure_manifest_url % '(format=mpd-time-csf)',
video_id, mpd_id='%s-dash' % cdn, fatal=False))
formats.extend(self._extract_ism_formats(
azure_manifest_url % '', video_id, ism_id='%s-mss' % cdn, fatal=False))
azure_progressive_base = get_cdn_shield_base('Prog', True)
azure_file_distribution = stream_data.get('azureFileDistribution')
if azure_file_distribution:
fds = azure_file_distribution.split(',')
if fds:
for fd in fds:
ss = fd.split(':')
if len(ss) == 2:
tbr = int_or_none(ss[0])
if tbr:
f = {
'url': '%s%s/%s_src_%s_%d.mp4' % (
azure_progressive_base, azure_locator, video_id, ss[1], tbr),
'format_id': '%s-http-%d' % (cdn, tbr),
'tbr': tbr,
}
width_height = ss[1].split('x')
if len(width_height) == 2:
f.update({
'width': int_or_none(width_height[0]),
'height': int_or_none(width_height[1]),
})
formats.append(f)
return formats
def _real_extract(self, url):
mobj = re.match(self._VALID_URL, url)
domain_id = mobj.group('domain_id') or mobj.group('domain_id_s')
@ -220,72 +356,15 @@ class NexxIE(InfoExtractor):
general = video['general']
title = general['title']
stream_data = video['streamdata']
language = general.get('language_raw') or ''
# TODO: reverse more cdns
cdn = stream_data['cdnType']
assert cdn == 'azure'
cdn = video['streamdata']['cdnType']
azure_locator = stream_data['azureLocator']
def get_cdn_shield_base(shield_type='', static=False):
for secure in ('', 's'):
cdn_shield = stream_data.get('cdnShield%sHTTP%s' % (shield_type, secure.upper()))
if cdn_shield:
return 'http%s://%s' % (secure, cdn_shield)
else:
if 'fb' in stream_data['azureAccount']:
prefix = 'df' if static else 'f'
else:
prefix = 'd' if static else 'p'
account = int(stream_data['azureAccount'].replace('nexxplayplus', '').replace('nexxplayfb', ''))
return 'http://nx-%s%02d.akamaized.net/' % (prefix, account)
azure_stream_base = get_cdn_shield_base()
is_ml = ',' in language
azure_manifest_url = '%s%s/%s_src%s.ism/Manifest' % (
azure_stream_base, azure_locator, video_id, ('_manifest' if is_ml else '')) + '%s'
protection_token = try_get(
video, lambda x: x['protectiondata']['token'], compat_str)
if protection_token:
azure_manifest_url += '?hdnts=%s' % protection_token
formats = self._extract_m3u8_formats(
azure_manifest_url % '(format=m3u8-aapl)',
video_id, 'mp4', 'm3u8_native',
m3u8_id='%s-hls' % cdn, fatal=False)
formats.extend(self._extract_mpd_formats(
azure_manifest_url % '(format=mpd-time-csf)',
video_id, mpd_id='%s-dash' % cdn, fatal=False))
formats.extend(self._extract_ism_formats(
azure_manifest_url % '', video_id, ism_id='%s-mss' % cdn, fatal=False))
azure_progressive_base = get_cdn_shield_base('Prog', True)
azure_file_distribution = stream_data.get('azureFileDistribution')
if azure_file_distribution:
fds = azure_file_distribution.split(',')
if fds:
for fd in fds:
ss = fd.split(':')
if len(ss) == 2:
tbr = int_or_none(ss[0])
if tbr:
f = {
'url': '%s%s/%s_src_%s_%d.mp4' % (
azure_progressive_base, azure_locator, video_id, ss[1], tbr),
'format_id': '%s-http-%d' % (cdn, tbr),
'tbr': tbr,
}
width_height = ss[1].split('x')
if len(width_height) == 2:
f.update({
'width': int_or_none(width_height[0]),
'height': int_or_none(width_height[1]),
})
formats.append(f)
if cdn == 'azure':
formats = self._extract_azure_formats(video, video_id)
elif cdn == 'free':
formats = self._extract_free_formats(video, video_id)
else:
# TODO: reverse more cdns
assert False
self._sort_formats(formats)

Loading…
Cancel
Save