Skip to content

Commit 3c58f9e

Browse files
goggleremitamine
goggle
authored and
remitamine
committed
[srgssr] improve extraction
- extract subtitle - fix extraction for new videos - update srf download domains closes ytdl-org#14717 closes ytdl-org#14725 closes ytdl-org#27231 closes ytdl-org#28238
1 parent ef28e33 commit 3c58f9e

File tree

2 files changed

+144
-79
lines changed

2 files changed

+144
-79
lines changed

‎youtube_dl/extractor/rts.py

Lines changed: 10 additions & 5 deletions
Original file line numberDiff line numberDiff line change
@@ -6,11 +6,12 @@
66
from .srgssr import SRGSSRIE
77
from ..compat import compat_str
88
from ..utils import (
9+
determine_ext,
910
int_or_none,
1011
parse_duration,
1112
parse_iso8601,
1213
unescapeHTML,
13-
determine_ext,
14+
urljoin,
1415
)
1516

1617

@@ -21,7 +22,7 @@ class RTSIE(SRGSSRIE):
2122
_TESTS = [
2223
{
2324
'url': 'http://www.rts.ch/archives/tv/divers/3449373-les-enfants-terribles.html',
24-
'md5': 'ff7f8450a90cf58dacb64e29707b4a8e',
25+
'md5': '753b877968ad8afaeddccc374d4256a5',
2526
'info_dict': {
2627
'id': '3449373',
2728
'display_id': 'les-enfants-terribles',
@@ -35,6 +36,7 @@ class RTSIE(SRGSSRIE):
3536
'thumbnail': r're:^https?://.*\.image',
3637
'view_count': int,
3738
},
39+
'expected_warnings': ['Unable to download f4m manifest', 'Failed to download m3u8 information'],
3840
},
3941
{
4042
'url': 'http://www.rts.ch/emissions/passe-moi-les-jumelles/5624067-entre-ciel-et-mer.html',
@@ -63,11 +65,12 @@ class RTSIE(SRGSSRIE):
6365
# m3u8 download
6466
'skip_download': True,
6567
},
68+
'expected_warnings': ['Unable to download f4m manifest', 'Failed to download m3u8 information'],
6669
'skip': 'Blocked outside Switzerland',
6770
},
6871
{
6972
'url': 'http://www.rts.ch/video/info/journal-continu/5745356-londres-cachee-par-un-epais-smog.html',
70-
'md5': '1bae984fe7b1f78e94abc74e802ed99f',
73+
'md5': '9bb06503773c07ce83d3cbd793cebb91',
7174
'info_dict': {
7275
'id': '5745356',
7376
'display_id': 'londres-cachee-par-un-epais-smog',
@@ -81,6 +84,7 @@ class RTSIE(SRGSSRIE):
8184
'thumbnail': r're:^https?://.*\.image',
8285
'view_count': int,
8386
},
87+
'expected_warnings': ['Unable to download f4m manifest', 'Failed to download m3u8 information'],
8488
},
8589
{
8690
'url': 'http://www.rts.ch/audio/couleur3/programmes/la-belle-video-de-stephane-laurenceau/5706148-urban-hippie-de-damien-krisl-03-04-2014.html',
@@ -160,7 +164,7 @@ def download_json(internal_id):
160164
media_type = 'video' if 'video' in all_info else 'audio'
161165

162166
# check for errors
163-
self.get_media_data('rts', media_type, media_id)
167+
self._get_media_data('rts', media_type, media_id)
164168

165169
info = all_info['video']['JSONinfo'] if 'video' in all_info else all_info['audio']
166170

@@ -194,6 +198,7 @@ def extract_bitrate(url):
194198
'tbr': extract_bitrate(format_url),
195199
})
196200

201+
download_base = 'http://rtsww%s-d.rts.ch/' % ('-a' if media_type == 'audio' else '')
197202
for media in info.get('media', []):
198203
media_url = media.get('url')
199204
if not media_url or re.match(r'https?://', media_url):
@@ -205,7 +210,7 @@ def extract_bitrate(url):
205210
format_id += '-%dk' % rate
206211
formats.append({
207212
'format_id': format_id,
208-
'url': 'http://download-video.rts.ch/' + media_url,
213+
'url': urljoin(download_base, media_url),
209214
'tbr': rate or extract_bitrate(media_url),
210215
})
211216

‎youtube_dl/extractor/srgssr.py

Lines changed: 134 additions & 74 deletions
Original file line numberDiff line numberDiff line change
@@ -4,16 +4,32 @@
44
import re
55

66
from .common import InfoExtractor
7-
from ..compat import compat_urllib_parse_urlparse
87
from ..utils import (
98
ExtractorError,
9+
float_or_none,
10+
int_or_none,
1011
parse_iso8601,
1112
qualities,
13+
try_get,
1214
)
1315

1416

1517
class SRGSSRIE(InfoExtractor):
16-
_VALID_URL = r'(?:https?://tp\.srgssr\.ch/p(?:/[^/]+)+\?urn=urn|srgssr):(?P<bu>srf|rts|rsi|rtr|swi):(?:[^:]+:)?(?P<type>video|audio):(?P<id>[0-9a-f\-]{36}|\d+)'
18+
_VALID_URL = r'''(?x)
19+
(?:
20+
https?://tp\.srgssr\.ch/p(?:/[^/]+)+\?urn=urn|
21+
srgssr
22+
):
23+
(?P<bu>
24+
srf|rts|rsi|rtr|swi
25+
):(?:[^:]+:)?
26+
(?P<type>
27+
video|audio
28+
):
29+
(?P<id>
30+
[0-9a-f\-]{36}|\d+
31+
)
32+
'''
1733
_GEO_BYPASS = False
1834
_GEO_COUNTRIES = ['CH']
1935

@@ -25,25 +41,39 @@ class SRGSSRIE(InfoExtractor):
2541
'LEGAL': 'The video cannot be transmitted for legal reasons.',
2642
'STARTDATE': 'This video is not yet available. Please try again later.',
2743
}
44+
_DEFAULT_LANGUAGE_CODES = {
45+
'srf': 'de',
46+
'rts': 'fr',
47+
'rsi': 'it',
48+
'rtr': 'rm',
49+
'swi': 'en',
50+
}
2851

2952
def _get_tokenized_src(self, url, video_id, format_id):
30-
sp = compat_urllib_parse_urlparse(url).path.split('/')
3153
token = self._download_json(
32-
'http://tp.srgssr.ch/akahd/token?acl=/%s/%s/*' % (sp[1], sp[2]),
54+
'http://tp.srgssr.ch/akahd/token?acl=*',
3355
video_id, 'Downloading %s token' % format_id, fatal=False) or {}
34-
auth_params = token.get('token', {}).get('authparams')
56+
auth_params = try_get(token, lambda x: x['token']['authparams'])
3557
if auth_params:
36-
url += '?' + auth_params
58+
url += ('?' if '?' not in url else '&') + auth_params
3759
return url
3860

39-
def get_media_data(self, bu, media_type, media_id):
40-
media_data = self._download_json(
41-
'http://il.srgssr.ch/integrationlayer/1.0/ue/%s/%s/play/%s.json' % (bu, media_type, media_id),
42-
media_id)[media_type.capitalize()]
43-
44-
if media_data.get('block') and media_data['block'] in self._ERRORS:
45-
message = self._ERRORS[media_data['block']]
46-
if media_data['block'] == 'GEOBLOCK':
61+
def _get_media_data(self, bu, media_type, media_id):
62+
query = {'onlyChapters': True} if media_type == 'video' else {}
63+
full_media_data = self._download_json(
64+
'https://il.srgssr.ch/integrationlayer/2.0/%s/mediaComposition/%s/%s.json'
65+
% (bu, media_type, media_id),
66+
media_id, query=query)['chapterList']
67+
try:
68+
media_data = next(
69+
x for x in full_media_data if x.get('id') == media_id)
70+
except StopIteration:
71+
raise ExtractorError('No media information found')
72+
73+
block_reason = media_data.get('blockReason')
74+
if block_reason and block_reason in self._ERRORS:
75+
message = self._ERRORS[block_reason]
76+
if block_reason == 'GEOBLOCK':
4777
self.raise_geo_restricted(
4878
msg=message, countries=self._GEO_COUNTRIES)
4979
raise ExtractorError(
@@ -53,53 +83,75 @@ def get_media_data(self, bu, media_type, media_id):
5383

5484
def _real_extract(self, url):
5585
bu, media_type, media_id = re.match(self._VALID_URL, url).groups()
86+
media_data = self._get_media_data(bu, media_type, media_id)
87+
title = media_data['title']
5688

57-
media_data = self.get_media_data(bu, media_type, media_id)
58-
59-
metadata = media_data['AssetMetadatas']['AssetMetadata'][0]
60-
title = metadata['title']
61-
description = metadata.get('description')
62-
created_date = media_data.get('createdDate') or metadata.get('createdDate')
63-
timestamp = parse_iso8601(created_date)
64-
65-
thumbnails = [{
66-
'id': image.get('id'),
67-
'url': image['url'],
68-
} for image in media_data.get('Image', {}).get('ImageRepresentations', {}).get('ImageRepresentation', [])]
69-
70-
preference = qualities(['LQ', 'MQ', 'SD', 'HQ', 'HD'])
7189
formats = []
72-
for source in media_data.get('Playlists', {}).get('Playlist', []) + media_data.get('Downloads', {}).get('Download', []):
73-
protocol = source.get('@protocol')
74-
for asset in source['url']:
75-
asset_url = asset['text']
76-
quality = asset['@quality']
77-
format_id = '%s-%s' % (protocol, quality)
78-
if protocol.startswith('HTTP-HDS') or protocol.startswith('HTTP-HLS'):
79-
asset_url = self._get_tokenized_src(asset_url, media_id, format_id)
80-
if protocol.startswith('HTTP-HDS'):
81-
formats.extend(self._extract_f4m_formats(
82-
asset_url + ('?' if '?' not in asset_url else '&') + 'hdcore=3.4.0',
83-
media_id, f4m_id=format_id, fatal=False))
84-
elif protocol.startswith('HTTP-HLS'):
85-
formats.extend(self._extract_m3u8_formats(
86-
asset_url, media_id, 'mp4', 'm3u8_native',
87-
m3u8_id=format_id, fatal=False))
88-
else:
89-
formats.append({
90-
'format_id': format_id,
91-
'url': asset_url,
92-
'preference': preference(quality),
93-
'ext': 'flv' if protocol == 'RTMP' else None,
94-
})
90+
q = qualities(['SD', 'HD'])
91+
for source in (media_data.get('resourceList') or []):
92+
format_url = source.get('url')
93+
if not format_url:
94+
continue
95+
protocol = source.get('protocol')
96+
quality = source.get('quality')
97+
format_id = []
98+
for e in (protocol, source.get('encoding'), quality):
99+
if e:
100+
format_id.append(e)
101+
format_id = '-'.join(format_id)
102+
103+
if protocol in ('HDS', 'HLS'):
104+
if source.get('tokenType') == 'AKAMAI':
105+
format_url = self._get_tokenized_src(
106+
format_url, media_id, format_id)
107+
formats.extend(self._extract_akamai_formats(
108+
format_url, media_id))
109+
elif protocol == 'HLS':
110+
formats.extend(self._extract_m3u8_formats(
111+
format_url, media_id, 'mp4', 'm3u8_native',
112+
m3u8_id=format_id, fatal=False))
113+
elif protocol in ('HTTP', 'HTTPS'):
114+
formats.append({
115+
'format_id': format_id,
116+
'url': format_url,
117+
'quality': q(quality),
118+
})
119+
120+
# This is needed because for audio medias the podcast url is usually
121+
# always included, even if is only an audio segment and not the
122+
# whole episode.
123+
if int_or_none(media_data.get('position')) == 0:
124+
for p in ('S', 'H'):
125+
podcast_url = media_data.get('podcast%sdUrl' % p)
126+
if not podcast_url:
127+
continue
128+
quality = p + 'D'
129+
formats.append({
130+
'format_id': 'PODCAST-' + quality,
131+
'url': podcast_url,
132+
'quality': q(quality),
133+
})
95134
self._sort_formats(formats)
96135

136+
subtitles = {}
137+
if media_type == 'video':
138+
for sub in (media_data.get('subtitleList') or []):
139+
sub_url = sub.get('url')
140+
if not sub_url:
141+
continue
142+
lang = sub.get('locale') or self._DEFAULT_LANGUAGE_CODES[bu]
143+
subtitles.setdefault(lang, []).append({
144+
'url': sub_url,
145+
})
146+
97147
return {
98148
'id': media_id,
99149
'title': title,
100-
'description': description,
101-
'timestamp': timestamp,
102-
'thumbnails': thumbnails,
150+
'description': media_data.get('description'),
151+
'timestamp': parse_iso8601(media_data.get('date')),
152+
'thumbnail': media_data.get('imageUrl'),
153+
'duration': float_or_none(media_data.get('duration'), 1000),
154+
'subtitles': subtitles,
103155
'formats': formats,
104156
}
105157

@@ -119,34 +171,26 @@ class SRGSSRPlayIE(InfoExtractor):
119171

120172
_TESTS = [{
121173
'url': 'http://www.srf.ch/play/tv/10vor10/video/snowden-beantragt-asyl-in-russland?id=28e1a57d-5b76-4399-8ab3-9097f071e6c5',
122-
'md5': 'da6b5b3ac9fa4761a942331cef20fcb3',
174+
'md5': '6db2226ba97f62ad42ce09783680046c',
123175
'info_dict': {
124176
'id': '28e1a57d-5b76-4399-8ab3-9097f071e6c5',
125177
'ext': 'mp4',
126178
'upload_date': '20130701',
127179
'title': 'Snowden beantragt Asyl in Russland',
128-
'timestamp': 1372713995,
129-
}
130-
}, {
131-
# No Speichern (Save) button
132-
'url': 'http://www.srf.ch/play/tv/top-gear/video/jaguar-xk120-shadow-und-tornado-dampflokomotive?id=677f5829-e473-4823-ac83-a1087fe97faa',
133-
'md5': '0a274ce38fda48c53c01890651985bc6',
134-
'info_dict': {
135-
'id': '677f5829-e473-4823-ac83-a1087fe97faa',
136-
'ext': 'flv',
137-
'upload_date': '20130710',
138-
'title': 'Jaguar XK120, Shadow und Tornado-Dampflokomotive',
139-
'description': 'md5:88604432b60d5a38787f152dec89cd56',
140-
'timestamp': 1373493600,
180+
'timestamp': 1372708215,
181+
'duration': 113.827,
182+
'thumbnail': r're:^https?://.*1383719781\.png$',
141183
},
184+
'expected_warnings': ['Unable to download f4m manifest'],
142185
}, {
143186
'url': 'http://www.rtr.ch/play/radio/actualitad/audio/saira-tujetsch-tuttina-cuntinuar-cun-sedrun-muster-turissem?id=63cb0778-27f8-49af-9284-8c7a8c6d15fc',
144187
'info_dict': {
145188
'id': '63cb0778-27f8-49af-9284-8c7a8c6d15fc',
146189
'ext': 'mp3',
147190
'upload_date': '20151013',
148191
'title': 'Saira: Tujetsch - tuttina cuntinuar cun Sedrun Mustér Turissem',
149-
'timestamp': 1444750398,
192+
'timestamp': 1444709160,
193+
'duration': 336.816,
150194
},
151195
'params': {
152196
# rtmp download
@@ -159,19 +203,32 @@ class SRGSSRPlayIE(InfoExtractor):
159203
'id': '6348260',
160204
'display_id': '6348260',
161205
'ext': 'mp4',
162-
'duration': 1796,
206+
'duration': 1796.76,
163207
'title': 'Le 19h30',
164-
'description': '',
165-
'uploader': '19h30',
166208
'upload_date': '20141201',
167209
'timestamp': 1417458600,
168210
'thumbnail': r're:^https?://.*\.image',
169-
'view_count': int,
170211
},
171212
'params': {
172213
# m3u8 download
173214
'skip_download': True,
174215
}
216+
}, {
217+
'url': 'http://play.swissinfo.ch/play/tv/business/video/why-people-were-against-tax-reforms?id=42960270',
218+
'info_dict': {
219+
'id': '42960270',
220+
'ext': 'mp4',
221+
'title': 'Why people were against tax reforms',
222+
'description': 'md5:7ac442c558e9630e947427469c4b824d',
223+
'duration': 94.0,
224+
'upload_date': '20170215',
225+
'timestamp': 1487173560,
226+
'thumbnail': r're:https?://www\.swissinfo\.ch/srgscalableimage/42961964',
227+
'subtitles': 'count:9',
228+
},
229+
'params': {
230+
'skip_download': True,
231+
}
175232
}, {
176233
'url': 'https://www.srf.ch/play/tv/popupvideoplayer?id=c4dba0ca-e75b-43b2-a34f-f708a4932e01',
177234
'only_matching': True,
@@ -181,12 +238,15 @@ class SRGSSRPlayIE(InfoExtractor):
181238
}, {
182239
'url': 'https://www.rts.ch/play/tv/19h30/video/le-19h30?urn=urn:rts:video:6348260',
183240
'only_matching': True,
241+
}, {
242+
# audio segment, has podcastSdUrl of the full episode
243+
'url': 'https://www.srf.ch/play/radio/popupaudioplayer?id=50b20dc8-f05b-4972-bf03-e438ff2833eb',
244+
'only_matching': True,
184245
}]
185246

186247
def _real_extract(self, url):
187248
mobj = re.match(self._VALID_URL, url)
188249
bu = mobj.group('bu')
189250
media_type = mobj.group('type') or mobj.group('type_2')
190251
media_id = mobj.group('id')
191-
# other info can be extracted from url + '&layout=json'
192252
return self.url_result('srgssr:%s:%s:%s' % (bu[:3], media_type, media_id), 'SRGSSR')

0 commit comments

Comments
 (0)