[rai] Improve extraction (closes #11790)

* Fix georestriction detection
* Detect live streams
+ Extract relinker metadata
* Improve ContentItem detection
+ Extract series metadata
* Fix tests
This commit is contained in:
Sergey M․ 2017-04-02 02:14:42 +07:00
parent 51342717cd
commit b8d8cced9b
No known key found for this signature in database
GPG key ID: 2C393E0F18A9236D

View file

@ -1,23 +1,40 @@
from __future__ import unicode_literals from __future__ import unicode_literals
import re
from .common import InfoExtractor from .common import InfoExtractor
from ..compat import compat_urlparse from ..compat import (
compat_urlparse,
compat_str,
)
from ..utils import ( from ..utils import (
ExtractorError, ExtractorError,
determine_ext, determine_ext,
find_xpath_attr, find_xpath_attr,
fix_xml_ampersands, fix_xml_ampersands,
GeoRestrictedError,
int_or_none, int_or_none,
parse_duration, parse_duration,
strip_or_none,
try_get,
unified_strdate, unified_strdate,
unified_timestamp,
update_url_query, update_url_query,
urljoin,
xpath_text, xpath_text,
) )
class RaiBaseIE(InfoExtractor): class RaiBaseIE(InfoExtractor):
def _extract_relinker_formats(self, relinker_url, video_id): _UUID_RE = r'[\da-f]{8}-[\da-f]{4}-[\da-f]{4}-[\da-f]{4}-[\da-f]{12}'
_GEO_COUNTRIES = ['IT']
_GEO_BYPASS = False
def _extract_relinker_info(self, relinker_url, video_id):
formats = [] formats = []
geoprotection = None
is_live = None
duration = None
for platform in ('mon', 'flash', 'native'): for platform in ('mon', 'flash', 'native'):
relinker = self._download_xml( relinker = self._download_xml(
@ -27,9 +44,27 @@ class RaiBaseIE(InfoExtractor):
query={'output': 45, 'pl': platform}, query={'output': 45, 'pl': platform},
headers=self.geo_verification_headers()) headers=self.geo_verification_headers())
media_url = find_xpath_attr(relinker, './url', 'type', 'content').text if not geoprotection:
geoprotection = xpath_text(
relinker, './geoprotection', default=None) == 'Y'
if not is_live:
is_live = xpath_text(
relinker, './is_live', default=None) == 'Y'
if not duration:
duration = parse_duration(xpath_text(
relinker, './duration', default=None))
url_elem = find_xpath_attr(relinker, './url', 'type', 'content')
if url_elem is None:
continue
media_url = url_elem.text
# This does not imply geo restriction (e.g.
# http://www.raisport.rai.it/dl/raiSport/media/rassegna-stampa-04a9f4bd-b563-40cf-82a6-aad3529cb4a9.html)
if media_url == 'http://download.rai.it/video_no_available.mp4': if media_url == 'http://download.rai.it/video_no_available.mp4':
self.raise_geo_restricted() continue
ext = determine_ext(media_url) ext = determine_ext(media_url)
if (ext == 'm3u8' and platform != 'mon') or (ext == 'f4m' and platform != 'flash'): if (ext == 'm3u8' and platform != 'mon') or (ext == 'f4m' and platform != 'flash'):
@ -53,11 +88,18 @@ class RaiBaseIE(InfoExtractor):
'format_id': 'http-%d' % bitrate if bitrate > 0 else 'http', 'format_id': 'http-%d' % bitrate if bitrate > 0 else 'http',
}) })
return formats if not formats and geoprotection is True:
self.raise_geo_restricted(countries=self._GEO_COUNTRIES)
return dict((k, v) for k, v in {
'is_live': is_live,
'duration': duration,
'formats': formats,
}.items() if v is not None)
class RaiPlayIE(RaiBaseIE): class RaiPlayIE(RaiBaseIE):
_VALID_URL = r'https?://(?:www\.)?raiplay\.it/.+?-(?P<id>[\da-f]{8}-[\da-f]{4}-[\da-f]{4}-[\da-f]{4}-[\da-f]{12})\.html' _VALID_URL = r'(?P<url>https?://(?:www\.)?raiplay\.it/.+?-(?P<id>%s)\.html)' % RaiBaseIE._UUID_RE
_TESTS = [{ _TESTS = [{
'url': 'http://www.raiplay.it/video/2016/10/La-Casa-Bianca-e06118bb-59a9-4636-b914-498e4cfd2c66.html?source=twitter', 'url': 'http://www.raiplay.it/video/2016/10/La-Casa-Bianca-e06118bb-59a9-4636-b914-498e4cfd2c66.html?source=twitter',
'md5': '340aa3b7afb54bfd14a8c11786450d76', 'md5': '340aa3b7afb54bfd14a8c11786450d76',
@ -65,110 +107,130 @@ class RaiPlayIE(RaiBaseIE):
'id': 'e06118bb-59a9-4636-b914-498e4cfd2c66', 'id': 'e06118bb-59a9-4636-b914-498e4cfd2c66',
'ext': 'mp4', 'ext': 'mp4',
'title': 'La Casa Bianca', 'title': 'La Casa Bianca',
'alt_title': 'S2016 - Puntata del 23/10/2016',
'description': 'md5:a09d45890850458077d1f68bb036e0a5',
'thumbnail': r're:^https?://.*\.jpg$', 'thumbnail': r're:^https?://.*\.jpg$',
'uploader': r're:^Rai.+', 'uploader': 'Rai 3',
'description': 're:^[A-Za-z]+' 'creator': 'Rai 3',
} 'duration': 3278,
}, { 'timestamp': 1477764300,
'url': 'http://www.raiplay.it/video/2016/11/gazebotraindesi-efebe701-969c-4593-92f3-285f0d1ce750.html?', 'upload_date': '20161029',
'md5': 'ed4da3d70ccf8129a33ab16b34d20ab8', 'series': 'La Casa Bianca',
'info_dict': { 'season': '2016',
'id': 'efebe701-969c-4593-92f3-285f0d1ce750', },
'ext': 'mp4',
'title': 'Gazebo - #gazebotraindesi',
'thumbnail': r're:^https?://.*\.png$',
'uploader': r're:^Rai.+',
'description': r're:^[A-Za-z]+'
}
}, { }, {
'url': 'http://www.raiplay.it/video/2014/04/Report-del-07042014-cb27157f-9dd0-4aee-b788-b1f67643a391.html', 'url': 'http://www.raiplay.it/video/2014/04/Report-del-07042014-cb27157f-9dd0-4aee-b788-b1f67643a391.html',
'md5': '8970abf8caf8aef4696e7b1f2adfc696', 'md5': '8970abf8caf8aef4696e7b1f2adfc696',
'info_dict': { 'info_dict': {
'id': 'cb27157f-9dd0-4aee-b788-b1f67643a391', 'id': 'cb27157f-9dd0-4aee-b788-b1f67643a391',
'ext': 'mp4', 'ext': 'mp4',
'title': 'Report - Report del 07/04/2014', 'title': 'Report del 07/04/2014',
'alt_title': 'S2013/14 - Puntata del 07/04/2014',
'description': 'md5:f27c544694cacb46a078db84ec35d2d9',
'thumbnail': r're:^https?://.*\.jpg$', 'thumbnail': r're:^https?://.*\.jpg$',
'uploader': r're:^Rai.+', 'uploader': 'Rai 5',
'description': r're:^[A-Za-z]+' 'creator': 'Rai 5',
} 'duration': 6160,
'series': 'Report',
'season_number': 5,
'season': '2013/14',
},
'params': {
'skip_download': True,
},
}, {
'url': 'http://www.raiplay.it/video/2016/11/gazebotraindesi-efebe701-969c-4593-92f3-285f0d1ce750.html?',
'only_matching': True,
}] }]
_RESOLUTION = '600x400'
def _real_extract(self, url): def _real_extract(self, url):
video_id = self._match_id(url) mobj = re.match(self._VALID_URL, url)
url, video_id = mobj.group('url', 'id')
# remove query and fragment part from url media = self._download_json(
canonical_url = compat_urlparse.urljoin(url, compat_urlparse.urlparse(url).path) '%s?json' % url, video_id, 'Downloading video JSON')
webpage = self._download_webpage(canonical_url, video_id)
media = self._download_json('%s?json' % canonical_url, title = media['name']
video_id, 'Downloading video JSON')
video = media['video']
relinker_info = self._extract_relinker_info(video['contentUrl'], video_id)
self._sort_formats(relinker_info['formats'])
thumbnails = [] thumbnails = []
if 'images' in media: if 'images' in media:
for _, value in media.get('images').items(): for _, value in media.get('images').items():
if value: if value:
thumbnails.append({ thumbnails.append({
'url': value.replace('[RESOLUTION]', self._RESOLUTION) 'url': value.replace('[RESOLUTION]', '600x400')
}) })
if 'video' not in media: timestamp = unified_timestamp(try_get(
raise ExtractorError('No video found') media, lambda x: x['availabilities'][0]['start'], compat_str))
video = media.get('video') info = {
duration = parse_duration(video.get('duration')),
formats = self._extract_relinker_formats(video.get('contentUrl'), video_id)
self._sort_formats(formats)
return {
'id': video_id, 'id': video_id,
'title': self._og_search_title(webpage).replace(' - video - RaiPlay', ''), 'title': title,
'description': self._og_search_description(webpage), 'alt_title': media.get('subtitle'),
'description': media.get('description'),
'uploader': media.get('channel'), 'uploader': media.get('channel'),
'duration': duration, 'creator': media.get('editor'),
'duration': parse_duration(video.get('duration')),
'timestamp': timestamp,
'thumbnails': thumbnails, 'thumbnails': thumbnails,
'formats': formats 'series': try_get(
media, lambda x: x['isPartOf']['name'], compat_str),
'season_number': int_or_none(try_get(
media, lambda x: x['isPartOf']['numeroStagioni'])),
'season': media.get('stagione') or None,
} }
info.update(relinker_info)
return info
class RaiIE(RaiBaseIE): class RaiIE(RaiBaseIE):
_VALID_URL = r'https?://.+\.(?:rai|rainews)\.it/dl/.+?-(?P<id>[\da-f]{8}-[\da-f]{4}-[\da-f]{4}-[\da-f]{4}-[\da-f]{12})(?:-.+?)?\.html' _VALID_URL = r'https?://[^/]+\.(?:rai\.(?:it|tv)|rainews\.it)/dl/.+?-(?P<id>%s)(?:-.+?)?\.html' % RaiBaseIE._UUID_RE
_TESTS = [{ _TESTS = [{
# subdomain test case # var uniquename = "ContentItem-..."
# data-id="ContentItem-..."
'url': 'http://www.raisport.rai.it/dl/raiSport/media/rassegna-stampa-04a9f4bd-b563-40cf-82a6-aad3529cb4a9.html', 'url': 'http://www.raisport.rai.it/dl/raiSport/media/rassegna-stampa-04a9f4bd-b563-40cf-82a6-aad3529cb4a9.html',
'info_dict': { 'info_dict': {
'id': '04a9f4bd-b563-40cf-82a6-aad3529cb4a9', 'id': '04a9f4bd-b563-40cf-82a6-aad3529cb4a9',
'ext': 'mp4', 'ext': 'mp4',
'title': 'TG PRIMO TEMPO', 'title': 'TG PRIMO TEMPO',
'upload_date': '20140612', 'thumbnail': r're:^https?://.*\.jpg$',
'duration': 1758, 'duration': 1758,
'thumbnail': r're:^https?://.*\.jpg$' 'upload_date': '20140612',
} }
}, { }, {
# rainews test case # with ContentItem in many metas
'url': 'http://www.rainews.it/dl/rainews/media/Weekend-al-cinema-da-Hollywood-arriva-il-thriller-di-Tate-Taylor-La-ragazza-del-treno-1632c009-c843-4836-bb65-80c33084a64b.html', 'url': 'http://www.rainews.it/dl/rainews/media/Weekend-al-cinema-da-Hollywood-arriva-il-thriller-di-Tate-Taylor-La-ragazza-del-treno-1632c009-c843-4836-bb65-80c33084a64b.html',
'info_dict': { 'info_dict': {
'id': '1632c009-c843-4836-bb65-80c33084a64b', 'id': '1632c009-c843-4836-bb65-80c33084a64b',
'ext': 'mp4', 'ext': 'mp4',
'title': 'Weekend al cinema, da Hollywood arriva il thriller di Tate Taylor \"La ragazza del treno\" ', 'title': 'Weekend al cinema, da Hollywood arriva il thriller di Tate Taylor "La ragazza del treno"',
'upload_date': '20161103', 'description': 'I film in uscita questa settimana.',
'thumbnail': r're:^https?://.*\.png$', 'thumbnail': r're:^https?://.*\.png$',
'description': r're:^[A-Za-z]+' 'duration': 833,
'upload_date': '20161103',
} }
}, { }, {
# with media information # with ContentItem in og:url
'url': 'http://www.rai.it/dl/RaiTV/programmi/media/ContentItem-efb17665-691c-45d5-a60c-5301333cbb0c.html', 'url': 'http://www.rai.it/dl/RaiTV/programmi/media/ContentItem-efb17665-691c-45d5-a60c-5301333cbb0c.html',
'md5': '11959b4e44fa74de47011b5799490adf', 'md5': '11959b4e44fa74de47011b5799490adf',
'info_dict': { 'info_dict': {
'id': 'efb17665-691c-45d5-a60c-5301333cbb0c', 'id': 'efb17665-691c-45d5-a60c-5301333cbb0c',
'ext': 'mp4', 'ext': 'mp4',
'title': 'TG1 ore 20:00 del 03/11/2016', 'title': 'TG1 ore 20:00 del 03/11/2016',
'description': 'TG1 edizione integrale ore 20:00 del giorno 03/11/2016',
'thumbnail': r're:^https?://.*\.jpg$', 'thumbnail': r're:^https?://.*\.jpg$',
'duration': 2214,
'upload_date': '20161103', 'upload_date': '20161103',
'description': r're:^[A-Za-z]+'
} }
}, { }, {
# drawMediaRaiTV test case # drawMediaRaiTV(...)
'url': 'http://www.report.rai.it/dl/Report/puntata/ContentItem-0c7a664b-d0f4-4b2c-8835-3f82e46f433e.html', 'url': 'http://www.report.rai.it/dl/Report/puntata/ContentItem-0c7a664b-d0f4-4b2c-8835-3f82e46f433e.html',
'md5': '2dd727e61114e1ee9c47f0da6914e178', 'md5': '2dd727e61114e1ee9c47f0da6914e178',
'info_dict': { 'info_dict': {
@ -176,83 +238,67 @@ class RaiIE(RaiBaseIE):
'ext': 'mp4', 'ext': 'mp4',
'title': 'Il pacco', 'title': 'Il pacco',
'description': 'md5:4b1afae1364115ce5d78ed83cd2e5b3a', 'description': 'md5:4b1afae1364115ce5d78ed83cd2e5b3a',
'thumbnail': r're:^https?://.*\.jpg$',
'upload_date': '20141221', 'upload_date': '20141221',
}, },
}, { }, {
# Direct relinker URL # initEdizione('ContentItem-...'
'url': 'http://www.tg1.rai.it/dl/tg1/2010/edizioni/ContentSet-9b6e0cba-4bef-4aef-8cf0-9f7f665b7dfb-tg1.html?item=undefined',
'info_dict': {
'id': 'c2187016-8484-4e3a-8ac8-35e475b07303',
'ext': 'mp4',
'title': r're:TG1 ore \d{2}:\d{2} del \d{2}/\d{2}/\d{4}',
'duration': 2274,
'upload_date': '20170401',
},
'skip': 'Changes daily',
}, {
# HDS live stream with only relinker URL
'url': 'http://www.rai.tv/dl/RaiTV/dirette/PublishingBlock-1912dbbf-3f96-44c3-b4cf-523681fbacbc.html?channel=EuroNews', 'url': 'http://www.rai.tv/dl/RaiTV/dirette/PublishingBlock-1912dbbf-3f96-44c3-b4cf-523681fbacbc.html?channel=EuroNews',
# HDS live stream, MD5 is unstable
'info_dict': { 'info_dict': {
'id': '1912dbbf-3f96-44c3-b4cf-523681fbacbc', 'id': '1912dbbf-3f96-44c3-b4cf-523681fbacbc',
'ext': 'flv', 'ext': 'flv',
'title': 'EuroNews', 'title': 'EuroNews',
}, },
}, { 'params': {
# Embedded content item ID 'skip_download': True,
'url': 'http://www.tg1.rai.it/dl/tg1/2010/edizioni/ContentSet-9b6e0cba-4bef-4aef-8cf0-9f7f665b7dfb-tg1.html?item=undefined',
'info_dict': {
'id': 'd80d4b70-3812-4501-a888-92edec729f00',
'ext': 'mp4',
'title': r're:TG1 ore \d{2}:\d{2} del \d{2}/\d{2}/\d{4}',
'upload_date': r're:\d{8}',
'description': r're:.+',
}, },
}, { }, {
# HLS live stream with ContentItem in og:url
'url': 'http://www.rainews.it/dl/rainews/live/ContentItem-3156f2f2-dc70-4953-8e2f-70d7489d4ce9.html', 'url': 'http://www.rainews.it/dl/rainews/live/ContentItem-3156f2f2-dc70-4953-8e2f-70d7489d4ce9.html',
# HDS live stream, MD5 is unstable
'info_dict': { 'info_dict': {
'id': '3156f2f2-dc70-4953-8e2f-70d7489d4ce9', 'id': '3156f2f2-dc70-4953-8e2f-70d7489d4ce9',
'ext': 'mp4', 'ext': 'mp4',
'title': 'La diretta di Rainews24', 'title': 'La diretta di Rainews24',
}, },
'params': {
'skip_download': True,
},
}] }]
def _real_extract(self, url):
video_id = self._match_id(url)
webpage = self._download_webpage(url, video_id)
iframe_url = self._search_regex(
[r'<iframe[^>]+src="([^"]*/dl/[^"]+\?iframe\b[^"]*)"',
r'drawMediaRaiTV\(["\'](.+?)["\']'],
webpage, 'iframe', default=None)
if iframe_url:
if not iframe_url.startswith('http'):
iframe_url = compat_urlparse.urljoin(url, iframe_url)
return self.url_result(iframe_url)
content_item_id = self._search_regex(
r'initEdizione\((?P<q1>[\'"])ContentItem-(?P<content_id>[^\'"]+)(?P=q1)',
webpage, 'content item ID', group='content_id', default=None)
if content_item_id:
return self._extract_from_content_id(content_item_id, url)
try:
return self._extract_from_content_id(video_id, url)
except ExtractorError:
# no media data, only direct relinker
pass
relinker_url = compat_urlparse.urljoin(url, self._search_regex(
r'(?:var\s+videoURL|mediaInfo\.mediaUri)\s*=\s*(?P<q1>[\'"])(?P<url>(https?:)?//mediapolis\.rai\.it/relinker/relinkerServlet\.htm\?cont=\d+)(?P=q1)',
webpage, 'relinker URL', group='url'))
formats = self._extract_relinker_formats(relinker_url, video_id)
self._sort_formats(formats)
title = self._search_regex(
r'var\s+videoTitolo\s*=\s*([\'"])(?P<title>[^\'"]+)\1',
webpage, 'title', group='title', default=None) or self._og_search_title(webpage)
return {
'id': video_id,
'title': title,
'formats': formats,
}
def _extract_from_content_id(self, content_id, url): def _extract_from_content_id(self, content_id, url):
media = self._download_json( media = self._download_json(
'http://www.rai.tv/dl/RaiTV/programmi/media/ContentItem-%s.html?json' % content_id, 'http://www.rai.tv/dl/RaiTV/programmi/media/ContentItem-%s.html?json' % content_id,
content_id, 'Downloading video JSON') content_id, 'Downloading video JSON')
title = media['name'].strip()
media_type = media['type']
if 'Audio' in media_type:
relinker_info = {
'formats': {
'format_id': media.get('formatoAudio'),
'url': media['audioUrl'],
'ext': media.get('formatoAudio'),
}
}
elif 'Video' in media_type:
relinker_info = self._extract_relinker_info(media['mediaUri'], content_id)
else:
raise ExtractorError('not a media file')
self._sort_formats(relinker_info['formats'])
thumbnails = [] thumbnails = []
for image_type in ('image', 'image_medium', 'image_300'): for image_type in ('image', 'image_medium', 'image_300'):
thumbnail_url = media.get(image_type) thumbnail_url = media.get(image_type)
@ -261,20 +307,6 @@ class RaiIE(RaiBaseIE):
'url': compat_urlparse.urljoin(url, thumbnail_url), 'url': compat_urlparse.urljoin(url, thumbnail_url),
}) })
formats = []
media_type = media['type']
if 'Audio' in media_type:
formats.append({
'format_id': media.get('formatoAudio'),
'url': media['audioUrl'],
'ext': media.get('formatoAudio'),
})
elif 'Video' in media_type:
formats.extend(self._extract_relinker_formats(media['mediaUri'], content_id))
self._sort_formats(formats)
else:
raise ExtractorError('not a media file')
subtitles = {} subtitles = {}
captions = media.get('subtitlesUrl') captions = media.get('subtitlesUrl')
if captions: if captions:
@ -287,14 +319,89 @@ class RaiIE(RaiBaseIE):
'url': captions, 'url': captions,
}] }]
return { info = {
'id': content_id, 'id': content_id,
'title': media['name'], 'title': title,
'description': media.get('desc'), 'description': strip_or_none(media.get('desc')),
'thumbnails': thumbnails, 'thumbnails': thumbnails,
'uploader': media.get('author'), 'uploader': media.get('author'),
'upload_date': unified_strdate(media.get('date')), 'upload_date': unified_strdate(media.get('date')),
'duration': parse_duration(media.get('length')), 'duration': parse_duration(media.get('length')),
'formats': formats,
'subtitles': subtitles, 'subtitles': subtitles,
} }
info.update(relinker_info)
return info
def _real_extract(self, url):
video_id = self._match_id(url)
webpage = self._download_webpage(url, video_id)
content_item_id = None
content_item_url = self._html_search_meta(
('og:url', 'og:video', 'og:video:secure_url', 'twitter:url',
'twitter:player', 'jsonlink'), webpage, default=None)
if content_item_url:
content_item_id = self._search_regex(
r'ContentItem-(%s)' % self._UUID_RE, content_item_url,
'content item id', default=None)
if not content_item_id:
content_item_id = self._search_regex(
r'''(?x)
(?:
(?:initEdizione|drawMediaRaiTV)\(|
<(?:[^>]+\bdata-id|var\s+uniquename)=
)
(["\'])
(?:(?!\1).)*\bContentItem-(?P<id>%s)
''' % self._UUID_RE,
webpage, 'content item id', default=None, group='id')
content_item_ids = set()
content_item_ids.add(content_item_id)
if video_id not in content_item_ids:
content_item_ids.add(video_id)
for content_item_id in content_item_ids:
try:
return self._extract_from_content_id(content_item_id, url)
except GeoRestrictedError:
raise
except ExtractorError:
pass
relinker_url = self._search_regex(
r'''(?x)
(?:
var\s+videoURL|
mediaInfo\.mediaUri
)\s*=\s*
([\'"])
(?P<url>
(?:https?:)?
//mediapolis(?:vod)?\.rai\.it/relinker/relinkerServlet\.htm\?
(?:(?!\1).)*\bcont=(?:(?!\1).)+)\1
''',
webpage, 'relinker URL', group='url')
relinker_info = self._extract_relinker_info(
urljoin(url, relinker_url), video_id)
self._sort_formats(relinker_info['formats'])
title = self._search_regex(
r'var\s+videoTitolo\s*=\s*([\'"])(?P<title>[^\'"]+)\1',
webpage, 'title', group='title',
default=None) or self._og_search_title(webpage)
info = {
'id': video_id,
'title': title,
}
info.update(relinker_info)
return info