[hrti] Improve and add support for playlists (Closes #9482)

This commit is contained in:
Sergey M․ 2016-07-03 02:22:14 +07:00
parent 95cf60e826
commit e3755a624b
No known key found for this signature in database
GPG Key ID: 2C393E0F18A9236D

View File

@ -2,148 +2,201 @@
from __future__ import unicode_literals from __future__ import unicode_literals
import json import json
import re
from .common import InfoExtractor from .common import InfoExtractor
from ..compat import ( from ..compat import compat_HTTPError
compat_HTTPError
)
from ..utils import ( from ..utils import (
clean_html,
ExtractorError,
int_or_none,
parse_age_limit,
sanitized_Request, sanitized_Request,
ExtractorError try_get,
) )
class HRTiIE(InfoExtractor): class HRTiBaseIE(InfoExtractor):
''' """
Information Extractor for Croatian Radiotelevision video on demand site Base Information Extractor for Croatian Radiotelevision
https://hrti.hrt.hr video on demand site https://hrti.hrt.hr
Reverse engineered from the JavaScript app in app.min.js Reverse engineered from the JavaScript app in app.min.js
''' """
_NETRC_MACHINE = 'hrti' _NETRC_MACHINE = 'hrti'
APP_LANGUAGE = 'hr' _APP_LANGUAGE = 'hr'
APP_VERSION = '1.1' _APP_VERSION = '1.1'
APP_PUBLICATION_ID = 'all_in_one' _APP_PUBLICATION_ID = 'all_in_one'
_API_URL = 'http://clientapi.hrt.hr/client_api.php/config/identify/format/json'
_VALID_URL = r'https?://hrti.hrt.hr/#/video/show/(?P<id>[0-9]+)/(?P<name>(\w|-)+)?'
_TEST = {
'url': 'https://hrti.hrt.hr/#/video/show/2181385/republika-dokumentarna-serija-16-hd',
'info_dict': {
'id': '2181385',
'ext': 'mp4',
'name': 'REPUBLIKA, dokumentarna serija (4_6)-2251938',
},
'skip': 'Requires login'
}
def _initialize_api(self): def _initialize_api(self):
'''Initializes the API and obtains the required urls''' init_data = {
api_url = 'http://clientapi.hrt.hr/client_api.php/config/identify/format/json' 'application_publication_id': self._APP_PUBLICATION_ID
app_data = json.dumps({ }
'application_publication_id': HRTiIE.APP_PUBLICATION_ID
})
self.uuid = self._download_json(api_url, None, note='Getting UUID',
errnote='Unable to obtain an UUID',
data=app_data)['uuid']
app_data = json.dumps({ uuid = self._download_json(
'uuid': self.uuid, self._API_URL, None, note='Downloading uuid',
'application_publication_id': HRTiIE.APP_PUBLICATION_ID, errnote='Unable to download uuid',
'screen_height': 1080, data=json.dumps(init_data).encode('utf-8'))['uuid']
'screen_width': 1920,
'os': 'Windows',
'os_version': 'NT 4.0',
'device_model_string_id': 'chrome 42.0.2311.135',
'application_version': HRTiIE.APP_VERSION
})
req = sanitized_Request(api_url, data=app_data) app_data = {
'uuid': uuid,
'application_publication_id': self._APP_PUBLICATION_ID,
'application_version': self._APP_VERSION
}
req = sanitized_Request(self._API_URL, data=json.dumps(app_data).encode('utf-8'))
req.get_method = lambda: 'PUT' req.get_method = lambda: 'PUT'
resources = self._download_json( resources = self._download_json(
req, None, note='Getting API endpoint and session information', req, None, note='Downloading session information',
errnote='Unable to get endpoint and session information', errnote='Unable to download session information')
headers={'Content-type': 'application/json'})
self._session_id = resources['session_id']
self.session_id = resources['session_id']
modules = resources['modules'] modules = resources['modules']
self.search_url = modules['vod_catalog']['resources']['search']['uri'] self._search_url = modules['vod_catalog']['resources']['search']['uri'].format(
self.search_url = self.search_url.format( language=self._APP_LANGUAGE,
language=HRTiIE.APP_LANGUAGE, application_id=self._APP_PUBLICATION_ID)
application_id=HRTiIE.APP_PUBLICATION_ID)
self.login_url = modules['user']['resources']['login']['uri'] self._login_url = (modules['user']['resources']['login']['uri'] +
self.login_url = self.login_url.format(session_id=self.session_id) '/format/json').format(session_id=self._session_id)
self.login_url += '/format/json'
self.logout_url = modules['user']['resources']['logout']['uri'] self._logout_url = modules['user']['resources']['logout']['uri']
def _login(self): def _login(self):
'''Performs a login to the webservice'''
(username, password) = self._get_login_info() (username, password) = self._get_login_info()
# TODO: figure out authentication with cookies
if username is None or password is None: if username is None or password is None:
self.raise_login_required() self.raise_login_required()
auth_data = json.dumps({ auth_data = {
'username': username, 'username': username,
'password': password, 'password': password,
}) }
try: try:
auth_info = self._download_json( auth_info = self._download_json(
self.login_url, None, note='Authenticating', self._login_url, None, note='Logging in', errnote='Unable to log in',
errnote='Unable to log in', data=auth_data) data=json.dumps(auth_data).encode('utf-8'))
except ExtractorError as ee: except ExtractorError as e:
if isinstance(ee.cause, compat_HTTPError) and ee.cause.code == 406: if isinstance(e.cause, compat_HTTPError) and e.cause.code == 406:
raise ExtractorError('Unable to login, ' + auth_info = self._parse_json(e.cause.read().encode('utf-8'), None)
'incorrect username and/or password') else:
raise raise
self.token = auth_info['secure_streaming_token'] error_message = auth_info.get('error', {}).get('message')
self.access_token = auth_info['session_token'] if error_message:
raise ExtractorError(
'%s said: %s' % (self.IE_NAME, error_message),
expected=True)
self.logout_url = self.logout_url.format(session_id=self.session_id, self._token = auth_info['secure_streaming_token']
access_token=self.access_token)
self.logout_url += '/format/json'
def _real_initialize(self): def _real_initialize(self):
'''Performs necessary operations so that the information extractor is
ready for operation'''
self._initialize_api() self._initialize_api()
self._login() self._login()
def _logout(self):
'''Performs logout from the webservice''' class HRTiIE(HRTiBaseIE):
self._download_json(self.logout_url, None, note='Logout', _VALID_URL = r'''(?x)
errnote='Unable to log out', fatal=False) (?:
hrti:(?P<short_id>[0-9]+)|
https?://
hrti\.hrt\.hr/\#/video/show/(?P<id>[0-9]+)/(?P<display_id>[^/]+)?
)
'''
_TESTS = [{
'url': 'https://hrti.hrt.hr/#/video/show/2181385/republika-dokumentarna-serija-16-hd',
'info_dict': {
'id': '2181385',
'display_id': 'republika-dokumentarna-serija-16-hd',
'ext': 'mp4',
'title': 'REPUBLIKA, dokumentarna serija (1/6) (HD)',
'description': 'md5:48af85f620e8e0e1df4096270568544f',
'duration': 2922,
'view_count': int,
'average_rating': int,
'episode_number': int,
'season_number': int,
'age_limit': 12,
},
'skip': 'Requires account credentials',
}, {
'url': 'https://hrti.hrt.hr/#/video/show/2181385/',
'only_matching': True,
}, {
'url': 'hrti:2181385',
'only_matching': True,
}]
def _real_extract(self, url): def _real_extract(self, url):
'''Extract the data necessary to download the video''' mobj = re.match(self._VALID_URL, url)
video_id = self._match_id(url) video_id = mobj.group('short_id') or mobj.group('id')
display_id = mobj.group('display_id') or video_id
metadata_url = self.search_url + \ video = self._download_json(
'/video_id/{video_id}/format/json'.format(video_id=video_id) '%s/video_id/%s/format/json' % (self._search_url, video_id),
display_id, 'Downloading video metadata JSON')['video'][0]
metadata = self._download_json(metadata_url, video_id, title_info = video['title']
note='Getting video metadata') title = title_info['title_long']
video = metadata['video'][0]
title_info = video.get('title', {})
title = title_info.get('title_long')
description = title_info.get('summary_long')
movie = video['video_assets']['movie'][0] movie = video['video_assets']['movie'][0]
url = movie['url'].format(TOKEN=self.token) m3u8_url = movie['url'].format(TOKEN=self._token)
formats = self._extract_m3u8_formats(
formats = self._extract_m3u8_formats(url, video_id, 'mp4') m3u8_url, display_id, 'mp4', entry_protocol='m3u8_native',
m3u8_id='hls')
self._sort_formats(formats) self._sort_formats(formats)
self._logout() description = clean_html(title_info.get('summary_long'))
age_limit = parse_age_limit(video.get('parental_control', {}).get('rating'))
view_count = int_or_none(video.get('views'))
average_rating = int_or_none(video.get('user_rating'))
duration = int_or_none(movie.get('duration'))
return { return {
'id': video_id, 'id': video_id,
'display_id': display_id,
'title': title, 'title': title,
'description': description, 'description': description,
'duration': duration,
'view_count': view_count,
'average_rating': average_rating,
'age_limit': age_limit,
'formats': formats, 'formats': formats,
} }
class HRTiPlaylistIE(HRTiBaseIE):
_VALID_URL = r'https?://hrti.hrt.hr/#/video/list/category/(?P<id>[0-9]+)/(?P<display_id>[^/]+)?'
_TESTS = [{
'url': 'https://hrti.hrt.hr/#/video/list/category/212/ekumena',
'info_dict': {
'id': '212',
'title': 'ekumena',
},
'playlist_mincount': 8,
'skip': 'Requires account credentials',
}, {
'url': 'https://hrti.hrt.hr/#/video/list/category/212/',
'only_matching': True,
}]
def _real_extract(self, url):
mobj = re.match(self._VALID_URL, url)
category_id = mobj.group('id')
display_id = mobj.group('display_id') or category_id
response = self._download_json(
'%s/category_id/%s/format/json' % (self._search_url, category_id),
display_id, 'Downloading video metadata JSON')
video_ids = try_get(
response, lambda x: x['video_listings'][0]['alternatives'][0]['list'],
list) or [video['id'] for video in response.get('videos', []) if video.get('id')]
entries = [self.url_result('hrti:%s' % category_id) for category_id in video_ids]
return self.playlist_result(entries, category_id, display_id)